RAVI SHARMA (KCS-711) Mobile Computing

Mobile Adhoc Network

(MANET)
RAVI SHARMA (KCS-711) Mobile Computing

RAVI SHARMA (KCS-711) Mobile Computing

 Introduction
RAVI SHARMA (KCS-711) Mobile Computing

RAVI SHARMA (KCS-711) Mobile Computing

Mobile Ad Hoc Networks (MANETs) refer to a class of wireless

R Anetworks
V I S H A Rthat
M A can
( K Cbe
S -formed
7 1 1 ) dynamically and
M orandomly
bile Comwithout
puting
the need for infrastructural setups. Such networks are able to
adapt and reconfigure themselves on the fly according to node
RAVI SHARMA (KCS-711) Mobile Computing
mobility and changing network topologies.

RAVI SHARMA (KCS-711) Mobile Computing

2
 MANET: Characteristics
RAVI SHARMA (KCS-711) Mobile Computing
• Infrastructure-less (Decentralized):
R A V–I Unlike
S H A the
R Mtraditional
A ( K C Smobile
- 7 1 1wireless
) M obase
networks in which b i lstations,
e Com p u tpoints
access ing

and servers have to be deployed before the networks can be used, MANET is
decentralised, with all mobile nodes functioning as routers and all wireless devices
being interconnected to one another.
R A V–I Intuitively,
S H A R Mthis
A means
( K C S that
- 7 1the
1 ) MANET is also a self-configuring
Mobile C o m p uintwhich
network ing
network activities, including the discovery of the topology and delivery of messages,
are executed by the nodes themselves. This adds robustness against topological and
communication failures.
RAVI SHARMA (KCS-711) Mobile Computing
• Dynamic Topology
– Nodes are free to move arbitrarily, causing the network topology to change rapidly
and unpredictably over time. Alternative paths are automatically found, after which
RAVI SHARMA (KCS-711) Mobile Computing
data packets are forwarded across the multi-hop paths of the network.
– MANETs use various routing mechanisms to adapt to dynamic topological changes.
3
 MANET: Characteristics
RAVI SHARMA (KCS-711) Mobile Computing
• Bandwidth-constrained variable-capacity links:
RAVI SHARMA (KCS-711) Mobile Computing
– In contrast to wired networks, which are characterised by high bandwidth links,
low bit error rates and stable and symmetric links, MANET’s wireless links have
significantly lower capacity in addition to having relatively low bandwidth links,
RAVI SHARMA (KCS-711) Mobile Computing
high bit error rates, and unstable and asymmetric links.

– One effect of having a low link capacity is that congestion is typically the norm
rather than the exception.
RAVI SHARMA (KCS-711) Mobile Computing

• Energy Constraints
– Since nodes in MANETs are often mobile having battery-powered devices, which
RAVI SHARMA (KCS-711) Mobile Computing
essentially are an exhaustible source of energy, power conservation /
management is an important consideration while designing a MANET System.
4
 MANET: Characteristics
RAVI SHARMA (KCS-711) Mobile Computing

R• A Security:
VI SHARMA (KCS-711) Mobile Computing
– Mobile wireless networks are more prone to the physical security threats of
eavesdropping, interception, denial-of-service and routing attacks as
compared to fixed-cable networks.
RAVI SHARMA (KCS-711) Mobile Computing
– To overcome these issues, MANET’s nodes must be configured to radiate as
little power as necessary and transmit as infrequently as possible. This will

R A V I decrease
S H A R Mthe
A probability
( K C S - 7 1of1detection
) and interception.
Mobile Computing

RAVI SHARMA (KCS-711) Mobile Computing

5
 MANET: Advantages and Dis-Advantages
RAVI SHARMA (KCS-711) Mobile Computing
• Advantages:
R A V–I Decentralized
S H A R M Ai.e. No
( Kcentral
C S - 7network
1 1 ) administration required.
Mobile Computing
– Each node can play both the roles i.e. of router and host showing autonomous nature.
– Self-configuring and self-healing nodes do not require human intervention.
– Highly scalability.
R A V–I Robustness
S H A R Mdue
A to( multi-hop
K C S - 7information
11) relay. Mobile Computing
• For example, in the cellular network setting, a drop in coverage occurs if a base station stops working, however the
chance of a single point of failure in a MANET is reduced significantly since the data can take multiple paths.

• Disadvantages:
R A V–I Battery
S H A Power
R M Aconstraints.
(KCS-711) Mobile Computing
– Lack of authorization facilities
• There are no central entities, so operations must be managed in a completely distributed manner.

– Limited physical security.

