【作者】 周江；

【导师】 王文博；

【作者基本信息】 北京邮电大学 ， 通信与信息系统， 2013， 硕士

【摘要】 随着移动互联网时代的到来,高清电视、视频会议、在线游戏等分组业务需求日益增长,这都要求未来移动通信能提供更高的传输速率,更广的信号覆盖,更稳定的传输质量。3GPP长期演进(LTE)作为第三代无线通信系统的演进,相比于3G系统LTE支持的系统带宽更大、带宽配置更灵活、上下行峰值速率更高、网络结构更扁平、接入时延更小。LTE带来高峰值速率的同时也将导致基带数字信号处理的高复杂度。通用处理器(GPP),以其强大的数据处理能力和灵活性,开始逐渐应用到高密集度数据处理的软件无线电中。本文研究的重点是设计与实现LTE下行接收的并行策略,以及均衡检测模块。理论方而主要研究均衡检测技术以对抗多径衰落、噪声以及干扰等带来的影响。实现方面主要研究优化策略以达到实时信号处理的高吞吐要求。通过利用多核处理器的多核以及寄存器等优势,在大粒度上实现多线程的并行计算,同时利用SIMD指令实现细粒度的指令并行,从而减少运算耗时,实现高速率的吞吐要求。此外本文还重点研究了均衡检测算法,在SISO、SIMO、MIMO三种传输模式下进行了均衡算法仿真,通过折中算法复杂度与误码率性能,最终选择合适的均衡算法在GPP上实现。本论文研究使用的SSE指令优化效果显著,对于在通用处理器上进行实时信号处理软件的开发有很大借鉴意义。更多还原

【Abstract】 With the arrival of the eva of mobile Internet, the damand of the packet services such as high-definition TV, video conferencing and online games grows faster and faster. And those growing demand requires the future mobile communication system to provide higher transmission rates, broader coverage and more stable transmission quality. As the evolution of the third generation wireless communication system,3GPP Long Term Evolution (LTE) supports a greater system bandwidth, more flexible bandwidth configuration, higher peak data rate of the uplink and downlink, more flat the network structure and less access delay. LTE system which brings the high peak rate will also lead to the high complexity of the baseband digital signal processing.General-purpose processor, which is suitable for software radio processing because of its powerful processing capability and flexibility, has aroused industry-wide attention and exploration. This paper mainly focuses on design and implementation of LTE downlink receiver parallel strategy and MIMO detection based on GPP. Theoretical aspects of the paper research detection technology to solve the impact of multipath fading, noise and interference. Development aspects of the paper sturdy how to design the optimization strategies to achieve high throughput requirements of real-time signal processing. Through the use of multi-core processor, we can both take advantage of multithreaded parallel computing and SIMD instructions to reduce processing time and realize high throughput requirements. This paper also focuses on the detection algorithm with transmission modes including SISO, SIMO and MIMO. Finally this paper chooses the appropriate algorithm and implements it on GPP after comparing the algorithm complexity and BER performance. The optimizing effect is obvious by using SSE instruction and it is of great significance for real-time signal processing based on GPP.更多还原