Over-the-Air Computation for Cooperative Wideband Spectrum Sensing and Performance Analysis

For sensor network aided cognitive radio, cooperative wideband spectrum sensing can distribute the sampling and computing pressure of spectrum sensing to multiple sensor nodes (SNs). However, this may incur high latency due to distributed data fusion, especially when the number of SNs is large. In this paper, we propose a novel cooperative wideband spectrum sensing scheme using over-the-air computation. Its key idea is to utilize the superposition property of wireless channel to implement the summation of Fourier transform. This avoids distributed data fusion by computing the target function directly. The performance of the proposed scheme is analyzed with imperfect synchronization between different SNs considered. Furthermore, a synchronization phase offset estimation and equalization method is proposed. The corresponding performance after equalization is also derived. Both a working prototype based on universal software radio periphera and Monte Carlo simulations are provided to verify the performance of the proposed scheme.