R A V–I All
SH network
A R Mentities
A ( Kcan
C be
S -mobile,
7 1 1 so
) a very dynamic topology
M isoneeded.
bile Computing
– Network functions must have a high degree of adaptability.
– High latency i.e. There is a huge delay in the transfer of data between two sleeping nodes. 6
 Ad hoc Networks: Applications
RAVI SHARMA (KCS-711) Mobile Computing
• Military environments (soldiers, tanks, planes)
RAVI SHARMA (KCS-711) Mobile Computing
– UAV Ad hoc networks
• These are also termed as Flying ad hoc networks (FANETs). FANETs are composed of unmanned aerial vehicles, allowing
great mobility and providing connectivity to remote areas.
• FANETS are typically used for data collection and situation sensing, without risking the pilot in a foreign unfriendly
environment.

– Tactical MANETs
RAVI SHARMA (KCS-711) Mobile Computing
• These are used by military units with emphasis on data rate, real-time requirement, fast re-routing during mobility, data
security, radio range, and integration with existing systems.

• Vehicular ad hoc networks (VANETs)

• Data monitoring and mining
RAVI SHARMA (KCS-711) Mobile Computing
• Emergency and rescue operations
• Meeting rooms
• Personal area networking, e.g. Bluetooth
RAVI SHARMA (KCS-711) Mobile Computing
• Wireless home networking
• Special applications (industrial control, taxis, boats)
7
 Ad hoc Routing Protocols
RAVI SHARMA (KCS-711) Mobile Computing
Ad-hoc routing protocols can be classified into three major groups based on
Rthe
A Vrouting
I S H Astrategy.
R M A ( These
K C S - 7are:
11) Mobile Computing

• Pro-active or Table driven

– Routes to a destination are determined when a node joins the network or changes
its location, and are maintained by periodic route updates
R A V–I InS Hother
A R Mwords,
A ( Keach
C S - node
7 1 1 )maintains one or more
M o tables
b i l e containing
C o m p u routing
ting
information to every other node in the network. When the network topology
changes the nodes propagate update messages throughout the network in order to
maintain a consistent and up-to-date routing information about the whole network.
• Reactive or On-demand
R A V–I When
S H A aR source
M A ( wants
K C S -to
7 1send
1 ) to a destination, it M o b i l the
invokes e Croute
o m request
p u t i n or
g
route discovery mechanisms to find the path to the destination. The route
remains valid till the destination is reachable or until the route is no longer
needed.
R• A Hybrid
VI SHARMA (KCS-711) Mobile Computing
– It combine the features of both pro-active and reactive routing protocols to scale
well with network size and node density.
8
 Ad hoc Routing Protocools
RAVI SHARMA (KCS-711) Mobile Computing

RAVI SHARMA (KCS-711) Mobile Computing

RAVI SHARMA (KCS-711) Mobile Computing

RAVI SHARMA (KCS-711) Mobile Computing

RAVI SHARMA (KCS-711) Mobile Computing

9
 Ad hoc Routing Protocools
RAVI SHARMA (KCS-711) Mobile Computing

RAVI SHARMA (KCS-711) GSR

Mobile Computing

Proactive FSR
RAVI SHARMA (KCS-711) Mobile Computing
DSDV
Adhoc Routing
Protocols
RAVI SHARMA (KCS-711) AODV
Mobile Computing

Reactive DSR
RAVI SHARMA (KCS-711) Mobile Computing
TORA
10
 Pro-active Routing Protocols
RAVI SHARMA (KCS-711) Mobile Computing
• Pro-active routing protocols require each node to maintain up-to-date
R A routing
V I S H Ainformation
R M A ( K C Sto- 7every
1 1 ) other node (orMnodes
o b i l elocated
C o m pwithin
u t i n ga
specific region) in the network.
• These routing protocols are based on the working principles of two
popular routing algorithms used in wired networks:
RAVI SHARMA (KCS-711) Mobile Computing
• Link-State routing (LS)
– In this approach, each node maintains at least a partial view of the whole
network topology. To achieve this, each node periodically broadcasts link-
R A V I state
S H Ainformation
R M A ( K Csuch
S - 7 as
1 1link
) activity and delay
M of
o bitsi loutgoing
e C o m links
p u t to
i nall
g
other nodes using network-wide flooding.
– When a node receives this information, it updates its view of the network
topology and applies a shortest-path algorithm to choose the next hop for
R A V I each
S H Adestination.
RMA (KCS-711) Mobile Computing
– The well-known routing protocol OSPF (open shortest path first) is an
example of a link-state routing protocol. 11
 Pro-active Routing Protocols
RAVI SHARMA (KCS-711) Mobile Computing
• Distance Vector routing (DV)
R A V–I In
SH A Rapproach
this M A ( Keach
C S -node
7 1 1periodically
) monitors M
theo cost
b i l of
e its
C ooutgoing
m p u t ilinks
ng
and sends its routing table information to all neighbours.
– The cost can be measured in terms of the number of hops or time delay or
other metrics.
– Each entry in the routing table contains at least the ID of a destination, the ID of
RAVI SHARMA (KCS-711) Mobile Computing
the next hop neighbour through which the destination can be reached at
minimum cost, and the cost to reach the destination.
– Thus, through periodic monitoring of outgoing links, and dissemination of the
routing table information, each node maintains an estimate of the shortest
RAVI distance
SHARM to Aevery
( K node
C S - in
7 1the
1 )network. Mobile Computing
– DBF (distributed Bellman Ford) and RIP (routing information protocol) are classic
examples of distance vector routing algorithms.

