A fuzzy metric for Collection Tree Protocol in Wireless Sensor Networks

Abstract

A fuzzy metric for the Collection Tree Protocol (CTP) algorithm in Wireless Sensor Networks (WSNs) is proposed. This metric incorporates node energy level and congestion along with the traditional link quality parameter used in CTP with an aim of making CTP energy and congestion aware. The performance of the fuzzy metric based CTP is evaluated and compared with the original CTP algorithm, in a TinyOS Simulator (TOSSIM), under different network conditions with short and long duration experiments. The two algorithms exhibit more or less the same Packet Delivery Ratio (PDR) for short duration experiments. In long duration experiments, the fuzzy metric based CTP maintains a PDR higher than 0.99 for all simulation times while the PDR of the traditional CTP goes down to 0.67 when the simulation time is 2000 seconds. In the new algorithm, all nodes die almost at the same time, maintaining high reachability and high PDR throughout the lifetime of the nodes. The opposite happens with the original algorithm thereby creating reachability problems and contributing to the low PDR. These results show that effective network lifetime can be improved by using the proposed fuzzy metric based CTP without jeopardizing the basic functionality of CTP. There is, however, a need to test the new approach under more network conditions. More aggregation techniques can also be used for comparison purposes.