Scalability of a novel shifting balance theory-based optimization algorithm: a comparative study on a cluster-based wireless sensor network

Scalability is one of the most important issues for optimization algorithms used in wireless sensor networks (WSNs) since there are often many parameters to be optimized at the same time. In this case it is very hard to ensure that an optimization algorithm can be smoothly scaled up from a low-dimensional optimization problem to the one with a high dimensionality. This paper addresses the scalability issue of a novel optimization algorithm inspired by the Shifting Balance Theory (SBT) of evolution in population genetics. Toward this end, a cluster-based WSN is employed in this paper as a benchmark to perform a comparative study. The total energy consumption is minimized under the required quality of service by jointly optimizing the transmission power and rate for each sensor node. The results obtained by the SBT-based algorithm are compared with the Metropolis algorithm (MA) and currently popular particle swarm optimizer (PSO) to assess the scaling performance of the three algorithms against the same WSN optimization problem.