Message Passing-Based Joint User Activity Detection and Channel Estimation for Temporally-Correlated Massive Access

This paper studies the user activity detection and channel estimation problem in a temporally-correlated massive access system where very large number of users communicate with base station sporadically each once activated can transmit probability over multiple consecutive frames. We formulate as dynamic compressed sensing (DCS) to exploit both sparsity temporal correlation activity. By leveraging hybrid generalized approximate message passing (HyGAMP) framework, we design computationally efficient algorithm, HyGAMP-DCS, solve this problem. In contrast only exploiting historical estimations, proposed algorithm performs bidirectional between neighboring frames for likelihood update fully activities. Furthermore, develop an expectation maximization HyGAMP-DCS (EM-HyGAMP-DCS) adaptively learn hyperparameters during procedure when statistics are unknown. particular, propose utilize analysis tool state evolution find appropriate hyperparameter initialization EM-HyGAMP-DCS. Simulation results demonstrate that our algorithms significantly improve accuracy reduce error.