A Novel Mechanism for Flooding Based Route Discovery in Ad Hoc Networks Jian Li and Prasant Mohapatra GlobeCom’03 Speaker ︰ CHUN-WEI

Outline  Introduction  PANDA Design  PANDA-LO  PANDA-LV  PANDA-TP  Simulation  Conclusion

Introduction  Flooding based route discovery  AODV,DSR…etc  Broadcast storm problem  In [11]”A Performance Comparison of Multi- hop Wireless Ad-hoc Networking Routing Protocol“ use Random Rebroadcast Delay (RRD) approach to solve the problem

Introduction  RRD approach is not the most suitable one in term of search for a better route  RRD approach may choice the worst next - hop candidate S I J K L M D

Introduction  We propose to use location and velocity information in determining the rebroadcast delay time  Positional Attribute based Next-hop Determination Approach (PANDA)

PANDA Design  Basic idea  Discriminate neighboring node as good or bad candidate for the next hop  Good candidate  Shorter rebroadcast delay  Bad candidate  Longer rebroadcast delay  Algorithm  PANDA-LO (Location Only)  PANDA-LV (Location & Velocity)  PANDA-TP (Transmission Power)

PANDA Design  Assume  Each mobile node is equipped GPS  Location and velocity information

PANDA Design – PANDA-LO  The farther away node form the upstream node, the shorter rebroadcast delay

PANDA Design – PANDA-LO Random value between 0 to t 1

PANDA Design – PANDA-LO  PANDA-LO may lead to fragile path  Not consider the link lifetime

PANDA Design – PANDA-LV  Use both location and velocity information  Estimating the link lifetime  Choosing stable links as the next hop

PANDA Design – PANDA-LV Θ

PANDA Design – PANDA-TP  In wireless sensor network, power conservation is more important than reduction of end-to-end delay  Break a big single hop into several small hops  Demonstrated the following example  Small hops power consumption is smaller than a big single hop

PANDA Design – PANDA-TP Assume ︰ P RXmin is the minimal receive power L is propagation loss

PANDA Design – PANDA-TP (α is between 2 and 4)

PANDA Design – PANDA-TP

Simulation  PANDA-LO and PANDA-LV  Use the codebase of DSR in ns-2 simulator  Simulation area is 1500×300 square meter  100 nodes uniformly deployed  A node’s speed is uniformly distributed in the range of (0, 20) meter per second  Transmission range is 250 meter

Simulation

 PANDA-TP  Simulation program by ourselves  Simulation area is 1500×300 square meter  Nodes can dynamically control their transmission range  In the route discovery phase, the nodes used a fixed transmission range of 250 meter

Simulation

Conclusion  PANDA approach can improving the performance, quality, and energy conservation of routing algorithm