Particle Swarm Optimized Antenna Reconfiguration in Multikernel Bayesian Learning Signal Beamforming Algorithm
Antenna reconfiguration for accurate beam forming is developed for Non-uniform Linear Array (ULA). Multi Kernel Bayesian Learning algorithm is adopted in the beam forming implementation. Undetermined source localization problem is solved using the Multi Kernel Sparse Bayesian Learning framework. Beam forming problem is considered the undetermined source localization problem and solved using the adaptive method. The Degree of Freedom (DOF) is increased using the adaptive nature of the manifold matrix while maintaining the same number of antennas. The response model that adaptively adjusts the manifold matrix in the Sparse Bayesian problem uses the Multi Kernel framework. MATLAB based implementation thus carried out on the ULA clearly exhibits better results over the single kernel model. The Mean Square Error (MSE) and Root Mean Square Error (RMSE) with Signal to Noise Ratio (SNR) variations are obtained to evaluate the performance of the proposed implementation. The performance obtained is found to be satisfactory and is at par with the recent previous implementation. Reconfiguration of distance between the antennas is implemented to obtain an adaptive beam forming with good performance comparison. Reconfiguration is optimized using the Particle Swarm Optimization (PSO) algorithm that controls the distance between the antennas for best Root Mean Square Error (RMSE) for different Signal to Noise Ratio(SNR).
Keywords-Direction of Arrival Estimation, Multi-kernel Bayesian Learning, Basis Pursuit Methods, Antenna Reconfiguration, Signal Beam-forming