![a. The pupa considered in reference to its adaptation to its
surroundings and its relation to phylogeny
The form of the pupa is a very variable one, as even in Lepidoptera
it is not entirely easy to draw the line between a pupa libera and a
pupa obtecta (Fig. 578); and though the period is one of inactivity,
yet when they are not in cocoons or in the earth in subterranean
cells, their form is more or less variable and adapted to changes in
their surroundings. Even in the obtected pupa of butterflies, there is,
as every one knows, considerable variability of shape and of
armature, which seems to be in direct adaptability to the nature of
their environment. Scudder has well shown that in certain
chrysalids, such as those of the Nymphalidæ, which are variously
tuberculated, and hang suspended by the tail, and often hibernate,
these projections serve to protect the body. All chrysalids with
projections or ridges on different parts of the body, being otherwise
unprotected, move freely when struck by gusts of wind, hence “the
greater the danger to the chrysalis from surrounding objects, the
greater its protection by horny tubercles and roughened callous
ridges.” The greater the protection possessed in other ways, as by
firm swathing or a safe retreat, the smoother the surface of the body
and the more regular and rounded its contours. The tendency to
protection by tubercles is especially noticeable in certain South
American chrysalids of nymphalid butterflies. This response to the
stimuli of blows or shocks is also accompanied by a sensitiveness to
the stimulus of too strong light.
Previously Scudder[103]
had made the important suggestion that the
smooth crescent-shaped belt of the “glazed eye” or “eyepiece” of
chrysalids is, as an external covering of the eye, midway between that
of the caterpillar and the perfect insect, and he asks: “May it not be a
relic of the past, the external organ of what once was? And are we to
look upon this as our hint that the archaic butterfly in its
transformations passed through an active pupal stage, like the lowest
insect of to-day, when its limbs were unsheathed, its appetite
unabated?” etc. Scudder also shows that “the expanded base of the
sheath covering the tongue affords protection also to the palpi which
lie beneath and beside the tongue.”](https://image.slidesharecdn.com/15902-250225231715-1b73df3e/85/Mobile-Computing-1st-edition-Edition-Safari-63-320.jpg)
![Fig. 590.
—Pupa of
Talæpori
a: a,
cocoon-
cutter;
with
vestiges
of four
pairs of
abdomin
al legs,
and the
cremaste
r.
“Each pupa first sawed through the cocoon near its juncture with the
leaf and worked its way through the gap, by means of the minute
backward-directed spines upon its back, until it reached the upper
cuticle of the leaf. Through this cuticle it sawed in the same way that it
did through the cocoon. The hole was in each case just large enough to
permit the chrysalis to work its way out, holding it firmly when partly
emerged. When half-way out it stopped, and presently the skin split
across the back of the neck and down in front along the antennal
sheaths, and allowed the moth to emerge.”[104]
We have observed and figured the cocoon-breaker in
Bucculatrix, Talæporia (Fig. 590, a), Thyridopteryx, and
Œceticus, and rough knobs or slight projection answering
the purpose in Hepialidæ, Megalopyge, Zeuzera, and in
Datana.[105]
See also the spine on the head of Sesia
tipuliformis (Fig. 578).
The imago of the attacine moths cuts or saws through its
cocoon by means of a pair of large, stout, black spines
(sectores coconis), one on each side of the thorax at the
base of the fore wings (Fig. 591), and provided with five or
six teeth on the cutting edge (C, D).
Our attention[106] was drawn to this subject by a rustling, cutting,
and tearing noise issuing from a cocoon of Actias luna. On
examination a sharp black point was seen moving to and fro, and then
another, until both points had cut a rough irregular slit, through which
the shoulder of the moth could be seen vigorously moving from side to
side. The hole or slit was made in one or two minutes, and the moth
worked its way at once out of the slit. The cocoon was perfectly dry.
The cocoon-cutter occurs in all the American genera, in Samia
cynthia, and is large and well marked in the European Saturnia
pavonia-minor and Endromis versicolora. In Bombyx mori the spines
are not well marked, and they are quite different from those in the Attaci. There
are three sharp points, being acute angles of the pieces at the base of the wing, and
it must be these spines which at times perform the cutting through of the threads
of the cocoon described by Réaumur, and which he thought was done by the facets
of the eyes. It is well known that in order to guard against the moths cutting the
threads, silkraisers expose the cocoon to heat sufficient to destroy the enclosed
pupa. In Platysamia the cocoon-cutters, though well developed, do not appear to
be used at all, and the pupa, like that of the silkworm and other moths protected by
a cocoon, moistens the silk threads by a fluid issuing from the mouth, which also
moistens the hairs of the head and thorax, together with the antennæ. It remains to
be seen whether these structures are only occasionally used, and whether the
emission of the fluid is not the usual and normal means of egress of the moth from](https://image.slidesharecdn.com/15902-250225231715-1b73df3e/85/Mobile-Computing-1st-edition-Edition-Safari-67-320.jpg)
![Fig. 591.—Cocoon-
cutter of the Luna
moth: front view of
the moth with the
shoulders elevated
and the rudimentary
wings hanging down:
s, cocoon-cutter; p,
patagium. B,
represents another
specimen with fully
developed wings: ms,
scutum; st, scutellum
of the mesothoracic
segment; s, cocoon-
cutter, which is
evidently a
modification of one
of the pieces at the
base of the fore
wings; it is
surrounded by
membrane, allowing
free movement. C
and D, different
views of the spine,
magnified, showing
the five or six
irregular teeth on the
cutting edge.
Fig. 592.—Larva and
pupa of a wood-wasp
(Rhopalum),
enlarged: h,
temporary locomotive
tubercles on head of
pupa.—Trouvelot del.
its cocoon. Dr. Chapman remarks
that throughout the obtected
moths “there are many devices for
breaking through the cocoon:
specially constructed weak places
in the cocoon, softening fluid,
applied by the moth, assisted by
special appliances of diverse sorts,
such as in Hybocampa[107] and
Attacus,” etc.
As to the fluid mentioned
above, Trouvelot states that it is
secreted during the last few days
of the pupa state, and is a
dissolvent for the gum so firmly
uniting the fibres of the cocoon.
“This liquid is composed in great
part of bombycic acid.” (Amer.
Naturalist, i, p. 33.)
The pupa of the dipterous genus
Sciara (S. ocellaris O. S.)
resembles a tineid pupa, and before transforming emerges
for about two-thirds of its length from the cocoon; the
pupa-skin remaining firmly attached in this position.[108]
Certain hymenopterous pupæ are provided with
temporary deciduous conical processes. Thus we have
observed in the pupa of Rhopalum pedicellatum two very
prominent acute tubercles between the eyes (h, Fig. 592).
As the cocoon is very slight, these may be of use either in
extracting itself from the silken threads or in pushing its
way along before emerging from the tunnel in the stem of
plants. (See also p. 611.)
c. The cremaster
Although this structure is in general confined
to lepidopterous pupæ, and is not always present
even in them, since it is purely adaptive in its
nature, yet on account of its singular mode of development from the
larval organs, and the accompanying changes in the pupal abdomen,
it should be mentioned in this connection. The cremaster is the stout,
triangular, flattened, terminal spine of the abdomen, which aids the](https://image.slidesharecdn.com/15902-250225231715-1b73df3e/85/Mobile-Computing-1st-edition-Edition-Safari-68-320.jpg)
![pupa in working its way out of the earth when the pupa is
subterranean, or in the pupa of silk-spinning caterpillars its
armature of secondary hooks and curved setæ enables it to retain its
hold on the threads of the interior of its cocoon after the pupa has
partially emerged from the cocoon, restraining it, as Chapman well
says, “at precisely that degree of emergence from the cocoon that is
most desirable.” He also informs us that while in the “pupæ
incompletæ the cremaster is attached to an extensible cable, which
always allows some emergence of the pupa, in the pupæ obtectæ
there is no doubt but that in such cases as the Ichthyuræ, Acronyctæ,
and many others, it retains the pupal case in the same position
within the cocoon that the living pupa occupied; this is also very
usually the case in the Geometræ and in the higher tineids (my
pyraloids).”
In many of the more generalized moths there is no cremaster (Micropteryx,
Gracilaria, Prodoxus, Tantura, Talæporia, Psychidæ, Hepialidæ, Zeuzera, Nola,
Harrisina), though in Tischeria and Talæporia (Fig. 590, but not in Solenobia) and
Psychidæ, two stout terminal spines perform the office of a cremaster, or there are
simply curved setæ on the rounded, unarmed end of the abdomen, as in Solenobia.
In the obtected Lepidoptera, for example in such a group as the Notodontidæ,
where the cremaster is present, though variable in shape, it may from disuse,
owing to the dense cocoon, be without the spines and hooks in Cerura, or the
cremaster itself is entirely wanting in Gluphisia, and only partially developed in
Notodonta. In the butterflies whose pupæ are suspended (Suspensi), the cremaster
is especially well developed. Reference might here be made to the temporary pupal
structures in certain generalized moths, which take the place of a cremaster, such
as the transverse terminal row of spines in Tinea, the two stout spines in Tischeria,
and the dense rough integument and thickened callosities of the pupal head and
end of abdomen of Phassus, which bores in trees with very hard wood; also the
numerous stout spines at the end and sides of the abdomen in Ægerians. These
various projections and spines, besides acting as anchors and grappling hooks, in
some cases serve to resist strains and blows, and have undoubtedly, like the
armature in the larvæ and imagines of other insects, arisen in response to
intermittent or occasional pressure, stresses, and impacts.
Mode of formation of the cremaster and suspension of
the chrysalis in butterflies.—We are indebted to Riley[109]
for an
explanation of the way the cremaster has originated, his observations
having been made on species of over a dozen genera of butterflies
(Suspensi).
He shows that the cremaster is the homologue of the suranal plate
of the larva.[110]
The preliminary acts of the larva have been observed](https://image.slidesharecdn.com/15902-250225231715-1b73df3e/85/Mobile-Computing-1st-edition-Edition-Safari-69-320.jpg)
![Fig. 593.—Shrunken larval skin
of Vanessa antiopa, cut open
from the back and showing (mr)
the retaining membrane, (rl)
the rectal ligament, and (tl) the
tracheal ligaments.
The structures in the chrysalis are, first, the cremaster, with its dorsal (Fig. 594,
dcr) and ventral (vcr) ridges, and the cremastral hook-pad (chp), said by Riley to
be “thickly studded with minute but stout hooks, which are sometimes compound
or furnished with barbs, very much as are some of our fishing-hooks, and which
are most admirably adapted to the purpose for which they are intended.”
Secondly, there are the other structures, viz., the sustainers (sustentors), two
projections which Riley states “homologize with the soles (plantæ) of the anal
prolegs, which take on various forms (3), but are always directed forward so as
easily to catch hold of the retaining membrane.” These sustentors are, however, as
Jackson[111] has shown, and as we are satisfied, the vestiges of the anal legs.](https://image.slidesharecdn.com/15902-250225231715-1b73df3e/85/Mobile-Computing-1st-edition-Edition-Safari-71-320.jpg)
![Malpighi had previously observed the same fact in the silkworm,
perceiving that before pupation the antennæ are concealed in the
head of the larva, where they occupy the place previously taken by
the mandibular muscles; also that the legs of the moth grew in those
of the larva, and that the wings developed from the sides of the
worm.
Even Réaumur (1734) remarked: “Les parties du papillon cachées
sous le fourreau de chenille sont d’autant plus faciles à trouver que la
transformation est plus proche. Elles y sont neanmoins de tout
temps.” He also believed in the simultaneous existence of two
distinct beings in the insect. “Il serait très curieux de connaître toutes
les communications intimes qui sont entre la chenille et le papillon....
La chenille hache, broye, digere les aliments qu’elle distribué au
papillon; comme les mères préparent ceux qui sont portés aux fœtus.
Notre chenille en un mot est destineé à nourrir et à defendre le
papillon qu’elle renferme.” (T. i, 8e
Mémoire, p. 363.)
Lyonet (1760), even, did not expose the error of this view that the
larva enveloped the pupa and imago, and, as Gonin says, it was
undoubtedly because he did not use for his dissections of the
caterpillar of Cossus any specimens about to pupate. Yet he detected
the wing-germs and those of the legs, stating that he presumed the
bodies he saw to be the rudiments of the legs of the moth (p. 450).
Herold, in his work on the development of the butterfly (1815), was
the first to object to this erroneous theory, showing that the wings
did not become visible until the very end of larval life; that as the
larval organs disappear, they are transformed or are replaced by
entirely new organs, which is not reconcilable with a simple putting
off of the outer envelope. The whole secret of metamorphosis, in
Herold’s opinion, consisted in this fact, that the butterfly in the larva
state increases and accumulates a supply of fat until it has reached
the volume of the perfect state; then it begins the chrysalis period,
during which the organs are developed and take their definite form.
[112]
(Abstract mostly from Gonin.) Still the old ideas prevailed, and
even Lacordaire, in his Introduction à l’Entomologie published in
1834, held on to Swammerdam’s theory, declaring that “a caterpillar
is not a simple animal, but compound,” and he actually goes so far as
to say that “a caterpillar, at first scarcely as large as a bit of thread,
contains its own teguments threefold and even eightfold in number,](https://image.slidesharecdn.com/15902-250225231715-1b73df3e/85/Mobile-Computing-1st-edition-Edition-Safari-78-320.jpg)
Mobile Computing 1st edition Edition Safari
- 1. Get the full ebook with Bonus Features for a Better Reading Experience on ebookgate.com Mobile Computing 1st edition Edition Safari https://ebookgate.com/product/mobile-computing-1st-edition- edition-safari/ OR CLICK HERE DOWLOAD NOW Download more ebook instantly today at https://ebookgate.com
- 2. Instant digital products (PDF, ePub, MOBI) available Download now and explore formats that suit you... Multimodality in Mobile Computing and Mobile Devices Methods for Adaptable Usability 1st Edition Stan Kurkovsky https://ebookgate.com/product/multimodality-in-mobile-computing-and- mobile-devices-methods-for-adaptable-usability-1st-edition-stan- kurkovsky/ ebookgate.com Mobile Commerce and Wireless Computing Systems 1st Edition Geoffrey Elliott https://ebookgate.com/product/mobile-commerce-and-wireless-computing- systems-1st-edition-geoffrey-elliott/ ebookgate.com HANDBOOK OF WIRELESS NETWORKS AND MOBILE COMPUTING 1st Edition Ivan Stojmenovic https://ebookgate.com/product/handbook-of-wireless-networks-and- mobile-computing-1st-edition-ivan-stojmenovic/ ebookgate.com Mobile Computing Concepts Methodologies Tools and Applications 1st Edition David Taniar https://ebookgate.com/product/mobile-computing-concepts-methodologies- tools-and-applications-1st-edition-david-taniar/ ebookgate.com
- 3. Programming in C 2nd Edition Safari https://ebookgate.com/product/programming-in-c-2nd-edition-safari/ ebookgate.com Separation Process Engineering Second Edition Safari https://ebookgate.com/product/separation-process-engineering-second- edition-safari/ ebookgate.com Context Aware Mobile and Ubiquitous Computing for Enhanced Usability Adaptive Technologies and Applications 1st Edition Dragan Stojanovic https://ebookgate.com/product/context-aware-mobile-and-ubiquitous- computing-for-enhanced-usability-adaptive-technologies-and- applications-1st-edition-dragan-stojanovic/ ebookgate.com Managing the Insider Threat No Dark Corners 1st edition Edition Safari https://ebookgate.com/product/managing-the-insider-threat-no-dark- corners-1st-edition-edition-safari/ ebookgate.com Documenting Software Architectures Views and Beyond Second Edition Safari https://ebookgate.com/product/documenting-software-architectures- views-and-beyond-second-edition-safari/ ebookgate.com
- 7. A01_GARGxxxx_01_SE_FM.qxd 4/5/10 6:26 PM Page ii This page intentionally left blank
- 8. Kumkum Garg Mobile Computing Theory and Practice A01_GARGxxxx_01_SE_FM.qxd 4/19/10 3:17 PM Page iii Delhi • Chennai • Chandigarh Department of Electronics and Computer Engineering Indian Institute of Technology Roorkee
- 9. Assistant Acquisitions Editor: Pradeep Banerjee Assistant Production Editor: Amrita Naskar Composition: Aptara®, Inc. Copyright © 2010 Dorling Kindersley (India) Pvt. Ltd This book is sold subject to the condition that it shall not, by way of trade or otherwise, be lent, resold, hired out, or otherwise circulated without the publisher’ s prior written consent in any form of binding or co ver other than that in which it is published and without a similar condition including this condition being imposed on the subsequent purchaser and without limiting the rights under cop yright reser ved above, no part of this publication ma y be reproduced, stored in or introduced into a retrie val system, or transmitted in an y form or by any means (electronic, mechanical, photocopying, recording or otherwise), without the prior written per mission of both the cop yright owner and the abo ve-mentioned publisher of this book. ISBN 978-81-317-3166-6 10 9 8 7 6 5 4 3 2 1 Published by Dorling Kindersley (India) Pvt. Ltd, licensees of P earson Education in South Asia. Head Office: 7th Floor , Knowledge Boule vard, A-8(A), Sector – 62, Noida, UP 201309, India. Registered Office: 11 Community Centre, Panchsheel P ark, New Delhi 110017, India. A01_GARGxxxx_01_SE_FM.qxd 4/19/10 3:05 PM Page iv
- 10. Contents Preface xiii 1 Introduction to Mobility 1 1.1 Process migration 1 1.2 Mobile computing 2 1.3 Mobile agents 3 1.4 Technical issues for mobility 4 1.5 Personal communication systems 4 1.6 Context-aware computing 5 1.7 Outline of the book 6 1.8 Summary 7 Problems 7 Multiple-choice questions 8 Further reading 9 2 Wireless and Cellular Communication 11 2.1 The electromagnetic spectr um 11 2.1.1 Radio waves 12 2.1.2 Microwaves 12 2.1.3 Infrared waves 12 2.1.4 Lightwaves 13 2.2 Communication satellites 13 2.2.1 Geostationar y satellites 14 2.2.2 Medium ear th orbit satellites 14 2.2.3 Low earth orbit satellites 14 2.3 Multiple-access schemes 15 2.3.1 FDMA—F requency division multiple access 16 2.3.2 TDMA—Time division multiple access 16 2.3.3 CDMA—Code division multiple access 17 2.4 Cellular communication 18 2.4.1 The first generation (1G): 1980 18 v A01_GARGxxxx_01_SE_FM.qxd 4/5/10 6:26 PM Page v
- 11. 2.4.2 The second generation (2G): 1992 19 2.4.3 The 2.5 generation (2.5G): 1996 20 2.4.4 The third generation (3G): 2000 ⫹ 20 2.4.5 The 3.5 generation (3.5G): 2000 ⫹ 21 2.4.6 The fourth generation (4G): 2002 ⫹ 21 2.5 Summary 22 Problems 22 Multiple-choice questions 23 Further reading 24 3 Wireless Networ ks 25 3.1 The need for ne w wireless standards 26 3.2 IEEE 802.11 WLAN standard 27 3.2.1 Physical la yer 27 3.2.2 MAC layer 29 3.2.3 Frame str ucture 32 3.2.4 Services 32 3.3 Bluetooth 33 3.3.1 Advantages of Bluetooth 35 3.3.2 Bluetooth applications 35 3.3.3 Bluetooth protocol stack 35 3.3.4 Bluetooth tracking ser vices 37 3.3.5 Bluetooth frame str ucture 38 3.4 Infrared systems 39 3.5 HiperLAN 40 3.6 The IEEE 802.16 WiMAX standard 41 3.7 Comparison of wireless technologies 42 3.8 Summary 43 Problems 44 Multiple-choice questions 44 Further reading 45 4 Logical Mobility I— Migrating Processes 47 4.1 What is a process? 47 4.2 Process migration 48 4.3 The steps in process migration 48 4.4 The advantages of process migration 52 4.5 Applications of process migration 53 4.6 Alternatives to process migration 53 4.7 Summary 54 Problems 54 Multiple-choice questions 55 Further reading 56 vi Mobile Computing A01_GARGxxxx_01_SE_FM.qxd 4/5/10 6:26 PM Page vi
- 12. 5 Physical Mobility 57 5.1 The requirements for ph ysical mobility 57 5.1.1 Wireless communication 57 5.1.2 Mobility 58 5.1.3 Portability 59 5.2 Overview of IPv4 and IPv6 61 5.2.1 IPv4 61 5.2.2 IPv6 62 5.3 Mobile IP 62 5.3.1 Goals of mobile IP 62 5.3.2 Applicability 63 5.3.3 Mobility suppor t in IPv4 63 5.3.4 Mobility suppor t in IPv6 66 5.4 Cellular IP 67 5.4.1 The cellular IP access networ k 68 5.4.2 Routing and paging cache 69 5.5 TCP for mobility 69 5.5.1 Indirect TCP 70 5.5.2 Snooping TCP 71 5.5.3 Mobile TCP 72 5.6 Mobile databases 73 5.6.1 Design issues 73 5.6.2 Problems in mobile databases 74 5.6.3 Commercially a vailable systems 74 5.7 The CODA file system—A case study 74 5.7.1 Cache manager V enus 75 5.7.2 Venus states 75 5.7.3 Design criteria 77 5.8 Summary 78 Problems 78 Multiple-choice questions 79 Further reading 80 6 Mobile Ad Hoc Networ ks 81 6.1 MANET characteristics 81 6.2 Classification of MANETs 82 6.3 Technologies for ad hoc networ ks 83 6.4 Routing in MANETs 83 6.4.1 Traditional routing protocols 83 6.4.2 Requirements for routing protocols 84 6.4.3 Classification of routing protocols 84 6.5 Proactive routing protocols — The DSDV protocol 85 6.5.1 Example of DSD V operation 86 6.6 Reactive routing protocols 88 Contents vii A01_GARGxxxx_01_SE_FM.qxd 4/5/10 6:26 PM Page vii