Due
RA V I toS the
HAR limitations
M A ( K C in
S -communication
711) resources
M such
o b i las
e battery
C o m ppower,
uting
the potentially very large number of nodes, network dynamics, and node
mobility, these protocols are not well suited for ad-hoc networks. 12
 Destination-Sequenced Distance-Vector (DSDV)
RAVI SHARMA (KCS-711) Mobile Computing
• Developed by Perkins & Bhagwat, 1994
RAVI SHARMA (KCS-711) Mobile Computing
• Based on the Bellman - Ford algorithm
• It ensures loop-free routing by tagging each route table entry with
a sequence number.
RAVI SHARMA (KCS-711) Mobile Computing
Destination
2
A X
Routing Table at S
RSource
AVI SHARMA (KCS-711) Mobile Computing
5 Destination Next Hop Distance
S B X A 3
7
Y C 4
RAVI SHARMA (KCS-711) Mobile Computing
C 3 Y
Destination 13
 Destination-Sequenced Distance-Vector (DSDV)
RAVI SHARMA (KCS-711) Mobile Computing

RAVI SHARMA (KCS-711) Mobile Computing

RAVI SHARMA (KCS-711) Mobile Computing

RAVI SHARMA (KCS-711) Mobile Computing

RAVI SHARMA (KCS-711) Mobile Computing

Routing Table Update 14

 DSDV Updates
RAVI SHARMA (KCS-711) Mobile Computing
• Each node periodically broadcasts updates to its neighbours.
RAVI SHARMA (KCS-711) Mobile Computing
• Updates are also broadcast to neighbours immediately when significant new
information, such as link breakage, is available.
• In order to reduce potentially large amounts of traffic generated by these
updates, two modes of updates can be employed:
R A V–I The
S H first
A R type
M A is( known
K C S -as
7 1“full
1 ) dump” where multiple
M network
o b i l e protocol
C o m pdata
u t iunits
ng
may be needed to carry all available routing information to the neighbours.
– The other mode of update is referred to as “incremental” where only routing
information changed since the last “full dump” is sent in a single network protocol
R A V I data
S H unit
ARM to the
A (neighbours.
KCS-711) Mobile Computing
• Each Routing Table will contain the following information:
1. All available destinations (Addresses)
2. Next node (to hop) to (move towards) a particular destination
R A V3.I The
S H number
A R M Aof (hops
K C to
S -reach
7 1 1that
) destination Mobile Computing
4. The sequence number assigned by the destination node
15
 DSDV Updates
RAVI SHARMA (KCS-711) Mobile Computing

RAVI SHARMA (KCS-711) Mobile Computing

RAVI SHARMA (KCS-711) Mobile Computing

RAVI SHARMA (KCS-711) Mobile Computing

@ 𝑀𝐻4
RAVI SHARMA (KCS-711) Mobile Computing

16
 DSDV Updates
RAVI SHARMA (KCS-711) Mobile Computing

RAVI SHARMA (KCS-711) Mobile Computing

RAVI SHARMA (KCS-711) Mobile Computing

@ 𝑀𝐻4
RAVI SHARMA (KCS-711) Mobile Computing

RAVI SHARMA (KCS-711) Mobile Computing

17
 Global State Routing (GSR)
RAVI SHARMA (KCS-711) Mobile Computing
• In a Link State (LS) based routing technique:
RAVI SHARMA (KCS-711) Mobile Computing
– A global network topology is maintained in all routers i.e. each router
possesses information about the complete network topology.
– Each router then independently calculates the best next hop from it for
every possible destination in the network using
RAVI SHARMA (KCS-711)
local information of the
Mobile Computing
topology.
– Any link change will be updated by flooding immediately

RAVI SHARMA (KCS-711) Mobile Computing

• In GSR each node maintains

A neighbour list
R A V–I The
SHA R M A ( Klist
neighbour C Sof- a
7 1node
1 ) X contains its neighbours
M o b i l ethat
C oare
m pwithin
u t i nits
g
radio transmission range.
18
 Global State Routing (GSR)
RAVI SHARMA (KCS-711) Mobile Computing
A topology table
RAVI SHARMA (KCS-711) Mobile Computing
– The topology table contains:
• The link-state information of each destination Y as reported by Y and
• A timestamp indicating the time Y has generated this information

