This solicitation closed on September 23, 2026
View current Department of Defense opportunities →Artificial Intelligence / Machine Learning (AI/ML)-Based Radar Data Compression
AI Overview
This RFP seeks AI/ML-based compression technology for high-rate radar data streams, addressing the challenge of massive synthetic aperture radar data volumes. The solution will develop neural network architectures to compress raw radar returns with minimal distortion, reducing bandwidth requirements for transmission and storage across defense and civilian sensor applications.
This summary is AI-generated from the official solicitation.
Key Details
Official Description
Next-generation radars (especially synthetic aperture radar (SAR)) collect data with massive data rates. Traditional image compression is not optimized for raw radar returns. Recent work extends neural compression to the complex SAR domain. Phase I is to explore autoencoder architectures for radar data using, for example, a complex-valued neural network encoder/decoder that learns to compress raw pulses or range-Doppler maps into a latent code with minimal distortion. The system could be trained...
Change History
Informational change · Status
Artificial Intelligence / Machine Learning (AI/ML)-Based Radar Data Compression
Status changed from status: Open to status: Closed.
Informational change · Description
Artificial Intelligence / Machine Learning (AI/ML)-Based Radar Data Compression
Description information changed.
Informational change · Description
Artificial Intelligence / Machine Learning (AI/ML)-Based Radar Data Compression
Description information changed.
Informational change · Description
Artificial Intelligence / Machine Learning (AI/ML)-Based Radar Data Compression
Description information changed.
Artificial Intelligence / Machine Learning (AI/ML)-Based Radar Data Compression
Q&A Changes Summary
New Question Added:
- Q1 (NEW): Clarifies that coherent SAR exploitation functions (coherent change detection, interferometric measurement) are in-scope examples of "radar utility" for Phase I evaluation.
Key Changes:
- All other questions (Q2–Q20) remain substantively unchanged in wording, but the addition of Q1 at the top signals Government emphasis on coherent SAR applications as a primary evaluation domain.
- No previously pending answers appear to have been formally resolved in this update; the Q&A still contains the same open technical questions regarding data provision, baseline selection, downstream task specification, and Phase II hardware targets.
Summary: One new clarification added on coherent SAR exploitation scope; no resolution of major pending technical or data-provision questions.
Artificial Intelligence / Machine Learning (AI/ML)-Based Radar Data Compression
Q&A Changes Summary
New Questions Added (3):
- Q1: Delivery format preferences (programming language, ML framework, containerization, open-source constraints)
- Q2: Mechanism and lead time for requesting Government datasets during Phase I performance period
- Q3: Fixed vs. variable-rate compression operating modes and which should govern primary evaluation
Reorganization: Previous Q15-Q19 (duplicate/fragmented questions on data furnishment, downstream tasks, and embedding targets) consolidated and renumbered as Q13-Q19 in updated version, maintaining content but improving clarity through consolidation.
No Material Answer Updates: Existing Q&As retained their previous response status (many still marked "Response Pending" or unanswered). The update primarily addresses implementation logistics and operational modes rather than substantive technical clarifications.
Artificial Intelligence / Machine Learning (AI/ML)-Based Radar Data Compression
Q&A Changes Summary
Significant expansion from 1 to 17 questions.
New topics added:
- Hardware constraints, memory, and throughput requirements (Q1)
- Data size ranges (Q2)
- Government data provision and acceptable datasets (Q3, Q7, Q9, Q15–16)
- Target metrics: compression ratio, time, success criteria (Q4, Q6)
- Primary radar representation/modality (Q8)
- Technical scope: compression point, radar modes, lossy vs. lossless, multi-channel (Q9)
- Evaluation baselines and thresholds (Q9)
- Phase II hardware targets and SWaP constraints (Q9, Q11)
- Transition sponsor and classification level (Q9)
- Downstream task specification and fidelity scoring (Q12–14)
- Alternative ML schemes beyond autoencoders (Q17)
- Dual-use applicability demonstration scope (Q10)
Original questions retained: Complex-valued architecture requirement and classical baselines (now Q5–6) remain substantively unchanged but repositioned.
Artificial Intelligence / Machine Learning (AI/ML)-Based Radar Data Compression
This Q&A clarifies technical flexibility for applicants: baseline compression comparisons can use program-specified methods with flexible matching criteria, and real-valued neural architectures treating I/Q channels are acceptable alternatives to fully complex-valued designs, provided they preserve phase information and meet performance targets.
Artificial Intelligence / Machine Learning (AI/ML)-Based Radar Data Compression
Status changed from Pre-Release to Open
Artificial Intelligence / Machine Learning (AI/ML)-Based Radar Data Compression
New opportunity: Artificial Intelligence / Machine Learning (AI/ML)-Based Radar Data Compression
Similar Opportunities
Get Alerts for Opportunities Like This
RallyProp monitors thousands of funding sources and sends instant alerts when opportunities match your technology and capabilities.
Start Free Trial