Cooperative Spectrum Sensing Using RMT For Cognitive Radio in Presence of Noise Correlation
This paper presents an eigen-value based cooperative spectrum sensing (SS) technique in the presence of noise
correlation for Cognitive Radio (CR). In this work, a new Standard Condition Number (SCN) based decision statistics is
defined using asymptotic random matrix theory (RMT) for the decision making process. First, the effect of noise correlation
under both noise only and signal plus noise hypothesis is defined. Then the new bounds for the correlated noise scenario are
defined and a the new SCN based threshold is derived for SS in CR. Simulation results show that sensing with the proposed
threshold gives better performance in the presence of noise correlation.
Index Terms—Random Matrix Theory (RMT), Noise Correlation, Standard Condition Number (SCN), Spectrum Sensing