|
|
以.pdf格式下载本文 This file type includes high resolution graphics and schematics when applicable. |
你们中的一些人可能还记得经典的音乐the 1960s and ‘70s titled “It’s a Bird...It’s a Plane…It’s Superman!” As the title suggests, it’s difficult to definitively categorize a flying object merely by observing it at a distance. In fact, this is particularly true when objects such as birds, airplanes, and superheroes are similar in both form and function.
二十年前,一个类似的可能行:“It’s a spectrum analyzer…It’s a software-defined radio (SDR)…It’s a PXI system!” Back then, such a line would have been accurate. Even at a distance, spectrum analyzers were clearly distinct from SDRs—and PXI was still as mythical as Superman. Today, however, each of these “instruments” are often virtually indistinguishable, since a PXI system can double as both an SDR and a spectrum analyzer.
The growing software capability of PXI instrumentation has blurred these once clear distinctions. More specifically, the ability for engineers to deploy intense signal-processing algorithms on PXI systems is allowing for the use of PXI in an ever-expanding range of applications—including those that we might classify as SDR.
As a result, engineers not only use PXI to replace their traditional spectrum analyzer; leading researchers use PXI to prototype advanced electronic systems ranging from passive radars to 5G test beds. In many of these applications, the requirements for high-performance transceivers combined with the need for intense signal-processing algorithms have ultimately led engineers to use FPGA technology in PXI.
Using FPGAs for Signal Processing
One of the unique characteristics of FPGAs is the ability to execute computationally intense and increasingly parallel signal-processing algorithms in real-time. In fact, today’s FPGA technology can deliver higher computational performance, and at lower cost, than digital signal processors (DSPs). Thus, FPGAs—often augmented with graphical programming units (GPUs)—have become a central element of advanced system design applications.
For example, in Figure 1, observe the difference in performance between state-of-the-art DSPs and FPGAs. According to the figure, FPGAs significantly outperform DSPs in terms of millions of multiple-accumulate operations (MMACs) per second. In addition, the FPGA holds the advantage both in MMACs per dollar, per chip, and per Watt of power.
随着计算能力的进步，FPGA技术的第二个关键进步是高级合成（HLS）软件的开发。从历史上看，大多数工程师使用传统的基于文本的方法（例如Verilog或VHDL）对FPGA进行了编程。对于这两种语言中的任何一种，开发计算密集的信号处理算法都是相对繁琐的，并且需要详细了解编程语言。此外，许多应用程序需要集成硬件I/O，从而为FPGA软件增加了更复杂的性能。
仪器中的新边界之一是高级抽象软件不仅能够合成FPGA IP，还可以简化IP与硬件I/O的集成。例如，图2比较了从ADC获取单个模拟数据样本的所需代码，然后将该样本写入直接内存访问（DMA）FIFO。如图所示，使用LabView FPGA的左侧代码比右侧的VHDL代码要简单得多。
By solving these challenges, our industry is enabling engineers to use instrumentation platforms in a much wider range of applications. For many engineers, the ability to execute custom, advanced signal-processing algorithms on the FPGA of an instrument eliminates the need for them to design additional hardware.
被动雷达
PXI中受益于用户可编程的FPGA受益的应用程序是无源雷达研究(Fig. 3). For example,SELEX系统Integratiin Rome, Italy uses你开发高级无源雷达系统的LabView FPGA软件。被动雷达是一种独特的应用程序，因为与传统雷达不同，该雷达使用板载“照明器”来产生刺激信号，被动雷达依赖于环境照明器，例如蜂窝塔，卫星传输和广播站。被动雷达的主要目的是通过小心地将入射信号与反射分开来测量信号的到达方向。
Prototyping a passive radar system introduces technical challenges both in the hardware and software implementations. On the hardware side, passive radar systems require tight phase synchronization between large numbers of receiver channels along with high-throughput data movement to a central processing unit. Due to these requirements, modular instrumentation platforms such as PXI are an ideal solution. That’s because PXI is inherently designed to handle both the synchronization and throughput requirements.
除了硬件外，诸如无源雷达之类的应用程序还需要强烈的信号处理，典型的算法包括多个通道之间的过滤和互相关。鉴于所需信号处理的大量并行性和复杂性，FPGA是一项关键技术。对于SELEX的团队，使用FPGA的路径涉及NI的LabView FPGA工具来开发自己的被动雷达IP，并将PXI系统部署为现场的被动雷达。
结论
如果您对RF信号分析器的想法是仅说明频谱显示的屏幕，那么您对仪器的期望太小了。借助不断发展的仪器平台，例如PXI和软件，使仪器的FPGA可以访问程序，如今的仪器正在经历其功能的急剧增长。下次您考虑设计自己的硬件以解决信号处理挑战时，请考虑使用仪器。您可能会对它可以做什么感到惊讶。
