Compressive Sensing for Uplink Cloud Radio Access Network with Limited Backhaul Capacity

Cloud radio access network (Cloud-RAN) is recognized as a promising cellular network architecture in which baseband signal processing is migrated to the cloud central unit (CU). This paper focuses on the multi-cluster Cloud-RAN where the whole network is divided into multiple clusters and each cluster of radio units (RUs) is connected to the cloud central unit (CU) via the cluster head (CH). As a result of the multiple access fading, the received signals at different RUs within one cluster are statistically correlated. Inspired by this observation, we propose a compressive sensing based uplink scheme in which the CH compresses the gathered data from the RUs within the same cluster and then transmits to the CU efficiently. And we derive the uplink rate, which increases with the user SNR, of the mobile stations (MSs) with successive interference cancellation (SIC) implemented in CU. We further prove that the SNR increases with the limited backhaul capacity and the maximum SNR can be reached when the measurement channel capacity equals to the backhaul capacity. Finally, numerical results are presented to demonstrate the efficacy of this compression scheme, the maximum SNR and achievable uplink rate under backhaul constraint, which complies with our previous analysis.