- 13. 6.6.1 Dynamic source routing (DSR) 89 6.6.1.1 Route disco very in DSR 89 6.6.1.2 Route maintenance in DSR 91 6.6.1.3 Route cache in DSR 91 6.6.2 Adaptive on-demand distance vector protocol 92 6.6.2.1 Route disco very in AODV 92 6.6.2.2 Route maintenance in A ODV 93 6.7 Comparison betw een DSR and A ODV 96 6.8 Summary 97 Problems 98 Multiple-choice questions 98 Further reading 100 7 Wireless Sensor Networ ks 101 7.1 Applications of wireless sensor networ ks 101 7.2 Differences from mobile ad hoc networ ks 103 7.3 Design issues 104 7.4 WSN architecture 104 7.4.1 Sensor hard ware components 105 7.4.2 WSN communications architecture 105 7.5 Routing protocols for WSN 106 7.5.1 Data-centric protocols 106 7.5.1.1 Flooding and gossiping 107 7.5.1.2 Sensor protocols for infor mation via negotiation (SPIN) 107 7.5.2 Hierarchical protocols 108 7.5.2.1 Low-energy adaptive clustering hierarch y 108 7.5.2.2 PEGASIS 109 7.5.2.3 TEEN and APTEEN 109 7.5.3 Location-based protocols 110 7.6 Case study 110 7.6.1 The MICA mote 110 7.6.2 TinyOS 111 7.7 Development wor k in WSN 112 7.8 Summary 112 Problems 113 Multiple-choice questions 113 Further reading 115 8 Mobile Handheld Devices 117 8.1 Characteristics of PD As 117 8.1.1 The ARM processor 119 8.1.2 Network connectivity 119 viii Mobile Computing A01_GARGxxxx_01_SE_FM.qxd 4/5/10 6:26 PM Page viii
- 14. 8.2 Palm handhelds 120 8.3 The Palm OS operating system 121 8.3.1 Memory management 121 8.3.2 Communication and networ king 122 8.4 HP handhelds 122 8.5 Windows CE 123 8.5.1 Memory architecture 124 8.5.2 Memory management 124 8.5.3 Processes and threads 124 8.5.4 Scheduling 125 8.5.5 Real-time perfor mance 125 8.6 The Windows Mobile operating system 125 8.7 Nokia handhelds 127 8.7.1 Specifications of Nokia 9210 127 8.7.2 Features 128 8.8 Symbian operating system 129 8.8.1 Design 129 8.8.2 Symbian str ucture 130 8.9 Summary 130 Problems 131 Multiple-choice questions 131 Further reading 132 9 The Mobile Inter net and Wireless W eb 133 9.1 The Web programming model 133 9.2 The WAP programming model 134 9.3 WAP protocol stack 135 9.4 Information-mode (I-mode) 136 9.5 WAP 2.0 136 9.6 WAP gateway 137 9.6.1 Push operation 138 9.6.2 Push message for mat (using PAP) 140 9.6.3 Pull operation 141 9.7 Summary 141 Problems 142 Multiple-choice questions 142 Further reading 144 10 Logical Mobility II — Mobile Agents 145 10.1 Mobile agents 146 10.2 Characteristics of mobile agents 146 10.2.1 Architecture 147 10.2.2 Mobile code and agents 147 Contents ix A01_GARGxxxx_01_SE_FM.qxd 4/5/10 6:26 PM Page ix
- 15. 10.2.3 Mobile agents and process migration 147 10.2.4 Client/ser ver and mobile agent architectures 147 10.3 Requirements for mobile agent systems 148 10.3.1 Portability 148 10.3.2 Ubiquity 148 10.3.3 Network communication 148 10.3.4 Server security 148 10.3.5 Agent security 149 10.3.6 Resource accounting 149 10.4 Mobile agent platfor ms 149 10.4.1 Aglets 150 10.4.1.1 The aglet object model 150 10.4.1.2 Aglet communication 151 10.4.1.3 The aglet e vent model 152 10.4.2 Agent Tcl 152 10.4.2.1 Agent Tcl architecture 152 10.4.2.2 Agent Tcl applications 155 10.4.3 PMADE 155 10.4.3.1 Agent submitter 156 10.4.3.2 Agent host 158 10.4.3.3 Communication manager s 158 10.4.3.4 State manager s 159 10.4.3.5 Persistence manager 160 10.4.3.6 Security manager 160 10.5 Java and mobile agents 161 10.5.1 Advantages of Ja va 161 10.5.2 Shortcomings of Ja va 161 10.6 Summary 162 Problems 162 Multiple-choice questions 163 Further reading 164 11 Security Issues in Mobile Computing 167 11.1 Security threats to wireless networ ks 168 11.2 IEEE 802.11 security through WEP 169 11.2.1 WEP security features of 802.11 wireless LANs 169 11.2.1.1 Authentication 169 11.2.1.2 Confidentiality 170 11.2.1.3 Integrity 171 11.3 Bluetooth security 172 11.4 WAP 2.0 security 174 11.5 Summary 174 Problems 175 Multiple-choice questions 175 Further reading 177 x Mobile Computing A01_GARGxxxx_01_SE_FM.qxd 4/5/10 6:26 PM Page x
- 16. 12 Design and Programming Projects 179 12.1 Implementation of mobile IP 179 12.2 Comparison betw een AODV and DSR protocols 182 12.3 Bluetooth application 184 12.4 Design of a W AP gateway 189 12.5 Mobile agents for networ k monitoring 190 12.6 An IEEE 802.11 LAN for a typical student hostel 194 12.7 An application using wireless sensor networ ks 196 12.8 Summary 198 Problems 198 Multiple-choice questions 198 Further reading 200 Appendix A—Ja va Networ k Programming 201 A.1 Java programming language 201 A.2 Socket programming 203 A.3 Remote procedure call (RPC) 205 A.4 Remote method in vocation (RMI) 207 Appendix B—Comparison Betw een Qualnet and NS2 211 Index 213 Contents xi A01_GARGxxxx_01_SE_FM.qxd 4/5/10 6:26 PM Page xi
- 17. A01_GARGxxxx_01_SE_FM.qxd 4/5/10 6:26 PM Page xii This page intentionally left blank
- 18. Preface M obile computing or computing-on-the-go is proving to be one of the most promising technological advances in computer science and engineering to date. With the advent and proliferation of portable, handheld hardware devices, equipped with wireless com- munication interfaces and carrying innovative applications and systems software, computing has now become truly ‘pervasive’ or ‘ubiquitous’. It is now commonplace to see people sitting in airport and hotel lounges, meeting rooms and even open spaces, keying away at their PDAs or laptops, checking e-mails and appointments, making to-do lists or just chatting with their friends. We are also looking at ‘smart dust’, in which thousands of miniature processing devices can be literally scattered in a battlefield or natural calamity areas to form a network and monitor the various activities therein, like movement of the enemy, management of bushfires, relief sup- plies and rehabilitation work, etc. Technological advances create newer and more innovative applications everyday, which in turn fuel the demand for new technology. This has become a not-so-vicious circle, keeping researchers and developers on their toes all the time. The beneficiary is of course the layman on the street, literally so in the case of mobile computing. It is important to note that mobile computing is not just mobile or wireless communica- tion, as some would believe. There is much more to mobile computing, and it is to remove this confusion that this book has been written. Of course, provision of higher and more wireless bandwidth is the driving force for mobile computing. But what is more important and challeng- ing is the design of various application protocols and algorithms, the small-footprint operating systems, efficient usage of the small-sized user interfaces and, above all, providing security of systems and applications. This book provides a focussed look at all the issues mentioned above and gives an insight into the large number of technologies available in these areas to the user today. Apart from the theory, which is presented in an easy-to-understand form, we have provided many examples and suggestions for hands-on programming to help understand better the underlying technologies. These have been actually undertaken by senior undergraduate and postgraduate students of com- puter science at IIT Roorkee. To assist the reader in programming applications, an appendix has been included which deals with some important aspects of Java network programming. This book is intended for both professionals and students of senior undergraduate- and postgraduate-level engineering courses in electrical, electronics and computer science who have a background in computer networks and Java programming. It can be used for a one-semester or a one-quarter course. It can also be adopted for short-term training courses for new employees or trainees. To make the new concepts easy to understand, each chapter ends with multiple-choice xiii A01_GARGxxxx_01_SE_FM.qxd 4/5/10 6:26 PM Page xiii
- 19. review questions. Other research-oriented and programming-type questions which exercise the readers’ mind are also included. Book organization This book has been organized into 12 chapters, covering the entire gamut of technologies rele- vant to mobile computing. These include wireless and cellular communication, wireless local area networks (WLANs), logical mobility consisting of process migration and mobile agents, handheld devices and their operating systems, physical mobility, mobile ad hoc networks, wireless sensor networks, wireless application protocol and the mobile Internet, security issues in mobile applications, etc. The last chapter gives a brief idea of some design projects that can be undertaken to better understand the theory. An appendix is also included for explaining the ba- sics of Java network programming. The material is just right for a four-month, one-semester course. For a short-term course for students who are familiar with the basics of wireless communica- tion, Chapters 2, 4, 10 and 11, which deal with wireless communication, migrating processes, mobile agents and security, respectively, can be omitted. The concepts discussed in this book can also be used for research in this fast-growing field, since most of the technologies that are used and are applicable today may not be relevant tomor- row as requirements for newer applications arise. Acknowledgements Over the entire duration of the writing and compiling of this book, many people have helped me; without them, this book would not have been possible. First and foremost, I would like to thank the many experts who reviewed drafts of this book. Their suggestions have certainly helped to improve the content and presentation of the book. I am grateful to my Ph.D. student R.B. Patel, who first suggested that I introduce a course on mobile computing at IIT Roorkee in 2003 in the postgraduate curriculum and write a book on this important topic. My heartfelt thanks to all my postgraduate and senior undergraduate students at IIT Roor- kee, who designed and developed various projects related to mobile computing. These projects provided the content for the last chapter of the book and helped tremendously in adding to the ‘practice’ part of the title of the book. I also thank IIT Roorkee and MIT Manipal for providing the working environment that made this book possible. Last but not the least, I thank my family members and friends whose support and constant encouragement during the three years of writing the book made this effort worthwhile and without their support, this book could never have been finished. KUMKUM GARG xiv Mobile Computing A01_GARGxxxx_01_SE_FM.qxd 4/5/10 6:26 PM Page xiv
- 20. M obility has been the hallmark of all animate and living entities in nature. Animals move from place to place, migrating to find food and shelter. Similarly, early humans migrated from their natural habitats in search for food. Today, humans move in search of better employment, entertainment, travel, etc. Thus, mobility stems from a desire to move towards resources and away from scarcity. As in nature, so also in the field of computer science, mobility is becoming important and necessary. Today, both physical and logical entities can move. Physical entities are small, mobile computers that can change their actual location, unlike early systems, which were bulky in size and therefore immobile. Logical entities may be either the running user applications (processes) that migrate within a local cluster of computers or mobile agents, which are net- work applications that migrate in a network and execute on behalf of their owners anywhere in the Internet. The concept of mobility in the field of computer science has thus been chronologically provided in process migration since the 1970s, in mobile computers since the 1980s and in mobile agents from the 1990s. In this chapter, we shall briefly discuss these concepts and their benefits and challenges for deployment. We shall come back to visit them in detail in subse- quent chapters. 1.1 Process migration Process migration is the act of transferring a process between two computers connected through a wired or wireless medium. A process is an operating system abstraction and has code, data and state, besides a unique identity in the system. Traditionally, process migration was used to achieve load distribution in a multiprocessor system like a cluster or network of computers, or it was resorted to for providing fault tolerance in such systems. Many research operating systems have implemented full-blown process migration mecha- nisms, as shown in Accent (Zayas 1987), Chorus (Rozier and Legatheaux 1986), Mach (Acetta et al. 1986) and VKernel (Cheriton 1984). On the other hand, commercial migration-related products provide a higher-level, checkpoint-like restart version of migration, as seen in Condor (Litzkow, Linvy, and Mutka 1988). The main benefits of process migration are that a process might move towards an under- loaded computer, a specific database, or some rare hardware device. Furthermore, it enables movement of the programming environment and application to a desired location. For example, Introduction to Mobility 1 1 M01_GARGxxxx_01_SE_C01.qxd 3/5/10 12:00 PM Page 1
- 21. if a computer has a partial failure or is about to shut down, a running process can migrate to another computer and continue execution there. The resulting flexibility and reliability are important and necessary. 1.2 Mobile computing Mobile computing is computing that allows continuous access to remote resources, even to small computing devices such as laptops, palmtops and other handheld devices like personal digital assistants (PDAs) and digital cell phones. Mobile computing has become possible with the rapid advances in very-large-scale integration (VLSI) and digital/wireless communication technologies. There are basically three issues of concern in physical mobility. These are given below and have been dealt with in various ways by various researchers. We shall introduce these issues in this chapter but discuss details in subsequent chapters. 1. Weak connectivity: It is a well-known fact that wireless communication suffers from fre- quent disconnection and slow speeds, as compared with wired communication. The challenge is how a computer can operate when disconnected from the network or intermittently connected or connected over very slow communication links. This issue has been taken up in the CODA file system, which will be discussed in detail in Chapter 5. 2. Wireless connectivity: When a computer moves between cells in a wireless network or from one computer network to another, it is required to continue operating without having to re-register in the new location. In other words, the handoff should be smooth. This issue has been dealt with admirably by the development of two protocols—mobile Internet protocol (IP) and cellular IP, both of which are discussed in detail in Chapter 5. 3. Ubiquitous computing: This is the term coined by Mark Weiser and refers to the scenario when computers are present everywhere around us but have been rendered so small and cheap that they fade into the background. This is also called pervasive computing. Wireless sensor net- works (WSNs) are examples of such ubiquitous or pervasive computing, and are discussed in de- tail in Chapter 7. Thus, mobility of physical devices can be viewed at three different levels of granularity. These are as follows: 1. Macro-mobility: This is mobility through a global network. While moving in such a network, it should be possible to communicate without breaking the existing access. In Chapter 5, we shall read about mobile IP, which is the protocol that takes care of macro- mobility. 2. Micro-mobility: This is mobility of a device in one single administrative domain of the global network. For cellular networks, this is the lowest level of mobility. Cellular IP is the protocol designed to take care of micro-mobility, and this will also be discussed in Chapter 5. 3. Ad hoc mobility: This is mobility within a mobile ad hoc network (MANET), caused by device mobility constantly changing the network topology. We shall study MANETs in Chapter 7 and visit several ad hoc routing protocols therein. Whatever the type of mobility, the benefits of mobile computing are obvious, since there is physical movement towards a desired resource. Here, both the owner and the computer move to 2 Mobile Computing M01_GARGxxxx_01_SE_C01.qxd 3/5/10 12:00 PM Page 2
- 22. provide both qualitative and quantitative benefits. Since it is possible to use computer resources while moving, users can take the computer away from its usual workplace and still be productive. Thus, mobile computing, like process migration, enables movement of the programming envi- ronment and application. If a wireless phone cannot connect from a specific area, moving to a new area can overcome natural obstacles. A major benefit with mobile computing is that the use of computers is increased, not only for computer professionals, but also for the lay person. This is very important, because in this information age, having continuous access is imperative for everyone on the go. 1.3 Mobile agents A mobile agent is a program that can move through a network and autonomously execute tasks on behalf of the users. An agent is different from a user application, as it represents and acts on the owner’s behalf by inheriting the owner’s authority. Unlike mobile code (applets), mobile agents carry data and thread of control. They require agent environments, acting like docking stations, to execute and are supported on top of a programming environment like a Java virtual machine (JVM). Mobile agents are used to great advantage in applications like e-commerce, software distribution, information retrieval, system administration, network management, etc. They are well suited for slow and unreliable links and also provide fault tolerance. Many mobile agent systems have been developed and reported in the literature. Some of the more well- known systems are Aglets, Agent Tcl and PMADE (platform for mobile agent development and execution). Since mobile agents also migrate towards a source of information or towards a computer that they manage, they provide great flexibility and can mean easier reconfiguration or improved reliability. Mobile agents may not have sufficient resources or connectivity from one host and may move to another host. It can be seen from the above that there is much commonality between the three kinds of mobility discussed above. Researchers have, over the years, developed various means and mech- anisms to deploy the above concepts to real-life situations. As a result, we have numerous tech- nologies that can be used to advantage. We discuss some of these briefly below. Detailed discussions are given in subsequent chapters. 1. Java as a language offers many concepts that are directly useful for mobile systems. For example, remote method invocation (RMI), object serialization and mobile code are all very useful for process migration and mobile agents. 2. Similarly, wireless technologies provide support for mobile computing, with the develop- ment of many wireless protocols like Bluetooth, the Infrared Data Association (IrDA) stan- dards, wireless access protocol (WAP), etc. 3. Infrastructure support for transparent movement of entities from one location to another on the Internet and for issues of performance, scalability and reliability have been pro- vided by the presence of numerous mobile agent systems that have been developed in re- cent years. 4. Standardization has been provided in the form of CORBA (common object request broker architecture) and the MASIF (mobile agent system interoperability facility) standard, which allow for interoperable systems to be built and used worldwide. Introduction to Mobility 3 M01_GARGxxxx_01_SE_C01.qxd 3/5/10 12:00 PM Page 3
- 23. 1.4 Technical issues for mobility Mobile systems, as we have seen from the above discussion, are being increasingly deployed world- wide. But there are many challenges and technical issues of concern here. These are as follows: 1. Security is the biggest challenge for mobility. Security includes user authentication, data integrity and privacy, prevention of denial of service and non-repudiation. It may be appre- ciated that it is easier to provide security for a stationary system as compared to a mobile one, since the former can be guarded by intrusion detection systems and firewalls. The latter provides more security holes that have to be plugged. These include problems like unautho- rized access, data corruption, denial of access/service, spoofing, Trojan horses, replaying and eavesdropping, among others. The existing security infrastructure is designed only to protect stationary systems and thus needs to be adapted or modified for use in mobile systems. Security of mobile systems is the subject of Chapter 11. 2. Reliability, in terms of availability of resources, in the presence of disconnection, is also a major issue for mobile systems. In fact, it is both a technical issue and a benefit for mobility. Reliability can be improved by mobility but needs additional support in the form of caching and loading of state. Similarly, replication and check-pointing, file-hoarding, message- queuing and fault-tolerance tools need to be provided. 3. Naming and locating are common issues for all forms of mobility. Without locating a mobile object, communication with it or its control is not possible. Communication channels must be reconstructed after every movement. Naming is to be associated with authentication, and all recycling is to be done with great care. Controlling a mobile entity is necessary to check its status or to suspend, kill or recall it. All three of the above issues and their implementation will be discussed in detail in sub- sequent chapters. 1.5 Personal communication sy stems A personal communication system (PCS) is a generic name for a commercial system that offers several kinds of personal communication services and extended mobility. The Federal Communi- cations Commission (FCC) defines PCS as a mobile and wireless service that can be integrated with different networks to provide a wide variety of mobile and wireless services to individuals and business organizations. It was deployed in the USA in 1996. A PCS employs a mobile station (MS), an inexpensive, lightweight and portable handset, to communicate with a PCS base station (BS). The common features of these systems are as follows: 1. They are based on a second-generation technology like GSM (global system for mobile communication), IS-136 or IS-95. 2. Since they use the higher 1900-MHz band, an MS needs more power. This is because higher frequencies have a shorter range than lower ones. Alternatively, it can be said that the BS and the MS need to be closer to each other; that is, use smaller cell sizes. 3. They offer a whole spectrum of communication services ranging from an ordinary cell phone, short message service (SMS), to cable TV and limited Internet access. A typical PCS architecture is shown in Figure 1.1. 4 Mobile Computing M01_GARGxxxx_01_SE_C01.qxd 3/5/10 12:00 PM Page 4
