Energy Efficient Routing Strategy for Dynamic Wireless Sensor Networks

Abstract

A wireless sensor network (WSN) is the network of large number of sensor nodes that connect to each other with or without help of existing infrastructure. A WSN base station is a huge database and receives data from sensor nodes. Each sensor node performs tasks of sensing, processing, data communication and can be de ployed in locations which are di cult for humans to access. WSN systems can be broadly categorized as either static networks or mobile/ dynamic networks. Static sensor network involves setting up a network where the sensor nodes are arranged in xed positions. Dynamic or Mobile WSNs (MWSNs) contain mobile nodes and are suitable for applications involving obtaining information from moving objects such as animals or humans where sensor nodes need to be attached directly to objects and move with them. MWSNs can be used for habitat monitoring, health-care, en vironmental monitoring, transportation networks, underwater surveillance, military applications and many more. A routing protocol helps to nd the path from sensor node to the sink. Routing in MWSNs has several challenges as against static sensor networks due to dynamic conditions. Energy is the most constrained resource as the sensor nodes are battery operated and preferably used in remote environments. The major contribution of the thesis is proposition of two new routing protocols namely Connectivity Based Energy E fficient Opportunistic Robust (CBEEOR) rout ing and Connectivity-based Cross-layer Opportunistic Forwarding (CCOF) for rout ing in MWSNs and an approach for predicting the network lifetime based on residual energy of node . Both the protocols use a neighbour node list prioritized on the basis of expected cost of forwarding the packet called as prioritized forwarder list. Conclu sions of the research work are crisply, pointing towards the possible areas of future research in this eld. In CBEEOR, expected cost of forwarding is determined as combination of packet transmission cost, expected packet relaying cost and network connectivity cost. CBEEOR implements a MAC layer two-tier delay mechanism to have forwarding agreement among the nodes in prioritized forwarder list. Whereas, CCOF computes the expected cost of forwarding as combination of packet trans mission cost and expected packet relaying cost. The packet transmission cost and expected relaying cost is determined jointly using energy consumption and algebraic connectivity of the network. In CCOF protocol , a single-tier forwarding agree ment is implemented for coordination among the nodes in the forwarder list. Both the protocols are simulated using ns2 and their performances are evaluated against the performance metrics like energy consumption, packet delivery, delay and over heads. An approach for network lifetime prediction based on highest value energy depletion rate is proposed. This approach is used to estimate the lifetime of WSN implemented using waspmote nodes. The results of network lifetime estimation are validated by measuring the lifetime until it becomes non-functional. Also CCOF is also implemented on waspmote hardware platform for measuring the lifetime of the network. The results of the above research work are published to research community in form of peer-reviewed journal publications and national conference publications.