Online Poweraware Routing in Wireless Adhoc Networks

1
Online Power-aware Routing in Wireless Ad-hoc
Networks
(Qun Li, Javed Aslam, Daniela Rus)

• Chong Jo Woon
• 2002. 11. 13

2
Contents

• Introduction
• Formulation
• Max-min algorithm
• Zone-based routing
• Simulation results
• Conclusion

3
Introduction

• Large ad-hoc networks
• The proliferation of low power analog and digital
  electronics
• Large ad-hoc network like a sensor network
  distributed over a large geographical area
• Power aware approach to routing messages
• Power consumption
• Minimizing power consumption during the idle mode
• Minimizing power consumption during the
  communication mode
• Metrics for power-aware routing
• Minimizing energy consumed
• Minimizing variance in each computer power level
• Minimizing ratio of cost/packet
• Minimizing maximum node cost
• Maximizing lifetime of network
• And so on

4
Introduction

• Approach
• Assumption
• Power limited
• Each message is important.
• Network is sparsely deployed.
• System is fast, scalable, and online.
• Network has a high degree of redundancy.
• Focus on
• Minimizing power consumption during communication
• Maximizing lifetime of network. the number of
  message sent until network partitioning
• Formulation power-aware routing
• Propose algorithms
• Simulation and analysis

5
Formulation The Model

• The power consumption at each node
• an aggregate between transit and receive
• Weighted graph G(V, E)
• Vertices computers in the network
• weight of vertex - the computers power level
• Edges pairs of computers that are in
  communication range
• weight of edge the power cost of sending a unit
  message between the two nodes
• - a host needs power e to transmit a
  message to another host who is d distance away
• k and c are constants for specific wireless
  system
• Lifetime of network
• The earliest time when a message cannot be sent
  due to saturated nodes

6
Formulation The Model

• s source
• t sink
• nij total number of messages from node vi to
  node vj
• eij power cost to send a message from node vi
  to node vj

7
Formulation - Competitive ratio

• Competitive ratio to the optimal off-line
  algorithm
• lifetime of network or number of messages sent

Network
Off-line
On-line

• the competitive ratio is small when n is large

• Theorem 1. No online algorithm for message
  routing has a constant competitive ratio in terms
  of the lifetime of the network or the number of
  messages sent

8
Max-min algorithm

• Online algorithm
• Message routing algorithm that do not assume
  prior knowledge of the message sequence
• In ad-hoc network, this sequence can be dynamic.
• Message routes should avoid nodes whose power is
  low.
• Overuse of those nodes will deplete their battery
  power.
• Is max-min path algorithm a solution?
• No, it can be bad.
• Algorithm that optimizes
• Minimizing power consumption
• Maximizing the minimal residual power in the
  network
• Relax the minimal power consumption condition !!
• -

9
Max-min algorithm

• Power of each node except for S 20
• Power of S infinite
• The weight of each edge on the arc is 1
• Max-min path
• Can transmit 20 messages
• The weight of each straight edge is 2
• Can transmit 10(n-4) messages !!!

10
Max-min algorithm

• utij (Pt(vi) eij) / P(vi) residual power
  fraction after sending a message from i to j
• P(vi) the initial power level of node vi
• eij the weight of the edge vivj
• Pt(vi) power of the node vi at time t

11
Max-min algorithm

• Adaptive Computation for z

• is estimation for the lifetime of that
  node

12
Max-min algorithm

• Theorem 2. The lifetime of algorithm M satisfies

• T_M the lifetime by max_min
• T_O the lifetime by optimal off-line (fixed)
• P_kO the remaining power at T_O (fixed)
• P_kM the remaining power at T_M (not fixed)
• P_0mk the minimal power consumption to send a
  message mk (fixed)
• time slot interval with cyclicity (fixed)

• provides a lower bound for the lifetime of the
  max-min zPmin algorithm as compared to the
  optimal solution
• shows tradeoff between z and

13
Max-min algorithm
14
Zone-Based Routing

• To implement max-min zP_min algorithm on large
  scale networks is hard
• Requires accurate power level information for all
  nodes in the network
• Zone-based routing
• Hierarchical approach
• Group all the nodes that are in geographic
  proximity as a zone !!
• Zone power estimation
• Global path selection
• Local path selection

15
Zone-Based Routing

• Zone Power Estimation
• neighboring zone to neighboring zone
• Estimated relative to the direction of message
  transmission
• Protocol
• Controlled by a node in the zone
• Polling each node for its power level followed by
  running the max-min algorithm
• Returned value is then broadcasted to all the
  zones in the system
• The frequency of this procedure is inversely
  proportional to the estimated power level
• By simulation

16
Zone-Based Routing

• Global Path Selection
• Use modified Bellman-Ford algorithm

17
Zone-Based Routing

• Local Path Selection
• Max-min zPmin algorithm is used directly to route
  a message within a zone

18
Simulation results(1)

• Simulation Environment
• Network scope 10 10
• Randomly generated position
• The number of hosts 20
• Initial power of each host 30
• The weigts generated 0.001 dij3

19
Simulation results(2)

• The effect of z on the maximal number of messages
  in a square network space.

• In some case, z leads to superior performance
• over the minimal power algorithm (z1) and the
  max-min
• algorithm (zinf)

• Adaptively selecting z with z_init 10 can
  increase 12,207

20
Simulation results(3)

• Compares the performance of max-min to
  the optimal solution

• Max-min zPmin performs better than 80 of
  optimal for 92 of the experiments and performs
  within more than 90 of the optimal for 53 of
  the experiments.

• Max-min zPmin has a good empirical competitive
  ratio

21
Simulation results(4)

• Zone-based routing algorithm performs better
  than 94.5 of max-min algorithm

• Simulation Environment
• Network scope 10 10
• Randomly generated position
• The number of hosts 20
• Initial power of each host 30
• The weigts generated 0.001 dij3

22
Conclusion

• Optimizing the performance of communication
  algorithms for power consumption and for the
  lifetime of the network is a very important
  problem.
• Max-min zPmin algorithm had a good empirical
  competitive ratio to the optimal off-line
  algorithm
• Max-min zPmin is practical for applications where
  there is no knowledge of the message sequence
• In large networks, Zone-based power-aware routing
  can be used

23
Proof of Theorem 2
24
Proof of Theorem 3
25
(No Transcript)