By Nancy Friedrich Industry solutions marketer for aerospace & defense at Keysight Technologies
Even if they only occur intermittently, interferers or other signal events can cause major problems in communications systems, making it necessary to detect, identify, and locate them in a timely manner.
In the real-world environment, the signal environment is unpredictable. Communications systems are deployed and tweaked if necessary to achieve expected performance levels. Yet problems still arise, putting communications at risk. In the worst cases, communications fails. More often, noise distortion or other issues occur. Elusive signals are often the cause of such issues, further complicating the problem because they are difficult to pinpoint, identify, and therefore eliminate. By working through the following considerations, you can figure out the best approach to tackle such surprises in your spectrum.
1. Determine the spectrum monitoring approach that best suits your needs.
Originally, spectrum monitoring used a swept heterodyne front end. It looked at one very narrow chunk of spectrum at a time, moving across the frequency band of interest. As a result, this approach limited the information gathered to the contents of the lens view. This partial view of the spectrum often failed to reveal problem signals. To overcome this issue, newer swept-mode analyzers use an increasingly wider stare. Yet the old approach still offers advantages. The swept heterodyne approach looks at a narrow chunk of spectrum, which contains less noise—making it possible to discern some lower-amplitude signals.
Spectrum monitoring requires constant tradeoffs between dynamic range, wider bandwidths, and sweep speeds. Just because a technology allows you to grab a gigahertz of spectrum, for example, it might not be the best approach. A larger and wider chunk of spectrum means more environmental noise. The noise floor rises as a result, limiting the selectivity or sensitivity of the front end. You will then struggle to pick out signals that might be miles away.
However, dynamic range is not an issue if the problem signal you are seeking is fairly high in amplitude and causing problems across the board. With a high-amplitude signal, you can simply grab a huge chunk of spectrum. The noise floor is not a concern, as the problem signal will show in that wide chunk of spectrum. In contrast, if a low-amplitude signal is causing problems, you can divide the spectrum into smaller steps. This approach reduces the noise floor, making the signal visible.
2. Need more processing power? Consider a Fast Fourier Transform (FFT) approach.
The process of detecting and identifying elusive signals becomes faster and more precise as technology evolves. Compared to their predecessors, current solutions look at a wider chunk of spectrum while reaching higher frequencies. Using an FFT approach with more processing power, you can realize a lower noise floor. An FFT is a series of mathematical equations used to emulate the frequency spectrum. A sufficiently large FFT reduces the spacing between each frequency point, lowering the resolution bandwidth and noise floor to provide high dynamic range and faster sweep speeds.
For example, with a 1 kHz resolution bandwidth (RBW), you can see signals spaced 1 kHz apart. Achieving a 1 kHz bandwidth across 1 GHz of stare bandwidth requires a huge FFT (100,000 points). You can increase the RBW to 1 MHz. However, all the signals in a 1 MHz band are combined into a single point on the spectrum. So instead of seeing 100,000 signals, you see only 100 signals. Larger FFTs can provide that narrower resolution bandwidth even with a wide bandwidth stare. To summarize, the lower the resolution bandwidth, the lower the noise floor and the more sensitive the front end.
To determine whether an FFT or other spectrum monitoring approach would be best, consider the following:
• How big of a chunk of spectrum can you look at instantaneously?
• How quickly can you step through that spectrum?
• What is your probability of intercept?
• Given the sweep speed, how quickly can you cover the area in the spectrum of interest?
3. Do an environmental scan or survey.
To detect and identify an elusive signal, it is key to begin by determining what signals are in the environment. Here, it is helpful to have already documented the signals you are broadcasting or using. Similarly, you should identify which channels are fine versus the ones that exhibit issues. Problems usually show up in a specific frequency or channel. You can use that information to focus your survey on that area of the spectrum.
You also can accomplish signal identification in the initial survey by looking at the frequency spectrum. If you know how the signal should appear, you can recognize a different-looking signal immediately. It might have a different modulation type or shape in the spectrum. Occasionally, such visual clues help easily discern the source of the signal—especially given the knowledge of the signals in the larger environment, thanks to the initial survey.
Note that you may not be able to see an interfering signal on a spectrum analyzer if it is riding right on top of the signal for which you are searching. When everything appears fine, but performance issues clearly exist, the next step is to deploy sensors around the area—for example, placing four or five sensors a mile or so apart. You can then perform the same survey again in that frequency band.
Sensor and geolocation data often reveal more details around what is occurring in the spectrum environment. For example, the data may show two locations for the same frequency, which makes it likely that one of them is causing the interference. Consider what is at each location, such as a tower or transmitter. Often, a location not documented as the source of any signals is very likely the source of interference. If these steps do not prove successful, you will have to physically go out to the point of interference to gather more information on the source.
4. Determine if you need to conduct a spectral search.
To maintain awareness of elusive or problem signals, users looking to perform spectrum monitoring generally use a survey or search approach or a combination of the two. Those coming to a new area to put up a tower, for example, commonly use the “spectral survey” discussed previously. With the search approach, in contrast, the engineer usually watches a specific signal because of information that is already known: who it belongs to or who is using it, for example. Search could focus on an individual’s cell phone number or devices in use. The user or agency performing the search preloads that information into a system and tells the system to find those elements. If the system detects that signal, it sends an immediate notification. The search approach is generally needed only if the elusive signal is a known or suspected threat.
Conclusion
Signals may be “elusive” because they are on the same frequency as the communications system, making them hard to detect with a spectrum analyzer. While nothing appears amiss on the screen, interference will likely affect audibility and clarity via noise or distortion. Other signal events are elusive because they rarely occur, maybe once a month or year. Yet they still can cause major problems, depending on their frequency. Take emergency services or early-warning systems, for example. Even if an interfering signal event occurs only once a year, it could negatively affect those mission-critical systems.
Engineers looking for elusive signals must consider everything from sweep speed to dynamic range, bandwidth, resolution, and environmental noise. An alternative to slowly sweeping across the spectrum is using the FFT approach to grab large blocks of data, do FFTs, and instantly obtain a wider chunk of spectrum. Whatever the approach, spectrum monitoring is a critical aspect of safeguarding spectrum and communications systems. Once you deploy radios and systems, your work is not done. You must pay attention to performance in the real-world environment.
Modern spectrum monitoring systems have moved from simple visual detection with a spectrum analyzer to automated detection, classification, identification, and location. Such systems send alerts if a signal looks distorted or an interfering signal appears in the area. As a result, you can act to eliminate the problem signal—even an elusive one—preventing or promptly terminating communications performance issues.
Nancy Friedrich is an Industry Solutions Marketer for Aerospace & Defense at Keysight Technologies. She joined Keysight after two decades working on engineering media brands, eventually serving as Executive Director of Content for a family of brands including Electronic Design, Microwaves & RF, and Machine Design. Nancy later served as Editor-in-Chief of Design News and Content Director for tradeshows including DesignCon, ESC, and the Smart Manufacturing shows.
