Local MTM-SVD based spectrum sensing in SIMO OFDM cognitive radio under bandwidth constraint

In this paper, we propose using local multitaper-singular value decomposition (Local-MTM-SVD) for spectrum sensing in OFDM-based cognitive radio (CR) systems. The spectrum sensing locality in this technique is enhanced by using single input at the primary (PR) transmitter and multi output antenna system (SIMO) at the different CR nodes (sensors). The global decision is declared using the OR rule at the CR basestation based on hard cooperation. This technique reduces the bandwidth of the feedback channel, and for powerful spectrum sensing technique; the MTM-SVD is used. Our proposed technique is examined using software simulation and theoretical analysis. The results show that there is a significant improvement in the probability of detection compared to using MTM only, or using the periodogram with single antenna in both AWGN and Rayleigh flat fading environments. We found the number of singular value squares that are useful in the technique is equal to or smaller than 2 when the total number of generated singular values is 4, and the MTM only technique tends to offer same performance as the Local-MTM-SVD technique using single antenna in the same conditions. The choice of the decision threshold level depends on the number of singular value squares which we used in the decision statistic.