Optimal power allocating for correlated data fusion in decentralized WSNs using algorithms based on swarm intelligence

In unstructured wireless sensor networks (WSNs), which consist of a dense collection of sensor nodes deployed randomly, the communication and processing capabilities of sensor nodes can be limited owing to their small embedded batteries and available bandwidth. Power management is therefore one of the most important issues to consider in the implementation of WSNs. As a result, decentralized detection, in which the fusion center makes the final decision to use data partially processed by local nodes, is more attractive than centralized detection in unstructured WSNs. This paper proposes a more efficient and effective method for solving the power allocation problem as a constrained optimization problem: to schedule power allocation in a distributed WSN using correlated observations and amplify-and-forward local processing at sensor nodes so that the WSN detects a constant signal while maintaining a sufficient fusion error probability threshold. To accomplish this goal, this paper proposes using Deb’s method, which does not require a penalty parameter when handling the constraints of the optimization problem. Additionally, representative optimization algorithms based on swarm intelligence are used, i.e., particle swarm optimization, ant colony optimization for continuous domains (ACO _R), and artificial bee colony. Through a simulation, their performance is compared for several different WSNs to determine the best algorithm for solving the power allocation problem.