ULA‐based near‐field source localisation in cognitive femtocell network: a comparative study of genetic algorithm hybridised with pattern search and swarm intelligence

Abstract

The rapidly proliferating wireless technology brings an overwhelming increase in wireless applications at a continuous pace. In this framework, a major proportion of overall voice and data traffic originates from indoor users, directing the research world towards the deployment of low-power, low-cost, and low-range femtocell networks in the existing macrocells. However, this development will result in cross-tier interference as well as interference to the primary or licensed users. Cognitive femtocell networks smartly address this challenge through their intelligent sensing and decision-making capabilities. This study proposes a joint spectrum sensing technique, which involves individual sensing by multiple cognitive femtocell base stations (CFBSs), each equipped with a uniform linear array (ULA) of passive sensors. The local observations of CFBSs are then evaluated at the fusion centre to make final decision applying majority rule. In addition to the detection of a number of active primary femtocell networks (PFNs), the proposed near-field source localisation technique provides four-dimensional parameter estimation of each detected PFN signal, i.e. amplitude, angle-of-arrival, frequency, and range. Finally, the proposal is supported by three different implementations, i.e. Hybrid Genetic Algorithm, Particle Swarm Optimization, and Artificial Bee Colony. It also outsmarts an existing single-ULA based spectrum sensing technique in the literature.