- 24. 1.6 Context-aware computing A context-aware computing system is one which has user, device and application interfaces which enable it to remain aware of various parameters like its surroundings, circumstances or actions. These parameters can be thought of as the present mobile network, surrounding devices or systems, changes in the state of the connecting network, etc. These could also mean physical parameters such as the present time of the day, presently remaining memory and battery power, presently available nearest connectivity, past sequence of actions, cached data records, etc. The context of a mobile device represents the circumstances, situations, applications or physical environment under which it is being used. For example, the context is student when the device is used to download faculty lectures. Context-aware computing leads to application-aware computing. This is because the appli- cation programming interfaces (APIs) are part of the context. For example, when using an e-mail ID, a mail-receiving or mail-sending application software is used for computing. An application can adapt itself to the context. For example, if context is a contact, the phone-talk application will adapt itself to use of the telephone number from the ‘contact’ and to the use of GSM or code division multiple access (CDMA) communication. Context-aware computing also leads to pervasive or ubiquitous computing. In mobile device data-communication, context includes the existence of the service discovery protocol, radio interface and corresponding protocol. If the service discovery protocol senses the context and finds Bluetooth, then the device uses Bluetooth to communicate. Use of context in computing helps in reducing the possibility of errors and ambiguity in the actions. It also helps in deciding the expected system response on computation. The five types of contexts that are important in context-aware computing are as follows: Physical context: The context can be that of the physical environment. The parameters for defining a physical context are service disconnection, light level, noise level and signal strength. Assume a mobile phone is operating in a busy, congested area. If the device is aware of the surrounding noises, it can raise the speaker volume. If there is intermittent loss of connectivity during the conversation, the device can introduce background noises so that the user does not feel discomfort due to intermittent periods of silence. Computing context: Computing context is defined by interrelationships and conditions of the network connectivity protocol in use. Examples of the latter could be Bluetooth, ZigBee, GSM, general packet radio service (GPRS) or CDMA. Computing context may also be bandwidth Introduction to Mobility 5 Specialized mobile network MS Cellular network Cable TV PCS base station PCS node PCS base station PCS node MS Figure 1.1 PCS Architecture M01_GARGxxxx_01_SE_C01.qxd 4/5/10 3:48 PM Page 5
- 25. and available resources. Examples of resources in a mobile device are keypad, display unit, printer and device cradle. User context: The user context is defined as user location, user profiles, and persons near the user. It is based on the condition of the user, the primary intent of the systems and all other elements that allow users and computing systems to communicate. Temporal context: Temporal context defines the interrelation between time and the occur- rence of an event or action. A group of interface components has an intrinsic or extrinsic temporal context. For example, when a user presses a key to add a contact in his mobile device, the device should prompt him to enter a number as an input. Structural context: It defines a sequence and structure formed by the elements or records. Graphical user interface (GUI) elements have structural context. Interrelation among the GUI elements depends on the structural positions on the display screen. For example, in a date, the hours are displayed on the left of the minutes. 1.7 Outline of the book This book discusses both the theory and practice of mobile computing, so that the reader gets a complete idea of not only the techniques available to facilitate mobile computing, but also how to program and implement applications based on them. Chapter 2 deals with the basics of wireless and cellular communication. The various wireless frequencies present in the electromagnetic spectrum, like radio, microwave, infrared and light, and their characteristic features and applications are presented. Satellite communication is dis- cussed with reference to geostationary, medium-orbit and low-orbit satellites. The various gener- ations of cellular phone communication are given in detail, as they form the basis of all communication for the handheld devices used in mobile computing. Chapter 3 discusses wireless local area networks (WLAN). The most popular WLAN is the IEEE Standard 802.11. Its various extensions and modifications are dealt with in detail. The Blue- tooth and infrared LANs are also presented. Both the versions of the European Standard HiperLAN are discussed. A comparison of all the above standards is presented to bring out their essential features and applications. Chapter 4 deals with logical mobility. It discusses in detail the concept of process migration, which is the forerunner of mobile computing. The need for process migration and its various steps are presented. Chapter 5 deals with physical mobility, its requirements and the challenges associated with it. It discusses the limitations of IP in providing for physical mobility. It shows how mobile IP and cellular IP overcome these problems in micro- and macro-mobility scenarios. The chapter also introduces mobile databases and their design issues. Finally, it looks in detail at the CODA file system developed to take into account disconnected operation in mobile computing. Chapter 6 is on MANETs. The characteristics and classification of MANET are discussed in detail, along with their application. The proactive and reactive routing strategies for MANET are introduced. Three popular MANET routing protocols, namely, destination sequenced distance vector (DSDV), dynamic source routing (DSR) and adaptive on demand distance vector (AODV) are discussed in detail, with examples of their routing mechanisms. A performance comparison of DSR and AODV is also given. 6 Mobile Computing M01_GARGxxxx_01_SE_C01.qxd 3/5/10 12:00 PM Page 6
- 26. Chapter 7 deals with WSNs. It shows how these are different from MANETs and gives their characteristics, architecture and some popular routing techniques developed for them. Case stud- ies of the Mica mote sensor node and the TinyOS operating system used for it are also presented. Chapter 8 discusses the handheld devices like PDAs and pocket computers used in mobile computing. It discusses the characteristics of various such devices, including Palm and HP devices. The operating systems used with such devices have certain special features. These are presented with respect to the Palm OS, the Windows CE and Windows Mobile operating systems. Wide area mobile computing is the subject of Chapter 9, which presents what is now called the mobile Internet and the WAP, used to access the Internet on the move. The traditional Web programming model is compared with the wireless Web programming model. The WAP protocol stack is introduced and the WAP Gateway is discussed in detail, together with its design. Chapter 10 revisits logical mobility in the form of mobile agents, their characteristics and architecture and highlights their differences with process migration, mobile codes and mobile objects. The two earliest and basic mobile agent platforms, namely, Aglets and Agent Tcl, are pre- sented in detail. PMADE, a mobile agent platform developed at IIT Roorkee, is also presented. A discussion on the advantages of Java as a programming language for mobile agents is also given. Chapter 11 discusses the most important and crucial issue of security in mobile computing systems. It highlights the security threats present in wireless systems. The security mechanisms present in IEEE 802.11, Bluetooth and WAP2.0 to take cognizance of and counter these threats are also discussed in this chapter. Since this book is about mobile computing practice, the last chapter, Chapter 12, presents in detail as many as seven programming projects that can be designed and implemented by readers in different aspects of mobile computing. It thus provides an opportunity to have hands-on experience in designing and coding such systems. The appendix gives some details of Java as a network programming language, and covers topics like socket programming, remote procedure call (RPC) and the Java RMI. Some examples are given to provide a clear understanding of these concepts. 1.8 Summary Mobility is the hallmark of all animate beings and represents the movement from scarcity to resource-rich locations. In computing, mobility is characterized by logical or physical mobility and is represented by process migration, mobile agents or handheld-device communication. Mobile computing includes all these concepts, and it gives rises to a number of benefits, together with many technical issues and challenges. These have been dealt with in various ways, as discussed in this chapter, and the details are the subjects of the ensuing chapters. In the next chapter, we shall concentrate on the various communication technologies that facilitate mobile computing. Problems 1. What are the most important challenges facing mobile computing today? Discuss each of them in detail. 2. Distinguish mobile computing from distributed computing. 3. Go on to the Web and find about the state-of-the-art in mobile computing. Introduction to Mobility 7 M01_GARGxxxx_01_SE_C01.qxd 3/5/10 12:00 PM Page 7
- 27. 4. Do you have a pocket computer or PDA? If so, list the facilities it provides that can be listed under mobile computing applications. 5. The computer-networking architecture consists of seven layers, as given in the ISO OSI reference model. In your opinion, in which layer(s) should mobility be incorporated and why? 6. Recent conferences on mobile computing, such as ACM Mobicom and MobiSys, have pub- lished many articles on the subject. Read them and identify some of the current research challenges being addressed by researchers. 7. Do you think Java is suited for programming mobile computing systems? Explain your answer. (Do not look ahead into the later chapters of the book!) 8. What is your idea of a ubiquitous computing scenario for the home? Elaborate on this. 9. Discuss why security concerns in traditional systems are simpler than those in mobile systems. Give one example of a security threat that is present in the latter but not in the former. 10. Give one example where ‘disconnected operation’ may become imperative in a mobile com- puting scenario. Multiple-choice questions 1. Which one of the following is ‘computing that allows continuous access to remote resources even with the physical mobility of small computing devices such as laptops’? (a) Soft computing (b) Mobile computing (c) Remote computing (d) Ubiquitous computing 2. Pervasive computing is also called by which one of the following names? (a) Soft computing (b) Mobile computing (c) Remote computing (d) Ubiquitous computing 3. Wireless sensor networks are examples of which one of the following? (a) Soft computing (b) Mobile computing (c) Remote computing (d) Ubiquitous computing 4. Which one of the following can be characterized as ‘mobility through a global network’? (a) Macro-mobility (b) Micro-mobility (c) Ad hoc mobility (d) None of the above 5. Mobility of a device in one single administrative domain of the global network is known as which one of the following? (a) Macro-mobility (b) Micro-mobility 8 Mobile Computing M01_GARGxxxx_01_SE_C01.qxd 3/5/10 12:00 PM Page 8
- 28. (c) Ad hoc mobility (d) None of the above 6. Which of the following is true for statements X and Y? X: A mobile agent is a program that can move through a network and autonomously execute tasks on behalf of the users. Y: Process migration is the act of transferring a process between two computers connected through a wired or wireless medium. (a) X is true but Y is false (b) X is false but Y is true (c) Both X and Y are true (d) Both X and Y are false 7. What is an Aglet? (a) A wireless protocol (b) A mobile agent (c) A pervasive computing technique (d) None of the above 8. Which one of the following is false for mobile agents? (a) They are well suited for slow and unreliable links (b) They cannot provide fault tolerance (c) Unlike mobile code (applets), mobile agents carry data and thread of control (d) They require agent environments 9. Which one of the following is not a wireless protocol? (a) Bluetooth (b) IrDA (c) WAP (d) CSMA/CD 10. Which one of the following is true for statements X and Y? X: It is easier to provide security for a mobile system as compared to a stationary system. Y: Security includes user authentication, data integrity and privacy, prevention of denial of service and non-repudiation. (a) X is true but Y is false (b) X is false but Y is true (c) Both X and Y are true (d) Both X and Y are false Further reading A.S. Tanenbaum (2005), Computer Networks, 4th ed. (Prentice Hall India). A.T. Campbell (2000), ‘Design, Implementation and Evaluation of Cellular IP’, IEEE Personal Communications, 7 (August): 42–49. C. Perkins (1998), Mobile IP: Design Principles and Practice (Addison-Wesley Longman). D. Kotz et al. (1997), ‘Agent Tcl: Targeting the Needs of Mobile Computing’, IEEE Internet Computing, 1(4): 58–67. Introduction to Mobility 9 M01_GARGxxxx_01_SE_C01.qxd 3/5/10 12:00 PM Page 9
- 29. D. Lange and M. Oshima (1998), ‘Mobile Agents with Java: The Aglet API’, World Wide Web, 1(3). D. Milojicic et al. (1998), ‘MASIF: The OMG Mobile Agent System Interoperability Facility’, in Proceedings of the International Workshop on Mobile Agents (MA ’98), Stuttgart. D. Milojicic, F. Douglis and R. Wheeler (eds) (2000), Mobility: Processes, Computers and Agents (Addison-Wesley). D.B. Lange and M. Oshima (1998), Programming and Deploying Java Mobile Agents with Aglets (Addison-Wesley). D.P. Agrawal and Q.A. Zeng (2003), Introduction to Wireless and Mobile Systems (Thomson). D.R. Cheriton (1984), ‘The V-kernel: A Software Base for Distributed Systems’, IEEE Software, 1(2): 19–42. E. Pitoura and G. Samaras (1998), Data Management for Mobile Computing (Norwell, MA: Kluwer Academic Publishers). E.R. Zayas (1987), ‘Attacking the Process Migration Bottleneck’, in Proceedings of the 11th ACM on Operating Systems Principles, pp. 13–24. F. Adelstein et al. (eds) (2005), Fundamentals of Mobile and Pervasive Computing (Tata McGraw- Hill). J. Kistler and M. Satyanarayan (1992), ‘Disconnected Operation in the CODA Distributed System’, ACM Transactions on Computer Systems, 10(1): 3–25. M. Rozier and J.M. Legatheaux (1986), ‘The Chorus Distributed Operating System: Some Design Issues’, Y. Parker et al (eds.), in Proceedings of the NATO Advanced Study Institute on Distributed Operating Systems: Theory and Practice, Springer-Verlag, New York, August 1986, pp. 261–289. M. Weiser (1991), ‘The Computer of 21st Century’, Scientific American, 265(3): 94–104. M.C. Powell and B.P. Miller (1983), ‘Process Migration in DEMOS/MP’, ACM SIGOPS OS Review, 17(5): 110–119. M.J. Acetta et al. (1986), ‘Mach, a New Kernel Foundation for UNIX Development’, in Proceed- ings of the Summer USENIX Conference, June 1986, pp. 93–113. M.J. Litzkow, M. Livny, and M.W. Mutka (1988), ‘Condor—A Hunter of Idle Workstations’, in Proceedings of the 8th International Conference on Distributed Systems, June 1988, pp. 104–111. R. Kamal (2007), Mobile Computing (Oxford University Press). R.B. Patel (2002), ‘Manual of PMADE’ (Internal Report, Department of E&CE, IIT Roorkee, Uttarakhand, India). Reza B’Far (2005), Mobile Computing Principles: Designing and Developing Mobile Applications with UML and XML (Cambridge University Press). T. Imielinski and H.F. Korth (eds) (1996), Mobile Computing (Norwell, MA: Kluwer Academic Publishers). U. Hausmann et al. (2003), Principles of Mobile Computing, 2nd ed. (Springer). V. Kumar (2006), Mobile Database Systems (John Wiley). W.R. Cockayne and M. Zyda (1998), Mobile Agents (Manning Publications). 10 Mobile Computing M01_GARGxxxx_01_SE_C01.qxd 3/5/10 12:00 PM Page 10
- 30. I n this chapter, we will discuss the transmission technologies that form the basis of all mobile computing. In particular, we study in detail mobile or wireless communication and the differ- ent protocols that have been developed to physically or logically connect two mobile devices. Thus, this chapter looks at the physical layer technologies used in mobile computing, using what is called unguided media, as opposed to guided media, which consist of copper fibres, twisted pairs and optical fibres, and which are used for wired communication. The basis for all wireless transmission is the electromagnetic spectrum, in which lie the different frequency bands that are used for wireless communication. We will discuss in detail the characteristics of each of these frequency bands and the wireless and cellular communication systems enabled by them. It is assumed here that the reader is familiar with the theoretical basis for data communica- tion, that is, the terms frequency, wavelength, channel, speed of light, bandwidth, the maximum data rate of a channel, etc., and the relation between them. For the sake of completeness, and because it is an important and relevant relation for this book, we must state here that the amount of information that a noisy channel can carry is gov- erned by its bandwidth. According to Shannon, the maximum data rate of a noisy channel whose bandwidth is H Hz and whose signal-to-noise ratio is S/N is given by Maximum data rate (bits/sec) H log2 (1 S/N) For example, a channel of 3,000 Hz bandwidth and signal-to-thermal noise ratio of 30 dB can never transmit more than 30,000 bps. For other related information, the uninitiated reader is referred to Tanenbaum (2003). 2.1 The electromagnetic spectr um The electromagnetic spectrum is shown in Figure 2.1. The radio, microwave, infrared and visible light portions of the spectrum can all be used for transmitting information by modulating the wave’s amplitude, frequency or phase. The higher frequencies, that is, ultraviolet light, X-rays and gamma rays, would give better results, but are normally not used because they are difficult to produce and modulate, do not propagate well through buildings and are harmful to humans. The various frequency bands have official International Telecommunication Union (ITU) names, as given in Figure 2.1, and are based on their wavelengths. In this section, we give, very briefly, the characteristics, advantages and disadvantages of each of the above wave bands and see how they are used for wireless transmission. Wireless and Cellular Communication 2 11 M02_GARGxxxx_01_SE_C02.qxd 3/5/10 12:05 PM Page 11
- 31. 12 Mobile Computing F(Hz) Band L MF H VH UH SHF EHF THF Twisted pair Coa Maritime Radio Microwave Infrared Visible UV X-ray Gamma AM radio FM radio T Terrestrial microwave Satellite Fiber optics 104 105 106 107 108 109 1010 1011 1012 1013 1014 1015 1016 F(Hz) 100 102 104 106 108 1010 1012 1014 1016 1018 1020 1022 1024 Figure 2.1 The Electromagnetic Spectr um 2.1.1 Radio waves Radio waves are present at the lower end of the spectrum and are widely used for both indoor and outdoor communication. They have the advantage that they are omnidirectional and are able to travel long distances, penetrating easily through buildings. Their disadvantages are that they suffer from interference between users and from electrical equipment. They also exhibit frequency- dependent properties; that is, at low frequencies, they pass through objects, but attenuation in power occurs as distance from the source increases. On the other hand, high-frequency radio waves travel in straight lines and cannot penetrate through obstacles. Furthermore, rain and sleet absorb such waves. 2.1.2 Microwaves Frequencies above 100 MHz are called microwaves. These have the advantage that they can be narrowly focused because they travel in straight lines. Thus, by properly aligning the sending and receiving antennae, they are able to give much higher signal-to-noise ratio. For the same rea- son, they are affected by the curvature of the earth if long-distance communication is to be used, making it necessary to build repeater towers for the transmitting antennae. Microwaves are less expensive to use than optical fibres and are therefore popular in mountainous and urban areas. Microwaves have the disadvantage that they suffer from multipath fading. This is because they do not pass easily through buildings and obstacles and are refracted by the atmospheric layer; some waves therefore arrive out-of-phase with the direct ones, resulting in cancellation of the signal. The effect of this type of fading changes with weather and frequency. 2.1.3 Infrared waves Unguided infrared and millimeter waves offer an alternative to the standard radio frequency communication for short ranges. However, they are subject to the following restrictions: • Transmission distance of less then 2 miles • Line-of-sight limitations • Restricted to 16 Mbps throughput • Presence of environmental disturbances, such as fog, dust and heavy rain M02_GARGxxxx_01_SE_C02.qxd 3/5/10 12:05 PM Page 12
