Modified Epc Global Network Architecture of Internet of Things for High Load Rfid Systems

ACEEE Int. J. on Network Security, Vol. 01, No. 03, Dec 2010

Modified Epc Global Network Architecture of
Internet of Things for High Load Rfid Systems
Atishay Jain1 , Ashish Tanwer2
Computer Science Department Thapar University, Patiala – 147001, India
Email: atishay811@gmail.com
Electronics and Communication Department, Thapar University, Patiala – 147001, India
Email: ashishtanwer@gmail.com

Abstract- This paper proposes a flexible and novel
architecture of Internet of Things (IOT) in a high density and TABLE I:
mobility environment. Our proposed architecture solves the RANGES AND APPLICATIONS OF RFID
problem of over-loading on the network by monitoring the
Frequency
total number of changed objects changing global location Characteristics Applications
Range
crossing the fringe boundaries rather than the actual number
of objects present or those that move within the local area. We Short range (To Livestock ID
Low Frequency
have modified the reader architecture of the EPCglobal 18 inches) Low Reusable
125 – 300 kHz
Architecture. The components and the working of the model reading speed containers
has been illustrated in detail. We have also discussed the Medium range Access Control
physical implementation of our model taking the examples of High Frequency (3-10 feet) Airline Baggage
a smart home sample application and the performance results 13.56 MHz Medium ID Library
have been tabulated and represented graphically.
reading speed automation
Index Terms-- Internet of Things, RFID, Smart Home High range (10
Supply chain
Application, Modified EPCGlobal Architecture Ultra High – 30 feet) High
management &
Frequency 400 reading speed
Container
MHz–1 GHz Orientation
I. INTRODUCTION Tracking
sensitive
The Barcode for tagging items is being replaced by Automated Toll
Microwave
RFID, the new generation Auto ID technology which is a Medium range Collection
Frequency > 1
realtime alternative that uses wireless communication to (10+ feet) Vehicle
GHz
uniquely identify and track an object. It was invented in Identification
1948 and was first-used during the IInd World War by the
US Army for identification of friend or foe (IFF) aircrafts. TABLE II:
RFID STANDARDS
The technology has found usage in many industry sectors
and application like airline baggage tracking, automated Specification Description Frequency
vehicle identification and toll collection. A RFID system ePC UHF 64-bit factory
basically consists of a Tag, a Reader and an antenna. The 900 MHz
Class O programmed
RFID Tag is a transponder with a silicon microchip for ePC UHF 96/128 bit one-time
storing large amounts of data which is used to uniquely 860-930 MHz
Class 1 programmable
identify the tagged item. Tags can be either active or ePC HF Class 96/128 bit one-time
passive. Passive tags are read only, gain their power from 1 programmable
13.56 MHz
that generated by a reader. The reading range is typically
ePC UHF 96/128 bit one-time-
shorter up to 30 feet (3 meters) and the data storage 860-960 MHz
Gen 2 Programmable
capacity is comparatively less (96/128 bits) as compared to
active tags. Active tags have both read/write capability and ISO 18000-3 Item Management 13.56 MHz
are powered by means of battery. This battery-supplied ISO 18000-4 Item Management 2.4 GHz
power enables data to be read and written on to a tag and
thus gives it a greater reading range up to 300 feet (100 ISO 18000-6 Item Management 860-960 MHz
meters) and large data storage capacity (128 KB). Some
popular frequency ranges or RFID and their applications
The RFID reader can be
are given in Table1.
• Fixed RFID reader like UHF standard
There are a number of existing (ISO) and proposed
• Multi antenna RFID reader for supporting several
RFID standards (EPC Global) that have different data
content, use different protocols and have different appliances and even can be
applications as shown in Table 2. With the adoption of Gen • Handheld mobile RFID (MRFID) readers
2 ePC (UHF) standards, the adoption of RFID systems is The internet of Things (IOT) is a networked
now a major tool for supply chain management. interconnection of objects. It is global expansion wireless
Electronic Product Code (EPC) network implemented
through RFID tags [3] or QR Codes.
7
© 2010 ACEEE
DOI: 01.IJNS.01.03.40


An EPC number essentially contains:
• Header, which identifies the length, type, structure,
version and generation of EPC
• Manager Number, which identifies the company or
company entity
• Object Class, refers to a stock keeping unit or product
SKU
• Serial Number, which identifies a specific item of the
Object Class being tagged.

