25 December 2020

More Bandwidth for Software Developers Means Better Apps for SIGINT/ISR/Comms

By BARRY ROSENBERG

Software Defined Radios (SDRs) will help transform the speed of military situational awareness, enabling faster decisions and action. SDRs do so by digitally converging the means by which the military communicates—data, voice and video over various spectrum assets—into a single comms device, ideally making the modes available simultaneously to both command and warfighters in the battlespace. Designing devices and applications which make that capability possible in an easy and cost-effective way is, however, a challenge that Motorola Solutions (MSI) has recently taken head on.

Applications that the military craves for software defined radios (SDRs) include better performing SIGINT (signals intelligence), ISR (intelligence, surveillance, reconnaissance), and communications techniques. But the introduction of advanced capabilities in these areas, as well as others like radar and satellite communications, require software developers to have the ability to access frequencies beyond 6 GHz and up to 30 GHz without the need for external hardware.

Without such capability, the ability of software developers to write new applications for SDRs is limited. And while software can be written for the processor fairly easily, the ability to alter the field-programmable gate array (FPGA) firmware so that it’s adaptable for developing sophisticated applications is often restricted for third-party developers.

Those drawbacks have now been erased with development of the NS-1 transceiver from Motorola Solutions. The properties of the NS-1 transceiver are based upon a new Microwave radio-frequency integrated circuit (RFIC) chipset developed by Applied Technology, which is an engineering group within Motorola Solutions. The in-house designed, custom Microwave RFIC has a frequency range from 2 MHz (or lower) to 30 GHz and a programmable instantaneous bandwidth of up to 1 GHz. Those specs provide a means for application development in a software-only environment, even when processing 1 GHz of spectrum.


Soldiers inventory radio equipment used to support system signal specialists during an exercise. Photo by Staff Sgt. Michael Crawford. The appearance of U.S. DoD visual information does not imply or constitute DoD endorsement.

It is anticipated that new, fully integrated, manpackable SDRs for military use will inevitably follow the NS-1. Innovations based on the NS-1 platform will yield: improved wideband intercept; low probability of intercept/low probability of detection (LPI/LPD) communications; advanced, portable, machine-learning-based signal classification tools for fieldable SIGINT applications and battlefield electro-magnetic management; and new gateway applications that can bridge standard communications waveforms to newly available, higher frequency links.

The State of SDR Today

The term “transceiver” is simply a combination of transmitter and receiver, and in modern contexts usually refers to a chip or chipset. In terms of an SDR, it’s one of the three main components: a general purpose processor to run software, an FPGA or similar device to handle time-sensitive tasks and hardware acceleration, and the transceiver to convert RF to digital data.

It’s common for software developers exploring concepts in SDR technology—including investigators in academia, engineers in government laboratories, Federally Funded Research and Development Centers () and University Affiliated Research Centers (UARCs), as well as other third-party radio software developers—to talk about two key parameters for the RF chipset. They are: 1) the tuning range for the center frequency; and 2) the instantaneous bandwidth.

Tuning range is analogous to turning the knob on the old AM/FM radio, indicating the minimum and maximum frequencies that the transceiver can serve. Instantaneous bandwidth defines how wide a space around that center frequency the receiver or transmitter can view at that specific moment in time. For example, if you have a receiver that is tuned to 900 MHz with an instantaneous bandwidth of 100 MHz, you can measure a response from 850 MHz to 950 MHz.

SDRs today are primarily based on only a few chipsets. Increasingly, though, we’re seeing a greater number of SDRs with single-channel frequency ranges up to 6 GHz with bandwidths typically around 50 MHz. A few newer devices are based on more recent chipsets capable of bandwidths of up to 200 MHz for receive and 450 MHz for transmit. These chipsets also have a lower frequency limit of 70 MHz, which leaves out critical HF spectrum, especially for DoD applications.

Recently, a Department of Homeland Security market survey report on radio frequency detection and spectrum analysis identified development of SDR platforms that can get 1 GHz of bandwidth with center frequency range from 2 MHz to 18 GHz in frequency, which is a step forward in capability of commercially available SDRs.

