Spectrum cartography based on dynamic compressed sensing by using multiple domains information

Radio maps have experienced their success in applications of wireless communications for years by offering metrics of radio frequency (RF) information, e.g., power spectral density (PSD), within a geographical region of interest. Spectrum cartography technique constructs radio maps to expand the abilities of RF awareness. However, seldom of existing methods aim at constructing radio maps by utilizing multiple domains information. In this paper, a novel framework inspired by dynamic compressed sensing (DCS) has been proposed firstly to solve this problem. This flexible framework first to apply joint group-Lasso for PSD map construction based on the different sparse patterns between space and frequency domains as well as innovatively utilizes transmitters' mobility patterns for support prediction of DCS. Simulation experiments have been processed to assess the performance of methods within the proposed framework and framework's superiority has been proven.