Reduction of radar band exclusion zones through database-aided beamforming

Coexistence between radar systems and low-power broadband devices has recently become a focus of research by regulators, industry and academic bodies, with the possibility of sharing the 3.5 GHz band indicated by the Federal Communications Commission (FCC) in 2012. Radar bands are seen as attractive candidates for Dynamic Spectrum Access (DSA) due to the large portion of spectrum they occupy and their spectral occupancy being fairly low. However, the current reliance on Geolocation Database (GL-DB) spectrum access techniques for protecting radar systems from interference leads to large exclusion zones and inevitably will block a great percentage of the world population from benefiting from this spectrum. In this paper, we analyse the adequacy of cognitive beamforming as a technique employed by small cells to reduce caused interference on radar systems and, consequently, decrease their exclusion zones´ sizes. We start by discussing the impact the radar systems´ antenna rotation has on the dynamics of the interference channel and what challenges it brings to the design of a cognitive beamforming technique with guaranteed performance. We then propose a GL-DB-aided beamforming scheme, where a centralized GL-DB defines an exclusion zone size matching the cognitive beamforming capabilities of the small cell evolved Node Basestation (eNB) querying for spectrum access. We show that when an eNB dedicates 4 antennas to cancelling its interference to a radar, exclusion zones can be reduced by approximately 40 km. We also studied the degrading effect cognitive beamforming may have on the small cells´ throughput and propose, as a solution, an adaptive scheme that selects the best beamforming strategy to use based on the distance between the radar and the small cell.