Architectures and Applications
for Wireless Sensor Networks
(01204525)
Routing
Chaiporn Jaikaeo
chaiporn.j@ku.ac.th
Department of Computer Engineering
Kasetsart University
Materials taken from lecture slides by Karl and Willig
Overview




Unicast routing in MANETs
Energy efficiency & unicast routing
Multi-/broadcast routing
Geographical routing
2
Unicast, ID-Centric Routing

Given: a network/a graph


Goal: Send a packet from one node to
another




Each node has a unique identifier (ID)
The routing & forwarding problem
Routing: Construct a table telling how can
reach a given destination
Forwarding: Consult this table to forward a
given packet to its next hop
Challenges
3
Challenges in WSNs/MANETs

Nodes may move around, neighborhood
relations change
C
B
A

Optimization metrics may be more
complicated

Not just “smallest hop count”
4
Ad hoc Routing Protocols

Because of challenges, standard routing
approaches not really applicable



Too big an overhead, too slow in reacting to
changes
Examples: Dijkstra, Bellman-Ford
Simple solution: Flooding




No routing table needed
Packets are usually delivered to destination
But: overhead is prohibitive
Usually not acceptable in most cases
5
Gossiping

Needs no routing table



Similar to flooding
Nodes forward packets with some
probability
Haas et al. studies gossiping behavior and
found that



There is a critical probability, x
p < x: gossip dies out very quickly
p > x: gossip reaches most nodes
6
Routing Protocol Classification


Main question: When does the routing
protocol operate?
Option 1: Always tries to keep routing data
up-to-date


Option 2: Route is only determined when
actually needed


Protocol is proactive / table-driven
Protocol operates on demand
Option 3: Combine these behaviors

Hybrid protocols
7
Routing Protocol Classification

Which data is used to identify nodes?


An arbitrary identifier?
The position of a node?


Can be used to assist in geographic routing
protocols
Identifiers that are not arbitrary, but carry
some structure?


As in traditional routing
Structure akin to position, on a logical level?
8
Proactive Protocols – Example

Fisheye State Routing (FSR)




Basic observation: When destination is far
away, details about path are not relevant
Look at the graph as if through a fisheye lens
Regions of different accuracy of routing
information
LS information about closer nodes is
exchanged more frequently
9
Reactive Protocols – Example

Recall reactive routing protocols




Initially, no information about next hop is
available at all
One possibility: Send packet to everyone
Hope: At some point, packet will reach
destination and an answer is sent pack – use
this answer for backward learning the route
from destination to source
Examples


Ad hoc On-demand Distance Vector (AODV)
Dynamic Source Routing (DSR)
10
DSR


Dynamic Source Routing protocol
Use separate route request/route reply
packets to discover route



Data packets only sent once route has been
established
Discovery packets smaller than data packets
Store routing information in the discovery
packets
11
DSR Route Discovery
Search for route from 1 to 5
[1]
[1,7]
2
1
[1]
7
5
4
3
6
7
[1,7]
5
4
2
1
3
6
[1,4]
[1,7,2]
7
[1,4,6]
4
6
2
1
2
1
5
3
[1,7,3]
7
5
4
6
3
[5,3,7,1]
Node 5 uses route information recorded in RREQ
to send back, via source routing, a route reply
12
AODV




Ad hoc On-demand Distance Vector
Very popular routing protocol
Same basic idea as DSR for discovery
procedure
Nodes maintain routing tables instead of
source routing
13
Alternative - Rumor Routing




Think of an “agent” wandering through the
network, looking for data/events
Agent initially
perform
random walk
Leave “traces”
in the network
Later agents
?
can use these
traces to find
data
14
Overview




Unicast routing in MANETs
Energy efficiency & unicast routing
Multi-/broadcast routing
Geographical routing
15
Energy-Efficient Unicast: Goals

Minimize energy/bit


Eg., A-B-E-H


A
3
2
1
Maximize
network
"lifetime"

4
1
2
3
B
D
Time until first
node failure
loss of coverage
partitioning
2
1
2
3
C
E
1
2
2
4
4
2 F
G
2
H
Example: Send data from node A to node H
16
Basic options for path metrics

Max total available
battery capacity





Sum of reciprocal
battery levels
Example: A-D-H
A
3
2
1
Largest reciprocal
level of all nodes in
path
1
2
3
D
2
1
2
3
C
B
E
Min-Max batt. cost


Sum of batt. levels
without needless
detours
Example: A-C-F-H
Min battery cost

4
1
2
2
4
4
2 F
G
2
H
Minimize variance in
power levels
17
Overview




Unicast routing in MANETs
Energy efficiency & unicast routing
Multicast/broadcast routing
Geographical routing
18
Broadcast & Multicast


Distribute a packet to all reachable nodes
(broadcast) or to a subgroup (multicast)
Basic options

Source-based tree: one tree per source




Minimize total cost
Minimize maximum cost to each destination
Shared, core-based trees
Mesh

Provides redundancy in data transfer
19
Goals for Source-Based Trees

For each source,
minimize total cost


Steiner tree
Src
2
The Steiner tree problem
2
For each source,
minimize maximum
cost to each destination