- 32. Wireless and Cellular Communication 13 However, the advantages of this technology are as follows: • Reasonable high bandwidth • No government license required for operation • Cost-effective • Capable of traversing multiple paths without interferences • More secure than radio • Immune to radio frequency interference and electromagnetic interference Infrared communication has very little use on the desktop. For example, it can be used for connecting notebook computers and printers, but is not used in computer-to-computer commu- nication. The Infrared Data Association (IrDA) has defined a number of standards governing infrared wireless communication. These include the IrDA-data and IrDA-control standards. These will be discussed in detail in the next chapter. 2.1.4 Lightwaves Unguided optical signalling has been around for many years. In recent years, coherent optical signalling using lasers mounted on rooftops has been used to connect the local area networks (LANs) in two buildings. The signals are inherently unidirectional, so each building requires a laser and photodetector. This scheme is very inexpensive and offers very high bandwidth. It is easy to install and does not require a license to operate. A major disadvantage is that laser beams cannot penetrate rain or thick fog. However, they work well on sunny days and can be effectively used for ‘wireless outdoors’. 2.2 Communication satellites Communication satellites have provided a very powerful wireless communication system since the first artificial satellite was put into orbit in 1962. A communication satellite is like a big microwave repeater in the sky. It consists of many transponders, each of which listens to some frequency spectrum, amplifies the incoming signal and rebroadcasts it at another frequency (to avoid interference with the incoming signal). Mobile satellite services allow global coverage, because in these systems satellites play the role of mobile base stations (BSs). Satellite-based systems are categorized according to the orbital altitude of the satellite. This is shown in Figure 2.2. Satellite-based systems Medium earth orbit satellites (MEOS) (widely varying altitudes between those of GEOS and LEOS) Geostationary satellites (GEOS) (altitude of 35,786 km) Low earth orbit satellites (LEOS) (altitude of the order of 1,000 km) Figure 2.2 Satellite Systems M02_GARGxxxx_01_SE_C02.qxd 3/5/10 7:33 PM Page 13
- 33. 14 Mobile Computing The major advantage of GEOS systems is that contiguous global coverage up to 75 degrees latitude can be provided with just three satellites. Their main drawback is that they have a large 240–270 ms round-trip propagation delay and need higher radio frequency (RF) power. On the other hand, LEOS require less power but frequent handoffs. We shall discuss the characteristics of each of these in some detail below. 2.2.1 Geostationary satellites Satellites at the altitude of 35,800 km in a circular equatorial orbit appear motionless in the sky. Such satellites are called geostationary satellites. With current technology, it is unwise to have geostationary satellites spaced much closer than 2 degrees in the 360-degree equatorial plane, to avoid interference. ITU has allocated certain frequencies to satellite users. The main ones are listed in Table 2.1. The C band was the first to be designed for commercial satellite traffic. This band is already over- crowded because it is also used by the common carriers for terrestrial microwave link. The L and S bands were added by an international agreement in 2000. However, they are narrow and crowded. The next higher band available to commercial communication carriers is Ku (K under) band. This band is not (yet) congested, and at these frequencies, satellites can be placed as close as 1 degree. However, another problem exists: rain. Water is an excellent absorber of these short microwaves. Bandwidth has also been allocated in the Ka (K above) band for com- mercial satellite traffic, but the equipment needed to use it is still expensive. A new development in the communication satellite world is the development of low-cost microstations, also called VSATs (very small aperture terminals). These tiny terminals have 1 m or smaller antennas (versus 10 m for a standard GEO antenna) and can put out about 1 watt of power. In many VSAT systems, the microstations do not have enough power to communicate directly with one another (via the satellite of course). Instead, a special ground station called the hub, with a large, high-gain antenna, is needed to relay traffic between VSATs. See Figure 2.3. 2.2.2 Medium ear th orbit satellites MEOs are deployed much lower than the GEOs, and must be tracked as they move through the sky, as they drift slowly in longitude, taking about 6 hours to circle the earth. They have a smaller footprint on the ground and require less powerful transmitters to reach them. The 24 GPS (global positioning system) satellites orbiting at about 18,000 km above the earth are an example of MEO satellites. 2.2.3 Low ear th orbit satellites Moving down in altitude, we come to the LEO satellites. Due to their rapid motion, large numbers of them are needed for a complete system. On the other hand, because the satellites are so close to earth, Band Downlink Uplink Bandwidth Problems L 1.5 GHz 1.6 GHz 15 MHz Low Bandwidth; crowded S 2.2 GHz 2.2 GHz 70 MHz Low Bandwidth; crowded C 4.0 GHz 6.0 GHz 500 MHz Terrestrial interference Ku 11 GHz 14 GHz 500 MHz Rain Ka 20 GHz 30 GHz 3,500 MHz Rain, equipment cost Table 2.1 The Main Satellite Bands M02_GARGxxxx_01_SE_C02.qxd 3/5/10 12:05 PM Page 14
- 34. Wireless and Cellular Communication 15 the ground does not need much power, and the round-trip delay is only a few milliseconds. The examples are Iridium, Globalstar and Teledisc, of which only the last one is briefly discussed here. Teledesic is targeted at bandwidth-hungry Internet users all over the world. The goal of the Teledesic system is to provide millions of concurrent Internet users with an uplink of as much as 100 Mbps and a downlink of up to 720 Mbps using a small, fixed, VSAT-type antenna, com- pletely bypassing the telephone systems. It uses 30 satellites with large footprints, using the high-bandwidth Ka band, and packet-switching in space, with each satellite capable of routing packets to its neighbours. Users who want to send packets request and get assigned bandwidth dynamically, in about 50 ms. 2.3 Multiple-access schemes In a wireless environment, there is a need to address the issue of simultaneous multiple access by many users or mobile stations (MSs) in the transmission range between the BS and themselves. Users are able to receive signals transmitted by others in the system. To accommodate a number of users, many traffic channels need to be made available. To provide simultaneous two-way com- munications (duplex communication), a forward (downlink) channel from BS to MS and a reverse (uplink) channel from MS to BS are necessary. Two types of duplex systems are used: frequency Hub Communication satellite VSAT 1 3 2 4 Figure 2.3 Hub and VSA Ts M02_GARGxxxx_01_SE_C02.qxd 3/5/10 12:05 PM Page 15
- 35. 16 Mobile Computing division duplexing (FDD) divides the frequency used, and time division duplexing (TDD) divides the same frequency by time. There are three basic ways in which many channels can be allocated within a given band- width. These are with respect to frequency, time and code division multiplexing, using three multiple-access techniques. These are frequency division multiple access, (FDMA), time division multiple access (TDMA) and code division multiple access (CDMA). FDMA mainly uses FDD, while TDMA and CDMA systems use either FDD or TDD. We will discuss these three techniques in this section along with their advantages and disadvantages. A multiple-access technique is important in mobile cellular systems, so that an MS can distinguish a signal from the serving BS, and also discriminate the signals from an adjacent BS. Multiple-access techniques are based on the orthogonalization of signals. An FDMA system is one which uses different carrier frequencies to transmit the signal for each user. If a system uses distinct time to transmit the signal for different users, it is a TDMA system. If a system uses different codes to transmit the signal for each user, it is a CDMA system. 2.3.1 FDMA—F requency division multiple access In FDMA, the allocation of frequencies to channels can either be fixed (as in radio stations) or dynamic (demand-driven). Furthermore, channels can be assigned to the same frequency at all times, that is, pure FDMA, or change frequencies according to a certain pattern, that is, FDMA combined with TDMA. The latter is done in many wireless systems to circumvent narrowband interference at some frequencies, known as frequency hopping. The sender and the receiver agree on a hopping pattern, so that the receiver can tune to the right frequency. Hopping patterns are normally fixed for a long period. As an example of FDMA, let us consider a mobile phone network based on the global system for mobile communication (GSM) standard for 900 MHz. There are 124 multiple-access channels per direction available at 900 MHz. The basic frequency allocation scheme is fixed and regulated by a national authority. All uplinks use the band between 890.2 and 915 MHz; all downlinks use 935.2 to 960 MHz. The BS allocates a certain frequency for uplink and downlink to establish a duplex channel with a mobile phone. Each channel (uplink and downlink) has a bandwidth of 200 KHz. Uplinks and downlinks have a fixed relation. For a certain channel n, if the uplink frequency is fu 890 MHz n 0.2 MHz, the downlink frequency is fd fu 45 MHz, i.e., fd 935 MHz n 0.2 MHz. 2.3.2 TDMA—T ime division multiple access TDMA offers a much more flexible scheme as compared with FDMA. Tuning to a certain fre- quency is not required, and the receiver can stay at the same frequency all the time. Very simple receivers and transmitters can thus be designed, since listening to many channels separated in time is easier than listening to different frequencies at the same time. Many different algorithms exist to control medium access using only one frequency. Almost all MAC schemes for wired networks like Ethernet, token ring, Asynchronous transfer mode (ATM), etc., work according to this principle. In TDMA, synchronization between receiver and sender has to be achieved in the time domain. This can be done either by using a fixed pattern or by using a dynamic allocation. Fixed allocation is not efficient in cases where bandwidth requirement is variable. M02_GARGxxxx_01_SE_C02.qxd 3/5/10 12:05 PM Page 16
- 36. Wireless and Cellular Communication 17 Many systems like IS-54, IS-136, GSM and digital European cordless telecommunications (DECT) use TDMA with fixed allocation. For example, for the DECT cordless phone system, the BS uses 1 out of 12 slots for the downlink, whereas the MS uses 1 out of 12 different slots for the uplink. Uplink and downlink are separated in time. Up to 12 different MS can use the same fre- quency without interference. Each connection is allotted its own uplink and downlink pair. The pattern is repeated every 10 ms; that is, each slot has a duration of 417 µsec. This repetition guar- antees access to the medium every 10 ms, independent of any other connection. A guard band is also used at the beginning and end of each slot to avoid collisions due to drifts in receiver and transmitter clock frequency or computational delays in placing the data in a slot. Fixed access patterns are efficient for connections with a constant data rate, as in classical voice transmission with 32 or 64 Kbps duplex. But they are inefficient for bursty data or asym- metric connections, as in Web browsing, where no data transmission occurs while the page is being read, whereas clicking on a hyperlink triggers data transfer from the MS to the BS, followed by a large volume of data returned from the Web server. In such cases, demand-oriented TDMA schemes are used. In demand-oriented TDMA, the allocation is traffic dependent, and the BS can reserve time slots for an MS on demand. 2.3.3 CDMA—Code division multiple access CDMA is the best technical solution available today and is the basis for the 3G mobile systems. It is also widely used in the United States in 2G mobile systems, competing with Digital advanced mobile phone system (D-AMPS). For example, Sprint personal communication services (PCS) uses CDMA, whereas ATT Wireless uses D-AMPS. CDMA is also known as International Stan- dard IS-95 or cdmaOne. CDMA is completely different from FDMA and TDMA. Instead of dividing the allowed fre- quency range into a few hundred narrow channels, CDMA allows each station to transmit over the entire frequency spectrum all the time. Multiple simultaneous transmissions are separated using codes. Codes used by users should have a good autocorrelation and should be orthogonal to other codes. For details of these two terms, the reader is referred to Tanenbaum (2003). Autocorre- lation helps a receiver to reconstruct the original data precisely even in the presence of distortion by noise, and orthogonality is necessary for two stations to share the medium without interference. Tanenbaum (2003) has given a very good analogy to explain the concept of CDMA: An airport lounge has many pairs of people conversing. TDMA is comparable to all the people being in the middle of the room but taking turns speaking. FDMA is comparable to the people being in widely separated clumps, each clump holding its own conversation at the same time as, but still independent of, the others. CDMA is comparable to everybody being in the middle of the room talking at once, but with each pair in a different language. The French-speaking couple just hones in on the French, reject- ing everything that is not French as noise. In CDMA we extract only the desired signal and reject everything else as random noise. Here, each bit time is subdivided into m short intervals called chips. There are normally 64 or 128 chips per bit or longer. Each station is assigned a unique m-bit code called a chip sequence. Chip sequences in IS-95, for example, are 242 1 chips long, and the chipping sequence is 1228800 chips/s; that is, the code repeats after 41.425 days. To transmit a 1 bit, a station sends its chip sequence. To transmit a 0 bit, it sends the 1-bit’s complement of its chip sequence. No other patterns are permitted. Thus, for m 6, if station A is assigned the chip sequence 010011, it sends a 1 bit by sending 010011 and a 0 bit by sending M02_GARGxxxx_01_SE_C02.qxd 3/5/10 12:05 PM Page 17
- 37. 18 Mobile Computing 101100. For pedagogical purposes, it is more convenient to use a bipolar notation, with binary 0 being ⫺1 and binary 1 being ⫹1. To synchronize the sender and the receiver, the sender transmits a long predefined chip sequence so that the receiver can lock onto it. Transmissions that are not synchronized are treated as noise. The longer the chip sequence, the higher the probability of detecting it correctly in the presence of noise. Implementation of the chip sequences and codes is complicated, and this is a major draw- back with the CDMA scheme. But it is used for wireless mobile communication, as it operates in a much higher (1.25 MHz) band than D-AMPS and GSM, where it can support many more users than either of these systems. For a good comparison of the above techniques, the reader is referred to Schiller (2006). 2.4 Cellular communication Wireless communication using unguided media, that is, radio and microwave frequencies or satellites, has found widespread use in mobile phones. These are currently being used for voice communication, but soon they will find use in data communication. Cellular communication has undergone many generations, in which the communication bandwidths and data speeds have continuously increased. This has given rise to many applications that have benefited mobility. In this section, we discuss briefly these generations and how they have revolutionized not only mobile phone communication, but also mobile computing. 2.4.1 The first generation (1G): 1980 Analog cellular systems were the first generation of mobile telephone communication systems. They used analog frequency modulation for only voice (speech) transmission. The various sys- tems that fall in this category are AMPS (Advanced Mobile Phone Service) (USA), Nordic Mobile Telephone (NMT)-900 (Sweden) and Cellular Digital Packet Data (CDPD), which is designed to provide packet data services on the top of existing AMPS. The system architecture is such that a geographic region is divided into cells. The size of the cells in AMPS is about 10–20 km across, but is lesser in digital systems. Each cell uses some set of frequencies not used by its neighbours. Transmission frequencies are reused in nearby but not adjacent cells. Figure 2.4a illustrates the concept of frequency reuse. The cells are normally circu- lar but are shown as hexagonal for ease of drawing. F6 F7 F1 F5 F2 F3 F4 F6 F7 F7 F6 F5 F1 F2 F1 F5 F2 F4 F3 F4 F3 Figure 2.4(a) Adjacent cells use different frequencies Figure 2.4(b) Microcells add more users (a) (b) M02_GARGxxxx_01_SE_C02.qxd 3/5/10 7:34 PM Page 18
- 38. Wireless and Cellular Communication 19 If the area is overloaded, the power is reduced and the overloaded cells are split into smaller microcells. This allows for more frequency reuse and is shown in Figure 2.4b. At the centre of each cell is a BS to which all the telephones in the cell transmit. In a small system, all the BSs are connected to a single device which is called an MTSO (mobile telephone switching office) or MSC (mobile switching centre). In a larger system, several MTSOs may be used in a hierarchical manner. Handoff: At any time instant, a mobile phone logically belongs to one cell and is under con- trol of its BS. When it moves physically from the cell, the BS notices the phone’s fading signal and finds out from other neighbouring BSs as to which one is getting the strongest signal. It then transfers ownership of the mobile to the BS of that cell. If a call is in progress, the mobile is asked to switch to the new channel used in that adjacent cell. This process is called handoff or handover. A BS is only a radio relay; the channel assignment is done by the MTSO. Handoffs can be either soft or hard. In a soft handoff, the mobile is acquired by the new BS before the old one signs off. Thus, there is no loss of continuity. But it requires the mobile to be able to tune to two frequencies at the same time. Neither first- nor second-generation devices can do this. 3G CDMA systems provide soft handover, resulting in seamless connectivity to the mobile. In a hard handoff, the old BS drops the mobile before the new one acquires it. The call is disconnected abruptly if there is no available frequency with the new BS, or there is a call drop till the new frequency is received. This is noticeable by the user but is typically of very short duration of about 60 ms in GSM systems. Different kinds of handover are possible when a mobile moves from one cell to another or when traffic through a specific stage becomes very high. Readers are referred to (Kamal, 2007) for details. The AMPS system uses 832 full-duplex channels, each consisting of a pair of simplex chan- nels (824–849 MHz for transmission and 869–894 MHz for reception). The individual cells use different frequencies, using a system referred to as FDMA. Here the maximum supported bit rate is 19.2 Kb/s. Although 1G communication provided a good start, its main disadvantage was low speed due to low available frequencies, interference due to frequency reuse and poor security. 2.4.2 The second generation (2G): 1992 The first generation of mobile phones was analog; the second generation was digital. The term PCS is sometimes used in the marketing literature to indicate the second-generation systems. Sometimes PCS is classified as a 2.5-generation (2.5G) system separately. The various advantages of digital cellular are as follows: 1. It is more robust as it displays resistance to noise and crosstalk and has efficient error correction. 2. It exhibits the intelligence of the digital network. 3. It is more flexible and can be integrated with the wired digital network. 4. Reduced RF transmission power is needed. 5. Encryption can be provided for communication privacy. 6. System complexity is reduced. 7. User capacity is increased. There are two basic technologies for managing shared access in digital cellular systems, which are further classified as shown in Figure 2.5. The IS-54 standard is a North American standard based on TDMA. It contains the 30 KHz spacing of AMPS to make the evolution from analog to digital easier. Each channel provides a raw bit rate of 48.6 Kb/s. M02_GARGxxxx_01_SE_C02.qxd 3/5/10 12:05 PM Page 19
