Optimum Diversity Combining and Equalization for Mobile Radio with Dispersive Fading and Co-channel Interference

• Main Authors: Meng-Chao Chang, 張孟超
• Other Authors: Jeng-Kuang Hwang
• Format: Others
• Language: en_US
• Online Access: http://ndltd.ncl.edu.tw/handle/46793375212041447345

碩士 === 元智大學 === 電資與資訊工程研究所 === 86 === In this thesis, we consider the problem of optimum MMSE receiver for mobile communication system using spatial diversity to combat multi-path fading channel and strong co-channel interference (CCI). Aware of the fact that original receiver proposed by Balaban performs poor in the presence of strong CCI, we derive the optimum receiver for the CCI case again and propose a new structure of optimum combining / equalization. According to this new structure of optimum combiner / equalizer, the channel impulse responses (CIR) for both the desired and interference signal are required. In practice, we assume that the training sequence of the CCI is also known, corresponding to a multi-user detection problem. And then, we obtain this CIR by channel estimation. Moreover, for tracking the time-varying channel, the system parameters should be updated several times during a data frame. We adopt the strategy of interpolating a set of estimated CIR values to determine these time-varying CIR values. Additionally, the structure of the MMSE receiver is linear and parallel, which can find applications in multi-user detection or signal separation problems. And, with this optimum receiver, several signals can use the same time slot simultaneously. Thus, system capacity can be increased significantly. With computer simulation, we first show the limitation of the original receiver, and then we demonstrate this new structure indeed attain significant performance improvement under the situation of signal fading 、inter-symbol interference and especially strong CCI. And the simulation results also reveal some relation between system performance and some system parameters. This provides some guidelines for system design.