II. BACKGROUND
EPCglobal has developed the Object Name Service
(ONS) [4], a mechanism which makes use of the Domain
Name System (DNS) protocol [5] to discover information
about a product and related services from the Electronic
Product Code (EPC) and is used for the resource
addressing of Internet of Things. The EPC is first encoded
to a Fully Qualified Domain Name (FQDN), then existing
DNS infrastructure is used to query for additional
information. This procedure makes use of the Name
Authority Pointer (NAPTR) DNS record [6], which is also
used by E.164 NUmber Mapping (ENUM) [7]. Ubiquitous Fig 1: EPCglobal Network Standards (Source: GS1 Germany/EPCglobal)
ID Center (uID Center) brings forward similar resource
addressing service named uCode Resolution Protocol C. Object Naming Service (ONS)
(uCodeRP) [8], which also utilizes the protocol similar to ONS is the network service system, similar to the DNS.
DNS. This server will contain all EPC numbers and their
We are proposing ONS based a novel architecture of associated IP addresses. ONS points out the specific EPC-
Internet of Things (IOT) implemented using RFID IS server where the information being queried. It has
network. Our proposed architecture solves the problem of standard security specifications and API.
over-loading as the load on the RFID Tag reader is the total
number of changed objects rather than the actual number of D. EPC Information Service (EPC-IS)
objects present as described in Section-II. EPC-IS is a software component to communicate with
the EPCglobal Network and the ONS server. It stores the
III. RFID NETWORK ARCHITECTURE information processed by EPC middleware and query
related information. EPCIS protocols manage storing and
EPCglobal is a joint venture between GS1 (formerly accessing of EPC information via the EPCglobal Network)
EAN International) and GS1 US (formerly Uniform Code
Council). The organization has created worldwide standard E. EPC middleware (Savant)
for EPC, RFID and the use of the Internet to share data via Program module or service with a series of specific
the EPCglobal Network. attributes, which is integrated by users to meet their
Fig 1 shows the EPCglobal Network Standards given specific needs. The most important part of EPC
by EPCglobal. EPCglobal Network has following middleware is Filter and collection Application Level
components Events (F&C ALE).
A. Tags (Transponder) with EPC Our proposed RFID based Internet of Things network
architecture is based on the original EPCglobal Network
Tags follow coding standard of EPC tag information. architecture and has some specializations as explained
The EPC coding scheme provides differentiating codes [1] below:
for each object of RFID network. Air interface protocol
(GEN 2 AIP) regulates communication between the reader RFID Tags: The RFID tags are attached to each object that
and the tag). Tag data translation protocols converts EPC is to be identified as shown in Fig 3 (a). The RFID tags
information to Internet compatible format used for our purpose are passive and are in inactive state
and need to be woken up by a wake-up call when in radio
B. EPC enabled Reader (Interrogator/Scanner) range of an active Tag Reader.
Reader follows standard reader protocol to exchange Tag readers (Master and Slave): Our system has two types
data between EPC-capable middleware Reader of Tag Readers –Master reader and Slave reader [2]. The
management specifications are used to configuration master reader is a conventional powerful fixed active
readers. reader with a direct fixed or wireless connection. It initiates
a read process in the slave reader and wakes up any passive
tags for power-up or any other service initiation. In
8
© 2010 ACEEE
DOI: 01.IJNS.01.03.40


addition, it collects the item-level information and forwards
it to the back-end for further processing. Fig 2 shows the
placement of Master and Slave Tag Readers. The Master
Tag readers are placed at entry/exit points while Slave Tag
Readers are in Blocks.
Apart form these, for monitoring precise location of each
object within the region a special transmitter is used to call
the wakeup calls on the tags that is placed at the boundary
of the block containing the Slave Readers. The transmitter
has a special antenna to focus the beam straight so that the
tags within the region of the slave reader only get activated
when they cross the block boundary. To minimize costs,
these transmitters do not have readers as more accurate Fig 3 (a) Block Structure (b) Communication of Master reader with ONS
location can be judged with te slave due to the focusing of and Inventory database
the beam in a transmitter(marked as red lining in figure
3(a). IV. NETWORK WORKING
The Master reader communicates with both ONS and
Inventory database in two different states as shown in Fig 3 The objects labeled with the RFID tags will be located
(b). The proposed database is used for inventory within the closed region divided into blocks each of which
management in the system with many records. It is having a slave reader to identify the items whose RFID tags
responsible for maintaining information as cache for can be activated by the slave reader as well as the
reading purposes. transmitter located at the block boundary. This slave reader
will in normal working situation be in inactive state. The
We have made following assumption in proposing our slave readers can be present alone in a block or networked
system architecture: as tiered arrangement of aggregation. The slave readers
1) The number of RFID objects in the block is very will be connected to a master reader which would act like
large. an aggregation server for the slave reader readings and will
pass the information the ONS and used the inventory
database for caching. The connection of the slave to the
master reader can be wireless, direct wired or switched
with redundant cabling based on the use case. The slave
readers located at the entry/exit point will remain active
and function as the primary readers for data update. The
transmitters located at the boundary of a block will remain
active but will have a limited range and will activate
objects that cross the block boundary.
A. Case I Initial Setup / Refresh
The initial setup will consist of the master reader
instructing the slave readers to active state (by sending
wake-up call) one by one which in return will force the
RFID tags of all the objects in the state space (region) into
the active state (by their wake-up calls). The slave readers
will transmit the read RFID tags to the master reader which
will, with the help of the ONS cache the location of the
Fig 2: Placement of Master and Slave Tag Readers at Floor 0 of a objects in the inventory server. The field will include the
Building having 6 Blocks (A-F)
RFID, the object description, the user identification number
2) Most of the objects are stationary for most of the time. (in case it is different from the RFID) as in the case of a
3) The block size is such that the movement of object local repository of the ONS. Alongside this, the location as
within the block is insignificant. well as the block number of the objects will be saved.
4) There are a fixed number of entry/exit points within
B. Case II Reading
the region. The objects can get in or go ou through the
region. The reading of a tag involves getting the data from the
5) The objects can move within the blocks of a region. inventory database and returning the information.
6) Each object has a unique RFID that can be interpreted C. Case III Movement of object within the Region
at the Object Name Server (ONS).
In case the object move within the region from one
block to another, the transmitter at the block periphery with
activate the tag and the slave reader in the new block will