- 39. 20 Mobile Computing The Pan-European GSM is based on TDMA with eight slots per radio channel. Each user transmits periodically in each of the slots with duration of 0.57 seconds. In the present version, GSM supports full-rate 22.8 Kb/s transmission. In spite of many improvements over 1G, 2G still has many shortcomings. First, it still focuses only on low data rate speech service. Second, the capacity still does not satisfy the ever-growing demand, and finally multimedia service is still not provided. 2.4.3 The 2.5 generation (2.5G): 1996 As an interim step towards the higher data rates of 3G, whose technology differs considerably from the current cellular technology, some new techniques are being deployed as stopgap meas- ures. They use incremental advances in cellular technology to increase the capacity of the currently deployed infrastructure. Enhanced data rates for GSM evolution (EDGE) is a 2.5G system, with a data rate of 384 kbps, which is higher than GSM. The errors introduced by the high speeds necessitate the use of nine different schemes for modulation and error correction. GPRS (general packet radio service) is another 2.5G scheme which is actually an overlay packet network over D-AMPS or GSM. In this scheme, voice Internet protocol (IP) packets can be exchanged between mobile senders and receivers, with speeds of 115 kbps. This is done by reserv- ing some time slots on some frequencies for packet traffic. The BS can dynamically vary the number and location of the time slots, based on how much voice traffic is to be sent in the cell. The BS sends the packet received from the mobile unit, to the Internet, through a wired connection. 2.4.4 The third generation (3G): 2000 ⴙ ⴙ In 1992, 3G was envisaged by ITU as International Mobile Telecommunication (IMT2000), but it still has not seen the light of day. Its aim is to implement true ‘anybody at any place’ communi- cation with ‘anyone at any time’. IMT2000 is defined as a system aimed at ‘the provision of worldwide mobile service through a limited number of wireless access points by combining vari- ous services and different systems’. It promises to connect up to 2 billion people worldwide by 2010 and offer data rates of up to 2 Mbps. The frequency bands identified for IMT2000 are 1885–2025 MHz and 2110–2200 MHz. Its goals are to • Support high mobile velocity (300–500 km/hour), compared with less than 100 km/hour in GSM. • Support global wandering, as opposed to district and country in GSM. • Support multimedia service, especially Internet service, 144 Kb/s (outdoor and higher velocity), 384 Kb/s (from outdoor to indoor, lower velocity), 2 Mb/s (indoor); speech with quality of service (QoS) and other services 4–100–200 Kbs/s (GSM, lower velocity). 2G Technologies TDMA European GSM IS-54 IS-95 CDMA Figure 2.5 The 2G Technologies M02_GARGxxxx_01_SE_C02.qxd 4/5/10 5:31 PM Page 20
- 40. Wireless and Cellular Communication 21 • Convenience for transition and evolvement or innovation, compatibility of services with various fixed/mobile networks. High quality and security comparable to the fixed network. • Highest spectrum availability, higher QoS, speech recognition technology, lower cost, higher security. • Use the advantages of technologies such as adversity transmitting and receiving, multi- path combining, turbo code, channel estimation, signal-to-interference power ratio (SIR) measurement and Transmit power control (TPC), space-time technology, multi-user de- tection and interference cancellation, beam forming and smart antennas, and soft hand- off, etc. Service targets for IMT-2000 are worldwide roaming, software radio and user identity mod- ule (smart card). Various services for users include multirate multimedia, that is, voice, image and high-speed data up to 2 Mbps. 2.4.5 The 3.5 generation (3.5G): 2000 ⴙ ⴙ The technical breakthrough towards 3.5G provides for an open architecture for service based on multimedia, and application of technologies such as smart antenna, software defined radio and TD-CDMA. The standardization focus has been moved from radio to network side, giving rise to the following advantages: • Increased possibility to accommodate different types of radio in one system. • A shift to the networking paradigm causes the problem of migrating legacy systems, so the effort for maintaining interoperability has been increased. • Rapid upgrade of the standard for 4G, giving rise to more attention on the system evolution scenario. 2.4.6 The four th generation (4G): 2002 ⴙ ⴙ Between 1992 and 1995, there was a project in the European Community that was called Mobile Broadband System (MBS), which targeted future outdoor, cellular scenarios with high mobility and high data rates, to provide mobile multimedia communications. These systems will be the fourth mobile generation. The European Radio communications Office (ERO) has proposed some features for 4G sys- tems which include high bandwidth, ubiquity (connectivity everywhere), seamless integration with wired networks (especially IP), adaptive resource and spectrum management, software radio, besides high quality of multimedia service. To implement the above features, innovative concepts are needed. The approach taken at the Mobile Multimedia Communication (MMC) project of the Delft University of Technology is to form a multi-disciplinary team in which user aspects get as much attention as the technologi- cal challenges. This MMC project has the following research goals: • User interface and transparency • Compression: Research has been carried out in two areas for source coding and two tech- niques have been proposed: 1. H.263 for mobile video communication 2. Compression of the shapes of video objects • Transmission protocols: The MMC project uses a hierarchical protocol structure that pro- vides different QoS to the various traffic streams in mobile multimedia communication. The hierarchy is realized with a hybrid TDM/FDM (time division multiplexing/frequency division M02_GARGxxxx_01_SE_C02.qxd 3/5/10 12:05 PM Page 21
- 41. 22 Mobile Computing multiplexing) technique in which frames (the largest unit of data) are composed of packets, fragments and radio data units (RDUs). • Broadband radio transmission: MMC transmission is located at the V band (from 40–75 GHz), centred at 60 GHz. New techniques for measurements have been proposed and carried out. The chosen modulation scheme is orthogonal frequency division multiplex (OFDM), which is specifically able to cope with the problems of the multipath reception. However, 4G is still a distant dream, since as of today even 3.5G systems are yet to take off. 2.5 Summary The electromagnetic spectrum contains all the frequencies that can be used in wireless commu- nication and is the basis of all mobile computing. The different portions of the spectrum com- prise radio waves, microwaves, infrared, and lightwaves, and their characteristics determine the data rates and applications in which each of these ‘unguided’ media can be used. Communication satellites are an upcoming and useful long-range transmission system. Depending on their height of deployment, these can classified as geostationary orbit, medium earth orbit and low earth orbit and can be used in different applications. Cellular communication has revolutionized the way mobile handhelds and phones are used. These handhelds are currently being used more for voice communication, but soon they will find widespread use for data. The first-generation systems were analog, and second-generation ones were digital with many options, like GSM, FDMA, TDMA and CDMA. There is a lot of talk about 3G, 3.5G and 4G systems, all of which are yet to take shape in reality. Each generation has improved on the capabilities of the older generation, with many new features added for broad- band applications. Handover is an important aspect of all mobile systems and must be handled with proper care to provide seamless connectivity to mobile devices. In the next chapter, we discuss wireless LAN (WLAN) standards, which are based on the short-range wireless communication technologies discussed in this chapter. Problems 1. If a binary signal is sent over a 4 KHz channel whose signal-to-noise ratio is 20 dB, what is the maximum data rate achievable? 2. In a tabular form, compare radiowaves, microwaves and infrared waves, with respect to their data rates, transmission distance, interference and cost. 3. Repeat Question 2 by comparing the three satellite communication types, namely, GEOS, MEOS and LEOS. 4. Give typical applications for each of the three satellite systems. 5. Discuss how digital communication is better than analog communication. 6. Compare and contrast FDMA, TDMA and CDMA techniques. 7. Elaborate on the goals of IMT2000. 8. Identify the generation of your own mobile phone. Do you think it has the functionality discussed in this chapter for the relevant generation? 9. Differentiate between the two types of handoffs. M02_GARGxxxx_01_SE_C02.qxd 3/5/10 12:05 PM Page 22
- 42. Wireless and Cellular Communication 23 10. What are the features proposed for 4G systems? 11. Assume that 4G mobile technology handhelds are already available. Give some ideas for new features that can be added in 4G systems of tomorrow. Multiple-choice questions 1. The higher frequencies, that is, ultraviolet light, X-rays and gamma rays, are normally not used for wireless transmission, because of which one of the following reasons? (a) They are difficult to produce and modulate (b) Do not propagate well through buildings (c) They are harmful to humans (d) All of the above 2. Which of the following is false for microwaves? (a) They travel in straight lines and are thus affected by the earth’s curvature (b) They are relatively inexpensive to use (c) They can propagate well through buildings (d) They are preferred over optic fiber, especially in harsh terrain or urban areas 3. According to Shannon’s theorem, the maximum data rate D of a noisy channel whose band- width is H Hz, and whose signal-to-noise ratio is S/N, is given by which one of the following formulae? (a) D H log2 (1 S/N) (b) D H (1 log2 S/N) (c) D 2H log2 (1 S/N) (d) None of the above 4. Which of the following is the correct sequence of waves in increasing order of frequencies? (a) Radio, microwaves, infrared, ultraviolet light, X-rays, gamma rays (b) Microwaves, radio, visible light, X-ray, ultraviolet light, gamma rays (c) Radio, microwaves, infrared, ultraviolet light, gamma rays, X-rays (d) Microwaves, radio, infrared, visible light, X-rays, gamma rays 5. Which one of the following is not true for infrared waves? (a) They are capable of traversing multiple paths without interferences (b) They are less secure than radio (c) They have reasonably high bandwidth (d) No government license is required for their operation 6. Globalstar satellites, which are close to the earth, do not need much power, and their round- trip delay is only a few milliseconds, are examples of which one of the following? (a) Geostationary satellites (GEOS) (b) Medium earth orbit satellites (MEOS) (c) Low earth orbit satellites (LEOS) (d) None of the above 7. To which one of the following generations does CDMA belong? (a) First generation (b) Second generation (c) Third generation (d) Fourth generation M02_GARGxxxx_01_SE_C02.qxd 3/5/10 12:05 PM Page 23
- 43. 24 Mobile Computing 8. Which one of the following is the multiple-access scheme used in GSM? (a) Time division multiple access (TDMA) (b) Frequency division multiple access (FDMA) (c) Code division multiple access (CDMA) (d) A combination of TDMA and FDMA 9. Which one of the following best characterizes IS 95? (a) a standard for cellular CDMA (b) a standard for cellular TDMA (c) a standard procedure for measuring indoor multipath propagation characteristics (d) a standard interconnecting base stations 10. The efficiency of a wireless system is given in which of the following units? (a) bits per second (b) bits per second per Hertz (c) bits per second per Hertz per km2 (d) None of the above Further reading A.S. Tanenbaum (2003), Computer Networks, 4th ed. (New Delhi, India: Pearson Education). C. Shannon (1948), ‘A Mathematical Theory of Communication’, Bell System Technical Journal, 27 (July, October): 379–423, 623–656. C.R. Casal, F. Schoute and R. Prasad, ‘A Novel Concept for Fourth Generation Mobile Multimedia Communication’, www.ubicom.tudelft.nl/MMC/Docs/VTC99.pdf (accessed November 2005) ———, ‘Evolution towards Fourth Generation Mobile Multimedia Communication’, www.ubicom. tudelft.nl/MMC/Docs/paper38.pdf (accessed March 2005) D.P. Agrawal and Q.A. Zeng (2003), Introduction to Wireless and Mobile Systems (Thomson, Singapore). J.F. Huber, D. Weiler and H. Brand (2000), ‘UMTS, the Mobile Multimedia Vision for IMT-2000: A Focus on Standardization’, IEEE Communications Magazine, 38 (September): 129–136. J.H. Schiller (2006), Mobile Communications, 2nd ed. (Pearson Education, USA). J.S. Lee and L.E. Miller (1998), CDMA Systems Engineering Handbook (London: Artech House). R. Kamal (2007), Mobile Computing (Oxford University Press). X. Zhou, ‘Overview of the Third Generation Mobile Communications’, www.meru.cecs.missouri. edu/workshop/zxb_pres1.ppt. (accessed January 2005) M02_GARGxxxx_01_SE_C02.qxd 3/5/10 12:05 PM Page 24
- 44. W ith the advent and recent proliferation of handheld devices, wireless local area net- works (WLANs) have become very popular. One can see them in offices, campus build- ings, airports, hotels, restaurants, etc., facilitating continuous access to the Internet, through what has come to be known as the wireless indoors. Recently, the concept of the wireless outdoors has also emerged, which is concerned with the so-called last mile technology or wireless local loop (WLL) or fixed wireless access. To provide connectivity to millions of homes and businesses one has to lay fibre, coax, or category 5 twisted pair, which is a very daunting and costly affair. The provider uses a directed antenna and a transmitter of predefined power to ensure stable reception of high-frequency signals within a limited coverage area, such as an individual building. WLL can be narrowband or wideband. Broadband wireless or wireless metropolitan area networks (WMANs) simply require erecting a big antenna on a hill just outside the town and installing antennas directed at it on customers’ rooftops. We shall study both WLAN and WMAN standards in this chapter. WLANs can operate in two configurations—with base stations or access points that are con- nected to the wired network, or without base stations, that is, mobile ad hoc networks (MANETs). MANETs are the subject of discussion in Chapter 6. Both configurations, however, use the short- range radio-wave transmission discussed in Section 2.2. See Figure 3.1. Wireless Networ ks 3 Base station To wired network (a) (b) Figure 3.1 Wireless Networ ks (a) With Base Station (b) Without Base Station 25 M03_GARGxxxx_01_SE_C03.qxd 3/8/10 3:33 PM Page 25
- 45. 26 Mobile Computing When wireless networks were first developed, there were many challenges that had to be met. Some of them have been mentioned in Chapter 2. These challenges included finding an available worldwide frequency band, dealing with the finite range of signals, maintaining user privacy, taking limited battery life into account, understanding the implications of mobility, making the system economically viable, etc. We shall be dealing with all these issues in the book. 3.1 The need for ne w wireless standards The main standard developed for WLANs is called the IEEE 802.11. So the question that arises here is: what is the need for a new standard, that is, why can’t the universal Ethernet be used for WLANs? The answer lies in the many ways in which wireless operation differs from the tradi- tional wired one. Some of these are discussed below. 1. Ethernet uses carrier sense multiple access with collision detection (CSMA/CD). An Ethernet station just waits until the ether is idle and starts transmitting. If it does not receive a noise burst back within the first 64 bytes, it assumes that the frame has been delivered correctly. But carrier sensing is not possible in the wireless environment. Also, not all stations are within the radio range of each other. Transmissions going on in one part of a cell may not be received elsewhere in the same cell. There are two problems encountered in this scenario—the problem of the hidden station and the problem of the exposed station. a. The hidden station problem: Shown in Figure 3.2 is a WLAN containing stations A, B and C. C, which is not in the radio range of A, is transmitting to station B. If station A senses the channel, it will not hear anything because it is hidden from C. It falsely con- cludes that it may now start transmitting to B, resulting in a collision. b. The exposed station problem: Consider the same WLAN, but now the scenario is as shown in Figure 3.3. A is transmitting to some station D not shown in the diagram. B is near A and can hear A sending. It falsely concludes that it cannot transmit to C, even though it can do so simultaneously. Thus, because of B’s exposed location to A, it defers its transmission even when it need not. 2. Multipath fading (interference). This is due to reflection of radio signals by solid objects, which results in signals being received along multiple paths. This may cause interference, leading to data becoming error-prone in the wireless environment. A B C C is transmitting Range of C’s radio A wants to send to B but cannot hear that B is busy Figure 3.2 The Hidden Station Problem M03_GARGxxxx_01_SE_C03.qxd 3/8/10 3:33 PM Page 26
- 46. Wireless Networ ks 27 3. No handoff in Ethernet. Handoff is the mechanism of allowing a mobile device contin- ued access even when it moves from one cell (network) to another. This is a major requirement in wireless networks, but is not needed in the wired Ethernet. 4. Half-duplex transmission. Most radios are half duplex. They cannot transmit and listen for noise bursts at the same time in a single frequency. Thus, carrier sensing is not possible. 5. Absence of mobility-aware software. Software that is mobility-based or mobility-aware is yet to be made universally available. Until that happens, WLANs of mobile, handheld comput- ers cannot be deployed as universally and simply as the standard Ethernet. The above limitations of the standard Ethernet necessitated the development of a new stan- dard for WLANs. 3.2 IEEE 802.11 WLAN standard The IEEE 802.11 WLAN Standard is popularly known as the Wi-Fi standard. We shall now study in detail its protocol stack, frame structure and services. The physical layer radio-transmission techniques are beyond the scope of this book, but we shall mention them briefly here. Figure 3.6 shows the lower two layers of the IEEE 802.11 protocol stack. Here the data link layer consists of two sublayers, called the logical link control (LLC) layer and the medium access control (MAC) layer. The IEEE 802.11 protocol stack is discussed in detail below. 3.2.1 Physical layer The 802.11 standard was developed in 1997 with data rates of 1 to 2 Mbps for WLANs. Initially, it had three possible modulation techniques for sending MAC frames from a sender station to a receiver station. Only some highlights of these techniques are given below and are as follows. For details, please refer to Tanenbaum (2003). • 802.11: Infrared, which uses diffused transmission at 0.85 or 0.95 microns. Two speeds are permitted, those of 1 Mbps and 2 Mbps. The advantage of infrared transmission, as seen in Chapter 2, is that infrared signals do not penetrate walls, so cells in adjacent rooms are well insulated from each other. But it is not good in sunlight, as sunlight swamps infrared signals. Further, bandwidth is limited. C A B A is transmitting Range of A’s radio B wants to send to C but mistakenly thinks the transmission will fail Figure 3.3 The Exposed Station Problem M03_GARGxxxx_01_SE_C03.qxd 4/5/10 3:53 PM Page 27
- 47. 28 Mobile Computing • 802.11: FHSS (frequency hopping spread spectrum), in which the transmitter hops from frequency to frequency hundreds of times per second. It uses 79 channels, each 1 MHz wide, starting at the low end of the 2.4 GHz ISM (industrial, scientific, medical applications) band. A pseudorandom generator is used to produce the sequence of hopped frequencies. Figure 3.4 shows the concept of FHSS. Stations need to use the same seed for the pseudorandom gener- ator and stay synchronized in time to hop to the same frequencies. The dwell time, which is the amount of time spent at each frequency, is adjustable, but must be less than 400 msec. Since the hopping sequence and dwell time are not known, FHSS provides security against eavesdropping. It is resistant to multipath fading and is relatively insensitive to radio inter- ference, which makes it popular for building-to-building links, that is, for wireless outdoors. Its main disadvantage is its low bandwidth and low power. • 802.11: DSSS (direct sequence spread spectrum) is like CDMA, but has some differences. It is also restricted to 1 or 2 Mbps. Each bit is transmitted as 11 chips in what is called a Barker sequence. Phase shift modulation is used at 1 or 2 Mbaud to transmit 1 or 2 bits per baud, when operating at 1 or 2 Mbps, respectively. The concept of DSSS is shown in Figure 3.5. Subsequently, these speeds were considered too slow, and in 1999, two new standards were proposed. These are as follows: • 802.11a: OFDM (orthogonal frequency division multiplexing), which uses the wider 5 GHz ISM frequency band to deliver up to 54 Mbps. In OFDM, which is a form of spread spectrum, but different from CDMA and FHSS, 52 different frequencies are used: four for synchroniza- tion and 48 for data. Splitting the signal into many narrow bands offers key advantages like better immunity to narrowband interference and the possibility of using non-contiguous bands. It also has good spectrum efficiency in terms of bits/Hz and good immunity to multi- path fading. • 802.11b: HR-DSSS (high-rate DSSS) is another spread-spectrum technique, which uses 11 million chips/second to deliver data rates up to 11 Mbps in the 2.4 GHz band. Data rates Spreading Transmitter Receiver Digital signal Digital signal Spreading signal Hopping pattern Power Frequency Power Frequency Power Frequency Despread Hopping pattern Figure 3.4 Concept of F requency Hopping Spread Spectr um (FHSS) M03_GARGxxxx_01_SE_C03.qxd 3/8/10 3:33 PM Page 28
