Chain Based Fault Tolerant Routing Protocols

Abstract

Sensor networks have emerged as a promising tool for monitoring (and possibly actuating) the physical world; utilizing self-organizing networks of battery-powered wireless sensors that can sense, process, and communicate. In wireless sensor networks (WSNs), energy is a critical resource; hence power efficient routing protocols is necessary for data transmission in order to extend the network lifetime. Recently number of efficient chain based protocols has been proposed for WSNs routing. These routing protocols have achieved lowest consumed energy and delay. However, fault-tolerance was not considered in these protocols. Since node failures are inevitable in WSNs due to the harsh deployment environment, nodes mobility, and nodes depleted of energy. Therefore, fault tolerance is a must for successful routing protocols. In this paper, fault tolerance is incorporated for chain based routing protocols. More specifically, two techniques for fault detection and fault recovery for chain based routing protocols are proposed. The two techniques employ the same strategy for fault detection; every node sends a notifying message to its successor neighboring node, if the successor node is alive it replies with a ready message else the node sending the notifying message deduces that its successor node is dead. However, the recovery strategy is different for the two techniques. The first technique overcomes the fault by having every predecessor node to a failed node instead of sending its data to the failed node forward it to the successor node of the failed node. This technique is used for both a single chain routing protocols as well as chain-chain routing protocols.  The second technique gets around the fault by choosing a backup node for the faulty from the neighboring chain closest the sink which satisfies minimum energy consumption.  However, the second technique is used only for chain-chain routing protocols. The fault detection phase of the two proposed protocols may be applied at each data round or at varying intervals as dedicated by the application and the environment in which the WSN is deployed. The simulation results indicate that the two proposed protocols achieved fault tolerance efficiently (energy and time wise) for a single node failure at each chain.