Trained Particle Swarm Optimization for Ad-Hoc Collaborative Computing Networks

Distributed processing is an essential part of collaborative computing techniques over ad-hoc networks. In this paper, a generalized particle swarm optimization (PSO) model for communication networks is introduced. A modified version of PSO, called trained PSO (TPSO), consisting of distributed particles that are adapted to reduce traffic and computational overhead of the optimization process is proposed. The TPSO technique is used to find the node with the highest processing load in an ad-hoc collaborative computing system. The simulation results show that the TPSO algorithm significantly reduces the traffic overhead, computation complexity and convergence time of particles, in comparison to the PSO.