Multi-core Signal Processing for Advanced Wireless Devices

本研究的目的是探讨高数据速率无线通信与先进的无线设备的有效信号处理算法。便携式无线设备需要支持诸如具有高数据速率和计算要求的实时高分辨率视频和移动的游戏之类的应用。移动的计算设备现在使用多个计算核心。这些处理器在单位时间内执行与单核处理器相同的计算量，同时需要更少的功率。要有效地利用这类新的处理器，需要仔细管理内核之间交换的信息。当前无线设备的信号处理算法由顺序计算组成，其不能有效地映射到多个核上。我们将开发具有并行计算的新信号处理算法，这些算法在多核上有效地执行。将研究计算持续时间、并行化和最终误差性能之间的权衡。在当前的文献中，接收器在并行化方面的权衡没有得到很好的探讨。新的信号处理算法将使小型低功耗先进无线设备能够支持高可靠性、高数据速率的通信应用。

The objective of this research is to investigate efficient signal processing algorithms for high data rate wireless communications with advanced wireless devices. Portable wireless devices need to support applications such as real-time high resolution video and mobile gaming with high data rate and computational requirements. Mobile computing devices now use multiple computing cores. These processors perform the same amount of computation per unit time as single core processors while requiring much less power. Making effective use of this new class of processors requires careful management of the information exchanged between the cores. Current wireless devices' signal processing algorithms consist of sequential calculations which do not map efficiently onto multiple cores. We will develop new signal processing algorithms with parallel calculations which execute efficiently on multiple cores. Trade-offs between calculation duration, parallelization, and final error performance will be investigated. The receiver trade-offs in terms of parallelization are not well explored in the current literature. The new signal processing algorithms will enable small low power advanced wireless devices to support highly reliable, high data rate communications applications.