manet and vanet routing

2011/2/27

1

MANET and VANET routing

2

Multi-hop Technology: Ad-hoc Networks

All hops are wirelessAll nodes are mobileMilitary Applications

Army Fleet of warships

Space ApplicationsA group of pathfindersA group of satellites

Commercial ApplicationsUsers with hand-held devices

Why Ad Hoc Networks ?Ease of deploymentSpeed of deploymentDecreased dependence on infrastructure

2011/2/27

2

The Challenges forMulti-hop Networks

� Channel Contention : reduces channel efficiency

How do you reduce packet-level contention ?

� Route Computation : message intensive

How do you design scalable routing protocols ?

� QoS Support : applications do not like fluctuations

What mechanisms are needed for QoS support on one-hop and QoS support on the full multi-hop route ?

3

Challenge 1: Channel Contention

� Channel Contention� Reduces efficiency of channel� Needs to be alleviated

� Contention in Single Hop Networks� 802.11 Medium Access (MAC) protocol

� Contention in Multi Hop Networks (proposed extensions of 802.11)� Quick Exchange� Fast Forward

4

2011/2/27

3

Contention in Single-hop Networks: 802.11

RTS

CTS

DATA

ACK

Sender Receiver

Tim

e

RTS: Request to sendCTS: Clear to send

� 802.11 DCF mode (Distributed Coordination Function)� CSMA/CA� RTS/CTS optional

� 802.11 PCF mode (Point Coordination Function)� Access point polls users to transmit� No contention� Very few companies market products for PCF

802.11 DCF

Contention in Multi-hop NetworksSelf-contention(the main type of contention in multi-hop networks)Contention between packets of same transport connection

Inter-stream contentionContention between DATA packet stream and ACK packet stream

Intra-stream contentionContention caused by packets of the same stream at different nodes

source destination

destination

source

DATA stream (TCP or UDP) TCP DATA stream

ACK stream

Contention for shared mediaContention for shared media

Self-contention is best resolved at the MAC layer b ecause…• Self-contention arises in the MAC layer• Requires no changes to widely deployed transport protocols• IEEE 802.11 is an evolving standard and is amenable to changes

prior MAC solutions: (none)prior MAC solutions: [Fu et. al., Infocom ’03]

2011/2/27

4

Routing

� For packet switched network� Goal

� path selection� Forwards the datagrams (packets) from

source to destination� destination address in packet determines

next hop

� routes may change during session

1

23

0111

value in arrivingpacket’s header

routing algorithm

local forwarding tableheader value

output link0100

010101111001

3221

Example of routing and forwarding

2011/2/27

5

Routing Algorithm classificationGlobal or decentralized

information?Global:

� all routers have complete topology, link cost info

� “link state” algorithms

Decentralized:

� router knows physically-connected neighbors, link costs to neighbors

� iterative process of computation, exchange of info with neighbors

� “distance vector” algorithms

Static or dynamic?Static: � routes change slowly

over timeDynamic: � routes change more

quickly� periodic update� in response to link

cost changes

Routing in Mobile Ad Hoc Networks (MANET)

� Formed by wireless hosts which may be mobile

� Without (necessarily) using a pre-existing infrastructure

� Routes between nodes may potentially contain multiple hops

� May need to traverse multiple links to reach a destination

� Mobility causes route changes� Dynamic topologies

2011/2/27

6

Flooding for Data Delivery

S D

P

P

P P

P

P

P

P

P

P

P D

S broadcasts data packet P to all its neighborsEach node receiving P forwards P to its neighbors

Sequence numbers used to avoid the possibilityof forwarding the same packet more than once

Node D does notforward the packet

Flooding for Data Delivery: Advantages

� Simplicity

� May be more efficient than other protocols when rate of information transmission is low enough that the overhead of explicit route discovery/maintenance incurred by other protocols is relatively higher

� this scenario may occur, for instance, when nodes transmit small data packets relatively infrequently, and many topology changes occur between consecutive packet transmissions

� Potentially higher reliability of data delivery� Because packets may be delivered to the destination on

multiple paths

2011/2/27

7

Flooding for Data Delivery: Disadvantages

� Potentially, very high overhead

� Data packets may be delivered to too many nodes who do not need to receive them

� Potentially lower reliability of data delivery� Flooding uses broadcasting -- hard to implement reliable

broadcast delivery without significantly increasing overhead

� Broadcasting in IEEE 802.11 MAC is unreliable

� In our example, nodes J and K may transmit to node D simultaneously, resulting in loss of the packet

� in this case, destination would not receive the packet at all

Why is Routing in MANET different ?

