| Future mobile communications system needs to provide various wireless data services across time and places. Multi-antenna system has attracted intensive re-search interests because it can promisingly meet the high date-rate requirements. As one kind of multi-antenna system, distributed antenna system (DAS) can im-prove the coverage and transmission capacity over limited spectrum, which has a brilliant application prospect. In the DAS, multiple remote antenna units (RAUs) are connected to the central processor (CP) via optical backhaul that forms a virtual cooperative network to acquire the space diversity gain. The transmit power can be saved while the reliability and effectiveness of information transmission can be improved, which becomes the green communication advocated in the modern soci-ety. On the other hand, massive multiple-input multiple-output (Massive MIMO) has unique advantages on interference control, spectrum efficiency, energy efficiency and etc., so it is the one of the key techniques of the fifth-generation (5G) mobile communication system. As the combination of Massive MIMO and DAS, massive DAS connects hundreds of antennas separated in a cell, which can reduce the aver-age distance between transceivers, and guarantee the fairness among all the users. Besides, massive DAS can be flexibly configured or implemented, so it can satisfy the stringent requirements.For the DAS, the majority of existing researches have treated the interference plus noise as a Gaussian random variable according to the central limit theorem (CLT). However, this analytical method is feasible only when there are tremendous amount of interference signals. When there are limited number of interfering sig-nals, the system performance analysed by this method deviates greatly from the real results. Further, the in-phase and the quadrature components of the higher-order modulated signals are correlated over Nakagami-m fading channels, which has not been studied for the DAS. For the communications system with strict accura-cy requirement, we need to efficiently and accurately predict the performance by resorting to the more suitable methods and mathematical tools.The co-channel interference (CCI) can severely affect the system performance. How to suppress or mitigate the CCI is an important issue of multicell DAS. Massive DAS contains a large number of radio frequency chains, which may restrict its application due to the expensive hardware cost. So, we can choose the suitable RAUs for the information transfer to reduce the signal overhead and hardware cost without losing diversity and multiplexing gain. In addition, since the antenna correlation exists in the massive MIMO system, the correlated channel fading should be properly modeled to study the diversity reception.To solve the above mentioned issues existing in the multicell and massive DAS, we have made deep and extensive studies, and the main contributions and innovations of the dissertation are listed as follows:1. We study the downlink performance of the multicell DAS. In order to reduce the CCI and transmit power, we design an antenna selection transmission scheme based on the maximal simultaneously signal-to-interference-plus-noise ratio (SINR). Firstly, the Gaussian and Q function approximation (GQA) are adopted in which the interference plus noise are treated as Gaussian distribut-ed with fixed variance. Based on these assumptions, we derive the closed-form error rate expression. Secondly, the interfering signal is assumed to be a ran-dom variable with random variance, and the characteristic function of the interfering signal is adopted to analyze the precise error rate performance over Nakagami-m fading channels. For the case of Rayleigh fading, the precise error rate expression is derived in closed-form. Compared with the GQA method, the precise error rate is not only determined by the desired signal, but it al-so depends on the channel power gain of the CCI. Numerical and simulation results verify that the proposed analytical method can quickly and accurately calculate the system performance. (The contributions correspond to Chapter 3 and the first published paper listed at the end of the dissertation.)2. The distributed antenna transmission schemes, and the error rate performance with dual-channel reception of quadrature phase shift keying (QPSK) are s-tudied in multicell environment. The RAU selection schemes are proposed based on the best channel quality and minimal link distance, respectively. The dependency between the in-phase and the quadrature components of the interfering signals are analyzed over Nakagami-m fading channels. We also treat the CCI as a random variable, and derive the precise error rate of the DAS with dual-channel reception. For the case of Rayleigh distribution, by omitting the dependence of the in-phase and the quadrature components of the CCI, we derive the precise error rate in closed-form. Numerical and simu-lation results verify the correctness of our theoretical analysis, and compares the performance of different transmission schemes. (The contributions corre-spond to Chapter 4 and the second published paper listed at the end of the dissertation.)3. For the wireless energy transfer (WET), we design a massive DAS model based on the multiple-circle layout, propose a reasonable energy and information transmission scheme, and derive the asymptotic throughput. This model sim-plifies the theoretical analysis and make the wireless-powered technique fea-sible. In the downlink WET phase, users select the RAU with the shortest distance for the energy harvesting to save the signal overhead and improve the EH efficiency. In the uplink information transmission phase, the transmis-sion probability of each user is analyzed, and the zero-forcing detection (ZFD) is employed at the receiver to separate the data among different users. By exploiting the advantages of massive MIMO, such as the channel hardening and the asymptotic orthogonality of user channels, the asymptotic throughput of a user is derived in closed-form based on the random matrix theory. The time fraction used for energy harvesting is optimized through maximizing the asymptotic throughput. (The contributions correspond to Chapter 5 and the fifth paper listed at the end of the dissertation.)4. We design a generalized correlated Weibull fading model with multivariate Gaussian random variables to study the diversity receptions of massively dis-tributed antenna cluster. Firstly, in the high regime of signal-to-noise ratio (SNR), the asymptotic outage probability and the error rate expressions are derived in closed-form for the selection combining (SC), equal gain combin-ing (EGC), and maximal ratio combining (MRC), respectively. Secondly, we derive the precise outage probability and the error rate for the selection com-bining in infinite series form with the transformation of the cumulative dis- tribution function of the average branch SNR, and we discuss the convergent of the infinite series. Numerical and simulation results show that the derived precise outage and error rate can improve the correctness of performance es-timation, which is meaningful for the optimization of the massive DAS. (The contributions correspond to Chapter 6, as well as the third and the fourth published papers listed at the end of the dissertation.)... |
PDF Full Text Request