Per Vices: Flexible radios help autonomous systems interface

An exclusive interview with Brandon Malatest, co-founder and  COO of Per Vices 

What is your title and role?

I’m one of the founders of Per Vices. My friend and I started the company a long, long time ago. I’m a physicist. I attended the University of Waterloo, graduated with an honors degree in physics, and have been working with Per Vices for more than ten years. My role has shifted from doing the actual engineering work to running all the other elements of the company, because as we continue to grow, we can’t be doing everything and that was the logical break. So, I definitely have a technical background, but I’m not the one who’s designing the products anymore.

When was the company founded?

In 2006. Our first commercial product was in 2012. We specialize in designing high performance software-defined radios. These are full transceivers that are used across a very wide range of markets: spectrum monitoring, electronic warfare, MRI, radar, test and measurement markets, communications, radio links — you name it and we’ve done some work in that space. The whole idea behind software-defined radios is that they are very flexible systems. So, with the same hardware platform, you can change the software or firmware, and have it used for a completely different application.

Wonderful. By the way, I grew up among physicists. My father was a physicist for 60 years, the last 35 of which at Brookhaven National Laboratory on Long Island. My paternal grandmother was one of the first women in Europe to get a Ph.D. in physics and math and Enrico Fermi was one of her thesis advisors.

 Why is SDR important for autonomous systems?

There are eight major points that I’d like to hit on.

Flexibility. Software-defined radios are reconfigurable, which means that the same hardware platform can be used for many different communication protocols, across different radio bands with varying bandwidth. That makes them very flexible for interfacing with other types of systems. So, in terms of the autonomous systems, there are a number of different wireless devices that need to be interoperable.

SDRs are very flexible radio platforms. They’re designed to have very wide operating frequencies with varying bandwidths so as to replace what used to be done in hardware through dedicated DSP chips, replacing dedicated hardware with a software-based architecture.

From a flexibility standpoint, that means that software-defined radios usually use some type some type of DSP mechanism, like a field programmable gate array (FPGA). That allows the SDR itself to process all types of different signals across varying frequencies and manipulate them in different ways. Anything that’s wireless is basically converting an analog signal to a digital signal and then performing some action on that digital signal. SDRs do that in a different way. They still have the hardware that’s used for tuning, but on the software side the decoding and processing happens. In a traditional FM radio, you have everything done in hardware: it has a dedicated tuning block, a dedicated DSP that does the demodulation, and it spits out the audio. So, the same way that computers, way back in the day, were designed and built from the ground up with one sole purpose — whether it be word processing or running complex trigonometry.

But now, if you look at the utility of computers, it is the fact that you can run different software applications on them. So, the same idea is with SDR. Traditional radio devices were built from the ground up for a single application. Now, SDRs are like a modern-day computer, where you can do basically anything within that tuning frequency. Also, just like with a computer, you change the software and use it for different applications.

Flexibility is definitely one of the most important elements. It allows the user or the system integrator to have an SDR that can adapt to different communication standards and frequency bands. This flexibility is crucial for autonomous systems operating in dynamic environments, where communication requirements may change. So, if you’re suddenly needing to change from operating at 2.4 GHz to operating at 5 GHz due to spectrum congestion or something along those lines, an SDR can do that with the same hardware platform.

YellowScan’s bathymetric lidar product, the Navigator, mounted on the Noa from Acecore. This full waveform lidar system ensures continuity between underwater points and the surrounding terrain.(Image: YellowScan)

Adaptive communication. Because SDRs tune to various frequency bands and then all the decoding is done in software, they can support different communication standards with the same hardware platform. That enables autonomous systems to communicate effectively with various entities in the environment, such as sensors and additional equipment.

Spectrum awareness. We can call it smart SDRs, or SDRs where you can integrate with AI or you can do your own pre-programming on it. You can monitor different parts of the spectrum to see which is the least congested, so that you can have a clear frequency band of operation to communicate or to use that information for passing data to and from sensors or a command and control system or anything.

Going back to the hardware component, the FPGA onboard acts as a digital signal processing unit. So, SDRs have those onboard DSP units — usually, FPGAs — and that allows for such things as signal modulation and demodulation filtering, waveform generation. All this can be done in real time. It also allows for all the sensor data to be processed very effectively and quickly, with significantly reduced latencies.

