Wireless Ad Hoc Rouing Without Explicit Control Messages

In this work, we investigate techniques for ad hoc wireless routing without explicit control messages. We develop a new wireless ad hoc routing protocol called Ad hoc Bridging Protocol (ABP). ABP distinguishes itself from earlier proposals in that it uses no explicit control messages for route discovery, setup, and maintenance, while making minimal use of implicit (data-like) control messages that require no special processing at intermediate nodes. ABP is inspired from the self-learning plug-and-play nature of traditional transparent bridges in wired LANs. Like wired bridges and many earlier ad hoc protocols, nodes in ABP learn and adapt routes to different destinations on-the-fly by observing the packets sent by those destinations. As a result, ABP is completely decentralized, self-starting, and automatically adapts to varying network conditions. However, a major difference is that ABP does not attempt to prevent routing loops at all costs. Consequently, ABP does not construct or maintain any overarching control infrastructure for loop prevention, such as extensive network-wide spanning trees in traditional LAN bridges, destination sequence numbers in AODV, or source-routing in DSR. The essential idea is to couple the backward learning technique from transparent bridging, with an associated refresh mechanism, and the use of an identification field to effectively limit the impact of any transient loops. We also prove that transient loops, if and when they occur, do not last longer than a bounded, short time interval and do not negatively impact the network performance. Results demonstrate that, even without explicit control messages, ABP can perform competitively in comparison to AODV and DSR protocols while significantly reducing the protocol complexity.


  1. Navodaya Garepalli, Kartik Gopalan, and Ping Yang, Control Message Reduction Techniques in Backward Learning Ad Hoc Routing Protocols, In the International Conference on Computer Communication Networks (ICCCN) 2008, U.S. Virgin Islands, August 2008. (Best Paper Candidate). [pdf] [bibtex]