Collaborative beamforming for wireless sensor network using particle swarm analysis

Abstract

In Wireless Sensor Network (WSN), nodes can collaborate to monitor, gather and select only the required data to transmit to the receivers. However, the nodes are working in uncertain hazardous environments that lead to undesirable high battery power consumption. Thus, it is desirable to improve radiation beampattern performance by introducing intelligent Collaborative Beamforming (CB) concept. It manages to increase the antenna gain and performance by aiming at desired objectives through intelligent capabilities. In this thesis, the nodes are designed to cooperate and collaborate among themselves and act as a collaborative antenna array. An optimal CB algorithm for intelligent sensor node array has been developed which combines CB and Particle Swarm Optimisation (PSO) in the presence of node geometry location uncertainties. The collaborative nodes are modelled in linear and circular array configurations. Firstly, a theoretical foundation employing CB inside WSN is developed consisting of three main stages: parameter initialisation, activation and optimisation setup. Then, newly proposed Intelligent Linear Sensor Node Array (ILSA) and Intelligent Circular Sensor Node Array (ICSA) are successfully optimised by applying Hybrid Least square improved PSO (HLPSO). The HLPSO has been developed using global constraint boundaries variables and, reinitialisation of particle’s position and velocity. It incorporates with Least Square approximation algorithm. For intereference occurence case at six unintended receivers, ILSA manages to significantly suppress Sidelobe Level (SLL) up to 85.54% in average. For null placement, the peak SLL within the null ranges angles have been greatly minimised up to 103%. The ICSA with multi-objective optimisation has outstandingly reduced SLL to 213% with 36° First Null Beamwidth size increment. Both ILSA and ICSA can effectively improve radiation beampattern performance and coverage by intelligently adjusting the shape of the beampatterns under different constraints as per desired usage. So, it accomplishes significant improvements compared to the referenced CB algorithm.