Forward-Error-Control Coded and Sparse-Modulated Preambles for Massive Connectivity

Massive connectivity (MC) is a key technique for supporting efficient communication with a vast number of users/devices in future wireless networks. For MC with sporadic traffic and short-length packets, user identification and channel estimation become bottlenecks in system performance. Utilizing the sparse pattern of active users, various compressed sensing-based designs have been proposed. However, the complexity of most existing designs is affected by the total number of users, making them unsuitable for future networks with an ever-increasing user scale. In this paper, we propose a forward-error-control (FEC) coded and sparse-modulated (FEC-SM) preamble design. By leveraging the FEC code structure and sparse features, we derive an efficient turbo-based Bayesian receiver whose complexity only scales linearly with the number of active users, regardless of the total number of users. Numerical results show that the proposed preamble outperforms existing alternatives in conventional scenarios, and performs well in scenarios with extremely massive numbers of users (e.g., 107 to 108).