� Host mobility� link failure/repair due to mobility may

have different characteristics than those due to other causes

� Rate of link failure/repair may be high when nodes move fast

� New performance criteria may be used� route stability despite mobility� energy consumption

2011/2/27

8

Challenge 2: Route Computation

� Key challenge in Routing in Multi-hop Wireless Networks� The destination may be anywhere in the network

� Routing Layer Solutions for Single-hop wireless networks� Macro Mobility: Mobile IP� Micro Mobility : Cellular IP

� Routing Layer Solutions for Multi-hop wireless networks� Query the whole network

� Proactive Routing� Reactive Routing

� Dynamic Source Routing (DSR)/Ad-hoc On Demand Distance Vector (AODV)

� Send broadcast only in the right direction (Scoped Broadcast)� Location Aided Routing (LAR)

� Explore only few select paths� Zone Routing Protocol (ZRP)

� Explore only a single path� Greedy Perimeter Stateless Routing (GPSR)

15

Proactive routing (Table-driven)

� Uses periodic route updates to maintain routing tables� Can either be link state or distance vector� Mobility is treated as link change� We introduce :

1.DSDV (Highly Dynamic Destination-Sequenced Distan ce-Vector Routing)

2.GSR (Global State Routing)

C.E. Perkins and P. Bhagwat, “Highly dynamic destina tion sequenced distance-vector routing (DSDV) for mobile computers,” in ACM SIGCOMM ’94, 1994, pp. 234–244.

T.-W. Chen and M. Gerla, “Global State Routing: A Ne w Routing Scheme for Ad-hoc Wireless Networks,” In Proceedings of IEEE ICC’98, A tlanta, GA, Jun. 1998, pp. 171-175.

2011/2/27

9

Table-driven, Proactive

� Link-State

A B

C D

Link-Seq Distance

(B,A) 10

(C,A) 8

A’ link-state

Link-Seq Distance

(A,B) 10

(D,B) 8

B’ link-state

Destination

next

A A

D D

B’ routing table

Destination

next

A A

D D

C’ routing table

Destination

next

C C

B B

D’ routing table

Destination

next

B B

C C

A’ routing tableDestination

next

A A

C A

D D

B’ routing table

Destination

next

A A

B A

D D

C’ routing table

Destination

next

A C

C C

B B

D’ routing table

Destination

next

B B

C C

D B

A’ routing table

10

8 8

7

� Distance-vector

Table-driven, Proactive

A B

C D

Destination

next

B B

C C

A’ update

Destination

next

A A

D D

B’ routing table

Destination

next

A A

D D

C A

B’ update

Destination

next

B B

C C

D’ routing table

Destination

next

A A

D D

C’ routing table

Destination

next

B B

C C

A’ routing table

Destination

next

A B

B B

C C

D’ routing table

10

8 8

7

Destination

next

A A

D D

C A

B’ routing table

Destination

next

A B

B B

C C

D’ updateDestination

next

A A

B D

D D

C’ routing table

Destination

next

A A

B D

D D

C’ update

Destination

next

B B

C C

D C

A’ routing table

2011/2/27

10

DSDV

� Destination Sequence Distance Vector Routing

� IBM 1996, simulated but not implemented� Uses modified Bellman-Ford Algorithm,

distance vector based, table driven� Route settling time and route may not

converge� Sequence number derived from dest. node

is used to keep the routing table up-to-date

DSDV

SD

B

E

C

A F

G

K

H

J

I

periodically sends its own route information to neighbors until stable

Dest next hop seq time

D D 1 D14 100

K H 2 K18 110

I H 2 I20 130

::

::

::

::

::


D J 2 D14 100

K H 2 K18 110

I H 2 I20 130

::

::

::

::

::

G’ routing tableJ’ routing table

2011/2/27

11

DSDV


D B 4 D20 210

K C 4 K22 200

I C 3 I40 150

::

::

::

::

::

