Article ID | Journal | Published Year | Pages | File Type |
---|---|---|---|---|
455775 | Computers & Electrical Engineering | 2013 | 10 Pages |
The performance of cellular uplink multiple-input multiple-output systems, where co-channel interference (CCI) at the base stations (BSs) is a major performance impairment, is investigated in this correspondence principally from a multi-user transmitter preprocessing (MUTP) perspective over correlated frequency-selective channels. Transmitter preprocessing at the mobile station is formulated by singular value decomposition (SVD) of only the individual user’s channel state information (CSI) while post-processing exploits all the users’ CSI using BS cooperation. Our simulations show that this approach completely removes CCI and also outperforms classical zero-forcing multi-user detection, which has about the same detection complexity. Further, the results show that our system, coupled with the optimal ‘water-filling’ strategy for power allocation, results in higher capacity than the maximum signal-to-noise ratio strategy. Furthermore, SVD-based MUTP outperforms dirty paper coding and Tomlinson–Harashima precoding in terms of achievable symbol error rate.
Graphical abstractFigure optionsDownload full-size imageDownload as PowerPoint slideHighlights► Performance of SVD aided UL multi-cell MU-MIMO with MUTP is investigated. ► BER performance over typical urban and vehicular channel models of LTE is presented. ► The achievable capacity in this setting is analyzed for two power allocation regimes.