RAVI SHARMA (KCS-711) Mobile Computing

A next-hop table
– For each destination Y, the next hop table contains the next hop Z, which is a
one-hop neighbour of X, to which packets must be forwarded from X
R A V I destined
SHARM forA Y. ( K C S - 7 1 1 ) Mobile Computing

A distance table
R A V–I The
S H distance
A R M A table
( K C contains
S - 7 1 1 the
) shortest distanceMtoo each
b i l edestination
C o m p u from
t i n gX
in terms of the number of hops.
19
 Global State Routing (GSR)
RAVI SHARMA (KCS-711) Mobile Computing
2
RAVI SHARMA (KCS-711) A Y M Destination
obile Computing

Source
5
X B
RAVI SHARMA (KCS-711) 7 Mobile Computing

C 3 Z Destination

RAVI SHARMA (KCS-711) Mobile Computing

Neighbour List Topology Table Next Hop Table Distance Table
A B C Link Time Node Next Node Next
Node
State Stamp Y A Y 3
R A V I S H A R M A Y( K C S -….7 1 1 ) 29 Z M oC b i l e C o Zm p u t i4n g
Z …. 7
20
 Global State Routing (GSR): Updates
RAVI SHARMA (KCS-711) Mobile Computing
• Whenever a node receives a routing message containing link-state updates
R Afrom
V I one
S H Aof
RMitsAneighbours,
( K C S - 7 1it1 updates
) M otable
its topology b i l eif the
Com puting
timestamp
is newer than the one stored in the table.

• After the node reconstructs the routing table it broadcasts the

information to its neighbours with other link-state updates.
RAVI SHARMA (KCS-711) Mobile Computing
• The key difference between GSR and traditional link-state algorithms is the
way routing information is disseminated.
– In LS, link state packets are generated and flooded into the network whenever a node
R A V Idetects
S H Atopology
R M A changes.
(KCS-711) Mobile Computing

– GSR doesn’t flood the link state packets. Instead, nodes in GSR maintain the link state
table based on the up to date information received from neighbouring nodes, and
periodically exchange it with their local neighbours only.
RAVI SHARMA (KCS-711) Mobile Computing
– Information is disseminated as the link state with larger sequence numbers replaces
the one with smaller sequence numbers. 21
 Global State Routing (GSR): Updates
RAVI SHARMA (KCS-711) Mobile Computing
– A node in GSR transmits longer packets containing multiple link-state updates to its
RAVI SHARMA (KCS-711) Mobile Computing
neighbours. Therefore GSR requires fewer update messages than a traditional link-
state algorithm in an ad-hoc network with frequent topology changes.

– Thus GSR can optimise MAC layer throughput since frequent smaller packets incur
RAVI SHARMA (KCS-711) Mobile Computing
higher MAC layer overhead than infrequent longer packets.

RAVI SHARMA (KCS-711) Mobile Computing

RAVI SHARMA (KCS-711) Mobile Computing

22
 Fisheye State Routing (FSR)
RAVI SHARMA (KCS-711) Mobile Computing

RAVI SHARMA (KCS-711) Mobile Computing

RAVI SHARMA (KCS-711) Mobile Computing

RAVI SHARMA (KCS-711) Mobile Computing

RAVI SHARMA (KCS-711) Mobile Computing

23
 Fisheye State Routing (FSR)
RAVI SHARMA (KCS-711) Mobile Computing

RAVI SHARMA (KCS-711) Mobile Computing

RAVI SHARMA (KCS-711) Mobile Computing

RAVI SHARMA (KCS-711) Mobile Computing

RAVI SHARMA (KCS-711) Mobile Computing

24
 Fisheye Technique
RAVI SHARMA (KCS-711) Mobile Computing
• FSR uses the "fisheye" technique proposed by Kleinrock and Stevens,
RAVI SHARMA (KCS-711) Mobile Computing
1971 (at the Dept. of Computer Science, UCLA)

• The eye of a fish captures with high detail the pixels near the focal point.
The detail decreases as the distance from the focal point increases.
RAVI SHARMA (KCS-711) Mobile Computing
• In routing, the fisheye approach translates to maintaining accurate
distance and path quality information about the immediate
R Aneighbourhood
V I S H A R M A of( KaCnode,
S - 7 1 with
1) M o detail
progressively less b i l e as
C othe
m pdistance
uting

increases.

