AEDB protocol tuning with a fast efficient parallel multi-objective local search

Abstract

Mobile ad hoc networks (MANETs) are infrastructure-less communication networks spontaneously created by a number of mobile devices. Due to its highly fluctuating topology, finding the optimal configuration of communication protocols is a complex and crucial task. Additionally, different objectives must be usually considered. In our previous work, we introduced a novel parallel multi-objective local search to optimise the AEDB broadcasting protocol in terms of coverage, energy used, broadcasting time, and network resources. In this paper, we extend our previous work by studying three different versions of the problem, that are restricting the minimum coverage value for valid solutions, in order to accelerate the search and make it focus on the desired area of the search space. Additionally, the convergence speed of the algorithm was analysed, and different population initialisation techniques were compared. The new algorithm is validated vs. efficient state-of-the art multi-objective evolutionary techniques, offering high quality results in much shorter times, especially for high coverage solutions.