Detection of Abrupt Change in AOA/AOD within Massive MIMO Systems

Acquiring the angle of arrival (AOA) and the angle of departure (AOD) is crucial for massive multiple-input multiple-output (MIMO) beamforming since they are the basis of channel modeling, antenna beamforming and positioning. In practice, abrupt changes of AOA/AOD may happen occasionally due to changes in the scattering environment. Although multiple studies have proposed precise angle estimation methods, they often do not consider how to detect such changes accurately and quickly. Due to the relatively long angle coherence time, it is not necessary to perform precise angle estimations frequently. In fact, rough estimations are sufficient to detect abrupt angle changes. Once such a change is detected, we can then employ a precise and high-complexity precise approach to re-estimate angles, significantly reducing the overall computational overhead of the system. To address this, we introduce an online detection framework that assesses changes at every channel coherence interval (CCI) within an angle coherence interval (ACI). Additionally, to reduce complexity, we propose an offline detection framework that makes decisions only at the final CCI of the ACI of a given length. Numerical results demonstrate that both the proposed online and offline schemes can detect changes with minimal delay and a low false alarm rate, verifying the feasibility of our proposed method for angle change detection.