SBIR Phase I: Massively Parallel Protocols for Software-based Wireless Systems

这一小型企业创新研究（SBIR）第一阶段项目的更广泛/商业影响在于它有可能更快、更便宜地实现最先进的无线电通信系统。无线电系统是日常生活中必不可少的一部分，从娱乐到公共安全。但是，现代无线电系统旨在有效地使用有限的可用频率，开发和部署成本很高。高成本的一个主要原因是需要定制的硅芯片来进行将微弱的无线电信号转换为有用数据的处理。 该项目旨在使无线电系统的建造成本更低。该团队没有构建定制芯片，而是使用商品计算机和专门设计的软件，可以高速运行无线电处理任务。这种速度是通过分析无线电处理任务并将其转换为在具有许多单独计算核心的处理器上运行的软件的技术实现的。 经济影响是双重的：该技术可以降低现有系统的成本，如蜂窝LTE和5G基站，它也使新的应用成为可能，这些应用过于昂贵，无法从定制硬件构建。SBIR第一阶段项目旨在了解如何快速有效地开发现代无线电系统所需的处理。 该团队还试图解决通信协议中难以实现的功能，因为计算过于复杂，或者在生成最终输出之前需要检查太多数据。他们还将讨论改变协议以消除瓶颈的机会。该技术将测量关键无线电算法在商用计算硬件上运行的速度，以及在存储器之间移动数据等基本但非生产性任务上花费的时间。目标是定量估计纯软件实现的无线电可以发送或接收多少数据。最终，该团队将设计与处理器内核数量成比例的协议：两倍的处理器提供两倍的数据吞吐量。该奖项反映了NSF的法定使命，并通过使用基金会的智力价值和更广泛的影响审查标准进行评估，被认为值得支持。

The broader/commercial impact of this Small Business Innovation Research (SBIR) Phase I project lies in its potential to implement state-of-the-art radio communication systems faster and less expensively. Radio systems are an essential part of everyday life, serving roles from entertainment to public safety. But modern radio systems, designed to use the limited number of available frequencies efficiently, are expensive to develop and deploy. A major reason for the high cost is that custom silicon chips are needed to do the processing that converts a weak radio signal into useful data. This project aims to make radio systems much cheaper to build. Instead of building custom chips, the team uses commodity computers and specially designed software that can run the radio processing tasks at high speed. This speed is enabled by technology that analyzes radio processing tasks and turns them into software which runs on a processor with many individual computing cores. The economic impact is twofold: that technology can reduce the cost of existing systems, such as cellular LTE and 5G base stations, it also makes possible new applications which are too expensive to build from custom hardware.This SBIR Phase I project seeks to understand how to develop the processing needed in modern radio system quickly and efficiently. The team also seeks to address the features of communications protocols that are hard to implement because the computations are too complicated or too much data needs to be examined before the final output is generated. They will also address the opportunities to change the protocol to eliminate the bottlenecks. The technology will measure the how fast key radio algorithms run on commodity computing hardware and how much time is spent on essential, but not productive, tasks such as moving data between memories. The objective is a quantitative estimate of how much data can be transmitted or received by a radio implemented purely in software. Ultimately, the team will design protocols that scale with the number of processor cores: twice as many processors giving twice the data throughput.This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.