Wideband Sensing and Optimization for Cognitive Radio Networks with Noise Variance Uncertainty

TL;DR

This paper introduces novel wideband spectrum sensing techniques for cognitive radio networks that effectively handle noise variance uncertainty, improving detection accuracy and throughput optimization.

Contribution

It proposes new ratio-based test statistics and a generalized energy detector that outperform existing detectors under noise uncertainty conditions.

Findings

GED does not suffer from SNR wall

Optimal sensing time depends on noise variance information

Significant throughput gains with proposed methods

Abstract

This paper considers wide-band spectrum sensing and optimization for cognitive radio (CR) networks with noise variance uncertainty. It is assumed that the considered wide-band contains one or more white sub-bands. Under this assumption, we consider throughput maximization of the CR network while appropriately protecting the primary network. We address this problem as follows. First, we propose novel ratio based test statistics for detecting the edges of each sub-band. Second, we employ simple energy comparison approach to choose one reference white sub-band. Third, we propose novel generalized energy detector (GED) for examining each of the remaining sub-bands by exploiting the noise information of the reference white sub-band. Finally, we optimize the sensing time ($T_o$) to maximize the CR network throughput using the detection and false alarm probabilities of the GED. The proposed GED does not suffer from signal to noise ratio (SNR) wall and outperforms the existing signal detectors. Moreover, the relationship between the proposed GED and conventional energy detector (CED) is quantified analytically. We show that the optimal $T_o$ depends on the noise variance information. In particular, with $10$TV bands, SNR=$-20$dB and $2$s frame duration, we found that the optimal $T_o$ is $28.5$ms ($50.6$ms) with perfect (imperfect) noise variance scenario.