Design of network coding based reliable sensor networks

Abstract

Network coding can be very useful in achieving a balance between energy efficiency and end-to-end packet error in sensor networks, particularly if only a subset of the nodes act as encoding nodes. In this article, a mathematical model and a heuristic algorithm are proposed to plan for the best placement of encoding nodes while ensuring reliability. These approaches are also able to address scenarios where sensor networks, using different gateways, are federated. In this case a distributed storage system is required to ensure the recovery of packets when related coded packets arrive to the different gateways. The experimental results show that the adequate number and placement of encoding nodes can be effectively determined, significantly enhancing the performance of constrained sensor networks using network coding.