Impact of Aggregation and Compression on Cluster-Based Wireless Sensor Networks

In Wireless Sensor Networks (WSNs), a clustered architecture is generally used to reduce energy consumption irrespectively of the application of the system. We prove in this work that, a clustered network only reduces energy consumption if aggregation or compression functions are enabled. Furthermore, a clustered network would consume much more energy if clustering techniques without the use of aggregation/compression (or a low aggregation coefficient) due to the extra consumption in the cluster formation phase. As an additional feature, a general energy consumption model based on the notion of energy units is developed that can be easily extrapolated to either theoretical or experimental values. Hence, the proposed analytical framework is valid for any commercial node or energy consumption model.