Dest 2
Obtained by overlapping
the individual shortest
paths
1
Dest 1
Shortest-path tree
Src
Dest 2
2
2
1
Dest 1
20
Broadcast/Multicast Classification
Broadcast
Multicast
One tree
per source
Minimize
total cost
(Steiner tree)
Minimize
cost to each node
(e.g., Dijkstra)
Shared tree
(core-based tree)
Single
core
Mesh
Multiple
core
21
Wireless Multicast Advantage

Wires



Locally distributing a packet to n neighbors
n times the cost of a unicast packet
Wireless: sending to n neighbors can incur costs
= tx to a single neighbor – if receive costs are ignored
= One tx, n rx – if receives are correctly tuned
= send n unicasts – if multicast not supported by MAC

If local multicast is cheaper, then wireless
multicast advantage is present

Can be assumed realistically
22
Steiner Tree Approximations


Computing Steiner tree is NP complete
A simple approximation


Pick some arbitrary order of all destination
nodes + source node
Successively add these nodes to the tree


For every next node, construct a shortest path to
some other node already on the tree
Performs reasonably well in practice
23
Steiner Tree Approximations

Takahashi Matsuyama heuristic




Similar, but let algorithm decide which is the
next node to be added
Start with source node, add that destination
node to the tree which has shortest path
Iterate, picking that destination node which
has the shortest path to some node already on
the tree
Problem: Wireless multicast advantage not
exploited!

And does not really fit to the Steiner tree
formulation
24
Broadcast Incremental Power


Or BIP
Exploits multicast wireless advantage



Goal: use as little transmission power as
possible
Based on Prim's MST algorithm
Once a node transmits and reaches some
neighbors, it becomes cheaper to reach
additional neighbors
25
BIP – Example
Round 1:
Round 2:
A
5
S
10
D
3
4
3
B
1
B
9
2
7
Round 4:
A
C
2
B
7
7
7
Round 5:
1
D
A
C
3
3
B
S (3)
3
S (3)
1
D
A
2
3
S (1)
1
C
Round 3:
A
B
S (5)
7
10
6
D
7
D
C (1)
C (1)
26
Multicast Incremental Power


Or MIP
Start with broadcast tree construction, then
prune unnecessary edges out of the tree
A
A
3
3
B
S
B
S
10
10
7
D
7
D
C
C
27
Mesh-Based Multicast

Example – ODMRP (On-Demand Multicast
Routing Protocol)
Sender NextHop
Sender NextHop
H
C
A
D
Sender NextHop
H
D
H
C
H
H
E
B
I
F
G
28
Overview




Unicast routing in MANETs
Energy efficiency & unicast routing
Multi-/broadcast routing
Geographical routing


Position-based routing
Geocasting
29
Geographic Routing

Implicitly infer routing information from physical
placement of nodes




E.g., position of current node, current neighbors,
destination known
Send to a neighbor in the right direction as next hop
Geographic routing
Position-based routing
 Use position information to aid in routing

Geocasting
 Send to any node in a given area
30
Overview




Unicast routing in MANETs
Energy efficiency & unicast routing
Multi-/broadcast routing
Geographical routing


Position-based routing
Geocasting
31
Position-Based Routing

“Most forward within range r” strategy


Nearest node with forward progress


Send to that neighbor that
realizes the most forward
progress towards destination
Idea: Minimize transmission power
Directional routing



Choose next hop that is angularly closest to
destination
Choose next hop that is closest to the connecting line
to destination
Problem: Might result in loops!
32
Problem: Dead Ends

Simple strategies might send a packet into
a dead end
33
Right Hand Rule

Basic idea to get out of a dead end: Put
right hand to the wall, follow the wall


Does not work if on some inner wall – will walk
in circles
Need some additional rules to detect such
circles
34
Right Hand Rule – GPSR

Greedy Perimeter Stateless Routing


Use greedy, “most forward” routing as long as
possible
If no progress possible: Switch to “face” routing




Face: largest possible region of the plane that is not
cut by any edge of the graph
Send packet around the face using right-hand rule
Use position where face was entered and
destination position to determine when face can be
left again, switch back to greedy routing
Requires: planar graph! (topology control can
ensure that)
35
Face Routing Example
36
GPSR – Example

Route packet from node A to node Z
Leave face
routing
I
E
B
F
H
K
Z
D
A
Enter
face
routing
J
C
L
G
37
GEM – Graph EMbedding




Geographical routing without positions
Construct virtual polar coordinates for nodes
Assign “virtual
angle range” to
neighbors of a
node, bigger
radius
Angles are
recursively
redistributed to
children nodes
38
Overview




Unicast routing in MANETs
Energy efficiency & unicast routing
Multi-/broadcast routing
Geographical routing


Position-based routing
Geocasting
39
Location-based Multicast (LBM)



Geocasting by geographically restricted
flooding
Define a “forwarding” zone – nodes in this
zone will forward the packet to make it
reach the destination zone
Packet is always forwarded by nodes within
the destination zone itself
40
Determining Next Hops

Use Voronoi diagram
B
Destination
Zone
C
S
D
A
41
Determining Next Hops

Use convex hulls
Destination
Zone
C
B
D
S
A
E
42
Trajectory-Based Forwarding (TBF)

Source routing + geocasting
43
Conclusion



Routing exploit various sources of
information to find destination of a packet
Routing can make some difference for
network lifetime
Non-standard routing tasks (multicasting,
geocasting) require adapted protocols
44