SD

B

E

C

A F

G

K

H

J

I


D G 3 D20 210

K C 3 K22 200

I C 3 I40 150

::

::

::

::

::


D D 1 D20 210

H H 1 K22 200

K H 2 I40 150

::

::

::

::

::

Find the path according tothe Routing Table (hop byhop)

S’ routing table

B’ routing tableJ’ routing table

GSR

� Link State Algorithm� Each node periodically sends its own route

information to neighbors� Each node notifies neighbors only when its

route information changes� neighbors then notify their neighbors if

necessary

� Using the Link State Algorithm to create the routing table

2011/2/27

12

GSR

SD

B

E

C

A F

G

K

H

J

I

Link-Seq Distance

(J,D) 5

(K,D) 10

D’ link-state

Link-Seq Distance

(D,J) 5

(G,J) 10

(H,J) 8

J’ link-stateLink-Seq Distance

(B,G) 10

(C,G) 8

(H,G) 6

(J,G) 10

G’ link-state

periodically sends its own route information to neighbors until stable

GSR

SD

B

E

C

A F

G

K

H

J

I

Shortest Path

path order

(S,A) S,A

(S,B) S,B

(S,C) S,C

(S,D) S,B,G,…

(S,E) S,A,E

(S,F) S,C,F

(S,G) S,B,G

(S,H) S,B,G,H

… …

Router-Table

Find the path according tothe Routing Table (source toany hosts)

2011/2/27

13

Proactive routing

� Drawbacks� Inefficient if there is few demands for

routes, and instability if there is high mobility

� Lots of signaling traffic� All of the routes may never be used

Reactive routing (On-demon)

� New nodes are added when needed� No periodic route updates� Find route when needed by the source node� Caching will help to improve the performance� We introduce :

1.DSR (Dynamic Source Routing )

2.AODV (Ad Hoc On-Demand Distance Vector Routing Pr otocol )

JOHNSON, D. B., AND MALTZ, D. B. Dynamic source rou ting in ad hoc wireless networks. In Mobile Computing, T. Imielinski and H. Korth, Eds. Kluwer AcademicPublishers, 1996, ch. 5, pp. 153–181.

Charles E. Perkins and Elizabeth M. Royer. “Ad hoc On -Demand Distance Vector Routing,” In Proceedings of the 2nd IEEE Workshop on Mobile Computing Systems and Applications (WMCSA’99), New Orleans, LA , Febr uary,1999, pp.90-100.

2011/2/27

14

Reactive routing (On-demon)

� Flood the whole network in search of the destination

� Destination or node knowing route to destination responds

� Two phases: � Routing discovery

� Route Request and Route Reply

� Routing maintenance

DSR: Dynamic Source Routing

� CMU 1996, simulated and implemented in 1999

� An extension of IP, it uses options field in IP

� Two phases: routing discovery and routing maintenance

� Caching and other features are used

2011/2/27

15

DSR

� When node S wants to send a packet to node D, but does not know a route to D, node S initiates a route discovery

� Source node S floods Route Request (RREQ)

� Each node appends own identifierwhen forwarding RREQ

Route Discovery in DSR

B

A

S E

F

H

J

D

C

G

IK

Z

Y

Represents a node that has received RREQ for D from S

M

N

L

2011/2/27

16


B

A

S E

F

H

J

D

C

G

IK

Represents transmission of RREQ

Z

YBroadcast transmission

M

N

L

[S]

[X,Y] Represents list of identifiers appended t o RREQ


B

A

S E

F

H

J

D

C

G

IK

• Node H receives packet RREQ from two neighbors:potential for collision

Z

Y

M

N

L

[S,E]

[S,C]

2011/2/27

17


B

A

S E

F

H

J

D

C

G

IK

• Node C receives RREQ from G and H, but does not for wardit again, because node C has already forwarded RREQ once

Z

Y

M

N

L

[S,C,G]

[S,E,F]


B

A

S E

F

H

J

D

C

G

IK

Z

Y

M

• Nodes J and K both broadcast RREQ to node D• Since nodes J and K are hidden from each other, their

transmissions may collide

N

L

[S,C,G,K]

[S,E,F,J]

2011/2/27

18


B

A

S E

F

H

J

D

C

G

IK

Z

Y

• Node D does not forward RREQ, because node Dis the intended target of the route discovery