With the NS-1, Motorola Solutions is providing up to 1 GHz of instantaneous bandwidth at center frequencies from 2 MHz to 30 GHz. The company also has technology roadmaps in place to at least double both of these parameters with chipset revisions.

Applications Development in a Software-Only Environment

“The NS-1 addresses the Joint Force’s need for better communications by providing a platform to investigate new communications methods and concepts that take advantage of the increased tuning range and bandwidth of the Microwave RFIC,” explained Dr. Robert Croswell, Distinguished Member of the Technical Staff at Motorola Solutions and Chief Scientist of Applied Technology’s Early Capture team. “Imagine Direct Sequence Spread Spectrum (DSSS) waveforms that are spread over hundreds of megahertz or fast-frequency hopping waveforms that could hop hundreds of thousands of times per second over one GHz of spectrum of choice, up to 30 GHz,” he said, referring to a modulation technique to reduce signal interference.

“Put multiple transceivers together and you could have waveforms that break up information in multiple gigahertz of non-contiguous spectrum, either through hopping or simultaneous transmission, all while providing access to additional frequencies that have not been available to SDR platforms in the past.”

Dr. Robert Croswell, Distinguished Member of the Technical Staff at Motorola Solutions and Chief Scientist of Applied Technology’s Early Capture team.

The NS-1 was designed to address the FPGA programming issue that is exacerbated by the increased amount of data that is generated by a 1 GHz wide transceiver—specifically the need for hardware acceleration to move the data so that the processor doesn’t bog down data transfer.

The transceiver provides a means for application development in a software-only environment, even when processing 1 GHz of spectrum. The NS-1 has a single transceiver card mated with commercial off-the-shelf FPGA hardware. It interfaces with standard computer hardware such as a desktop or laptop PC through a Thunderbolt 3 interface, which writes sample blocks directly to PC memory, so there is no transfer limitation due to processor loading. The user can then develop software for the PC, whether for a General Purpose Processor or Graphic Processor Unit, and the PC hardware can be scaled as needed for application development solely in software.

The NS-1’s ability to monitor more bandwidth translates to more capability for a given transceiver. Software developers can break out and decode more standard communications channels with a single receiver. They can dwell on a full GHz of bandwidth and not miss an intermittent signal if they happened to be scanning a different block of spectrum the moment that signal was present.

“With NS-1, I can hop frequencies within its instantaneous bandwidth without the need to retune the center frequency,” said Croswell. “In our system, hopping time within that 1 GHz of bandwidth is on the order of a few sample time periods, typically tens of nanoseconds. Wider bandwidth enables fast hopping on both the receiver and transmitter, and with that bandwidth I can extend techniques like DSSS to spread the information over wider bandwidths than in previous radios.

“Wider spread bandwidth translates to additional coding gain, allowing for better LPI/LPD waveform properties by smoothing out the peaks associated with the waveform over additional spectrum or allowing reduced transmitter power to complete a link.”

The Takeaway

The NS-1 is targeted at the R&D community, and is designed to give software engineers access to the new, wideband, broad frequency range transceiver chipset in a development environment that they know. With a single radio channel available in NS-1, the entire 1 GHz can be sampled at once. The receiver can dwell continuously on the full bandwidth so that no intermittent signal is lost. The transceiver can also output or receive information anywhere in the 1 GHz bandwidth without the need to retune the center frequency, so narrowband operations in contiguous spectrum can be accelerated, as in the case of wideband fast-hopping protocols.

Quite simply, wider bandwidth allows a single transceiver to gaze at more spectrum at a given time. It enables wider bandwidth protocols for receive or transmit, enabling the user to push more data (think the “Shannon limit” noisy-channel coding theorem) or use additional coding gain to reduce observability. The center frequency range extension up to the K-band enables SDR applications for those that are traditionally realized in specialized hardware.

A wide base of developers should yield new innovations that are appropriate for the battlefield. While the NS-1 platform isn’t likely to become standard warfighter gear, the applications that are built for it can be transitioned to embeddable versions of the SDR based on the Microwave RFIC.

No comments: