Why Digital Beamforming Is Useful for Radar

In this video, we’re going to talk about digital beamforming for radar systems. Now, before we begin, I want to remind you that we did a whole series already on the basics of radar. In that series we talk about how we can use radar to determine range, range rate, and angle of arrival of a target, and even multiple targets. And we did it using an array of antenna elements.

We also have videos talking about the basics of digital beamforming where we covered how we can use sensor arrays with multiple RF chains to have more control over the shape of the beam in ways that maximize signal and minimize interference. Plus, it also allows for multiple independently steered beams from the same array.

Now, in this video I want to discuss how we can take advantage of those aspects of digital beamforming to improve radar systems. That is, what benefits can we get out of having a radar system that has the ability to generate multiple beams of various different shapes? So, I hope you stick around for it. I’m Brian, and welcome to a MATLAB Tech Talk.

Imagine you have a ground-based radar system and you want to observe objects within a designated airspace.

For example, think about an airport radar system. Part of its job is to track cooperative objects like expected airplanes that take off and land at the airport. Part of its job is also to search for uncooperative objects as well like drones that are flying illegally within the airspace. And, further more, part of its job is monitoring the weather for approaching storms and gust fronts. Now, these functions of tracking, searching, and monitoring all might be done with separate radar systems, or in some cases they might be done from the same system with a multi-function radar system. But in either case, digital beamforming can improve the way the radar accomplishes these functions.

To understand how, let’s take a step back. Radar uses radio frequencies to sense objects from a distance. In a typical case, a transmit antenna sends out a signal, that signal reflects off an object and returns to a receive antenna.

Now, in its most basic form, a radar could send out a signal equally in all directions, and if there is a reflective object near by, then after some time, the radar would receive an echo signal. The presence of an echo means that there is an object in the vicinity.

Of course, this doesn’t tell us a whole lot about the object, like where it is or if multiple objects are at the same distance. So, to answer these questions, we turn to a directional antenna. This is an antenna that radiates more power or receives more power in certain directions over others. And as we talked about in previous MATLAB Tech Talks, we can use an array of antennas to get produce a more directed beam. Plus, by adjusting the phase to each element in the array, we have the ability to steer the beam. This is a phased array.

So, with a linear array like I’m showing here, we can adjust phase to steer the beam back and forth and search for objects. When we get a detection, we can know approximately where it is in the sky, and if objects are farther away from each other than the resolution of the radar, then we can distinguish them.

Of course, this is just a 1-dimensional scan back and forth. But for the airport radar, we want to scan across azimuth and elevation. So, instead of a linear phased array, we can use a 2-dimensional planar phased array.

In this way, we can steer the beam to a specific azimuth and elevation, dwell there long enough to determine if there is a detection, and then move over to a new area of the sky and dwell there. And, by scanning through the entire airspace, we can keep track of where detections are made, and therefore where the objects are. And then by revisiting the detection later, and noting how they have moved, we can track the object.

Now a phased array is all well and good, but there are some problems that a fixed beam shape like this can’t easily address, even if that beam is steerable. And I want to talk about two of them in this video. The first is related to this simultaneous track and search problem that we just walked through and the second is in regards to dealing with other interferences and radar scatters in the environment. Both of these, can be addressed with digital beamforming.

Digital beamformers, just like phased arrays, use an array of sensor elements. In the case of radar, these are an array of antennas. The difference between phased arrays and digital beam formers is that with the former, we only have a single RF chain, which means each antenna is fed the same signal and the only variation between them is their phase, and their gain.

However, with digital beamformers, we have multiple RF chains feeding the array. This means that we have full control over the phase, gain and signal shape for each RF chain. And as we showed in the beamformer MATLAB Tech Talk, this allows for a wider range of beam shapes and multiple beams from the same array. So, with digital beamformers we can switch between different beam shapes, where they’re pointed, and how many there are nearly instantaneously.

Now, why is this important for radar? Well, let’s go back to the search and track problem.

The accuracy of a detection is a function of the resolution of the radar, which itself is a function of beam width. So, if we want to know more accurately where the objects are, we could use a narrower beam. However, if we use a phased array with a narrow beam, it takes more time to scan the sky since each dwell period is looking at a smaller overall angular space. Therefore, the revisit time between observations is longer and so a fast object might be difficult to track since you might only get one or two detections while it’s in the airspace.

Now, if we use a digital beamformer instead, we can set the array weights such that we have a really wide beam when searching a large volume, and then when a detection is made, we change the weights to create a more narrow beam that scans the area around the detection in order to produce a more accurate measurement.