M

N

L

[S,E,F,J,M]


� Destination D on receiving the first RREQ, sends a Route Reply (RREP)

� RREP is sent on a route obtained by reversing the route appended to received RREQ

� RREP includes the route from S to D on which RREQ was received by node D

2011/2/27

19

Route Reply in DSR� Route Reply can be sent by reversing the route in Route

Request (RREQ) only if links are guaranteed to be bi-directional� To ensure this, RREQ should be forwarded only if it received on a

link that is known to be bi-directional

� If unidirectional (asymmetric) links are allowed, then RREP may need a route discovery for S from node D � Unless node D already knows a route to node S

� If a route discovery is initiated by D for a route to S, then the Route Reply is piggybacked on the Route Request from D.

� If IEEE 802.11 MAC is used to send data, then links have to be bi-directional (since Ack is used)

Route Reply in DSR

B

A

S E

F

H

J

D

C

G

IK

Z

Y

M

N

L

RREP [S,E,F,J,D]

Represents RREP control message

2011/2/27

20

Dynamic Source Routing (DSR)

� Node S on receiving RREP, caches the route included in the RREP

� When node S sends a data packet to D, the entire route is included in the packet header� hence the name source routing

� Intermediate nodes use the source routeincluded in a packet to determine to whom a packet should be forwarded

Data Delivery in DSR

B

A

S E

F

H

J

D

C

G

IK

Z

Y

M

N

L

DATA [S,E,F,J,D]

Packet header size grows with route length

2011/2/27

21

DSR Optimization: Route Caching

� Each node caches a new route it learns by any means� When node S finds route [S,E,F,J,D] to node D, node S

also learns route [S,E,F] to node F� When node K receives Route Request [S,C,G] destined

for node, node K learns route [K,G,C,S] to node S� When node F forwards Route Reply RREP [S,E,F,J,D],

node F learns route [F,J,D] to node D� When node E forwards Data [S,E,F,J,D] it learns route

[E,F,J,D] to node D� A node may also learn a route when it overhears Data

packets

Use of Route Caching� When node S learns that a route to node D is

broken, it uses another route from its local cache, if such a route to D exists in its cache. Otherwise, node S initiates route discovery by sending a route request

� Node X on receiving a Route Request for some node D can send a Route Reply if node X knows a route to node D

� Use of route cache � can speed up route discovery

� can reduce propagation of route requests

2011/2/27

22

Use of Route Caching

B

A

S E

F

H

J

D

C

G

IK

[P,Q,R] Represents cached route at a node(DSR maintains the cached routes in a tree format)

M

N

L

[S,E,F,J,D][E,F,J,D]

[C,S]

[G,C,S]

[F,J,D],[F,E,S]

[J,F,E,S]

Z

Use of Route Caching:Can Speed up Route Discovery

B

A

S E

F

H

J

D

C

G

IK

Z

M

N

L


[C,S]

[G,C,S]

[F,J,D],[F,E,S]

[J,F,E,S]

RREQ

When node Z sends a route requestfor node C, node K sends back a routereply [Z,K,G,C] to node Z using a locallycached route

[K,G,C,S]RREP

2011/2/27

23

Use of Route Caching:Can Reduce Propagation of Route Requests

B

A

S E

F

H

J

D

C

G

IK

Z

Y

M

N

L


[C,S]

[G,C,S]

[F,J,D],[F,E,S]

[J,F,E,S]

RREQ

Assume that there is no link between D and Z.Route Reply (RREP) from node K limits flooding of RREQ.In general, the reduction may be less dramatic.

[K,G,C,S]RREP

Route Error (RERR)

B

A

S E

F

H

J

D

C

G

IK

Z

Y

M

N

L

RERR [J-D]

J sends a route error to S along route J-F-E-S when its attempt to forward the data packet S (with route SEFJD) on J-D fails

Nodes hearing RERR update their route cache to remo ve link J-D

2011/2/27

24

Dynamic Source Routing: Advantages

� Routes maintained only between nodes who need to communicate� reduces overhead of route maintenance

� Route caching can further reduce route discovery overhead

� A single route discovery may yield many routes to the destination, due to intermediate nodes replying from local caches

