Joint Beamforming, Terminal Scheduling, and Adaptive Modulation with Imperfect CSIT in Rice Fading Correlated Channels with non-persistent Co-channel Interference
Ref: CISTER-TR-171103 Publication Date: 31, Dec, 2017
Joint Beamforming, Terminal Scheduling, and Adaptive Modulation with Imperfect CSIT in Rice Fading Correlated Channels with non-persistent Co-channel InterferenceRef: CISTER-TR-171103 Publication Date: 31, Dec, 2017
This paper presents a resource allocation algorithm for multi-user wireless networks affected by non-persistent co-channel interference. The analysis considers a network with one base station (BS) that uses a multiple antenna transmitter (beamformer) to schedule (in a time-division manner) transmissions towards a set of $J$ one-antenna terminals in the presence of $K$ non-persistent interferers. The transmitter is assumed to employ Maximum-Ratio Combining (MRC) beamforming with spatially-correlated branches and channel envelopes modelled as Rice-distributed processes. The BS has access to an imperfect (outdated) copy of the instantaneous Channel State Information (CSI) of each terminal. Based on this CSI at the transmitter side (CSIT), the BS proceeds to select (at each time interval or time-slot) the terminal with the highest measured channel strength for purposes of transmission. This imperfect CSIT is also used to calculate the coefficients of the beamformer that will be used to transmit information towards the scheduled terminal, as well as for selecting the most appropriate modulation format (via threshold-based decision). The main merits of this work are the following: 1) joint analysis of MRC-based beamforming, terminal scheduling based on maximum channel strength, and modulation assignment, and 2) joint modelling of the effects of spatial correlation, co-channel interference and imperfect CSIT. Results suggest that scheduling helps in rejecting co-channel interference and the degrading effects of imperfect CSIT. Spatial correlation could some times lead to better performance than the uncorrelated case, particularly in the low SNR (Signal-to-Noise Ratio) regime. Conversely, uncorrelated branches always outperform the correlated case in the high SNR regime. Spatial correlation tends to accumulate over the antenna array thus leading to a more noticeable performance degradation and more allocation errors due to the outdated CSIT assumption. The line of sight component is found to contribute to a better reception, but it also reduces the ability to counteract the degrading effects of imperfect CSIT due to the lack of diversity combining.
Published in International Journal of Advances in Telecommunications, Electrotechnics, Signals and Systems, IARIA, Volume 10, Issue 3 and 4, pp 186-195.