- 48. Wireless Networ ks 29 supported are 1, 2, 5.5 and 11 Mbps. These rates may be dynamically adapted during opera- tion to achieve the optimum speed possible under current conditions of load and noise. Although it is incompatible with 802.11a and is much slower, its range is 7 times greater. In 2001, another standard was proposed, which is • 802.11g: This uses the modulation technique of 802.11a, that is, OFDM, and the frequency band of 802.11b, so it theoretically delivers up to 54 Mbps data rates. 3.2.2 MAC layer To overcome the hidden and exposed terminal problems of the CSMA/CD-based Ethernet, the MAC sublayer of 802.11 supports two modes of operation. These are the DCF and the PCF (which is optional) and are discussed below. 1. Distributed Coordination Function (DCF): As the name suggests, this mode does not use any central control like the Ethernet. But it uses CSMA/CA, that is, CSMA with collision avoid- ance, which itself supports two methods of operation. Spreading Transmitter Receiver Digital signal s(t) Digital signal s(t) Spreading signal m(t) Code c(t) Power Frequency Power Frequency Power Frequency Despread Code c(t) Figure 3.5 Direct Sequence Spread Spectr um (DSSS) Upper layers Logical link control MAC sublayer 802.11 Infrared 802.11 FHSS 802.11 DSSS 802.11a OFDM 802.11b HR-DSSS 802.11g OFDM Physical layer Data link layer Figure 3.6 The Lower Layers of the IEEE 802.11 Protocol Stack M03_GARGxxxx_01_SE_C03.qxd 3/8/10 3:33 PM Page 29
- 49. 30 Mobile Computing The first method uses physical channel sensing. When a station wants to transmit, it senses the channel. If it is idle, it starts transmitting. It does not continue to sense the channel while transmitting, but sends the complete frame, which may be destroyed at the receiver due to inter- ference there. If a collision occurs, the colliding stations wait for a random time, using the Ether- net binary exponential backoff (BEB) algorithm, and then try again later. If the receiver does not send an acknowledgement, the transmitter knows that a collision has occurred. There is no colli- sion detection at the transmitter. The second method is based on multiple access with collision avoidance for wireless (MACAW) and uses virtual channel sensing. It works as shown in Figure 3.7. Suppose there are four stations A, B, C and D in a network, such that B and C are within the range of A. D is not within A’s range but is within the range of B. Suppose A decides to send data to B. The protocol works as follows: 1. A sends a small 30 byte RTS (request to send) frame to B. 2. If B is ready to receive data, it responds with a CTS (clear to send) frame. 3. When A receives the CTS, it sends its data frame and starts an acknowledgement (ACK) timer. 4. If B correctly receives the data frame, it responds with an ACK frame and terminates the exchange. 5. In case A’s ACK timer expires before it receives the ACK, the whole protocol is repeated. 6. C also receives the RTS frame, as it is in the range of A. It realizes that someone else wants to send data, so it stops transmitting till the data exchange is done. 7. D receives the CTS frame as it is in the range of B. Thus, it also maintains the same state as C. Note that the signals shown in Figure 3.7 for C and D, called network allocation vector (NAV), are not transmitted. They are internal reminders to indicate that no data can be transmit- ted during that time. This is a kind of virtual channel busy signal, asserted by the stations them- selves, using the NAV. The time for which they must wait can be calculated using the information present in the RTS and CTS frames. Because of the noisy, wireless channel, the probability of the frame reaching the destination successfully decreases with frame length. For noisy channels, 802.11 allows frames to be frag- mented into smaller pieces, each with its own checksum. Once the channel has been acquired using RTS and CTS, multiple fragments can be sent in a row (see Figure 3.8). The sequence of fragments is called a fragment burst. Fragmentation increases the throughput by allowing only bad fragments to be retransmitted, not the whole frame. A B C D RTS CTS ACK Data NAV NAV Time Figure 3.7 CSMA/CA V irtual Channel Sensing M03_GARGxxxx_01_SE_C03.qxd 3/8/10 3:33 PM Page 30
- 50. Wireless Networ ks 31 2. Point Coordination Function (PCF): This mode of operation uses a central base station which polls other stations, asking them if they have any frames to send. No collisions can occur here. The base station broadcasts a beacon frame periodically, with the necessary system para- meters, viz., hopping sequence, dwell times, clock synchronization, etc. It also invites new stations to sign up for the polling service. During signing up, the station is guaranteed a certain fraction of bandwidth to maintain quality of service (QoS). To save battery life, the base station can direct a mobile station to go into sleep state until explicitly awakened by the base station or the user. While the mobile station is asleep, the base station buffers any frames directed to it. 802.11 allows both PCF and DCF to coexist within one cell by carefully defining the inter- frame time interval. After a frame has been sent, a certain amount of time is required before any station may send another frame. Four different intervals are defined, each for a specific purpose. These are shown in Figure 3.9. 1. SIFS (short interframe spacing) is used to allow the parties in a single dialog to go first. This includes sending a CTS frame, ACK frame and fragment bursts. 2. PIFS (PCF interframe spacing) is used by exactly one station to respond after an SIFS interval. If the station fails to make use of its chance and the time PIFS elapses, the base station sends a beacon frame or poll frame. 3. DIFS (DCF interframe spacing) is the time after which the base station does not respond. Then any station may attempt to acquire the channel to send a new frame. The usual contention rules apply here. 4. EIFS (extended interframe spacing) is the time used by a station to report error if it has a received a bad or unknown frame. A B C D RTS Frag 1 Frag 2 Frag 3 CTS ACK ACK Fragment burst ACK NAV NAV Time Figure 3.8 IEEE 802.11 F ragment Burst ACK Time SIFS PIFS DIFS EIFS Figure 3.9 IEEE 802.11 Interframe Spacing M03_GARGxxxx_01_SE_C03.qxd 3/8/10 3:33 PM Page 31
- 51. 32 Mobile Computing 3.2.3 Frame structure The frame structure for the 802.11 data frame is shown in Figure 3.10. • The frame control field is 2 bytes long and has various subfields. • Version field gives the version of the protocol as usual. • Type indicates whether it is a data, control or management frame. • Subtype is concerned with the type of control frame. • To DS and From DS indicate that the frame is going to or coming from the intercell distribution system, that is, the Ethernet. • MF indicates more frames will follow. • Retry indicates retransmission of a frame sent earlier. • More indicates that the sender has additional frames to send. • Pwr is the power management bit used by the base station to put the receiver into or take it out of sleep mode. • W specifies that the wireless equivalent privacy (WEP) algorithm is used for encryption. This algorithm will be discussed in detail in Chapter 9. • specifies to the receiver that a sequence of frames with this bit on must be processed strictly in order. • Duration field tells for how long the frame and its ACK will occupy the channel. • Address 1 is the sender’s address. • Address 2 is the receiver’s address. • Address 3 is the sender’s base station address. • Address 4 is the receiver’s base station address. • Sequence field allows fragment numbering—12 bits for frame and 4 bits to identify fragment. • Management frames have a similar format, except that they have only one base address, because they are restricted to a single cell. • Control frames are shorter. They have only one or two addresses and no data and sequence fields. Subtype is important here. 3.2.4 Services A WLAN must provide nine types of services, five for distribution and four for the station. The distribution services are concerned with managing membership within a cell and for interacting with stations outside the cell. They are as follows: 1. Association: This service is used when a mobile station wants to connect to the base station. The mobile station must identify itself and indicate the data rates supported by it, whether Version 2 Bits 2 4 1 1 1 1 1 1 1 1 Type Subtype To DS From DS MF Retry Pwr More W O Frame control 2 Bytes 2 6 6 6 2 6 0-2312 4 Duration Address 1 Address 2 Address 3 Seq. Address 4 Data Checksum Figure 3.10 Data Frame Format for IEEE 802.11 M03_GARGxxxx_01_SE_C03.qxd 3/8/10 3:33 PM Page 32
- 52. Wireless Networ ks 33 it needs PCF services (polling) and what its power requirements are. The base station may accept or reject the mobile station. However, if it is accepted, the mobile station has to authen- ticate itself. 2. Disassociation: This service is used when either the mobile station or the base station wants to break the connection. The base station may do so for maintenance purposes or if it wants to go down. The mobile station may disassociate when it is leaving or shutting down. 3. Re-association: This service is used when a mobile station wants to change the base station, as when it moves from one cell to another. 4. Distribution: This service deals with routing of frames that are sent to the base station. If the destination of the frame is local, it is sent directly over the air; else, it is forwarded over the wired network. 5. Integration: This service is used if the frame has to be sent to a non-802.11 network. The frame must be translated from the 802.11 format to the format of the destination network. The station services are concerned with working within the same cell. They are used after a mobile station has associated with a base station and are as follows: 1. Authentication: This service is used by the base station to check the identity of the mobile station. Initially, the standard did not require the base station to prove its own identity to the mobile station, but this defect in the standard is being corrected. 2. De-authentication: When a mobile station that has been authenticated wants to leave the network, it is de-authenticated by the base station. After de-authentication, it cannot use the network anymore. 3. Privacy: This service is used to ensure that the data in the wireless network is confidential. Encryption is used for this purpose. 4. Data delivery: This service is the one used by mobile stations to send and receive data. It is a reliable service requiring the higher layers to provide for error detection and correction. 3.3 Bluetooth Bluetooth was developed in 1994 by a study interest group (SIG) consisting of IBM, Intel, Nokia and Toshiba for connecting mobile phones or computing and communication devices without the use of cables. In 2002 it was taken up by the IEEE wireless personal area network (WPAN) Committee as the IEEE 802.15 standard, for the physical and data link layers. It is a short-range, low-cost and power-efficient radio-frequency-based wireless technology that supports both point-to-point and point-to-multipoint connections. It connects one hand- held device to another Bluetooth-enabled device(s) within a 30-foot or 10-meter radius, such as mobile phones, laptops, printers and other accessories. It is like having a universal remote for the kinds of devices one uses every day and is oriented towards the mobile consumer wanting to do digital imaging and multimedia applications. Bluetooth operates in the unlicensed ISM band, with slight locational variations. Its essen- tial characteristics are summarized in Table 3.1. Bluetooth-enabled devices can automatically locate each other, but user action is necessary to make connections with other devices and to form networks. Eight devices can be connected in a Bluetooth network, known as a piconet. One of them acts as the master and the others are called slaves. A scatternet is formed when two or more piconets connect via a bridge node. This is shown in Figure 3.11. M03_GARGxxxx_01_SE_C03.qxd 3/8/10 3:33 PM Page 33
- 53. 34 Mobile Computing Laptop C (Master of piconet 3) Laptop Laptop Laptop A (Master of piconet 1) Piconet 1 Piconet 3 Laptop B (Master of piconet 2) Laptop D User C’s PDA Piconet 2 User B’s PDA User B’s mobile phone Figure 3.11 Bluetooth Scatter net In addition to the seven active slaves, there can be up to 255 parked nodes (in low power state) in the net that can only respond to a beacon signal from the master. Slaves are dumb devices, doing what the master tells them to do. The piconet is a TDM system, with the master controlling the clock and determining which slave gets to communicate in which time slot. All communication is between master and slave, not between slave and slave. Characteristics Description Physical La yer FHSS Frequency Band 2.4–2.4835 GHz Hop Frequency 1,600 hops/s Data Rate 1 Mbps Data and Networ k Security Provides three le vels of security , two levels of de vice trust and three le vels of ser vice security. Operating Range 10 m Throughput Around 720 Kbps Advantages No wires and cables for man y interfaces, can penetrate w alls and other obstacles, uses low pow er and minimal hard ware. Disadvantages May interfere with other ISM band technologies, has low data rates. Table 3.1 Bluetooth Features M03_GARGxxxx_01_SE_C03.qxd 3/8/10 3:33 PM Page 34
- 54. Wireless Networ ks 35 3.3.1 Advantages of Bluetooth Bluetooth offers many advantages to users, both the home user and the small business user. This is because of its simple connectivity, which provides increased efficiency and reduced costs. These advantages are as follows: 1. Non-cable connections: Bluetooth technology has replaced cables by wireless for many short-range interconnections. These include mouse and keyboard computer connections; 12 Mbps and 480 Mbps USB (USB 1.1 and 2.0); printers and modems, at 4 Mbps; and wireless headsets and microphones for interfacing with PCs and mobile phones. 2. File sharing: File sharing between Bluetooth-enabled devices has become very simple, enabling Bluetooth-compatible laptops to share files with each other or Bluetooth-compatible mobile phones to act as wireless modems for laptops. Thus Bluetooth provides the laptop with complete networking facilities without using an electrical interface for the same. 3. Wireless synchronization: Bluetooth-enabled devices can automatically synchronize with each other wirelessly. This facilitates personal information present in address and appoint- ment sheets to be transferred between PDAs, laptops, palmtops and mobile phones. 4. Wireless Internet connectivity: Since Bluetooth is supported by a variety of devices, Internet connectivity is possible when these devices connect together and use each other’s cap- abilities. Thus a laptop with a Bluetooth connection can request a mobile phone to establish a dial-up connection and then access the Internet through that connection. Bluetooth will soon be available in office appliances like PCs, faxes, printers and laptops; communication devices like cell phones, handsets, pagers, and headsets; and home systems like DVD players, cameras, refrigerators and microwave ovens. Many other exciting applica- tions for Bluetooth include vending machines, banking and other electronic payment sys- tems, wireless office and conference rooms, smart homes and in-vehicle communications and parking. 3.3.2 Bluetooth applications The Bluetooth standard provides for 13 specific applications to be supported by Bluetooth V1.1. These are given in Table 3.2. 3.3.3 Bluetooth protocol stack The Bluetooth protocol stack defines the software layers used for communication on top of the radio link. The lower layer defines the Bluetooth-specific components. The middle layer consists of the industry standard protocols that were adapted for Bluetooth use so that applications can be ported to Bluetooth easily. The top layer is the application layer. See Figure 3.12. The Bluetooth radio layer defines the transmission characteristics. It sends bits from the master to the slave and vice versa. It is a low-power system with a range of 10 m. Bluetooth trans- ceivers use Gaussian frequency shift keying (GFSK) modulation and employ FHSS with a hopping pattern of 1,600 hops/sec over 79 frequencies in a quasi-random fashion. Although the theoretical maximum data rate of a Bluetooth network is 1 Mbps, in reality it cannot support such data rates because of communication overhead. Second-generation Bluetooth technology is likely to go up to a data rate of 2 Mbps. Bluetooth networks can support both data and speech channels. One asynchronous data channel can be combined with up to three simultaneous synchronous speech channels. A combination of packet-switching technology and circuit-switching technology is used in M03_GARGxxxx_01_SE_C03.qxd 3/8/10 3:33 PM Page 35
- 55. Random documents with unrelated content Scribd suggests to you:
- 56. Fig. 578.—Pupa obtecta: a, of Sesia, with its cocoon-cutter on the head; b, of Tortrix vacciniivorana. THE PUPA STATE The word pupa is from the Latin meaning baby. Linnæus gave it this name from its resemblance to a baby which has been swathed or bound up, as is still the custom in Southern Europe. The term pupa should be restricted to the resting inactive stage of the holometabolous insects. Lamarck’s term chrysalis was applied to the complete or obtected pupa of Lepidoptera and of certain Diptera, and mumia, a mummy, to the pupæ of Coleoptera, Trichoptera, and most Hymenoptera. Latreille (1830) also restricted the term pupa to the “oviform nymph,” or puparium, of Diptera. Brauer applies the term nymph to the pupa of metabolous insects. The typical pupa is that of a moth or butterfly, popularly called a chrysalis. A lepidopterous pupa in which the appendages are more or less folded close to the body and soldered to the integument, was called by Linnæus a pupa obtecta; and when the limbs are free, as in Neuroptera, Mecoptera, Trichoptera, and the lepidopterous genus Micropteryx it is called a pupa libera (Fig. 579). When the pupa is enclosed in the old larval skin, which forms a pupal covering (puparium), the pupa was said by Linnæus to be coarctate. The pupa of certain Diptera, as that of the orthoraphous families, is nearly as much obtected as that of the tineoid families of moths, especially as regards the appendages of the head; the legs being more as in pupæ liberæ (Fig. 580). The male Coccid anticipates the metabolous insects in passing through a quiescent state, when, as Westwood states, it is “covered by the skin of the larva, or by an additional pellicle.” The body appears to be broad and flat, the antennæ and fore legs resting under the head, while the two hinder pairs of legs are appressed to the
- 57. Fig. 579.—Pupa libera of neuropterous insects a, Corydalus cornutus; b, Sialis; c, Hemerobius. Fig. 580.—Pupa obtecta of Diptera: a, Ptychoptera; b, Tabanus atratus; c, Proctacanthus philadelphicus; d, Midas clavatus. Fig. 581.— Pupa libera of Icerya purchasi, ventral view.—After Riley, Insect Life. under side of the body. There is but a slight approach to the pupa libera of a metabolous insect. Riley states that the male larva of Icerya purchasi forms a cocoon waxy in character, but lighter, more flossy, and less adhesive than that of the female egg- cocoon. It melts and disappears when heated, proving its entirely waxy nature. When the mass has reached the proper length, the larva casts its skin, which remains in the hind end of the cocoon, and pushes itself forward into the middle of the cocoon. The pupa (Fig. 581) is of the same general form and size as the larva. All the limbs are free and slightly movable, so that they vary in position, though ordinarily the antennæ are pressed close to the side, as are the wing-pads; the front pair of legs are extended forward. “If disturbed, they twist and bend their bodies quite vigorously.” The pupa state lasts two or three weeks. A similar pupa is that of Icerya rosæ. (Riley and Howard.) The metamorphosis of Aspidiotus perniciosus is of interest. The male nymph differs much after the first moult from the female, having large purple eyes, while the female nymph loses its eyes entirely. It passes into what Riley terms the pro-pupa (Fig. 582, b), in which the wing-pads are present, while the limbs are short and thick. The next stage is the “true pupa” (Fig. 582, c, d), in which the antennæ and legs are much longer than before. There is no waxy cocoon, but only a case or scale composed of the shed larval skin, i.e. “with the first moult the shed larval skin is retained beneath the scale, as in the case of the female; with the later moultings the shed skins are pushed out from beneath the scale,” and when they transform into the imago they “back out from the rear end of their scale.”
