Applying REECHD to non-uniformly distributed heterogeneous devices

Heterogeneous Wireless Sensor Networks (WSNs) are essential to implement the IoT vision. They are a virtual skin that allows the gathering of environmental data. This can be used to enhance human life and build innovative applications such as smart environments, advance health care systems and smart cities. WSNs are composed of nodes that can have dierent initial energy, dierent node transmission rate and dierent hardware. Energy eciency is one of the most important challenges when building WSNs since nodes are battery powered. Clustering is an extensively used approach to enhance the heterogeneous WSN lifetime. Clustering partitions the WSN nodes into a set of clusters. Each cluster contains a cluster head (CH) that collects data from its member nodes and forwards them to a centralised Base Station (BS). Although a plethora of dierent clustering approaches have been proposed CH selection is usually based on the node residual energy. Some approaches can take into account the node transmission rate but only for cluster formation. Rotating Energy Ecient Clustering for Heterogeneous Devices (REECHD) is our novel clustering algorithm that considers both node residual energy and node transmission rate for cluster head election. REECHD also proposes the introduction of the intra-trac rate limit (ITRL). This limits the amount of intra-trac data that a CH can receive. ITRL can improve energy eciency. In this work we apply REECHD to a WSN where devices are not uniformly distributed. More precisely, we consider WSNs where some subareas generate a higher volume of trac. We show how the use of ITRL improves energy eciency by adaptively generating a dierent amount of clusters in dierent WSN subareas. REECHD outperforms the state of art clustering protocols of 220% when rst node die lifetime measure is considered. Our results show that REECHD enhances on average the network lifespan when compared to the state of art protocols.