Sensor Network Routing
Romit Roy Choudhury
and
Pradeep Kyasanur
(Some slides are based on Dr. Nitin Vaidyas tutorial)
A Review of Current Routing
Protocols for Ad Hoc Mobile
Wireless Networks
Elizabeth M Royer, Chai-Keong Toh
Mobile Ad Hoc Wireless Networks
Unreliable wireless medium
Mobile nodes
No central authority
Traffic patterns application specific
Energy constraints
More information in MANET Charter
Example Ad Hoc Network
S E
F
B
C
D
A G
H
Nodes have unique identifiers
Routing problem find path between S and D
Classification of routing protocols
Table-driven (proactive)
Up-to-date routing information maintained
Routing overhead independent of route usage
Source-initiated (demand-driven / reactive)
Routes maintained only for routes in use
Explicit route discovery mechanism
Hybrid Protocols
Combination of proactive and reactive
Classification (cont.)
Ad Hoc Routing Protocols
Proactive Hybrid Reactive
Table driven Hybrid Source-initiated
on-demand
DSDV OLSR WRP ZRP
CGSR AODV DSR TORA ABR
SSR
Table-driven Routing Protocols
Each node maintains a routing table
Contains routes to all nodes in the network
Changes to network topology is immediately
propagated
Protocols differ in mechanisms used to
propagate topology information
Destination Sequenced Distance
Vector (DSDV)
Based on Bellman-Ford algorithm
Enhanced with sequence number to avoid
loops
Fresher routes have higher sequence numbers
Optimizations added to reduce routing
overheads incremental data exchange,
delayed exchange of updates
DSDV Example
Destination Next Metric Seq. Nr
Routing Table of A A 0 A-550
Node A B B 1 B-102
C B 2 C-588
D B 3 D-312
A
B C D
Route information is exchanged periodically
Clusterhead Gateway Switch
Routing (CGSR)
Nodes organized into hierarchy of clusters.
Each node has a clusterhead, selected
using an election.
Nodes send packet through clusterheads.
Clusterheads communicate amongst
themselves using DSDV.
Two clusters are connected through a
gateway node
Wireless Routing Protocol (WRP)
Maintains multiple tables
Distance, routing, link-cost, etc.
Link change messages exchanged only
between neighbors
Loop freedom using novel algorithm
Uses predecessor hop information
Other Table-Driven Protocols
Optimized Link State Routing Protocol
(OLSR) RFC 3626
Optimization of link-state routing to wireless
Topology Dissemination Based on Reverse
Path Forwarding (TBRPF) - RFC 3684
Also based on link-state routing
Source-Initiated On-Demand
Routing
Create routes only when needed
Routes found using a route discovery
process
Route maintenance procedure used to
repair routes
Ad Hoc On-Demand Distance
Vector Routing (AODV)
Now RFC 3561, based on DSDV
Destination sequence numbers provide
loop freedom
Source sends Route Request Packet
(RREQ) when a route has to be found
Route Reply Packet (RREP) is sent back
by destination
Route Error messages update routes
Route Requests in AODV
S E
F
B
C
D
A G
H
Represents a node that has received RREQ for D from S
Route Requests in AODV
Broadcast transmission
S E
F
B
C
D
A G
H
Represents transmission of RREQ
Route Requests in AODV
S E
F
B
C
D
A G
H
Represents links on Reverse Path
Reverse Path Setup in AODV
S E
F
B
C
D
A G
H
Node C receives RREQ from G and H, but does not forward
it again, because node C has already forwarded RREQ once
Route Reply in AODV
S E
F
B
C
D
A G
H
Represents links on path taken by RREP
Dynamic Source Routing (DSR)
Similar to AODV in route discovery
Full source-route is aggregated in RREQ,
and sent back in RREP
Each data packet has full source route
Route table overhead only at source node
However, overhead with each data packet
Route Requests in DSR
S E
F
B
C
D
A G
H
Represents a node that has received RREQ for D from S
Route Requests in DSR
Broadcast transmission
S E
F
B
C
D
A G
H
Represents transmission of RREQ
Route Requests in DSR
S E
F
B
C
D
A G
H
RREQ keeps a list of nodes on the path from the source
Route Reply in DSR
S E
F
B
C
D
A G
H
Represents links on path taken by RREP
Associativity-Based Routing
Defines metric Degree of Association
Stability
This metric used instead of shortest hop
Nodes with less mobility/better links have
higher stability value
DSR-like protocol is used for routing
Signal Stability Routing
Signal strength of links is used as metric
DSR-like routing is used
RREQ is forwarded only if packet is
received over a link with good signal
strength
Other metrics
Expected Transmission Time (ETT) metric
Easier to compute, and more useful than
signal strength
Weighted Cumulative Expected
Transmission Time
Better for multi-radio, and asymmetric rate
links
Temporally Ordered Routing Algorithm
Directed Acyclic Graph (DAG) rooted at
destination is used to route packets
Link Reversal algorithm used to update
DAG (along with notion of height)
Algorithm is distributed and loop-free
Recent result - Link reversal takes O(n2)
time and message complexity to stabilize
TORA Example
A B F
C E G
DAG maintained to destination D
TORA Example
A B F
C E G
Link (G,D) broke
Node G has no outgoing links
TORA Example
A B F
C E G Represents a
link that was
reversed recently
Now nodes E and F have no outgoing links
TORA Example
A B F
C E G Represents a
link that was
reversed recently
Nodes E and F do not reverse links from node G
Now node B has no outgoing links
TORA Example
A B F
C E G Represents a
link that was
reversed recently
Now node A has no outgoing links
TORA Example
A B F
C E G Represents a
link that was
reversed recently
Now all nodes (except destination D) have outgoing links
TORA Example
A B F
C E G
DAG has been restored with only the destination as a sink
Other routing protocols
Geographic Routing Protocols
Location Aided Routing (LAR)
Distance Routing Effect Algorithm for Mobility
(DREAM)
Greedy Perimeter Stateless Routing (GPSR)
Hybrid Routing Protocols
Zone Routing Protocol (ZRP)
Discussion
Proactive routing protocols suitable for
high traffic load, low mobility
On-demand routing protocols suitable for
low traffic load and/or moderate mobility
With high mobility, flooding of data packets
may be the only option
Locating and Bypassing Routing
Holes in Sensor Networks
Qing Fang, Jie Gao and Leonidas J. Guibas
GPSR
Location of the destination node is
assumed to be known
Each node knows location of its neighbors
Each node forwards a packet to its
neighbor closest to the destination
If routing holes are found, uses perimeter
routing (right-hand rule)
Routing Holes
C E
F
B
J HOLE
D
S A G
I
H
Problem with GPSR Approach
Maintaining perimeter graph expensive,
especially in sensor networks
Identifying holes (and boundary around
holes) useful for routing around them
Also useful for path migration, information
storage
Node where packets get stuck (due to a
hole) define the boundary around holes
Definitions
Weak stuck node P P is the closest node
to node Q (among Ps neighbors), and Q
is out of range of P
Q is called black node
J
P
Q
H
Definitions
Strong stuck node P P is closest node to
point Q, and Q is out of range of P
Collection of Qs is called black region
P
Q
Black Region
H
Proposed Algorithms
TENT rule enables detection of strongly
stuck nodes
P
O
H
Proposed Algorithms
BOUNDHOLE- identifies the boundary of a
hole
Start with a stuck node, and sweep
counter-clockwise
Move from stuck node to stuck node till the
originating node is reached, completing
loop
Discussion
Identifying holes useful for many
applications
Hole identification assumes circular radio
transmission pattern
Can a similar algorithm be designed using
connectivity properties alone?
