Routing in Delay Tolerant Networks

Seminar by
Anubhav Mahajan
06111502709
C.S.E. 7th Sem










Delay Tolerant Networks(DTNs) are promising new
development in network research field.
DTNs are those networks where instantaneous end-to-end
paths among nodes may not be possibly established or are
absent for a long period of time.
Purpose of this study was to survey DTN routing problem.
It includes study of DTN routing protocols each of which
aims at increasing the message delivery probability as well
as reducing , as much as possible, the overheads incurred
during the whole delivery process of the message.
Henceforth, performance of these routing protocols will be
evaluated using the ONE simulator which is designed
specifically to simulate these routing protocols.









Delay Tolerant Networks are wireless networks where
disconnections and delays may occur frequently due to
propagation phenomena such as node mobility, power
outages etc.
DTN comes into play when delays are observed due to nonexistence or long absence of instantaneous end-to-end
paths among nodes.
One of the main reasons behind these are untimely and
random movement of intermediate nodes (mobile hosts)
that have to carry data from any source to destination.
In order to achieve data delivery, a “store-and-forward”
approach is taken where data is incrementally moved and
stored throughout the network in “hopes” that it will
eventually reach its destination.



To survey DTN routing problem.



To study different DTN routing protocols.



To study ONE simulator.



Performance evaluation of DTN routing protocols using the
ONE simulator.





DTN comes into play when instantaneous end-to-end paths
among network nodes may not be possible to establish or
are absent for a very long time.
A “store-and-forward” approach is taken where data is
incrementally moved and stored throughout the network in
“hopes” that it will eventually reach its destination.









DTN can be rendered as a special case of Ad-hoc networks
where network partitions occurs very frequently and that
too for a long period of time.
Traditional Ad-hoc networking routing protocols requires an
pre-existing instantaneous end-to-end path between
source and destination nodes either directly or indirectly
through intermediate mobile nodes.
But if an instantaneous end-to-end connectivity is failed to
get established, these protocols fails along with them.
These failures also includes the cases where at the time of
origination of a message there was an instantaneous endto-end connectivity between source and destination which
may not be present at the time when the message is
travelling through the path and results in loss of message.



Interplanetary Internet.



Mobile posts around the planet including space stations or
posts on terrain such as Antarctica.



Deep sea communication.



Mobile military deployments across seas and desserts.



Routing protocols will be used to move messages end-toend across the network which is time varying as well as
whose dynamics are not known in advance.



Goals of a DTN routing protocol are to:

1.) maximize message delivery rate
2.) minimize message latency
3.) minimize the total resources consumed in message
delivery such as buffer sizes within mobile hosts(carriers)
devoted to store other host’s messages ,energy of the
host(carrier) consumed in storage and transmission of that
message etc.






A flooding based DTN routing protocol.
Simply based on random pair-wise exchanges of messages
that takes place among mobile hosts and among islands of
mobile hosts to ensure eventual message delivery.
Copies of a message will spread like some epidemic of a
disease in the network, ultimately reaching to its
destination.






SV : Summary vector
Summary vector of a mobile host is a brief summary of
messages that are currently present in the buffer of a
mobile host.
When two mobile hosts come into contact with each other,
they decide which messages are to be exchanged between
them by looking at other host’s summary vector.



A controlled flooding based DTN routing protocol but with
no random pair-wise exchanges.



Message exchanges are based on delivery predictability
which indicates how likely is that this node will be able to
deliver a message to its particular destination.



Delivery predictability is measured on the basis of the
number of times two hosts comes into contact or the
number of times a host has successfully delivered a
message to a particular destination through transitivity.



A message will be transferred from first node to second
node if and only if the second node have a higher value of
delivery predictability than the first node for the
destination of that message.









Not a flooding based technique.
It consists of two phases :Spray phase and Wait phase.
Spray phase: Source node will spray a limited number of
copies of the message into those nodes that have much
higher delivery predictability than the source node in hope
that one of the sprayed node is destination itself.
Wait phase : If the destination is not found in the spraying
phase, each of the nodes carrying a message copy tries to
performs direct transmission (i.e. will forward the message
only to its destination).
If still not successful then spraying phase will begin again
after waiting for some pre-specified time period.