9
© 2010 ACEEE
DOI: 01.IJNS.01.03.40


read the tag values. Only the location will be updated B. Each room as a RFID segment
within the database. In this arrangement, each entry point within a room acts
D. Case IV Removal of Object as an entry point and each room itself acts as a complete
RFID system. The main server that accounts for all the
In case of removal of object, the active readers the exit
rooms can be located anywhere and can run separate blocks
points will activate the RFID tags. These RFID tags will
of code for each room. The transmitters will be located
send the information to the readers which will match it
within blocks inside each rom giving very precise location
inside the inventory database. The deletion of information
of an object within a room. Though, it could have been
can occur directly from the inventory database for the
done in the previous architecture increasing the block
object. In case of an object absent, it indicates that the data
density but having a huge number of blocks in a system
inside the database needs refresh. The primary difference
would increase the calculations costing heavily in case of a
between the reading in case of movement and removal is
requirement of a refresh. This architecture, though will be
that, in case of removal, the transmitter cum receiver makes
able to handle very large number of objects within a room
sure that the object is removed, while in case of movement,
as well as an account for a more accurate position of an
the trade-off between accuracy and precision is applied by
object within the room the object was detected in.
moving the slave reader away from the transmitter for more
The second architecture can be used to provide nearly
precision.
accurate information for all the devices in a smart home. If
E. Case V Addition of Object the RFID objects are equipped with sensors that detect
The addition of object to the system through the entry movement and transmit information, the system though
points will involve the wakeup of the RFID tag by the entry optimal, would get out of practical considerations for size
points. The entry point reader will inform the master to and cost.
wake up the corresponding slave to a reading state where it Fig 4 shows the possible arrangements of RFID Tags
reads active tags but does not transmit the Wake-up Calls. readers based upon our architecture. Arrangement 1 used
The newly entered object will be in the active state and will some master tags (colored red) and some slave tags
be transmitting the RFID information. This can be used to (colored orange). If we add the master readers which are
track its new location by the slave readers without indicated in orange, it turns into the arrangement 2. The
activating the other objects and thus preventing energy densite of the slave tags would have to be increased in that
wastage or clatter. The fringe transmitters will continue to case within the rooms. The RFID tag of the object gets
function normally to locate the exact location within the activated as soon as it changes room/enter/exits the house.
region. This is then recorded as it moves across the house. The
sensitivity of the system can be increased by increasing the
V. SMART HOME APPLICATION number of density of readers at strategic locations which
slows the performance in case of large amount of
The high level architecture of the smart home is similar movements of objects.
to figure 2 and 3. There are two possible arrangements for
the architecture in a smart home.
A. Each room as a block
In this arrangement, the assumptions that hold are that
the RFID objects in a room are not of a very large number
and there is no major ovement of objects between the
rooms. The RFID tagging is used to detect whether the
object is at home and in which room was it taken last. The
advantage of this structure is lower costs though it is
impractical for real life implementation. In figure 4, the red
RFIS reader is the active reader that will activate the RFID
tag when the item enters the house. The green slave readers
are placed such that the entire house is in the reading range
of the slave readers. It does not matter in which room they
are placed as long as we know the entire house is covered. Fig 4 Arrangement of RFID Tags in Sweet Home
The active RFID tag will provide its details and location
An example of the strategic location could be above the
while the object is moved to its final position. This position
dining table as it marks an entrance of a different part of
will be indicated as the objects’ position in the reader. It is
the main hall – the kitchen. The correct room as well as the
assumed that the entry/exit point in the house is only the
last known position of the object is available in this setup.
main door and not the windows. The transmitters wil be
placed within the doors of the room. These will enable us
VI. PERFORMANCE RESULTS
to get an idea regarding which room the ojecy currently is.
For the implementation of our architecture the
performance characteristics of master and slave RFID tag
readers should be shown in Table 4 and corresponding Fig
10
© 2010 ACEEE
DOI: 01.IJNS.01.03.40


