Abstract: Developments in wireless technologies and MEMS have led to much ease of use of wireless sensor networks. The energy issue is still an open challenge. Clustering is a good solution if it ensures even load distribution. Recently fuzzy logic is being used to combine different characteristics of nodes to decide the cluster structure of the network. This paper proposes how node differentiation can be done in heterogeneous networks based on position of the node, and then fuzzy logic is used for the cluster formation. The fuzzy system has two outputs – node selection probability and cluster size bound. Thus our proposal not only elects CH but also determines that cluster stricture should be such that potential hot spots are not given much load. Potential hot spot nodes are identified using the proposed node differentiation technique. Connectivity of the network for longer time is our main QoS parameter.

Keywords: WSN, clustering, connectivity, heterogeneous, fuzzy logic, hotspot problem