Research On Inter-Cell Interference Cancellation In LTE System

Inter-Cell Interference (ICI) is an inherent problem which is critical to system capacity. Cell-edge users in cellular systems suffer from the interferences of adjacent cells and their throughputs are typically very low. Inter-cell interference cancellation technique can eliminate co-channel interferences and enhance cell-edge performance. Orthogonal Frequency Division Multiple Access (OFDMA) is chosen as multiple-access scheme for downlink of LTE due to its high spectral efficiency and some other superiorities. Different from Code Division Multiple Access (CDMA), OFDMA can not mitigate interferences by using spread. Combining with Probabilistic Data Association (PDA) and Interleave Division Multiple Access (IDMA), iterative interference cancellation is mainly researched in LTE. PDA is capable to perform the iterative multi-user (cell) detection and IDMA can make a substantial performance gain for suppressing the inter-cell interference.Firstly, a MIMO system with multi-cell interferences is established and three types of existing PDA detection algorithms are considered in this thesis, i.e. Minimum Mean Square Error Detection (MMSED) algorithm, mean Soft Decision Detection (SDD) algorithm and Maximum-A-Posteriori Detection (MAPD) algorithm. The MMSED and the SDD algorithms are Soft Interference Cancellation Detection (SoICD) algorithms, while the MAPD algorithm is generalized Turbo detection algorithm. The simulation results demonstrate that the MMSED algorithm is worse than the SDD algorithm in complexity and performance even though they both are based on the same detection principle. The advantages of the SDD algorithm are low complexity, good stability and excellent performance on dealing with weak interferences, whose disadvantage is a failure capacity of coping with strong interferences. The advantage of the MAPD algorithm is excellent performance on dealing with strong interferences, whose disadvantages are low complexity and failure stability of coping with weak interferences.Secondly, in order to attain their mutual complements, an improved PDA detection algorithm is proposed, i.e. grouping MAPSDD algorithm which integrates the SDD and MAPD algorithms very well. In terms of the differences of interferences intensities, three different kinds of schemes are proposed. The simulation results demonstrate that it is feasible and flexible to utilize different detection schemes for different intensity interferences. According to the analysis of simulation results, a synthesized detection scheme is obtained as follows. If the interference signals are strong, they would be divided into the same group with desired signals and detected by the grouping MAPSDD algorithm. If the interference signals are middle, they would be detected by the SDD algorithm. And if the interference signals are weak, they would be processed by gaussian approximation. The synthesized detection scheme not only improves performance but also lowers complexity.Finally, two kinds of MIMO-OFDM systems with multi-cell interferences are established and the performances of algorithms in the three systems consisting of MIMO, BLAST-OFDM and STBC-OFDM are simulated and analyzed. The simulation results demonstrate that the schemes of the improved algorithm exhibit similar properties and apply to MIMO-OFDM system as well.