Detecting Aerial Radars from Massive MIMO Base Stations: Design and Experimental Evaluation

Dynamic access to spectrum traditionally reserved for static incumbents could greatly expand the capabilities of next-generation cellular networks. However, many incumbent users, especially military radars, are sensitive and classified, making coexistence viable only if the cellular system can robustly detect their presence and activate interference-mitigation strategies. In this article, we examine the problem of initial detection of an aerial radar platform operating within a wide frequency band by a cellular base station that takes advantage of massive multiple-input multiple-output (MIMO) antenna arrays as the key enabling technology. We develop a general framework wherein the individual antennas of the massive MIMO array are dynamically tuned to smaller subbands within the accessible spectrum to locate the unknown operational band of the aerial radar. To this end, we derive a detector based on the generalized likelihood ratio for the proposed framework. We provide an extensive experimental analysis of a first-of-its-kind over-the-air dataset collected using the Rice University massive MIMO Reconfigurable Ecosystem for Next-generation End-to-end Wireless (RENEW) platform and a drone-borne radar emulator. We characterize the detector's performance with respect to the number of antennas in the receiver array and examine the influence of real-world multipath on its reliability using RENEW data.