Improvement of Multiple Antenna Sensing Technique for Detecting the White Space in a Spectrum Sharing System

Abstract

Exact detection of White Space (WS) is one of the actions in a Spectrum Sharing System (SSS) to determine unused spectrum for proper utilization. However, exact detection of WS is being affected by channel impairments, resulting in harmful interference. The Existing Multiple Antenna Spectrum Sensing (EMASS) technique used in addressing this effect is characterized with noise uncertainty leading to low detection rate due to setting of thresholds that is based on noise variance. Hence, this paper proposes an Improved Multiple Antenna Spectrum Sensing (IMASS) for detecting the WS in a SSS. Various copies of licensed user’s signals are received through the unlicensed user antennas over different antenna configuration. The received signals are combined using a modified equal gain combiner and energy of the combined signal is determined using Parseval’s relation for a discrete time signal. The received signal is used to form a square matrix which is converted to covariance matrix. Characteristic equation is obtained from covariance matrix to determine the minimum eigenvalue. The ratio of energy to minimum eigenvalue of the received signal is obtained and used as test statistics. The IMASS technique is evaluated using Probability of Detection (PD) and Total Error Probability (TEP) by comparing with EMASS. The proposed IMASS technique gives better performance with higher PD and lower TEP values than EMASS at all different antenna configurations.