Dynamic Source Routing: Disadvantages

� Packet header size grows with route length due to source routing

� Flood of route requests may potentially reach all nodes in the network

� Care must be taken to avoid collisions between route requests propagated by neighboring nodes� insertion of random delays before forwarding RREQ

� Increased contention if too many route replies come back due to nodes replying using their local cache� Route Reply Storm problem

� Reply storm may be eased by preventing a node from sending RREP if it hears another RREP with a shorter route

2011/2/27

25

Ad Hoc On-Demand Distance Vector Routing (AODV) [Perkins99Wmcsa]

� DSR includes source routes in packet headers� Resulting large headers can sometimes degrade

performance� particularly when data contents of a packet are small

� AODV attempts to improve on DSR by maintaining routing tables at the nodes, so that data packets do not have to contain routes

� AODV retains the desirable feature of DSR that routes are maintained only between nodes which need to communicate

AODV

� Route Requests (RREQ) are forwarded in a manner similar to DSR

� When a node re-broadcasts a Route Request, it sets up a reverse path pointing towards the source� AODV assumes symmetric (bi-directional) links

� When the intended destination receives a Route Request, it replies by sending a Route Reply

� Route Reply travels along the reverse path set-up when Route Request is forwarded

2011/2/27

26

Route Requests in AODV

B

A

S E

F

H

J

D

C

G

IK

Z

Y

Represents a node that has received RREQ for D from S

M

N

L


B

A

S E

F

H

J

D

C

G

IK

Represents transmission of RREQ

Z

YBroadcast transmission

M

N

L

2011/2/27

27


B

A

S E

F

H

J

D

C

G

IK

Represents links on Reverse Path

Z

Y

M

N

L

Reverse Path Setup in AODV

B

A

S E

F

H

J

D

C

G

IK

• Node C receives RREQ from G and H, but does not for wardit again, because node C has already forwarded RREQ once

Z

Y

M

N

L

2011/2/27

28


B

A

S E

F

H

J

D

C

G

IK

Z

Y

M

N

L


B

A

S E

F

H

J

D

C

G

IK

Z

Y

• Node D does not forward RREQ, because node Dis the intended target of the RREQ

M

N

L

2011/2/27

29

Route Reply in AODV

B

A

S E

F

H

J

D

C

G

IK

Z

Y

Represents links on path taken by RREP

M

N

L

Route Reply in AODV

� An intermediate node (not the destination) may also send a Route Reply (RREP) provided that it knows a more recent path than the one previously known to sender S

� To determine whether the path known to an intermediate node is more recent, destination sequence numbers are used

� The likelihood that an intermediate node will send a Route Reply when using AODV not as high as DSR� A new Route Request by node S for a destination is assigned a

higher destination sequence number. An intermediate node which knows a route, but with a smaller sequence number, cannot sendRoute Reply

2011/2/27

30

Forward Path Setup in AODV

B

A

S E

F

H

J

D

C

G

IK

Z

Y

M

N

L

Forward links are setup when RREP travels alongthe reverse path

Represents a link on the forward path

Data Delivery in AODV

B

A

S E

F

H

J

D

C

G

IK

Z

Y

M

N

L

Routing table entries used to forward data packet.

Route is not included in packet header.

DATA

2011/2/27

31

Timeouts� A routing table entry maintaining a reverse

path is purged after a timeout interval� timeout should be long enough to allow RREP to

come back

� A routing table entry maintaining a forward path is purged if not used for a active_route_timeout interval� if no data is being sent using a particular routing

table entry, that entry will be deleted from the routing table (even if the route may actually still be valid)

Link Failure Reporting

� A neighbor of node X is considered active for a routing table entry if the neighbor sent a packet within active_route_timeout interval which was forwarded using that entry

� When the next hop link in a routing table entry breaks, all active neighbors are informed

� Link failures are propagated by means of Route Error messages, which also update destination sequence numbers

2011/2/27

32

Route Error

� When node X is unable to forward packet P (from node S to node D) on link (X,Y), it generates a RERR message

� Node X increments the destination sequence number for D cached at node X

� The incremented sequence number N is included in the RERR

� When node S receives the RERR, it initiates a new route discovery for D using destination sequence number at least as large as N

Destination Sequence Number

� Continuing from the previous slide …

� When node D receives the route request with destination sequence number N, node D will set its sequence number to N, unless it is already larger than N

