DARPA’s Microsystems Technology Office is seeking information on learning effective control strategies for teams of multi-function autonomous radio frequency (RF) systems.
The goal of this RFI is to gather new insights on the challenges preventing continuous learning of real-time strategies for collaborative RF teams, and identify candidate solutions that could overcome these challenges.
DARPA is interested in responses that address the following two topics; all responses are expected to address the first topic and may optionally address the second topic.
Topic 1: Challenges and opportunities in learning for collaborative RF control: Key areas of interest include real-time RF control algorithms, lifelong learning under edge processing constraints, and training methods that close the sim-to-real gap.
Responses may highlight promising approaches from other domains along with rationale that such solutions are applicable to the RF problems. Responses are encouraged to focus on the key challenges that prevent real-world applications of learning for autonomous radios; discussion of toy problems without providing insight on how such problems scale is discouraged.
Topic 2: Panelist for ERI Summit workshop: Responders may recommend a panelist and up to one alternate for participation in a related workshop at the upcoming ERI Summit. Recommended panelists are expected to be available to attend the workshop in-person, which is planned for August 24, 2023, in Seattle, Washington.
Responses are due 28 June 2023.
Supporting document
