| Within the family of WiMAX standards, IEEE 802.16e is the only one that can support both the non-line of sight (NLOS) and mobile broadband wireless communications. It can meet the requirements of high transmission rates and mobility of the future wireless communications. On the other hand, smart antenna technology is an important part of the future communications because of its good ability to improve the communication quality, to increase the system capacity and to expand the cell coverage. In the IEEE 802.16e system, smart antenna, defined as Advanced Antenna System (AAS), is supported as an optional part to improve the system performance. This work focused on implementing AAS in the uplink of IEEE 802.16e OFDMA system. The main work is as follows:Firstly, the signal model of antenna array in the condition of multi-user communication and multi-path fading was analyzed. The smart antenna technology was also classified from the point of view of the system framework, the signal bandwidth and the adaptive algorithms.My understanding of the IEEE 802.16e OFDMA system, including that of the uplink transmission process and that of the related specifications for AAS, was outlined in Chapter Third. A basic uplink employing the PUSC sub-carrier permutation was constructed. Through the analysis of the channel conditions in the OFDMA system with AAS, a multi-path Rayleigh fading channel and a space-time vector channel were modeled.At the beginning of the fourth chapter, the theories and the application constraints of several conventional Direction of Arrival (DoA) estimation algorithms were summarized. Utilizing the correlation characteristics of the preambles in AAS, a new DoA estimation method that is particularly suitable for the multi-user and multi-path transmission was proposed. It has good estimation precision, low computational complexity and can estimate the larger number of DoAs than that of the array antennas.Regarding to beamforming, this work put emphasis on the switched-beam method. The general switched-beam algorithm in the IEEE 802.16e system was simulated, and its performance bottleneck was analyzed. Then a modified switched-beam algorithm based on the DoA information was proposed by incorporating the idea of multi-user detection. Simulation results showed that this modified method could provide better system performance. When the users are far away in space, the frequency efficiency is increased greatly while the final whole capacity gain is less than the number of beams.By investigating into the modified switched-beam algorithm, an adaptive beamforming method using the generalized inverse of array response matrix was described at the end of Chaper Four. Its effect is equal to that of the modified switched-beam systems. The LCMV beamforming was also studied as a comparison. Simulation results showed that it performs worse than the modified switched-beam method, because we use the correlation matrix of the received signals instead of that of the interference plus noise to calculate the weight in the simulation.Finally, the scheduling problem on the MAC layer for OFDM/SDMA systems was studied. With the simplified OFDM/SDMA system model and some assumptions of the physical layer, two scheduling schemes were proposed. The first one only considers increasing the system capacity, while the second onefurther considers guaranteeing the fairness among users by borrowing the thought of proportional fair scheduling. Both the two algorithms sequentially decide users served on a sub-carrier, so they are heuristic schemes and have low complexity.Although I have been devoting myself to the research of smart antenna that is applied in IEEE 802.16e OFDMA system, this dissertation is merely a preliminary work. There is much other work to do in the future, including the implementation of SDMA using AMC sub-carrier permutation, the simulation of the AAS ranging process, the research on new DoA estimation algorithm with higher accuracy, the investigation into other adaptive beamforming algorithms suitable for OFDMA systems and the study on the more practical scheduling algorithms for AAS.This work is supported by the Natural Science Foundation of China (60572115), the Natural Science Foundation of Zhejiang Province (Z104252) and Key Science & Technology Projects of Zhejiang Province (2005C21005). |
PDF Full Text Request