Distributed Linear Summing in Wireless Sensor Networks

The usage of mechanisms for data aggregation becomes an important part of many real-life applications due to a reduction of negatives factors. In this paper, we examine the applicability of average consensus with a varying mixing parameter for distributed summing in wireless sensor networks. A stopping criterion proposed for these networks is assumed to be implemented in order to bound the algorithm execution. The parameters of the implemented stopping criterion and the mixing parameter of average consensus are varied in order to find their most suitable initial configuration in terms of the estimation precision and the convergence rate over 60 random geometric graphs with either a dense or a sparse connectivity.