• The fisheye scope is defined as the set of nodes that can be reached
RAVI SHARMA (KCS-711) Mobile Computing
within a given number of hops.
25
 Fish Eye Scope
RAVI SHARMA (KCS-711) Mobile Computing

RAVI SHARMA (KCS-711) Mobile Computing

L I
B G
F
RAVI SHARMA (KCS-711) Mobile Computing

C
K
A J
RAVI SHARMA (KCS-711) MHo b i l e C o m p u t i n g
E D

RAVI SHARMA (KCS-711) Mobile Computing

A’s 2-hop Fisheye Scope
26
 Fisheye State Routing (FSR): Protocol Operation
RAVI SHARMA (KCS-711) Mobile Computing
• Each node running FSR is required to maintain following tables:
RAVI SHARMA (KCS-711) Mobile Computing

Neighbour List
– When a node receives a link state message, it records/updates the sender's
R A V Ilink
SHstate
A R information
M A ( K C Sin- 7its1 neighbour
1) list. Mobile Computing

– If the node does not receive link state updates from a neighbour after a
timeout interval, the entry for that neighbour will be deleted from the
R A V Ineighbour
S H A R Mlist.
A (KCS-711) Mobile Computing

– The following information is maintained in the list for each neighbour node:
• Neighbour ID
RAVI SHARMA (KCS-711) Mobile Computing
• Latest time receiving from the neighbour

27
Fisheye State Routing (FSR): Protocol Operation
RAVI SHARMA (KCS-711) Mobile Computing
Topology Table
RAVI SHARMA (KCS-711) Mobile Computing
– Topology table records the topology information obtained from the link state
message.
– Each destination has an entry in the table. The entry contains three parts:
• Destination information
R A V I •S Link
H A state
R M Ainformation
(KCS-711) Mobile Computing
• Variables for selecting effective entries.
– Based on this table, the routing table is calculated.
– The distance information is maintained in the routing table and is used to
R A V Iclassified
S H A R the
M Anode
( K CtoSa- fisheye
7 1 1 ) scope. Mobile Computing
– The topology table has following fields for every link state entry:
• Destination Address
• Destination sequence number
R A V I •S Last
H A heard
R M Atime (KCS-711) Mobile Computing
• A list of its neighbours
• Previous sequence number
28
• A flag for "NeedToSend"
Fisheye State Routing (FSR): Protocol Operation
RAVI SHARMA (KCS-711) Mobile Computing
Routing
R A V I S HTable
ARMA (KCS-711) Mobile Computing

– The routing table of FSR provides the next hop information to

forward the packets for the other destinations in the network.
R A V IThe
S H entries
A R M A are
( K C updated
S-711) when topologyMtable
o b i l is
e changed.
C o m p u t iThe
ng

routing table has the following fields:

• Destination Address
RAVI SHARMA (KCS-711) Mobile Computing
• Next hop address

• Distance

RAVI SHARMA (KCS-711) Mobile Computing

29
Fisheye State Routing (FSR): Protocol Operation
RAVI SHARMA (KCS-711) Mobile Computing
Link State Messages
RAVI SHARMA (KCS-711) Mobile Computing
1. Originating Link State Message
– Each node broadcasts the latest link state information to its neighbours. Link
state messages are not flooded in the neighbours.
R A –V IFSR
S Huses
A R different
M A ( K update
C S - 7 1intervals
1) M o b in
for different entries i l the
e Ctable
o m paccording
uting
to their distance from the node.
• To be precise, entries corresponding to nodes that are nearby (within a predefined scope)
are propagated to the neighbours more frequently than entries of nodes that are far away.
RAVI SHARMA (KCS-711) Mobile Computing

2. Receiving Link State Message

– When a node receives a link state message, it first checks its neighbour link
R A V Istate
S H list
A Rfor
M the
A (sender.
KCS-711) Mobile Computing
• If the sender is a new neighbour, it will insert the sender into the list.

• Otherwise, it will update the timestamp of the sender in the neighbour list. 30
Fisheye State Routing (FSR): Protocol Operation
RAVI SHARMA (KCS-711) Mobile Computing
Routing Table Calculation
RAVI SHARMA (KCS-711) Mobile Computing

• After all incoming Link State Message entries are examined, if there are
any changes in the topology table, routing table is recomputed.

•R AThe
V I topology
S H A R M Atable
( K CisS -checked
7 1 1 ) to remove stale
M o bentries
i l e C prior
o m p utot i the
ng

calculation.

• Based on the latest topology table, the Dijkstra's algorithm is performed

RAVI SHARMA (KCS-711) Mobile Computing
to find the shortest paths from current node to all the destinations known
to it, i.e., in the topology table.