2011/2/27

33

Link Failure Detection

� Hello messages: Neighboring nodes periodically exchange hello message

� Absence of hello message is used as an indication of link failure

� Alternatively, failure to receive several MAC-level acknowledgement may be used as an indication of link failure

Why Sequence Numbers in AODV� To avoid using old/broken routes

� To determine which route is newer

� To prevent formation of loops

� Assume that A does not know about failure of link C-D because RERR sent by C is lost

� Now C performs a route discovery for D. Node A receives the RREQ (say, via path C-E-A)

� Node A will reply since A knows a route to D via node B

� Results in a loop (for instance, C-E-A-B-C )

A B C D

E

2011/2/27

34

Summary: AODV

� Routes need not be included in packet headers� Nodes maintain routing tables containing entries

only for routes that are in active use� At most one next-hop per destination maintained at

each node� Multi-path extensions can be designed

� DSR may maintain several routes for a single destination

� Unused routes expire even if topology does not change

Clustering / Hierarchical

� Avoid drastic network-wide changes� Small clusters converge fast� May use hybrid design� Compromised Method� Reduce disadvantages and promote advantages� On theory better but IMHO it’s a bit too complicated

and “smart”, the rule of the real world is “The simpler the better.”

� We introduce :1. ZRP (Zone Routing Protocol)

HAAS, Z., AND PEARLMAN, M. The performance of query control schemes for the zone routing protocol. In Proceedings of the SIGCOMM ’98 Conference on Communications Architectures, Protocols and Applica tions (Sept. 1998).

2011/2/27

35

ZRP: Zone Routing Protocol

� Cornell 1998, simulated only� Zone based and hybrid of proactive and

reactive� Proactive intra-zone and reactive inter-zone� Each node has complete topology up to k hops� For destination outside k hops, query

forwarded to the boundary nodes� Problem: How to decide the appropriate radius

of the zone

� Group

� Border Node� Discovery Process

ZRP

Each node has complete topology up to k hops, and become a group

1. When node S wants to send data to node D,

If D is in the group, send it

Else, send Route Request to Border Node

2. When receives a Route Request

If D is in the group, send Route Reply to source no de

Else, send Route Request to Border Node

The edge node in a group

2011/2/27

36

ZRP

SD

B

E

C

A F

G

K

H

J

I群組的領導群組的領導群組的領導群組的領導。。。。

表示藍色節點所管轄表示藍色節點所管轄表示藍色節點所管轄表示藍色節點所管轄的中間節點的中間節點的中間節點的中間節點。。。。

表示藍色節點所管表示藍色節點所管表示藍色節點所管表示藍色節點所管轄的邊界節點轄的邊界節點轄的邊界節點轄的邊界節點。。。。

表示藍色節點管轄表示藍色節點管轄表示藍色節點管轄表示藍色節點管轄之外的其他節點之外的其他節點之外的其他節點之外的其他節點。。。。

node border

A x

B x

C x

E o

F o

G o

Neighbor Table of “S”

“S” 將將將將RREQ送往自送往自送往自送往自己管理的己管理的己管理的己管理的Border 。。。。

ZRP

SD

B

E

C

A F

G

K

H

J

I

“G” 將將將將RREP由原由原由原由原路徑送回路徑送回路徑送回路徑送回””””S”。。。。

node border

A O

B x

C X

D O

F X

… …

Neighbor Table of “G”

Path order

(S,D) S,G,D

… …

Routing Table of “S”

群組的領導群組的領導群組的領導群組的領導。。。。




2011/2/27

37

ZRP

SD

B

E

C

A F

G

K

H

J

I

Path order

(S,D) S,G,D

… …

Routing Table of “S”





ZRP

SD

B

E

C

A F

G

K

H

J

I

node border

A O

B x

C X

D O

F X

… …

Neighbor Table of “G”





If “D” not the Destination ：：：：

2011/2/27

38

75

VANETs vs. MANETs

� A VANET consists of vehicles to form a network which is similar to a Mobile Ad Hoc Network (MANET). However, there are following differences between these two networks.� Vehicles mobility