- 58. Fig. 582.—Aspidiotus perniciosus, development of male insect: a, ventral view of larva after first moult; b, the same, after second moult (pro-pupa stage); c and d, true pupa, ventral and dorsal views. All greatly enlarged.—After Riley. The pupæ of Coleoptera and of Hymenoptera, though there is, apparently, no near relationship between these two orders, are much alike in shape, and, as Chapman pertinently suggests, those of both orders are helpless from their quiescence, and hence have resorted for protection to some cocoon or cell. But it is quite otherwise with the pupæ of Lepidoptera and Diptera, which vary so much in adaptation to their surroundings, and hence afford important taxonomical and phylogenetic characters. This, as regards the Lepidoptera, was almost wholly overlooked until Chapman called attention to the subject, and showed that the pupæ had characters of their own, of the greatest service in working out the classification, and hence the phylogeny, of the different lepidopterous groups. We have, following the lead of Chapman, found the most striking confirmation of his views, and applied our present knowledge of pupal structures to dividing the haustellate Lepidoptera into two groups,—Paleolepidoptera and Neolepidoptera. The pupæ of the Neuroptera, Coleoptera, and Hymenoptera differ structurally from the imago, in the parts of the head and thorax being less differentiated. Thus in the head the limits or sutures between the epicranium and clypeus, and the occiput and gula, are obscurely marked, while the tergal and pleural sclerites of the imago are not well differentiated until the changes occurring just before the final ecdysis.
- 59. It is easy, however, to homologize the appendages of the pupæ with those of the imago of all the holometabolous orders except in the case of the obtected pupa of the Lepidoptera (and probably of the obtected dipterous pupæ), where the cephalic appendages are soldered together. That the appendages of the lepidopterous pupa are, as generally supposed, merely cases for those of the imago has been shown by Poulton to be quite erroneous. He says: “If we examine a section of a pupal antenna or leg (in Lepidoptera), we shall find that there is no trace of the corresponding imaginal organ until shortly before the emergence of the imago. In the numerous species with a long pupal period, the formation of imaginal appendages within those of the pupa is deferred until very late, and then takes place rapidly in the lapse of a few weeks. This also strengthens the conclusion that such pupal appendages are not mere cases for the parts of the imago, inasmuch as these latter are only contained within them for a very small proportion of the whole pupal period.” On the other hand, Miall and Hammond claim that there is a strong superficial contrast as to the formation of the imaginal organs, between Lepidoptera and tipularian Diptera, the appendages, wings, and compound eyes being substantially those of the imago. “With the exception of the prothoracic respiratory appendages and the tail-fin, there is little in the pupa of Chironomus which does not relate to the next stage.” The exact homology of the “glazed eye” of the lepidopterous pupæ and of the parts under the head, situated over the maxillæ, is difficult to decide upon, and these points need farther examination. In the dipterous pupa it is interesting to observe that the halteres are large and broad, which plainly indicates that they are modified hind wings. The number and arrangement of the spiracles is different in pupæ from those of the larva and imago.
- 60. Fig. 583.—Simulium piscicidium: a, larva; b, c, d, pupa; e, thoracic leg; f, row of bristles at end of body. A, S. pecuarum, pupa; a, b, c, adminicula.—After Riley. There are also secondary adaptive structures peculiar to the pupa, which are present and only of use in this stage. These are the thoracic, spiracular, or breathing appendages of the aquatic Diptera (Fig. 583), the various spines situated on the head or thorax, or on the sides, or more often at the end of the abdomen, besides also the little spines arranged in more or less circular rows around the abdominal segments, the cocoon-breaker, and the cremaster of many pupæ. In the pupa of certain Diptera, there is a terminal cremaster-like spine, as in that of Tipula eluta (Fig. 584), Tabanus lineola (Fig. 585), besides adminicula or locomotive spines like those of lepidopterous pupæ (Fig. 580, a, b, c).
- 61. Fig. 584. —Pupa of Tipula eluta. Fig. 585. —Pupa of Tabanus lineola.— This and Fig. 584 after Hart.
- 62. Fig. 586.—Pupa of Galerita lecontei, and of Adelops hirtus (a, b, c).—After Hubbard. The pupæ of Coleoptera are variously spined or hairy (Fig. 586). Those of Hydrophilus and of Hydrobius are provided with stout spines on the prothorax and abdomen which support the body in its cells, so that, as Lyonet first showed, though surrounded on all sides by moist earth, it is kept from contact with it by the pupal spines; other pupæ of beetles, such as that of the plum weevil, which is also subterranean, possess similar spines. The abdomen of many coleopterous pupæ, such as those of Carabidæ, end in two spines, to aid them in escaping from their cells in wood or in the earth; others have stiff bristles, and others spines along each side of the abdomen (Fig. 586). All these structures are the result of a certain amount of activity in what we call quiescent pupæ, but most of these are for use at the end of pupal life, at the critical moment when by their aid the insect escapes from its cocoon or subterranean cell, or if parasitic, bores out of its host. If we are to account for the causes of their origin, we are obliged to infer that they are temporary deciduous structures due to the need of support while the body is subjected to unusual strains and stresses in working its way out of its prison in the earth, or its cell within the stems and trunks of plants and similar situations. They are pupal inheritances or heirlooms, and well illustrate the inheritance of characters acquired during a certain definite, usually brief, period of life, and transmitted by the action of synchronous heredity. The pupæ of certain insects are quite active, thus that of Raphidia, unlike that of Sialis, before its final ecdysis regains its activity and is able to run about. (Sharp, p. 448.)
- 63. a. The pupa considered in reference to its adaptation to its surroundings and its relation to phylogeny The form of the pupa is a very variable one, as even in Lepidoptera it is not entirely easy to draw the line between a pupa libera and a pupa obtecta (Fig. 578); and though the period is one of inactivity, yet when they are not in cocoons or in the earth in subterranean cells, their form is more or less variable and adapted to changes in their surroundings. Even in the obtected pupa of butterflies, there is, as every one knows, considerable variability of shape and of armature, which seems to be in direct adaptability to the nature of their environment. Scudder has well shown that in certain chrysalids, such as those of the Nymphalidæ, which are variously tuberculated, and hang suspended by the tail, and often hibernate, these projections serve to protect the body. All chrysalids with projections or ridges on different parts of the body, being otherwise unprotected, move freely when struck by gusts of wind, hence “the greater the danger to the chrysalis from surrounding objects, the greater its protection by horny tubercles and roughened callous ridges.” The greater the protection possessed in other ways, as by firm swathing or a safe retreat, the smoother the surface of the body and the more regular and rounded its contours. The tendency to protection by tubercles is especially noticeable in certain South American chrysalids of nymphalid butterflies. This response to the stimuli of blows or shocks is also accompanied by a sensitiveness to the stimulus of too strong light. Previously Scudder[103] had made the important suggestion that the smooth crescent-shaped belt of the “glazed eye” or “eyepiece” of chrysalids is, as an external covering of the eye, midway between that of the caterpillar and the perfect insect, and he asks: “May it not be a relic of the past, the external organ of what once was? And are we to look upon this as our hint that the archaic butterfly in its transformations passed through an active pupal stage, like the lowest insect of to-day, when its limbs were unsheathed, its appetite unabated?” etc. Scudder also shows that “the expanded base of the sheath covering the tongue affords protection also to the palpi which lie beneath and beside the tongue.”
- 64. Fig. 587.—Pupa of Micropteryx purpuriella, front view: md, mandibles; mx. p, maxillary palpus, end drawn separately; mx.′ p, labial palpi; lb, labrum. All this tends to show the importance of studying the structure of the pupa, in order to ascertain how the pupal structures have been brought about, with the final object of discovering whether the pupæ of the holometabolic insects are not descended from active nymphs, and if so, the probable course of the line of descent. b. Mode of escape of the pupa from its cocoon “In all protected pupæ,” as Chapman says, “the problem has to be faced, how is the imago to free itself from the cocoon or other envelope protecting the pupa.” In the Coleoptera and Hymenoptera the imago becomes perfected within the cocoon or cell, as the case may be, and as Chapman states, “not only throws off the pupal skin within the cocoon, but remains there till its appendages have become fully expanded and completely hardened, and then the mandibles are used to force an outlet of escape,” and he calls attention to the fact that “in many cases, even in some entire families, they are of no use whatever to the imago except in this one particular,” and he cites the Cynipidæ as perhaps the most striking instance of this circumstance. In those Neuroptera which spin a silken cocoon, e.g. the Hemerobiidæ, the Trichoptera, and in Micropteryx (Fig. 588), the jaws used by the pupa for cutting its way out of the cocoon are even larger in proportion than in the pupa of caddis-flies (Fig. 588), being of extraordinary size.
- 65. Fig. 588.—Mandibles (md) of Micropteryx purpuriella, enlarged.— Author del. A, pupal head of a hydropsychid caddis-fly, showing the large mandibles.—After Reaumur, from Miall. In Myrmeleon the pupa pushes its way half out of the cocoon, and then remains, while the imago ruptures the skin and escapes (Fig. 589, a). Thus in the Neuroptera and Trichoptera we have already established the more fundamental methods of escape from the cocoon, which we see carried out in various ways in the more generalized or primitive Lepidoptera. The most primitive method in the Lepidoptera of escaping from the cocoon seems to be that of Micropteryx.
- 66. Fig. 589.—Larva of Myrmeleon with (a) its cocoon and cast pupa-skin. “In this genus,” says Chapman, “though it is nominally the pupa that escapes from the cocoon, it is in reality still the imago, the imago clothed in the effete pupal skin. To rupture the cocoon it uses not its own jaws, but those of the pupal skin, energizing them, however, in some totally different way from ordinary direct muscular action, their movements being the result of the vermicular movements of the pupa, acting probably by fluid pressure on the articular structure of the jaws, by some arrangement not altogether different perhaps from the frontal sac of the higher Diptera. In the Micropteryges the jaws of the pupa not only rupture the cocoon, but appear to be the most active agents in dragging the pupa through the opening in the cocoon and through any superincumbent earth, being merely assisted by the vermicular action of the abdominal segments, and we find in accordance with this circumstance that the pupal envelope is still very thin and delicate, and has little or no hardening or roughness by which to obtain a leverage against the walls of the channel of escape.” (Trans. Ent. Soc. London, 1896, pp. 570, 571.) Some sort of a beak or hard process, more or less developed, according to Chapman, adapted for breaking open the cocoon exists in nearly all the Lepidoptera with incomplete pupæ (pupæ incompletæ), except the limacodid and nepticulid section. “In all these instances the pupa emerges from the cocoon precisely as in the Micropteryges, that is, the moth it really is that emerges, but does so encased in the pupal skin. To achieve this object, it seems to have been found most efficient to have three, four, or five abdominal segments capable of movement, but to have the terminal sections (segments) soldered together.” This cocoon-breaker, as we may call it, is especially developed in Lithocolletis hamadryadella. As described by Comstock, it forms a toothed crest on the forehead which enables it to pierce or saw through the cocoon.
- 67. Fig. 590. —Pupa of Talæpori a: a, cocoon- cutter; with vestiges of four pairs of abdomin al legs, and the cremaste r. “Each pupa first sawed through the cocoon near its juncture with the leaf and worked its way through the gap, by means of the minute backward-directed spines upon its back, until it reached the upper cuticle of the leaf. Through this cuticle it sawed in the same way that it did through the cocoon. The hole was in each case just large enough to permit the chrysalis to work its way out, holding it firmly when partly emerged. When half-way out it stopped, and presently the skin split across the back of the neck and down in front along the antennal sheaths, and allowed the moth to emerge.”[104] We have observed and figured the cocoon-breaker in Bucculatrix, Talæporia (Fig. 590, a), Thyridopteryx, and Œceticus, and rough knobs or slight projection answering the purpose in Hepialidæ, Megalopyge, Zeuzera, and in Datana.[105] See also the spine on the head of Sesia tipuliformis (Fig. 578). The imago of the attacine moths cuts or saws through its cocoon by means of a pair of large, stout, black spines (sectores coconis), one on each side of the thorax at the base of the fore wings (Fig. 591), and provided with five or six teeth on the cutting edge (C, D). Our attention[106] was drawn to this subject by a rustling, cutting, and tearing noise issuing from a cocoon of Actias luna. On examination a sharp black point was seen moving to and fro, and then another, until both points had cut a rough irregular slit, through which the shoulder of the moth could be seen vigorously moving from side to side. The hole or slit was made in one or two minutes, and the moth worked its way at once out of the slit. The cocoon was perfectly dry. The cocoon-cutter occurs in all the American genera, in Samia cynthia, and is large and well marked in the European Saturnia pavonia-minor and Endromis versicolora. In Bombyx mori the spines are not well marked, and they are quite different from those in the Attaci. There are three sharp points, being acute angles of the pieces at the base of the wing, and it must be these spines which at times perform the cutting through of the threads of the cocoon described by Réaumur, and which he thought was done by the facets of the eyes. It is well known that in order to guard against the moths cutting the threads, silkraisers expose the cocoon to heat sufficient to destroy the enclosed pupa. In Platysamia the cocoon-cutters, though well developed, do not appear to be used at all, and the pupa, like that of the silkworm and other moths protected by a cocoon, moistens the silk threads by a fluid issuing from the mouth, which also moistens the hairs of the head and thorax, together with the antennæ. It remains to be seen whether these structures are only occasionally used, and whether the emission of the fluid is not the usual and normal means of egress of the moth from
- 68. Fig. 591.—Cocoon- cutter of the Luna moth: front view of the moth with the shoulders elevated and the rudimentary wings hanging down: s, cocoon-cutter; p, patagium. B, represents another specimen with fully developed wings: ms, scutum; st, scutellum of the mesothoracic segment; s, cocoon- cutter, which is evidently a modification of one of the pieces at the base of the fore wings; it is surrounded by membrane, allowing free movement. C and D, different views of the spine, magnified, showing the five or six irregular teeth on the cutting edge. Fig. 592.—Larva and pupa of a wood-wasp (Rhopalum), enlarged: h, temporary locomotive tubercles on head of pupa.—Trouvelot del. its cocoon. Dr. Chapman remarks that throughout the obtected moths “there are many devices for breaking through the cocoon: specially constructed weak places in the cocoon, softening fluid, applied by the moth, assisted by special appliances of diverse sorts, such as in Hybocampa[107] and Attacus,” etc. As to the fluid mentioned above, Trouvelot states that it is secreted during the last few days of the pupa state, and is a dissolvent for the gum so firmly uniting the fibres of the cocoon. “This liquid is composed in great part of bombycic acid.” (Amer. Naturalist, i, p. 33.) The pupa of the dipterous genus Sciara (S. ocellaris O. S.) resembles a tineid pupa, and before transforming emerges for about two-thirds of its length from the cocoon; the pupa-skin remaining firmly attached in this position.[108] Certain hymenopterous pupæ are provided with temporary deciduous conical processes. Thus we have observed in the pupa of Rhopalum pedicellatum two very prominent acute tubercles between the eyes (h, Fig. 592). As the cocoon is very slight, these may be of use either in extracting itself from the silken threads or in pushing its way along before emerging from the tunnel in the stem of plants. (See also p. 611.) c. The cremaster Although this structure is in general confined to lepidopterous pupæ, and is not always present even in them, since it is purely adaptive in its nature, yet on account of its singular mode of development from the larval organs, and the accompanying changes in the pupal abdomen, it should be mentioned in this connection. The cremaster is the stout, triangular, flattened, terminal spine of the abdomen, which aids the
- 69. pupa in working its way out of the earth when the pupa is subterranean, or in the pupa of silk-spinning caterpillars its armature of secondary hooks and curved setæ enables it to retain its hold on the threads of the interior of its cocoon after the pupa has partially emerged from the cocoon, restraining it, as Chapman well says, “at precisely that degree of emergence from the cocoon that is most desirable.” He also informs us that while in the “pupæ incompletæ the cremaster is attached to an extensible cable, which always allows some emergence of the pupa, in the pupæ obtectæ there is no doubt but that in such cases as the Ichthyuræ, Acronyctæ, and many others, it retains the pupal case in the same position within the cocoon that the living pupa occupied; this is also very usually the case in the Geometræ and in the higher tineids (my pyraloids).” In many of the more generalized moths there is no cremaster (Micropteryx, Gracilaria, Prodoxus, Tantura, Talæporia, Psychidæ, Hepialidæ, Zeuzera, Nola, Harrisina), though in Tischeria and Talæporia (Fig. 590, but not in Solenobia) and Psychidæ, two stout terminal spines perform the office of a cremaster, or there are simply curved setæ on the rounded, unarmed end of the abdomen, as in Solenobia. In the obtected Lepidoptera, for example in such a group as the Notodontidæ, where the cremaster is present, though variable in shape, it may from disuse, owing to the dense cocoon, be without the spines and hooks in Cerura, or the cremaster itself is entirely wanting in Gluphisia, and only partially developed in Notodonta. In the butterflies whose pupæ are suspended (Suspensi), the cremaster is especially well developed. Reference might here be made to the temporary pupal structures in certain generalized moths, which take the place of a cremaster, such as the transverse terminal row of spines in Tinea, the two stout spines in Tischeria, and the dense rough integument and thickened callosities of the pupal head and end of abdomen of Phassus, which bores in trees with very hard wood; also the numerous stout spines at the end and sides of the abdomen in Ægerians. These various projections and spines, besides acting as anchors and grappling hooks, in some cases serve to resist strains and blows, and have undoubtedly, like the armature in the larvæ and imagines of other insects, arisen in response to intermittent or occasional pressure, stresses, and impacts. Mode of formation of the cremaster and suspension of the chrysalis in butterflies.—We are indebted to Riley[109] for an explanation of the way the cremaster has originated, his observations having been made on species of over a dozen genera of butterflies (Suspensi). He shows that the cremaster is the homologue of the suranal plate of the larva.[110] The preliminary acts of the larva have been observed
- 70. by various authors since the days of Vallisneri, i.e. the larva hanging by the end of the abdomen, turning up the anterior part of the body in a more or less complete curve, and the skin finally splitting from the head to the front edge of the metathoracic segment, and being worked back in a shrivelled mass toward the point of attachment. The critical feat, adds Riley, which has most puzzled naturalists, is the independent attachment of the chrysalis and the withdrawal from and riddance of the larval skin which such attachment implies. Réaumur explained this in 1734 by the clutching of the larval skin between sutures of the terminal segments of the chrysalis, and this is the case, though the sutures act in a somewhat different way. Before pupation the larva spins a mass or heap of silk, the shape of which is like an inverted settee or a ship’s knee, and “one of the most interesting acts of the larva, preliminary to suspension, is the bending and working of the anal parts in order to fasten the back of the (suranal) plate to the inside of the back of the settee, while the crotchets of the legs are entangled in the more flattened position or seat.” In shedding the larval skin, the following parts are also shed, and have some part to play in the act of suspension: i.e. 1st, the tracheal ligaments (Fig. 593, tl), or the shed tracheæ from the last or 9th pair of spiracles; 2d, the rectal ligament (Fig. 593, rl), or shed intestinal canal; 3d, the Osborne or retaining membrane (membrana retinens, Fig. 593, mr), which is the stretched part of the membrane around the rectum and in the anal legs, and which is intimately associated with the rectal ligament.