Based on formation of social groups.



Assumes that nodes that are assigned to the same social
network (classroom, project team) will regularly interact
with members of that social group.



With these groups identified, consistent routes towards
particular destinations are identified based on the delivery
history of a group or node.



To deliver a message to a member or members of a
particular group , simply transfer the message to a node
which regularly interacts with any member of that group.



Rest, that member of the group will spread the message to
all the members of that group.





Simulation tool, developed in Java, used for processing of
DTN routing protocols.
It has been designed specifically for DTN routing protocols
and applications.
Provides interactive visualization and post-processing
tools.






First we write implementation of a DTN routing protocol in
the form of JAVA classes which are fed into ONE simulator,
which itself is a collection of java classes and packages.
ONE simulates the protocol as per our specifications.
As a result, it generates a detailed report on performance
evaluation of the protocol as per simulation.

Feeded Java Classes

Simulation Data

Specifications

Report



Performance of DTN routing protocols will be evaluated on
the basis of some specified parameters.



These are :
1.) Percent messages delivered using that protocol.
2.) Average latency of delivered messages.
3.) Total number of transmissions.
4.) Hop count per message.
5.) Average overhead per message.



After simulating a DTN protocol, as per our pre-specified
configurations, the report of performance of the protocol
will be generated in terms of these parameters.

First we provide a configuration to ONE over which these
protocols will be simulated turn by turn.
 This configuration includes :
1.) The map/scenario over which nodes will roam.
2.) The number of nodes in the scenario.
3.) The buffer size at each node.
 Then on the basis of following two criteria, performance of
each DTN routing protocol will be evaluated and
compared. These are :
1.) Parameter vs. Variable Buffer Size keeping number of
nodes fixed in the scenario.
2.) Parameter vs. Variable node count keeping buffer size
of each node fixed in the scenario.




Successful survey of DTN routing problem.



Successful study of different DTN routing protocols.



Successful study of the ONE simulator and its working.



Successful performance evaluation of DTN routing
protocols using the reports generated by the ONE
simulator.



Investigations for other forwarding strategies that can
enhance the performance of DTN further.



There is much scope for addition of more maturity to the
ONE simulator . This includes more functionalities such as
automatic comparisons among DTN protocols on the basis
of parameters rather than making us do manual
comparisons.



Standardization of data which is used for simulating DTN
protocols on the ONE simulator.



[1] A. Vahdat and D. Becker, “Epidemic routing for partially
connected adhoc networks,” Duke University, Tech. Rep.,
2000.



[2] Roy Cabaniss, Sanjay Madria, George Rush, Abbey Trotta
and Srinivasa S. Vulli Department of Computer Science,
“Dynamic Social Grouping Based Routing in a Mobile Ad-Hoc
Network”Missouri University of Science and Technology, Rolla,
Missouri 65401, School of Computing and
Engineering,University of Missouri, Kansas City, Missouri 64110



[3] Thrasyvoulos Spyropoulos, Konstantinos Psounis and
Cauligi S. Raghavendra, “Spray and Wait: An Efficient Routing
Scheme forIntermittently Connected Mobile Networks”



[4] Sushant Jain, Kevin Fall, Rabin Patra “Routing in a Delay
Tolerant Network”



[5] Anders Lindgren, Avri Doria, Olov Schel´en “Probabilistic
Routing in Intermittently Connected Networks”

[6] https://www.netlab.tkk.fi/tutkimus/dtn/theone/
[7] https://crawdad.cs.dartmouth.edu/data.php
[8] https://www.wikipedia.org

Routing in Delay Tolerant Networks

More Related Content

What's hot (20)

Viewers also liked (19)

Similar to Routing in Delay Tolerant Networks (20)

Recently uploaded (20)

Routing in Delay Tolerant Networks