Detecting the Number of Active Users in Coded Random Access Systems

This paper addresses the problem of detecting the number of unrecovered active users in a coded random access system with successive interference cancellation (SIC) at the receiver. The considered transmission protocol is contention resolution diversity slotted Aloha, where each user generates and transmits in random slots multiple packet replicas. A maximum a posteriori detector, capable to track the a posteriori probability mass function of the number of not yet recovered active users over SIC iterations, is developed for the case of availability of a priori information on the active users' distribution. Numerical results are provided, illustrating the detection algorithm capability to achieve low mean square error values taking advantage of new observations gathered throughout the SIC process.