Two efficient and low complexity multiuser scheduling algorithms are proposed for the uplink multi- ple-input multiple-output systems in this paper. Conventionally, the exhaustive search algorithm （ESA） can give the...Two efficient and low complexity multiuser scheduling algorithms are proposed for the uplink multi- ple-input multiple-output systems in this paper. Conventionally, the exhaustive search algorithm （ESA） can give the optimal performance; however, it is complexity prohibitive for practical implementation. Aiming at re- ducing the complexity while keeping the achievable sum rate performance, two heuristic algorithms are proposed for the multiuser scheduling problems： the improved genetic algorithm and simplified norm-based greedy algo- rithm. Moreover, we also consider the heterogeneity scenario where a modified grouping-based user selection al- gorithm is given to guarantee the user＇ s fairness. Specifically, the asymptotic behavior of the norm-based greed- y algorithm is given when each user is equipped with one antenna. Numerical examples demonstrate the superi- ority of our proposed schedulin~ and ~rouoin~ algorithms.展开更多
基金Sponsored by the Technology Specific Project(Grant No. 2010ZX03002-003-01)
文摘Two efficient and low complexity multiuser scheduling algorithms are proposed for the uplink multi- ple-input multiple-output systems in this paper. Conventionally, the exhaustive search algorithm （ESA） can give the optimal performance; however, it is complexity prohibitive for practical implementation. Aiming at re- ducing the complexity while keeping the achievable sum rate performance, two heuristic algorithms are proposed for the multiuser scheduling problems： the improved genetic algorithm and simplified norm-based greedy algo- rithm. Moreover, we also consider the heterogeneity scenario where a modified grouping-based user selection al- gorithm is given to guarantee the user＇ s fairness. Specifically, the asymptotic behavior of the norm-based greed- y algorithm is given when each user is equipped with one antenna. Numerical examples demonstrate the superi- ority of our proposed schedulin~ and ~rouoin~ algorithms.
