Influence of Sample Vector Normalization on Cooperative Spectrum Sensing Under Direct-Conversion Receiver Effects

This paper investigates sample vector normalization as a statistical preprocessing technique for cooperative spectrum sensing under realistic direct-conversion receiver (DCR) impairments. A comprehensive receiver model is considered, including residual DC offset, filtering, automatic gain control, quantization, and whitening. Extensive Monte Carlo simulations demonstrate that receiver impairments severely distort the sample covariance matrix, break the CFAR property, and degrade detection performance. Normalization effectively mitigates these effects, restoring threshold stability, monotonic behavior with respect to SNR, and robust detection across different blind sensing algorithms. The results establish normalization as a key enabler for practical and reliable cooperative spectrum sensing.