Energy Efficient Clustering and Routing Protocols for Wireless Sensor Networks: a Simulation Approach

This dissertation presents an approach to solve the Wireless Sensors Networks (WSNs) energy depletion problem. WSNs devices are usually battery-powered thus gathering data from a WSN in an energy efficient way is challenging. Each device is a node of the network. Clustering, an approach which organises nodes into sets (clusters), is a widely used solution proposed by researchers. Each cluster has a coordinator, called cluster head (CH), that gathers data from its neighbours and communicates with near CHs. Nodes elect to become CH basing on their hardware capabilities and some other network metrics. Most of the works in the literature use a cluster election that only considers the nodes battery capacity and residual energy. Such behaviour can cause a quick depletion of nodes energy level. CHs should be chosen considering other features in addition to residual energy. For this reason, Rotating Energy Efficient Clustering for Heterogeneous Devices (REECHD) has been conceived. REECHD is a clustering protocol which considers both node residual energy and node work for leader election. REECHD reduces the number of leader elections by rotating the CH role among nodes in the same cluster. REECHD also introduces the concept of intra-traffic rate limit (ITRL) which defines a limit on the communication inside the clusters and can be used to further improve energy efficiency. REECHD is then further enhanced by introducing a novel routing approach (CER-CH) which combines a novel rotation heuristic with a novel top-down CH routing tree definition in order to balance the node energy consumption and generate more energy efficient routing trees. CER-CH defines the rotation and the routing amongst CHs, regardless of the strategy chosen for the initial CHs election and cluster formation. Any clustering algorithm that produces a CH per cluster can be used. REECHD and CER-CH have been fairly compared with the state of art routing protocols through simulation. Simulations show that they ensure a gain with respect to the state of art clustering protocols.