Cooperative Processing for the WLAN Uplink

We study the potential of cooperative processing in WLAN uplink scenarios where the access points are connected to a backhaul with limited capacity. We evaluate different decoding strategies based on an outage criterion and discuss the influence of the available backhaul rate. Compared to the ergodic analyses in the literature we observe three major differences: Firstly, outage rates are much more robust to the backbone limitation. Secondly, Slepian-Wolf coding (compress & forward processing) does not improve the outage performance significantly compared to pure quantization and forwarding. Finally, quantization based approaches offer substantially higher gains compared to decoding based strategies. Because of the large number of existing WLANs, we conclude that cooperative processing is a key enabler for high rates on future WLAN uplinks: in cases of limiting backhaul capacities (e.g. ADSL broadband connections in residential environments) as well as in cases of large backhaul capacities (e.g. LAN connections in office environments).