摘要
This article puts forward two novel user-grouping algorithms for grouped multi-carrier （MC）-code division multiple access （CDMA） systems. As is well known, the adaptive assignment for user-grouping plays an important role for link quality of multi-access transmissions. In the study, the capacity-maximizing problem of user-grouping is formulated. By using the Kuhn-Tucker condition, the optimal criterion is deduced and found to have a similar form with signal to noise plus interference （SINR）. However SINR includes the signal power that can only be determined after user-grouping. Therefore the optimal criterion will lead to an impractical application. To deal with it, the user＇s equivalent SINR for minimum mean square error （MMSE） detector is proposed and served as a suboptimal assignment criterion, based on which two kinds of user-grouping algorithms are proposed. In the algorithms, only partial channel information is needed at the base station, which saves a large part of the bandwidth occupied by feedback information. Computer simulations have evaluated an excellent performance of the proposed algorithms at both link quality and data rate. Meanwhile, the proposed algorithms have lower implementation complexity for practical reality.
This article puts forward two novel user-grouping algorithms for grouped multi-carrier （MC）-code division multiple access （CDMA） systems. As is well known, the adaptive assignment for user-grouping plays an important role for link quality of multi-access transmissions. In the study, the capacity-maximizing problem of user-grouping is formulated. By using the Kuhn-Tucker condition, the optimal criterion is deduced and found to have a similar form with signal to noise plus interference （SINR）. However SINR includes the signal power that can only be determined after user-grouping. Therefore the optimal criterion will lead to an impractical application. To deal with it, the user＇s equivalent SINR for minimum mean square error （MMSE） detector is proposed and served as a suboptimal assignment criterion, based on which two kinds of user-grouping algorithms are proposed. In the algorithms, only partial channel information is needed at the base station, which saves a large part of the bandwidth occupied by feedback information. Computer simulations have evaluated an excellent performance of the proposed algorithms at both link quality and data rate. Meanwhile, the proposed algorithms have lower implementation complexity for practical reality.
基金
the DoCoMo Beijing Labs Co. Ltd., Program for New Century Excellent Talents in Beijing University of Posts and Telecommunications (04-0112)