Reduced hardware complexity. This relates to putting it all together. When you have the DSP unit and the flexible hardware platform, you don’t really need anything else. So, you can use that SDR to minimize the complexity associated with the overall system. When you’re using multiple disparate technologies, it does become challenging to make sure that they’re all integrated well with one another and work well together. With SDRs, you can really simplify that, that hardware complexity. Then, because SDRs are programmable and customizable, they can be used all the way from prototyping to production. By changing different software or firmware elements associated with the SDR, you can have it operate in different ways. So, if you want to prototype a different communication band that may work better for some environments, you can do that very easily without needing to re-spin new hardware upon new hardware upon new hardware, which gets to be costly and time-consuming.

Remote monitoring and control. SDRs can be set for what we’ll call static operation, where they will perform only one task, to prevent tampering. You can also set them to be updated over the air or through some type of network. So, again, the flexibility is quite significant and it’ll allow you to mitigate different challenges where some of the systems might not be able to be controlled or interfaced regularly from a hardware perspective. That can all be encrypted.

How does your module interface with autonomous platforms?

The analog connection — to external amplifiers or filters, if you need to have very clean use of a particular part of the spectrum — is via SMA connections, which are pretty standard in the industry. On the digital side, there are two different ports. One is an out-of-band communication interface. That’s just a 1 GB Ethernet port that is very common across the entire industry and is used for configuring the SDR. Again, that’s out of band, so you’re not causing any interference with the actual operation. The other one is the digital interface for sending data to and from the system. That can be done over a 10 G interface, a 40 G interface or a 100 G interface, depending on the platform, how much bandwidth you need, etc.

What is your market at this point?

We don’t discuss our customers specifically, because often they are either very large commercial entities that don’t like to have their names disclosed or they are defense prime contractors that don’t want their names to be disclosed either. So, I can’t really get into customer specifics. What I can say, however, is that our SDRs have been deployed in support of many applications, across such systems as radars, early warning systems, MRIs, signals intelligence, spectrum monitoring, low latency wireless links, and test and measurement. There’s definitely been interest in having our systems deployed for a variety of spectrum monitoring applications on UAVs or other autonomous systems where dynamic spectrum control is important.

That’s exactly where our high-performance products fit the bill, because they have a very wide tuning range — from near DC up to 18 GHz, and it offers up to 16 radio chains, where you can monitor every different part of the spectrum continuously.

Are your systems deployed primarily on land, sea, or air platforms?

Primarily on land and air vehicles. Our system is not miniaturized for robots. When you get into sea, it does become a little bit more challenging.

What about small UAVs?

It depends on how small you’re talking about. Often, those small UAVs will use a card, as opposed to an entire system. So, it’ll be a special PCB, that performs just one dedicated function. For some of the UAVs that demand the highest performance, they usually can support a payload of one of our SDRs. They’re not small, but they definitely can support the size, weight and power of one of our higher end SDRs, which are 19” 3U form factor rack-mountable solutions.

What are a couple of use cases or scenarios for autonomous systems?

Interoperability would probably be one of the biggest that keeps creeping up.

Swarms?

Yes. Basically, the ability for the different elements within a system to communicate with one another. The idea would be having some sort of AI or machine learning applied to autonomous systems, which is the future. The problem right now with it is that many of the autonomous systems utilize different communication protocols. That makes it very challenging to have a single set of controls to interface with them.

For us, one of the use cases would be for SDR to be the intermediary. So, capturing the digital data from each of those different frequencies and combining them into one source for that machine learning or AI to utilize. Imagine that you had sensors that were being used for short range radar, long range radar, communications, etc. You will be across the L band, the C band, the 2.8 GHz band, the 5.8 GHz band. You will be across several different protocols. Not all those systems will play well with one another. So, where we fit in is communicating with all of those disparate devices and converting all that data into a digital domain for additional processing to take place.

Another scenario is what we see a lot in our customers who are doing command and control for various tactical systems and want a single platform on their side to interface with all these separate RF devices.