Source-Initiated Geographical Data Flow for Wireless Ad Hoc and Sensor Networks

Abstract

A novel source-initiated geographical data flow technique, called stateless weighted routing (SWR), is presented in this paper. Nodes keep only their own virtual geographical position and require no local topological information. Each node calculates the weight of its own. Initially, this value is its relative distance to the sink. Each node decides to retransmit or to drop the packet received by comparing its own weight to the weight of the sender and the weight of the sink, i.e, destination, which are contained in the received packet. The comparison actually provides the mean for stateless routing. Although the weight parameter includes only the distance information for the time being, it may also include QoS (quality of service) parameters such as the energy left at the node. QoS will definitely help to increase the lifetime of the system. Having had the feature of being stateless, the technique is made free from the use of excessive communications to handle the routing tables. The SWR provides braided-paths if not multi-paths, naturally which is essential to improve reliability and to serve for the time-critical data. The use of thresholds in retransmissions provides the system with a flexible and energy-efficient data flow. Moreover, to the best of our knowledge, the SWR is known to be the first stateless routing technique running independent of the MAC-layer underneath.