TALONet: A power-efficient grid-based congestion avoidance scheme using multi-detouring technique in Wireless Sensor Networks

Traffic congestion avoidance is a critical issue associated with data routing in Wireless Sensor Networks (WSNs). Previous researches employ either traffic or resource control strategies as solutions. In this paper, we propose an energy-efficient grid-based traffic congestion avoidance scheme in WSNs, termed as TALONet. With our mechanism, the approaches applied to avoid congestion have three folds: 1) we use two different transmission power levels to alleviate the link-level congestion (LLC); 2) we also employ an effective buffer management approach to avoid node-level congestion (NLC); and 3) based on virtual grid frameworks, a multi-path detouring technique is used to increase the channel capacity for the congested traffic flows. Extensive simulation results show that our proposed TALONet scheme is superior to other existing protocols. Comparing to TARA, Backpressure, and naive (without congestion control) schemes, the improvement on the number of dropped packets can be achieved to about 45%, 51%, and 80%, respectively.