Universal Partitioning of a Large Array for Communications in Environments with Limited Spatial Coherence

Large aperture arrays improve communication performance with higher gain, but are susceptible to limited spatial coherence of signals, especially in shallow water acoustic environments. An acoustic communication receiver's ability to combine the signal wavefront coherently degrades when there are losses across the array in either phase coherence or the ability to track phase variation. To mitigate such coherence issues, the array can be separated into smaller segments of sensors known as subapertures. The phase variations can be tracked and mitigated on each subaperture. Subsequently, each subaperture can be coherently combined to realize the full array gain, through this two-stage process. However, subaperture size selection remains an open problem, especially in dynamic, uncertain environments like those experienced in underwater acoustic communications. In this paper, a universal algorithm is proposed for realizing the performance of the best possible partitioning among a collection of partitions of the receiver into subapertures, such that the overall receiver performance is as good as if the best partitioning were known a priori. This work builds on previous work in universal adaptive filtering and beamforming.