Massive Unsourced Random Access: Exploiting Angular Domain Sparsity

This paper investigates the unsourced random access (URA) scheme to accommodate numerous machine-type users communicating a base station equipped with multiple antennas. Existing works adopt slotted transmission strategy reduce system complexity; they operate under framework of coupled compressed sensing (CCS) which concatenates an outer tree code inner for slot-wise message stitching. We suggest that by exploiting MIMO channel information in angular domain, redundancies required encoder/decoder CCS can be removed improve spectral efficiency, thereby uncoupled protocol is devised. To perform activity detection and estimation, we propose expectation-maximization-aided generalized approximate passing algorithm Markov field support structure, captures inherent clustered sparsity structure domain channel. Then, reconstruction form clustering decoder performed recognizing slot-distributed channels each active user based on similarity. put forward slot-balanced $ K $ -means as kernel decoder, resolving constraints collisions specific application scene. Extensive simulations reveal proposed achieves better error performance at high efficiency compared CCS-based URA schemes.