Analysis on Data Compression of Two-stage Sensing for Cognitive Radio

In wireless network systems that use cognitive radio, massive amounts of sensing data are collected by cooperative sensing and wideband sensing. In such a system requires high resolution sensing data for high signal detection accuracy to protect the primary users and analysis of the received signals. In order to achieve compressing sensing data and high resolution, two-stage sensing: sub-Nyquist-based wideband sensing with data compression and high-resolution narrowband sensing has been considerd. In this paper, we present an analysis on the data compression of two-stage sensing. Several methods have been proposed for the compressive architecture and reconstruction algorithm of sub-Nyquist-based wideband sensing. However, compression performance due to differences in methods has not been compared. We evaluated the compression performance of the whole two-stage sensing process for three compressive architectures that are Ramdom Demodulator, Rondom Filtering and Modulated Wideband Converter and two reconstruction algorithms that are Basis Pursuit and Orthogonal Matching Pursuit by changing the signal sparsity and channel placement. The results yielded a combination of methods with good compression performance.