� Vehicles move at high speed but mobility is regular and predictable

� Network topology� High speed movement makes network topology dynamic

� No significant power constraint� Recharging batteries from vehicle

� Localization� Vehicles position estimate accurately through GPS systems

or on-board sensors

76

Features of VANETs

� The characteristics of VANETs can be summarized after comparing with the MANETs.� Dynamic topology

� Nomadic nodes with very high speed movement cause frequent topology variation

� Mobility models� Vehicles move along original trajectories completely different

from typical MANET scenarios� Infinite energy supply

� Power constraint can be neglected thanks to always recharging batteries

� Localization functionality� Vehicle can be equipped with accurate positioning systems

(GPS and GALILEO) integrated by electronic maps

2011/2/27

39

77

Preliminaries in Routing Protocols (1/2)� In VANETs, the following characteristics

affect the design of the routing protocols.� High-speed node movement � Frequent topology change� Short connection lifetime especially with multi-

hop paths

� Traditional topological routing protocols for ad hoc networks are not suitable.

78

Preliminaries in Routing Protocols (2/2)

� In the following, the routing protocols in the recent years that address the characteristics of VANETs are presented with three basic categories: � Position-based Routing � Geocasting Routing � Broadcast Routing

� MANET� Table-driven, Proactive� On-demon, Reactive� Clustering / Hierarchical

2011/2/27

40

Scoped Broadcast(Location Aided Routing)

� Obtain the location and speed and direction of motion of the destination using GPS

� Based on the time elapsed since last update, restrict the direction of route query flood

� Issues:� Requires GPS support

79

sourcedest

Request ZoneDetermined by source

Number of nodes

Rou

ting

pack

ets/

data

pac

ket

Flooding

LAR

1000x1000 regionNumber of nodes 15, 30, 50Various tx range

200, 300, 400 and 500

Position-based Routing� The routing decision at each node is based on

the destination’s position contained in the packet and the position of the forwarding node’s neighbors.

� Assumption� Position-based routing does not keep global network

information but requires information on physical locations of nodes.

� Every vehicle has an on board GPS (or some other type of positioning service) for routing.

2011/2/27

41

Exploring a single path (Greedy Perimeter Stateless Routing)

� Each node learns the location of all its neighbors

� Greedy mode forwarding (default): pick the neighbor closest to the destination

� Perimeter mode forwarding: If a node itself is closest to the destination, it then tries to forward the packet around the void

� Issues:� Requires GPS support

81

void

x

D

Node x’s void with respect to destination D

Motivation

� Source routing vs. routing table� Hop-by-hop routing table: small number of entries� Source routing: high bit overhead

� Global protocol � Cannot scale� End-to-end route maintenance� Flooding� Overhead proportional to network size & mobility

2011/2/27

42

10/8/2002 CS537S Sensor Networks 83

GPSR: Answer is location !� Local state: neighbor table

� All I need to know is all my neighbors locations� #State proportional to density instead of

#destinations

� Local protocol� Route discovered and fixed locally� Overhead independent from network size & mobility


Assumptions� All wireless routers know their own positions

� Sources can determine the locations of destinations� local directory service� Bonus: location-based communication

make directory service unnecessary

� To handle void:� Perfect radio model: identical, circular radio

range. � nodes are roughly in a plane.

2011/2/27

43


Location-based Communication

ID-based� From ID to ID� What is the reading of

sensor 125.111.1.5?� Rely on unreliable

individual sensors

Location-based� From location to location� What is the virus density in

south terminal of airport?� Individual sensors NOT

important

� Local coordination :Sensors in interested area aggregate data

� Sensor-base comm .: Send aggregated result to base station


Exchange Location

� Proactive, periodic beaconing� Waste bandwidth in static network� Piggyback location in data packet

� On-demand beaconing� Before send, request for all neighbor

locations if not known� More efficient in static sensor networks

2011/2/27

44


Greedy Forwarding

A C

Closest to C

E

� � “Optimal” path� Efficient: O(1) on sorted neighbor table� The denser a network, the more likely to work


But We Still Need to Solve Void …

� Especially in network that is deployed in a ad hoc fashion

2011/2/27

45


Right-Hand Rule

Go around the perimeter of a void!


