Navy STTR DON26TZ05-NV023: Detection and Classification of Low Probability of Intercept Radar Waveforms Using FPGA with Cognitive Techniques

Below is a brief summary. Please check the full solicitation before applying (link in resources section).

Quick Answer

DON26TZ05-NV023 is a Navy STTR topic seeking an FPGA based electronic warfare capability that detects and classifies Low Probability of Intercept radar waveforms in near real time. It falls under NAVAIR and the Quantum and Battlefield Information Dominance critical technology area. Phase I awards go up to 315,000 dollars across a six month base and six month option. Proposals are due September 23. The topic is ITAR and EAR restricted.

What This Topic Is Looking For

Digital Radio Frequency Memory systems, known as DRFMs, are used to characterize and jam adversarial radar signals. This topic focuses specifically on the detection and classification piece of that problem, using the Field-Programmable Gate Array inside a Radio Frequency System-on-Chip, or RFSoC, device.

Low probability of intercept signals are hard to catch because they run low power, high duty cycle, and wideband, which demands long integration times and sensitive receivers. The topic scopes the work around signal processing and FPGA implementation rather than the RF hardware itself. Acceptable approaches include traditional methods like adaptive matched filtering with FFT frequency detection, filter banks, autocorrelation, and Wigner-Ville or cyclostationary processing, as well as more advanced cognitive techniques such as convolutional neural networks and LSTM based deep learning. The FPGA is specifically valuable here because hardware emulation of these algorithms allows real time or near real time performance.

The Navy's Airborne Threat Simulation Organization at NAWCWD Point Mugu will make real LPI waveform data available to train, validate, and benchmark whatever approach is proposed, and can also provide ground truth data for measuring performance. The target evaluation hardware is an AMD, formerly Xilinx, RFSoC kit. Because real time hyperparameter tuning on the FPGA is not practical, any neural network or cognitive model has to be designed and trained on a high compute PC or Linux platform first, then transitioned to hardware. Non deep learning algorithms can be developed directly in HDL and simulated in tools like ModelSim.

Phase I, II, and III Expectations

Phase I is about building the model and proving feasibility. Deliverables include a large training and validation database, which the topic notes can be built using tools like MATLAB, identification of candidate algorithms, and a feasibility demonstration before anything moves to actual RFSoC hardware.

Phase II moves into hardware implementation. That means additional HDL development, testbenches to track performance as the model transitions from MATLAB and ModelSim into Xilinx Vivado, and deployment onto the actual RFSoC. Initial testing happens on a laboratory benchtop using LPI waveforms transmitted by RF instrumentation such as Keysight equipment. Further validation happens at NAWCWD Point Mugu facilities, including over the air testing in an anechoic chamber. If additional funding becomes available later, the detector could be integrated onto a target platform such as the GQM-163A for live, virtual, and constructive training realism, though that would require added systems engineering support.

Phase III focuses on finalizing hardware and firmware for integration into an operational target platform, ruggedizing the design, running comprehensive operational test and evaluation across live, virtual, and constructive scenarios, and building a transition and manufacturing plan for fleet deployment.

Beyond the military application, this technology has broad commercial relevance anywhere weak or complex signals need to be detected in a crowded spectrum. The topic specifically calls out cognitive radio and cell tower connectivity in low signal environments, automotive radar for ADAS and autonomous vehicles, spectrum monitoring for regulators and private companies, and scientific applications in radio astronomy and atmospheric sensing.

Funding and Timeline

Phase I Base: up to 200,000 dollars for six months of work.

Phase I Option: up to 115,000 dollars for an additional six months.

Combined Phase I maximum: 315,000 dollars, not including TABA.

TABA: up to 6,500 dollars additional if requested.

Phase II maximum: up to 2,000,000 dollars including TABA, though the actual amount depends on the awarding command's available funding.

Proposal deadline: September 23. Submission is only accepted through the DoW SBIR/STTR Innovation Portal, DSIP, and must be certified by a Corporate Official before the BAA closes.

STTR Requirements

Because this is an STTR topic, the proposing small business must partner with a qualifying research institution. The small business must perform at least 40 percent of the work and the research institution at least 30 percent, measured across both base and option costs. Navy laboratories do not satisfy this requirement on their own, though they can be added alongside a qualifying university, FFRDC, or nonprofit research institution.

Frequently Asked Questions

What is the deadline for DON26TZ05-NV023? September 23. Proposals must be certified in DSIP before that date.

What kind of company or team fits this topic best? Teams with strength in FPGA and RFSoC development, digital signal processing, and ideally experience with cognitive or deep learning techniques applied to RF signals, paired with a research institution that can support algorithm development.

Does the Navy provide test data? Yes. The Airborne Threat Simulation Organization at NAWCWD Point Mugu will provide access to real LPI waveform data for training, testing, and validating whatever algorithm or model is proposed.

What hardware platform is this built for? An AMD, formerly Xilinx, RFSoC evaluation kit, which is representative of the DRFM RFSoC hardware used in test and evaluation threat simulation systems.

Is a research partner required? Yes. This is an STTR topic, so a formal partnership with a qualifying research institution performing at least 30 percent of the work is mandatory.

Is this topic export controlled? Yes. It is restricted under ITAR and may also fall under EAR. Any planned use of foreign nationals must be disclosed, including country of origin, visa status, and specific tasks.

Does this technology have non-defense applications? Yes. The topic specifically names cognitive radio, cell phone connectivity improvements, automotive radar for autonomous vehicles, spectrum monitoring, and radio astronomy as commercial and scientific applications.