Network Diversity Multiple Access with Imperfect Channel State Information at the Transmitter Side
Ref: CISTER-TR-161004 Publication Date: 3 to 4, Nov, 2016
Network Diversity Multiple Access with Imperfect Channel State Information at the Transmitter SideRef: CISTER-TR-161004 Publication Date: 3 to 4, Nov, 2016
Network diversity multiple access (NDMA) is the family of algorithms with the highest potential throughput in the literature of signal-processing assisted random access protocols. NDMA uses the concept of protocol-induced retransmissions to create an adaptive source of diversity. This diversity is used to resolve packet collisions employing signal separation tools without the explicit need (or as a complement) of a multiple antenna receiver. This paper proposes a further improvement on the performance of NDMA by allowing each terminal access to an outdated copy of its individual channel state information (CSI). Based on this decentralized CSI, each terminal conveniently decides to transmit only if the estimated channel gain surpasses a threshold that is optimized to maximize performance. This ensures that the probability of terminal presence detection, and thus the probability of correct estimation of the collision multiplicity are considerably improved at the receiver end. The paper is focused on the modelling of the receiver operational characteristic (ROC) of the terminal presence detector considering that the CSI used by each terminal is potentially inaccurate (outdated) due to feedback delay. The results indicate that when the correlation coefficient that describes the accuracy of the available CSI tends to zero, the scheme degrades into the conventional NDMA. By contrast, when the quality of the channel state information improves, the throughput can nearly achieve the nominal channel rate (minimum throughput penalty). The selection of the detector thresholds for channel gain and terminal presence is optimized to maximize system performance.
RTUWO Advances in Wireless and Optical Communications 2016 (RTUWO2016).