OFDMA 기반 하향링크 네트워크 시스템의 기지국간 간섭 신호의 통계적인 모델링 및 성능 개선 방안 연구

Abstract

An accurate statistical inter-cell interference (ICI) model is required in order to accurately assess and improve the performance of a downlink cellular orthogonal frequency division multiple access (OFDMA) network. First, the probability density function (pdf) of the ICI for downlink cellular OFDMA networks using QAM modulation under log-normal shadowing and multipath Rayleigh fading is derived. The derived pdf is observed to highly deviate from the Gaussian distribution, especially at moderate to small network loads. Based on the derived ICI statistics, the binary channel decoder performance can be significantly improved compared to the case when the traditional decoder optimized for the AWGN channel is employed. A simple suboptimum decoding algorithm that closely approximates the derived optimum decoder performance is also proposed. In order to investigate the statistics of the ICI and the performance of the maximum likelihood (ML) decoder under more general channel models, the results for Rayleigh fading are extended to include the case where the signals from the BSs undergo Ricean fading. The decision variables are observed to significantly deviate from the Gaussian distribution as with the results under Rayleigh fading. Also, it is observed that, though the ML decoder derived under Rayleigh fading is not optimum under Ricean fading, it offers performance very close to the optimum decoder across a very wide range of Ricean factors. This implies that receivers designed under the Rayleigh fading channel model are quite robust and should perform very close to the ML decoder under Ricean fading channels.Finally, in order to improve the network throughput of an OFDMA based network, employing coherent $M$-ary frequency shift keying (MFSK) modulation combined with $M$-ary channel code is considered. The performance of soft decision decoded, synchronous, OFDMA networks using MFSK with simplified system model in Rayleigh fading channels is analyzed under the assumption that the number of frequency hopping slots is sufficiently large. Based on this assumption, the joint pdf of the correlator outputs are derived which is used to compute the channel capacity and evaluate the soft decision decoding performance of the 3GPP Turbo and convolutional codes. Results indicate that soft decision decoding based on the derived pdf drastically outperforms traditional decoders based on the Gaussian approximation of the effect of the multiple access interference as with the results with downlink cellular OFDMA networks using QAM. The performance comparison between the OFDMA networks using QAM and coherent MFSK with either binary channel code or $M$-ary channel code is also presented and the results show that coherent MFSK combined with $M$-ary channel code gives increased network throughputs.