•R AOld
V I routing
S H A Rtable
M A is
( Kreplaced
C S - 7 1 1completely
) by the newly
M o b computed
i l e C o m next
p u t ihop
ng

information.
31
 Fisheye State Routing (FSR)
RAVI SHARMA (KCS-711) Mobile Computing
• FSR is a table-driven routing protocol which is adapted to the wireless ad
R Ahoc
V I environment.
SHARMA (KCS-711) Mobile Computing

• Through updating link state information with different frequencies

(depending on the fisheye scope distance), FSR scales well to large
network size and keeps overhead low (without compromising route
R Acomputation
V I S H A R Maccuracy
A ( K C when
S - 7 1 the
1 ) destination is near).
Mobile Computing

• It provides an implicit hierarchical routing structure.

•R AAlthough, as FSR retains a routing entry for eachMdestination,

VI SHARMA (KCS-711)
it avoids the
obile Computing
extra work of "finding" the destination (as in on-demand routing) and thus
maintains low single packet transmission latency.

• As the mobility increases, routes to remote destinations become less

R Aaccurate.
V I S H A RHowever,
M A ( K C with
S - 7 1FSR
1 ) in play, when M
a opacket
b i l e approaches
C o m p u t i nits
g
destination, it finds increasingly accurate routing instructions as it enters
sectors with a higher refresh rate. 32
 Fisheye State Routing (FSR)
RAVI SHARMA (KCS-711) Mobile Computing
• FSR is more desirable for large mobile networks where:
R A –V IMobility
S H A RisMhigh
A (KCS-711) Mobile Computing
• A change on a link far away from the source does not necessarily cause a change in the
routing table at the source.
– Bandwidth is low
• Scope update scheme has larger advantages in reducing routing update packet size and
R A V I S achieve
HARM high
A date
( K packet
C S - 7to1routing
1 ) packet ratio. Moreover, M othe
b ifact
l e that
C othemroute
p u terror
i n gis
weighted by distance obviously reduces the sensitivity to network size.

• By choosing proper number of scope levels and radius size, FSR proves to
be a flexible solution to the challenge of maintaining accurate routes in ad
R Ahoc
V I networks.
SHARMA (KCS-711) Mobile Computing

• Advantages of FSR:
– Usage of up-to-date shortest routes
R A –V IRobustness
S H A R M to
A host
( K Cmobility
S-711) Mobile Computing
– Exchange Partial Routing Update with neighbours
– Reduced Routing Update Traffic 33
 FSR: progressive accuracy
RAVI SHARMA (KCS-711) Mobile Computing

RAVI SHARMA (KCS-711) Mobile Computing

RAVI SHARMA (KCS-711) Mobile Computing

RAVI SHARMA (KCS-711) Mobile Computing

S D

R A V I S H A RImprecise
M A ( Kknowledge
C S - 7 1 1of) the best path to a distant
M o destination
bile Computing
is compensated by the fact that
the route becomes progressively more accurate as the packet gets closer to destination
34
 Reactive Routing Protocols
RAVI SHARMA (KCS-711) Mobile Computing
• Unlike pro-active routing protocols, reactive routing protocols find and
R A maintain
V I S H A Rroutes
M A (when
K C S needed
- 7 1 1 ) (on-demand) Mobile Computing
– This reduces the routing overheads for the networks where the rate of topology
change is very high.
• Route discovery usually involves flooding route request packets through
R A the
V I network.
SHARMA (KCS-711) Mobile Computing
– When a node that is a destination or has a route to the destination is reached, a
route reply is sent back to the source of the request.
• Reactive routing protocols can be classified into two groups based on the
R A way
V I Srouting
H A R M information
A ( K C S - 7 1is 1stored
) at each node
M oand
b i l carried
e C o minp urouting
ting
packets:
– source routing
• Each data packet contains a list of node addresses known as the source route that constitutes the
complete path from the source to the destination.
R A V–I hop-by-hop
S H A R M routing
A (KCS-711) Mobile Computing
• Each data packet carries only the destination address and the next hop address, and each intermediate
node in the routing path uses its routing table to forward the data packet to the next hop towards the
destination. 35
 Dynamic Source Routing (DSR) RFC: 4728
RAVI SHARMA (KCS-711) Mobile Computing
• DSR (Johnson & Maltz, 1996) is based on the concept of source routing.
R• A Each
V I S node
H A R in
M DSR
A ( KisCrequired
S - 7 1 1 )to maintain a route
M cache
o b i l ethat
C ocontains
m p u t i the
ng
source routes to the destinations the node has learned recently.
• An entry in the route cache is deleted when it reaches its timeout.
• Route Discovery
RAVI SHARMA (KCS-711) Mobile Computing
– When a source node needs to send a data packet to a destination node, it searches
its route cache to determine if it already has a route to the destination.
– If there is a route to the destination, it uses the route to send the data packet.
R A V I Otherwise
S H A R MitA initiates
( K C S a- 7route
1 1 )discovery process by Mbroadcasting
o b i l e C ao route
m p urequest
ting
(RREQ) packet to its neighbours.
– The RREQ contains the address of the source, the address of the destination, a
request id, and a route record.
RAVI SHARMA (KCS-711) Mobile Computing
• The request id is a sequence number maintained locally by the source node.
• The route record is the addresses of the intermediate nodes through which the RREQ will pass to reach
the destination. At the source the route record does not contain anything.
36
 Dynamic Source Routing (DSR) RFC: 4728
RAVI SHARMA (KCS-711) Mobile Computing
– When a node receives a copy of the RREQ, it checks the <source address,
RAVI SHARMA (KCS-711) Mobile Computing
request id> pair in its list of recently seen route requests.
• If there is a match or the route record contains the address of the node, the RREQ is dropped.

