Joint block sparse signal recovery-based active user detection in 5G cloud radio access networks

The Cloud Radio Access Network (C-RAN) is a state-of-the-art system paradigm that simultaneously improves spectral and energy efficiency. Capacity constraints of the fronthaul links connecting Remote Radio Heads (RRH) to the Cloud Unit are notable limitations of these networks. The multitude of RRHs and users make active user estimation and calculating Channel Side Information (CSI) between active users and RRHs necessary for implementing these networks. Moreover, in C-RAN, user activity detection is essential for energy-efficient resource allocation, calculating CSI, optimal precoder design, interference management, and multi-user detection. This study investigates active user detection in C-RAN as a joint block sparse signal recovery problem and evaluates the impact of fronthaul limitations, sparsity level, and other network parameters for different sparse signal reconstruction methods. We introduce an efficient method that is based on recovering multiple sparse signals sharing the same sparsity pattern or the same support set of non-zero entries. This method is developed using 5G training signals for user activity detection in C-RAN with fronthaul capacity limitations and without prior knowledge of the sparsity of users. In the end, we compare active user detection results for different sparse signal recovery methods, namely joint block sparse signal and block sparse signal algorithms, with different network specifications.