From: Yuhan Cai (yuhancai@cs.washington.edu)
Date: Sun Nov 21 2004 - 22:39:06 PST
Title: GPSR: Greedy Perimeter Stateless Routing for Wireless Networks
Authors: Brad Karp and H. T. Kung
Reviewed by: Yuhan Cai
Main results of the paper:
， A routing protocol called Greedy Perimeter Stateless Routing (GPSR) for packet forwarding in wireless datagram networks is presented.
， The superiority of the approach is demonstrated by extensive simulation of mobile wireless networks.
Strengths of the paper:
， The protocol uses only information from a router¨s immediate neighbors in the networks to make greedy forwarding decisions, and therefore is efficient and cost-effective.
， The protocol is able to recover when greedy choices are not available.
， The protocol is more scalable with respect to the number of network destinations and mobility rate, compared with both shortest-path and ad-hoc protocols.
， The protocol is stable and converges quickly and therefore is suitable for mobile environment where there are frequent topology changes.
， Geographic routing allows routers to be almost stateless and it possesses the key feature of self-describing.
Key limitations:
， Only 2-dimensional surfaces are dealt with in this paper, while perimeter forwarding can be in 3-dimensional space.
， The performance is compared with only Dynamic Source Routing. There may be other routing techniques that are worth exploring.
Relevance of the paper:
， Existing routing algorithms such as DV and LS algorithms have a inherent problem with scalability with respect to the network diameter and mobility. GPSR provides a powerful way to achieve more scalable networks.
Future work:
， Future research work is expected on the combination of GPSR and a distributed database system.
， It might be useful to compare the uses of RNG and GG planarizations in GPSR.
， The assumption that a node can reach all other nodes within its radio range could be relaxed by a more efficient way.
This archive was generated by hypermail 2.1.6 : Sun Nov 21 2004 - 22:39:12 PST