• Otherwise the node checks whether it is the destination or contains a route to the destination.
R–AIfV the
I Snode
H A RisM A ( K aC destination
neither S - 7 1 1 ) nor has a route toMthe
o bdestination:
ile Computing
• It appends its address to the route record and rebroadcasts the RREQ to its neighbours.
• A copy of the RREQ thus propagates through the network until it reaches the destination or a
node that has a route to the destination.
R–AOtherwise
V I S H A Ra M A (reply
route K C S (RREP)
- 7 1 1 )is generated when the
M oRREQ
b i l ereaches
C o m either
p u t i nthe
g

destination or an intermediate node that contains a route to the destination.

• If the node generating the RREP is the destination itself, it copies the route record from the
RREQ to the RREP.
R A V• I If Sthe
H responding
A R M A node
( K CisSan- 7 11)
intermediate M cached
node, it appends its o b i l route
e C toom
thepincomplete
uting
route record and puts the complete route record in the RREP.

37
 Ad hoc On-Demand Distance Vector Routing RFC: 3561

RAVI SHARMA (KCS-711) Mobile Computing

• Developed by C. Perkins, E. Belding-Royer and S. Das (2003)
RAVI SHARMA (KCS-711) Mobile Computing
• Joint Effort of Nokia Research Center, University of California, Santa
Barbara and University of Cincinnati

RAVI SHARMA (KCS-711) Mobile Computing

Operation:

• AODV requires each node to maintain a list of its active neighbours by

R A sending
V I S H Aperiodic
R M A ( HELLO
K C S - 7packets
11) or by listeningMtoo data
b i l e transmissions
C o m p u t i nof
g

neighbouring nodes in promiscuous mode.

• When a source node needs to send a data packet to a destination node

RAVI SHARMA (KCS-711) Mobile Computing
and does not contain any route to the destination, it initiates a path
discovery process to find a route to the destination. 38
 Ad hoc On-Demand Distance Vector Routing RFC: 3561

RAVI SHARMA (KCS-711) Mobile Computing

• Every node maintains two monotonically increasing counters: a sequence
RAVI SHARMA (KCS-711) Mobile Computing
number and a broadcast id.

• The source broadcasts a route request (RREQ) packet to its neighbours

R A containing
V I S H A R the
M A address
( K C S - of
7 1the
itself, its sequence
1) M onumber,
b i l e C its
o mbroadcast
puting
id, the address of the destination, its last known sequence number of
the destination, and a hop-count with a value of zero.
R A V–I The
S Hpair
A R <address
M A ( KofCthe
S - source,
7 1 1 ) broadcast id> uniquely
M identifies
o b i l e a CRREQ.
omputing

– The destination sequence number is used to determine the relative freshness of two
pieces of routing information generated by two nodes for the same destination.

R A V I S• H ItAmeans
R Mthe
A packet
( K Cwith
S -the
7 highest
1 1 ) destination sequence number
M ois more
b i l recent.
e Computing
– The broadcast id is incremented by the source every time it broadcasts a RREQ.
39
 Ad hoc On-Demand Distance Vector Routing RFC: 3561

RAVI SHARMA (KCS-711) Mobile Computing

R• A When
V I S Ha
ARnode
M A receives
( KC S - 7a
1 1RREQ,
) it checks if Mit ohas
b i lreceived
e C o m pau RREQ
ting
with the same <address of the source, broadcast id> pair before.
– If there is a match, the node drops the RREQ and thereby limits the number

R A V I of
SHbroadcast
A R M A packets.
(KCS-711) Mobile Computing
– Otherwise it accepts the RREQ for further processing.
• If the node has a route to the destination with a destination sequence number

R A V I S H greater
A R M than
A ( the
K C destination
S - 7 1 1 ) sequence number ofMthe
o breceived
i l e CRREQ,
o m piturespond
ting
to the RREQ by sending a route reply (RREP) packet to the source.

