LAUSR

Given the explosive growth of wireless devices ranging from the 2D ground to the 3D space, the first task preceding any form of dynamic spectrum management for the smart city infrastructures is the sensing of heterogeneous spectrum situation in wireless environments. However, challenge arises when increasing wide bandwidth and enormous geographic space meet power-limited CR receivers with constrained hardware capability. In this paper, a 3D compressed wideband spectrum mapping model is established by exploiting the underlying sparse nature of wideband spectrum situation based on compressed sensing, where joint time-frequency-space compression is executed at sub-Nyquist sampling rate. Due to the ineffectiveness of traditional algorithms, a novel two-phase 3D compressed wideband spectrum mapping scheme is proposed which is composed of sampling points optimization and 3D wideband spectrum situation recovery. Therein, quadrature and right-triangular (QR) block pivoting based spatial sampling matrix optimization algorithm considers every frequency priority of each location jointly to select dominant sampling locations. Alternating direction method of multipliers is further applied to iteratively solve the situation recovery. Lastly, in-depth numerical simulations demonstrate the spectrum situation recovery performance, which proves the proposed 3D compressed wideband spectrum mapping scheme greatly reduces the number of measurements, while achieving a high level of wideband spectrum mapping accuracy.