Assembling GPSR Together� Maintenance

� All nodes maintain a single-hop neighbor table

� Chop links to construct planar graph

� At source: mode = greedy

� Intermediate node:

Greedy Forwarding Perimeter Forwarding

greedy fails

have left local maximagreedy

worksgreedy fails

2011/2/27

46


a

d

b

c

x

e

f

z

a

x

c

e

z

Example

92

Location-Based Multicast (LBM) (3/4)

S

Source/Sink Forwarding Zone

Geocast Region

� Scheme 1: Box Forwarding Zone� When a node receives

a Geocast packet, it will forward the packet to its neighbors if it is within a forwarding zone; otherwise, it will discard the packet.

2011/2/27

47

93

Location-Based Multicast (LBM) (4/4)

� Scheme 2: CONE Forwarding Zone

� Unlike scheme 1, Scheme 2 does not have a forwarding zone explicitly.

� A Goecast packet that should be forwarded is based on the position of the sender node and receiver node.

H

S

Source/Sink

Multicast Region

Z

XF

A

B

G

D

Y

EC (XC ,YC )

A (XA ,YA )

D (XD ,YD )

B (XB ,YB )

E (XE ,YE )

H (XH ,YH )

Broadcast Routing� In Inter-Vehicle Communication Systems

(IVC)，broadcasting is an efficient method to spread messages.

� The reasons of occurring broadcast storm� In a broadcasting network, the situations of

contentions and collisions often take place if an efficient broadcasting scheme is not used.

� The result incurred by broadcasting is called broadcast storm.

2011/2/27

48

Broadcast Storm� In VANETs, broadcast is used for disseminating

the traffic information：� Detour route� Accident alert� Construction warning� etc…

� Some messages will be periodically broadcasted by roadside unit (RSU) for several hours or even some days.� The problem of broadcast storm in VANET is more

serious than that in MANET

Broadcast Routing� Message Dissemination

� Ideal solution: Minimum Connected Dominating Set, which minimizes packet rtx and preserves network connectivity.

� Realistic solutions: trade-off between robustness and redundancy.

� The important concern in designing a broadcast scheme in VANET.� How to design broadcast algorithm to efficiently transmit

messages to the target nodes.

� To design a broadcast algorithm to make the desired vehicles to receive the message as soon as possible.

2011/2/27

49

Four Broadcasting Strategies

� Different broadcasting strategies to select the forwarding nodes:� Probability-based� Location-based � Neighbor-based� Cluster-based

Broadcast Routing

� 1. Probability-based: � A given PDF determines the decision, for

example depending on the number of copies a node has received.

� The strategy is often dynamic.

� PDF = probability distribution function

2011/2/27

50

Broadcast Routing

� Probability-basedC ar A

P D F = 0 .8

C ar BP D F = 0 .5

F o rw ard in g N o d e ch o o se

Broadcast Routing

� Location-based � The selection criterion is the amount of

additional area that would be covered by enabling a node to forward.

� Some proposal also computes position prediction as useful input information.

2011/2/27

51

Broadcast Routing

� Location-based Target

Forwarding Node choose

Car Bwants to turn right

Car A

Broadcast Routing

� Neighbor-based� A node is selected depending on its

neighbors status (for instance, the status concerns how a neighbor is connected to the network).

2011/2/27

52

Broadcast Routing

� Neighbor-basedTarget


Car B

Car ACollect the information of neighbors

Broadcast Routing

� Cluster-based� Nodes are grouped in clusters

represented by an elected cluster-head. Only cluster-heads forward packets.

� Nodes in the same cluster share some features (e.g., relative speed in VANETs).

� Reclustering on-demand or periodically.

2011/2/27

53

Broadcast Routing

� Cluster-based

Cluster-Header

Cluster-Header

Gateway-Node


106

VADD: Vehicle-Assisted Data Delivery in VehicularAd Hoc Networks

Jing Zhao

In Proceedings of Annual Joint Conference of theIEEE Computer and Communications Societies (INFOCOM ), pp.1-12,2006.

Department of Computer Science & EngineeringThe Pennsylvania State University

Guohong Cao

2011/2/27

54

107

VADD

108

VADD

� 1. Transmit through wireless channels as much as possible.

� 2. If the packet has to be carried through certain roads, the road with higher speed should be chosen.

� 3. Dynamic path selection should continuously be executed throughout the packet forwarding process.

2011/2/27

55

109

VADD

manet and vanet routing

Documents