- 71. Fig. 593.—Shrunken larval skin of Vanessa antiopa, cut open from the back and showing (mr) the retaining membrane, (rl) the rectal ligament, and (tl) the tracheal ligaments. The structures in the chrysalis are, first, the cremaster, with its dorsal (Fig. 594, dcr) and ventral (vcr) ridges, and the cremastral hook-pad (chp), said by Riley to be “thickly studded with minute but stout hooks, which are sometimes compound or furnished with barbs, very much as are some of our fishing-hooks, and which are most admirably adapted to the purpose for which they are intended.” Secondly, there are the other structures, viz., the sustainers (sustentors), two projections which Riley states “homologize with the soles (plantæ) of the anal prolegs, which take on various forms (3), but are always directed forward so as easily to catch hold of the retaining membrane.” These sustentors are, however, as Jackson[111] has shown, and as we are satisfied, the vestiges of the anal legs.
- 72. Fig. 594.—Ideal representation of the anal subjoint of Vanessa antiopa, from behind, with the spines removed, and all parts forced apart by pressure so as to show the homologies of the parts in the chrysalis which are concerned in pupation: homologies indicated by corresponding letters in Fig. 595, except that r (the rectum) corresponds with pr in Fig. 595. Fig. 595.—Anal parts of chrysalis of Vanessa antiopa, just prior to final extraction from shrunken larval skin: c, cremaster; chp, cremastral hook-pad; h, one of the hooks, more enlarged; vcr, ventral cremastral ridge; dcr, dorsal cremastral ridge; lr, larval rectum; pr, pupal rectum; rp, rectal plate; sr, sustentor ridges; mr, membrana retinens; rl, rectal ligament; tl, tracheal ligament; the 11th or last spiracle- bearing joint and the 12th joint being numbered.
- 73. Fig. 596.—A, chrysalis of Terias. B, posterior end of chrysalis of Paphia. C, posterior end of chrysalis of Danais. E, one of the sustainers of Terias, greatly enlarged to show its hooked nature. All the parts of subjoint lettered to correspond with Fig. 595. Thirdly, the sustentor ridges, which, as Riley states, may be more or less obsolete in some forms, in Paphia (Fig. 596, B) and Limenitis form “quite a deep notch, which doubtless assists in catching hold of the larval skin in the efforts to attach the cremaster.”
- 74. Fig. 597.—Pupation of butterflies: a, attachment of larva of Danais archippus; p, attachment of larva of Paphia glycerium; b, ideal larva soon after suspension; d, ideal larva a few hours later, the needle (n) separating the forming membrane from the sustainers; l, ideal larva just before splitting of larval skin, with retaining membrane loosened from the sustainers and showing its connection both with the larval and pupal rectum. In all the figures the joints of the body are numbered; the forming chrysalis is shaded in transverse lines; the intervening space between it and larval skin is dotted: h, is the hillock of silk; hl, hooks of hind legs; ap, anal plate; lr, larval rectum; pr, pupal rectum; mr, retaining membrane; c, cremaster; s, sustainers.—This and Figs. 593–596 after Riley. “It is principally,” adds Riley, “by the leverage obtained by the hooking of the sustainers in the retaining membrane, which acts as a swimming fulcrum, that the chrysalis is prevented from falling after the cremaster is withdrawn from the larval skin. It is also principally by this same means that it is enabled to reach the silk with the cremastral hook-pads.” “Dissected immediately after suspension, the last abdominal segment of the larva is found to be bathed, especially between the legs and around the rectum, in an abundance of translucent, membranous material.” “An hour or more after suspension the end of the forming chrysalis begins to separate from the larval skin, except at the tip of the cremaster (Fig. 597, b). Gradually the skin of the legs and of the whole subjoint (10th segment) stretches, and with the stretching, the cremaster elongates, the rectal piece recedes more and more from the larval rectum, and the sustentor ridges diverge more and more from the cremaster, carrying with them, on the sustainers, a part of the soft membrane.”
- 75. The rectal ligament will sustain at least 10 or 12 times the weight of the chrysalis. That of Apatura seems to rely almost entirely on the rectal ligament, assisted by the partial holding of the delicate larval skin.
- 76. FORMATION OF THE PUPA AND IMAGO IN THE HOLOMETABOLOUS INSECTS (THE DIPTERA EXCEPTED) We have seen that in the incomplete metamorphosis, although there may be as many as five, and possibly seven moults, and in Chloëon as many as 20, and in Cicada septemdecim perhaps 25 or 30, there is but a slight change of form from one stage to another, and no period of inactivity. And this gradual outer transformation is so far as yet known paralleled by that of the internal organs, the slight successive changes of which do not differ from those observed in the growth of ametabolous insects. With the growth of the internal organs there probably goes on a series of gradual regenerative processes, and Korschelt and Heider state that we may venture to assume that each changed cell or group of cells which have become exhausted by the exercise of the functions of life are reabsorbed and become restored through the vital powers of the tissues, so that as the result there goes on a constant, gradual regeneration of the organs. While the Hemiptera have only an incomplete metamorphosis, the males of the Coccidæ are, as shown by O. Schmidt, remarkable for passing through a complete or holometabolous development, with four stages, three of which are pupal and inactive. Hence, as Schmidt observes, there is here a hypermetamorphosis, like that of the Meloidæ, Stylopidæ, etc. Shortly before the end of the larval stage of the male appear the imaginal buds of the eyes, legs, and wings. In the 2d or 1st pupal stage there is an atrophy of the antennæ and legs. On the other hand, at this stage the female completes its metamorphosis. The rudiments of the wings arise on the edge of the dorsal and ventral side of the 2d thoracic segment, and this, we would remark, is
- 77. significant as showing a mode of origin of the wings intermediate between that of the manometamorphic and holometamorphic insects. (See pp. 137–142.) While Schmidt could not ascertain the exact structure of the imaginal buds, he says “in general the process of formation of the extremities is exactly as Weismann has described in Corethra.” The two later pupal stages are “as in other metabolic insects.” (See p. 690, Fig. 637.) Thus far the internal changes in the metamorphosis of the Coleoptera have not been thoroughly studied. They are less complete than in the other holometabolous insects, the differences between the larva and imago being much less marked than in the more specialized orders, and so far as known all the larval organs pass, though not without some great changes, directly into the imaginal ones, the only apparent exception being the mid-intestine, which, as stated by Kowalevsky, undergoes a complete transformation during metamorphosis. The following account, then, refers almost wholly to the Lepidoptera, Hymenoptera, and Diptera. a. The Lepidoptera The first observations on the complete metamorphosis of insects which were in any way exact were those of Malpighi, in 1667, and of Swammerdam, in 1733. While the observations of Swammerdam, as far as they extended, were correct, his conclusions were extraordinary. They were, however, accepted by Réaumur and by Bonnet, and generally held until the time of Herold in 1815, and lingered on for some years after. The rather famous theory of incasement (“emboîtement”) propounded by Swammerdam was that the form of the larva, pupa, and imago preëxisted in the egg, and even in the ovary; and that the insects in these stages were distinct animals, contained one inside the other, like a nest of boxes, or a series of envelopes one within the other, or, to use his own words: “Animal in animali, seu papilio intra erucam reconditus.” This theory Swammerdam extended to the whole animal kingdom. It was based on the fact that by throwing the caterpillar, when about to pupate, in boiling water, and then stripping off the skin, the immature form of the butterfly with its appendages was disclosed.
- 78. Malpighi had previously observed the same fact in the silkworm, perceiving that before pupation the antennæ are concealed in the head of the larva, where they occupy the place previously taken by the mandibular muscles; also that the legs of the moth grew in those of the larva, and that the wings developed from the sides of the worm. Even Réaumur (1734) remarked: “Les parties du papillon cachées sous le fourreau de chenille sont d’autant plus faciles à trouver que la transformation est plus proche. Elles y sont neanmoins de tout temps.” He also believed in the simultaneous existence of two distinct beings in the insect. “Il serait très curieux de connaître toutes les communications intimes qui sont entre la chenille et le papillon.... La chenille hache, broye, digere les aliments qu’elle distribué au papillon; comme les mères préparent ceux qui sont portés aux fœtus. Notre chenille en un mot est destineé à nourrir et à defendre le papillon qu’elle renferme.” (T. i, 8e Mémoire, p. 363.) Lyonet (1760), even, did not expose the error of this view that the larva enveloped the pupa and imago, and, as Gonin says, it was undoubtedly because he did not use for his dissections of the caterpillar of Cossus any specimens about to pupate. Yet he detected the wing-germs and those of the legs, stating that he presumed the bodies he saw to be the rudiments of the legs of the moth (p. 450). Herold, in his work on the development of the butterfly (1815), was the first to object to this erroneous theory, showing that the wings did not become visible until the very end of larval life; that as the larval organs disappear, they are transformed or are replaced by entirely new organs, which is not reconcilable with a simple putting off of the outer envelope. The whole secret of metamorphosis, in Herold’s opinion, consisted in this fact, that the butterfly in the larva state increases and accumulates a supply of fat until it has reached the volume of the perfect state; then it begins the chrysalis period, during which the organs are developed and take their definite form. [112] (Abstract mostly from Gonin.) Still the old ideas prevailed, and even Lacordaire, in his Introduction à l’Entomologie published in 1834, held on to Swammerdam’s theory, declaring that “a caterpillar is not a simple animal, but compound,” and he actually goes so far as to say that “a caterpillar, at first scarcely as large as a bit of thread, contains its own teguments threefold and even eightfold in number,
- 79. besides the case of a chrysalis, and a complete butterfly, all lying one inside the other.” This view, however, we find is not original with Lacordaire, but was borrowed from Kirby and Spence without acknowledgment. These authors, in their Introduction to Entomology (1828), combated Herold’s views and stoutly maintained the old opinions of Swammerdam. They based their opinions on the fact, then known, that certain parts of the imago occur in the caterpillar. On the other hand, Herold denied that the successive skins of the pupa and imago existed as germs, holding that they are formed successively from the “rete mucosum,” which we suppose to be the hypodermis of later authors. In a slight degree the Swammerdam-Kirby and Spence doctrine was correct, as the imago does arise from germs, i.e. the imaginal disks of Weismann, while this was not discovered by Herold, though they do at the outset arise from the hypodermis, his rete mucosum. Thus there was a grain of truth in the Swammerdam-Kirby and Spence doctrine, and also a mixture of truth and error in the opinions of Herold. The real nature of the internal changes wrought during the process of metamorphosis was first revealed by Weismann in 1864. His discovery of the germs of the imago (imaginal buds) of the Diptera, and his theory of histolysis, or of the complete destruction of the larval organs by a gradual process, was the result of the application of modern methods of embryology and histology, although his observations were first made on the extremely modified type of the Muscidæ or flies, and, at first, he did not extend his view to include all the holometabolous insects. Now, thanks to his successors in this field, Ganin, Dewitz, Kowalevsky, Van Rees, Bugnion, Gonin, and others, we see that metamorphosis is, after all, only an extension of embryonic life, the moults and great changes being similar to those undergone by the embryo, and that metamorphosis and alternation of generations are but terms in a single series. Moreover, the metamorphoses of insects are of the same general nature as those of certain worms, of the echinoderms, and the frog, the different stages of larva, pupa, and imago being adaptational and secondary. While the changes in form from the larva to the pupa are apparently sudden, the internal histogenetic steps which lead to them are gradual. In the Lepidoptera a few days (usually from one to three) before assuming the pupa stage, the caterpillar becomes
- 80. restless and ceases to take food. Its excrements are now hard, dry, and, according to Gonin, are “stained carmine red by the secretions of the urinary tubes.” Under the microscope we find that they are almost exclusively composed of fragments of the intestinal epithelium. These red dejections were noticed by Réaumur, and afterwards by Herold, and they are sure indications of the approach of the transformations. It now wanders about, and, if it is a spinner, spins its cocoon, and then lies quietly at rest while the changes are going on within its body. Meanwhile, it lives on the stores of fat in the fat-body, and this supply enables it to survive the pupal period. The amount of fat is sometimes very great. Newport removed from the larva of Cossus ligniperda 42 grains of fat, being more than one-fourth of the whole weight of the insect, he adds that the supply is soon nearly exhausted during the rapid development of the reproductive organs, “since, when these have become perfected, the quantity that remains is very inconsiderable.” Although the larval skin of a lepidopterous insect is suddenly cast off, the pupa quickly emerging front it, yet there are several intermediate stages, all graduating into each other. If a caterpillar of a Clisiocampa, which, as we have observed, is much shortened and thickened a day or two before changing to a pupa, is hardened in alcohol and the larval skin is stripped off, the semipupa (pro-nymph, pro-pupa of different authors) is found to be in different stages of development, and the changes of the mouth-parts are interesting, though not yet sufficiently studied. Newport attributes the great enlargement and changes in the shape of the thoracic segments of the larva of Vanessa urticæ at this time, to the contraction or shortening of the muscles of the interior of those segments, “which are repeatedly slowly extended and shortened, as if the insect were in the act of laborious respiration.” This, he adds, generally takes place at short intervals during the two hours immediately preceding the change to the pupa, and increases in frequency as that period approaches. He thus describes the mode of moulting the larval skin: “When the period has arrived, the skin bursts along the dorsal part of the 3d segment, or mesothorax, and is extended along the 2d and 4th, while the coverings of the head separate into three pieces. The insect then exerts itself to the utmost to extend the fissure along the segment of the abdomen, and, in the meantime, pressing its body through the opening, gradually withdraws its antennæ and legs, while the skin, by successive
- 81. contortions of the abdomen, is slipped backwards, and forced towards the extremity of the body, just as a person would slip off his glove or his stocking. The efforts of the insect to get entirely rid of it are then very great; it twirls itself in every direction in order to burst the skin, and, when it has exerted itself in this manner for some time, twirls itself swiftly, first in one direction, then in the opposite, until at last the skin is broken through and falls to the ground, or is forced to some distance from it. The new pupa then hangs for a few seconds at rest, but its change is not yet complete. The legs and antennæ, which when withdrawn from the old skin were disposed along the under surface of the body, are yet separate, and do not adhere together as they do a short time afterwards. The wings are also separate and very small. In a few seconds the pupa makes several slow, but powerful, respiratory efforts; during which the abdominal segments become more contracted along their under surface, and the wings are much enlarged and extended along the lateral inferior surface of the body, while a very transparent fluid, which facilitated the slipping off of the skin, is now diffused among the limbs, and when the pupa becomes quiet dries, and unites the whole into one compact covering.” The changes in the head and mouth-parts.—The changes of form from the active mandibulate caterpillar to the quiescent pupa, and then to the adult butterfly, are, as we have seen, in direct adaptation to their changed habits and surroundings, and they differ greatly in details in insects of different orders. In many Lepidoptera and certain Diptera the pupa and imago are without the mandibles of the larva, and, instead, the 1st maxillæ in the former order, and the 2d maxillæ in the latter, are highly developed and specialized. The changes in the shape of the head, with the antennæ, the latter rudimentary in the larvæ of the two orders named, are noteworthy, and will be referred to under those orders. The same may be said of the thorax with the legs and wings, and the abdomen with the ovipositor. Every part of the body undergoes a profound change, though in the Coleoptera, Trichoptera, and the more generalized and primitive Diptera, each segment and appendage of the larva are directly transformed into the corresponding parts of the pupa, and subsequently of the imago. We shall see, however, beyond, that this general statement does not apply to the Hymenoptera, in which there
- 82. is a process of cephalization or transfer of parts headward, peculiar to that order.
- 83. Fig. 598.— Internal organs of Sphinx ligustri: 1, head; 2–4, thoracic, 5–13, abdominal segments; V, fore-, M, mid-, E, hind- intestine; gs, brain; gi, infraœsophage al ganglion; n, ventral ganglion; vm, urinary tubes;
- 84. Welcome to Our Bookstore - The Ultimate Destination for Book Lovers Are you passionate about books and eager to explore new worlds of knowledge? At our website, we offer a vast collection of books that cater to every interest and age group. From classic literature to specialized publications, self-help books, and children’s stories, we have it all! Each book is a gateway to new adventures, helping you expand your knowledge and nourish your soul Experience Convenient and Enjoyable Book Shopping Our website is more than just an online bookstore—it’s a bridge connecting readers to the timeless values of culture and wisdom. With a sleek and user-friendly interface and a smart search system, you can find your favorite books quickly and easily. Enjoy special promotions, fast home delivery, and a seamless shopping experience that saves you time and enhances your love for reading. Let us accompany you on the journey of exploring knowledge and personal growth! ebookgate.com