Intelligent Sensing for Automated Spectrum Assignment
David Kidston, Maoyu Wang · 2019
As the number of wireless devices explodes in the lead up to the release of 5G communications technology, it is expected that the demands for usage of the wireless spectrum will increase to the point where current spectrum allocation methods will no longer be sufficient. Dynamic Spectrum Management (DSM) uses cognitive radio methods to sense the current state of spectrum usage by other devices, and then makes use of that data to allocate spectrum in time and space to best meet users communication requirements in near real time. Our previous work on a Spectrum Environment Awareness (SEA) sensor system suggests that static tasking of the large number of sensors required for 5G would consume large amounts of storage and communications resources. Intelligent and collaborative tasking of sensors would reduce both these overheads as well as the burden on data analysis. In this paper, we describe an intelligent distributed collaborative sensing system with centralized control for use in DSM-based automated spectrum assignment. We use simulation results to estimate the amount of sensor reading duplication in such a system. These results provide direction on how intelligent sensing systems can optimize the tasking of sensors in this domain.