Traffic-adapted load balancing in sensor networks employing geographic routing

Load balancing is an important issue in sensor networks especially when geographic routing is employed since in this case, a node often forwards its packet to a certain neighbor. As a result, nodes located at the intersection of multiple routes to the base station tend to be placed under great stress and drain its energy quickly. This can potentially lead to disconnection of the network. Current solutions to this problem fail to take the traffic load at nodes into account and can potentially forward packets toward the heavy trafficked region and create collisions and network congestion. In this paper, we propose a load balancing scheme, namely Traffic Adaptive Routing (TAR), which works with any geographic routing protocols. TAR relieves the aforementioned issues by having a node evaluate the level of traffic at its neighbors and decide where to forward the packet based on the distance to the base station and the traffic load. Through simulations, we have shown that when compared against the existing load balancing techniques, TAR is able to deliver 10% more packets to the base station using approximately the same amount of energy due to the reduction of the collisions and congestion.