6. As shown in figure, the number of tags read by slave Internet of Things (IOT) is future network of all objects
reader should decrease steeply with increase in reading and it can be implemented using RFID. Its standards are
distance so that two slave tag reader cannot interfere with given by EPCglobal. Our proposed architecture is based on
each other. On the other hand, the master tag reader should EPCglobal Network with some assumptions and
have long range so that it can make connection with all the modifications.
slave readers.
TABLE III:
READING PERFORMANCE (%) OF MASTER AND SLAVE TAG READERS

Tags Read by Tags Enabled
Reading Tags Read by
Master for Slave
Distance Slave Reader
Reader Reader
(in ft) (On
(Entry Exit (Using
activation)
Points) Transmitter)
1 100% 100% 100%
2 100% 80% 60%
3 95% 55% 40%
4 90% 45% 15%
5 85% 35% -
6 80% 20% - Fig 5: Reading Performance comparison of master and Slave Tag Readers

It solves the problem of over-loading as the load on the
The higher range of master readers as illustrated in the RFID Tag reader is the total number of changed objects
figure is a prime requirement for the system to be effective. rather than the actual number of objects present. We have
No object entering the system should be able to skip the also discussed the way to implement our architecture for
master reader. Ideal placement for the master reader would sample applications smart home. Future work in this
be overhead or along the sides of the entry/exit points. The system could be to incorporate a means for tracking
sphere of 100% reading performance should cover the movement of objects within the system. Another extension
entire area of these points. Additional readers networked as could be to provide a placement criterion for the slave
one can be used to boost performance. The performance readers, which in the current architecture use the
curve indicates that the slave reader should also be placed overlapping structure similar to the placement of wireless
to make blocks (Fig 3a) keeping them at 4-5 feet from each access point for internet access. The major difference in the
other. This range largely depends on the tag density (here systems is the size of the slave readers which is much
20 tags/feet) for higher density increases clutter and smaller, with smaller ranges as compared to access points.
disturbance. Also for slave readers, the density refers to the Also, slave readers can be arranged without wired
number of active tags within range rather than all tags, in connections to contact and share information wirelessly
case of which, there could be 1-2 tags in a slow moving over the network. Awareness of RFID technology and
system within the range of a tag reader. The transmitters Internet of Things concept can benefit industry and home
have to be placed to form an interlocking block. They are automation globally.
made especially to have little effect on object. It is worth
mentionaing that the object may be able to bypass the REFERENCES
transmitter in rare cases because of the vertical range, but
would never be able to pass out of the system unnoticed [1] Supply Chain Information Transmission based on RFID and
from the master readers at the entry and exit points. The Internet of Things, Bo Yan, Guangwen Huang, see its ref 1, 6,
7.
density of the mesh of transmitters and the slave readers
[2]Smart Home Mobile RFID-based Internet-Of-Things Systems
would largely depend on the usiness case and the accuracy and Services, Mohsen Darianian, Martin Peter Michael
requirement of the system, which is not that critical in a [3] A Model Supporting Any Product Code Standard for the
smart home as in stock movement tracking applications. Resource Addressing in the Internet of Things
[4] EPCglobal Inc, Object Name Service (ONS) Version 1.0,
VII. CONCLUSION EPCglobal, US, Oct.2005.
[5] P. Albitz and C. Liu, DNS and BIND, 4th ed. O’Reilly &
RFID offers new levels of visibility for companies that Associates, 2001.
want to track physical items between locations. Our model [6] M. Mealling and R. Daniel, The Naming Authority Pointer
with some modifications can be used in the retail supply (NAPTR) DNS Resource Record, RFC 2915, IETF,
chain where the goods tagged at the point of manufacture September 2000.
can now be traced from the factory to the shop floor, [7] P. Faltstrom, “E.164 number and DNS”, IETF RFC2916,
providing a real time view of inventory for all supply chain September 2000.
[8] Koji Minegishi, on ucode Resolution Server Connection Tests,
partners.
TRONWARE, 2003, V.84, PP. 71-73.

11
© 2010 ACEEE
DOI: 01.IJNS.01.03.40

Modified Epc Global Network Architecture of Internet of Things for High Load Rfid Systems

More Related Content

What's hot (20)

Viewers also liked (7)

Similar to Modified Epc Global Network Architecture of Internet of Things for High Load Rfid Systems (20)

More from IDES Editor (20)

Recently uploaded (20)

Modified Epc Global Network Architecture of Internet of Things for High Load Rfid Systems