And so a digital beamformer takes advantage of both being able to scan the sky quickly and have high resolution. And if we are tracking an object, the radar could visit that object more frequently with a really directed high bandwidth beam so that we can update the track more often and lower our uncertainty in where it is and where it’s going.

In this way, we have the ability to switch between a volume search task that scans the entire area quickly, a cued search task, that will produce better accuracy for a detected object, and a tracking task, where a tracked object is revisited more frequently. This a multifunction phased array radar system or MPAR.

But, using the multi-beam capabilities of a digital beamformer means that a radar could also do all three of these functions in parallel.

Now, my drawings can’t really get across how cool and complex multifunction radar can be. So, instead let me show you this much better animation. This comes from a MATLAB example on multi beam radar, which I’ve linked to below.

Up at the top is the range of elevation and azimuth angles that the radar is scanning within and it is doing it by breaking up the area into a grid of these blue circles and then dwelling at each of those grid points to see if a detection is made. Each filled in blue circle is where the volume search task is currently looking and you can see how that moves across the scan area to search for new targets.

Occasionally, an object is detected and so the radar is also performing a higher accuracy cued search in those regions to get a better fix on the object.

And, the radar is performing the occasional track update looks, and track confirmation looks. And it’s doing all of this simultaneously.

Well, it is but there is a catch. The array only has a limited amount of bandwidth, power-aperture product, and time resources. So, while the radar system is performing each of these tasks in parallel, they each require some of the limited resources and therefore, they must compete with each other for access to them. And this produces a rather interesting trade between these tasks over time as the radar has to balance volume searching, cued searching, and track confirmations and updates.

It’s kind of mesmerizing to watch how this radar manages each of these tasks. And on this note, if you want to play around with developing your own multi beam radar plan, then I recommend you check out the MATLAB example. This example covers setting up the requested tasks and scheduling those tasks. However, it also walks through a lot more. Specifically, it covers the system resources and how to allocate them between the different tasks. It also simulates the environment with multiple maneuvering targets and it shows the impact that false positives have on the array resource management. It’s all really cool and I think it’ll help you visualize some of the concepts we talked about here.

Ok, so this multifunction radar system is one benefit of digital beamforming. The other benefit that I want to mention in this video is in regards to shaping the beam in a way that maximizes the signal to noise plus interference ratio.

To understand that, let’s use this airport radar example again. Here, my drawing of this beam is a bit misleading since it looks like the radar has gain only in the direction of the object. However, as we know from the radar series videos, there are side lobes that are a byproduct of the array interference patterns and these lobes could have significant gain, both transmit and receive, in directions where we actually want to minimize gain.

For example, there could be interferences in the area like other nearby radar, maybe intentional jammers, or other sources of radio wave interferences that we want to avoid picking up on the airport radar. Or on the flip side, there may be other sensors and radars that are sensitive to our radiation and so to be a good neighbor, we may choose to not radiate in their direction.

If you recall from the video we did on beamforming, we have a lot of control over the shape of the beam pattern simply by adjusting the phase and gain independently to each element. We don’t need to have a symmetric beam pattern, but can tweak it and morph it in a way that maintains a main lobe - even if it’s a funky shaped one, but shrinks some of the side lobes, or places a null in a chosen direction.

In this way, the shape of the beam could always be changing between dwell periods, even for the same function. For example, if all we’re doing is searching with a wide beam, then for each grid point, the beam could be formed in a way that moves the main beam to the new location but maintains the null in the direction of the stationary interference.

In fact, in the video on radar for communication, I showed you a particular adaptive algorithm called MVDR which maximizes the signal to noise plus interference ratio in an environment with interferences. That kind of algorithm could also be used for radar applications as well. However, there are other algorithms that you can use to help optimize the array weights beyond just MVDR.

One such algorithm is called linear-constraint minimum variance or LCMV and I’ve left a link to a MATLAB example of it if you’d like to try it out yourself.

Alright, like all Tech Talks, this was a fast introduction into a topic. But what I hope you take away from this video is how digital beamformers can open up more functions and more capabilities for radar systems. We can use them to work around interferences, we can use them to avoid being the cause of interferences on other systems, and we can perform multiple radar functions simultaneously, from the same array.

So, this is where I’m going to leave this video. If you don’t want to miss any future Tech Talk videos don’t forget to subscribe to this channel. Also, if you want to check out my channel control system lectures you can find more control theory topics there as well. Thanks for watching and I’ll see you next time.