Hardware-accelerated prototyping, characterization and optimization of multiple-antenna communication transmitters and receivers

多天线通信发射机和接收机的硬件加速原型设计、表征和优化

• 批准号: 105567-2011
• 负责人: Cockburn, Bruce
• 金额: $ 1.6万
• 依托单位: University of Alberta
• 依托单位国家: 加拿大
• 项目类别: Discovery Grants Program - Individual
• 财政年份: 2011
• 资助国家: 加拿大
• 起止时间: 2011-01-01 至 2012-12-31
• 项目状态: 已结题
• 来源: https://www.nserc-crsng.gc.ca/ase-oro/Details-Detailles_eng.asp?id=475326
• 关键词: Hardware; accelerated; prototyping; characterization; optimization

The main objective of the proposed research programme is to develop, extend and demonstrate a methodology for using field-programmable gate arrays (FPGAs) to greatly accelerate the simulation of wireless communication systems. The designers of multiple-antenna wireless communication systems face a rapidly escalating challenge where the designs of proposed transmitters and receivers must be verified and optimized over a wide range of posssible design configurations and wireless channel conditions. The performance of each candidate design must be verified using long-running simulations with pseudorandomly generated data communicated over realistic but complex models of the wireless channel. To ensure statistically meaningful bit and and symbol error rate (BER and SER) measurements, the simulated communication systems must process an extremely large number of transmitted data bits. This situation is made worse by the use of powerful error correcting codes, such as Turbo codes, which further reduce the effective BER and SER for the same channel conditions and hence lengthen the required simulations. Consequently conventional software-based simulations can easily take many days and even weeks of time on even the fastest computer workstations.

拟议的研究计划的主要目的是开发，扩展和演示一种使用现场可编程栅极阵列（FPGA）的方法，以极大地加速无线通信系统的模拟。 多 - 安德滕纳无线通信系统的设计师面临着迅速升级的挑战，在这些挑战中，必须在广泛的可使用设计配置和无线通道条件下验证和接收器的设计。 必须使用长期运行的仿真来验证每个候选设计的性能，并具有通过无线通道的现实但复杂模型传达的伪态生成的数据。 为了确保统计上有意义的位和符号错误率（BER和SER）测量，模拟通信系统必须处理大量的传输数据位。 通过使用强大的错误校正代码（例如Turbo代码），这种情况使这种情况变得更糟，从而进一步降低了在相同的频道条件下的有效BER和SER，从而延长了所需的模拟。 因此，在最快的计算机工作站上，基于软件的传统模拟很容易花费数天甚至数周。