• Otherwise it rebroadcasts the RREQ to its neighbours after increasing the value
of the hop count. This process repeats and eventually a RREQ is assumed to
RAVI SHARMA (KCS-711) Mobile Computing
arrive at a node that is either the destination itself or has a current route to the
destination 40
Temporally Ordered Routing Algorithm (TORA)
RAVI SHARMA (KCS-711) Mobile Computing
• TORA (Park & Corson, 1997) uses a Directed Acyclic Graph (DAG), rooted at a
RAVI SHARMA (KCS-711) Mobile Computing
destination, to represent multiple routes for a source and destination pair.
• TORA assigns directions to the links between routers to form a routing structure
that is used to forward datagrams to the destination.
• A router assigns a direction ("upstream" or "downstream") to the link with a
RAVI SHARMA (KCS-711) Mobile Computing
neighbouring router based on the relative values of a metric associated with each
router.
• The metric maintained by a router can conceptually be thought of as the router's
R A "height"
V I S H A(i.e.,
R Mlinks
A are
( K Cdirected
S - 7 1 from
1 ) the higher router to
M the
o b lower
ile C omputing
router).
– The significance of the heights and the link directional assignments is that a router may only
forward datagrams downstream.
– Links from a router to any neighbouring routers with an unknown or undefined height are
R A V I considered
S H A R Mundirected
A ( K Cand
S - cannot
7 1 1 )be used for forwarding. M o b i l e C o m p u t i n g
– Collectively, the heights of the routers and the link directional assignments form a loop-free,
multipath routing structure in which all directed paths lead downstream to the destination.
41
Temporally Ordered Routing Algorithm (TORA)
RAVI SHARMA (KCS-711) Mobile Computing

RAVI SHARMA (KCS-711) Mobile Computing

H=1

RAVI SHARMA (KCS-711) Mobile Computing

H=? H=2 H=1

RAVI SHARMA (KCS-711) Mobile Computing

H=?
RAVI SHARMA (KCS-711) Mobile Computing

42
 TORA: Protocol Operation
RAVI SHARMA (KCS-711) Mobile Computing
• TORA can be separated into four basic functions:
RAVI SHARMA (KCS-711) Mobile Computing

Creating routes (Query)

– Creating routes corresponds to the selection of heights to form a directed
sequence of links leading to the destination in a previously undirected
R A V I network
S H A R or
M portion
A ( K C of
S -the
7 1network.
1) Mobile Computing

Maintaining routes
– Maintaining routes refers to the adapting the routing structure in response to
RAVI SHARMA (KCS-711) Mobile Computing
network topological changes.
– For example:
• Following the loss of some router's last downstream link, some directed paths may
temporarily no longer lead to the destination.
R A V I S• HThis
A R event
M A triggers
( K C S - a7 sequence
11) of directed linkMreversals
o b i l e (caused
C o m pbyu tthe
i n re-
g
selection of router heights), which re-orients the routing structure such that all
directed paths again lead to the destination.
43
 TORA: Protocol Operation
RAVI SHARMA (KCS-711) Mobile Computing
Erasing routes
R A V– I InS cases
H A R where
M A ( the
K C network
S - 7 1 1 )becomes partitioned,
M links
o b i line the
C oportion
m p u tof
i nthe
g
network that has become partitioned from the destination must be marked as
undirected to erase invalid routes.
– During this erasing routes process, routers set their heights to null and their
adjacent links become undirected.
RAVI SHARMA (KCS-711) Mobile Computing
Optimizing routes
– TORA includes a secondary mechanism for optimizing routes, in which routers
re-select their heights in order to improve the routing structure.
RAVI SHARMA (KCS-711) Mobile Computing
TORA accomplishes these four functions through the use of four
distinct control packets:
• query (QRY)
R A •V I update
S H A R(UPD)
MA (KCS-711) Mobile Computing
• clear (CLR)
• optimization (OPT) 44
 TORA: Protocol Operation
RAVI SHARMA (KCS-711) Mobile Computing

RAVI SHARMA (KCS-711) Mobile Computing

RT_RQ
flag set

QRY QRY QRY

RAVI SHARMA (KCS-711) Mobile Computing
S D
𝒌-hops
𝑯𝒆𝒊𝒈𝒉𝒕 = 𝟎
UPD UPD UPD
RAVI SHARMA (KCS-711) Mobile Computing

𝑯𝒆𝒊𝒈𝒉𝒕 = 𝒌 + 𝚫 𝑯𝒆𝒊𝒈𝒉𝒕 = 𝒌

RAVI SHARMA (KCS-711) Mobile Computing

45