Waveforms like Wi-Fi, 4G/5G, and Bluetooth satisfy the general population’s basic wireless communication and information sharing needs, but the static nature of these fixed waveforms is vulnerable to instability. Decades often pass between updated versions of wireless protocols.
The ability to quickly conduct point to point communications across dynamic environments – from secure facilities to “the wild” – is integral to the national security mission. While U.S. government (USG) waveform names are lesser known to the general population, USG invests heavily in waveform development to meet these needs, but due to the broad nature of missions, waveform development can be slow, expensive, and repetitious. The waveform development and acquisition cycle may take years, which can make new waveforms obsolete before becoming fully operational in a rapidly changing world. Therefore, communications frameworks that can generate on-demand waveforms are especially attractive when the time and cost of developing new waveforms to meet fluid mission needs are prohibitively high.
To meet the Intelligence Community’s (IC) data transmission and reception needs, the Intelligence Advanced Research Projects Activity (IARPA) will launch the Endless Generative Waveforms (End-Gen) program. End-Gen seeks to develop optimization frameworks, such as generative artificial intelligence (GAI) and machine learning (ML), capable of consuming input mission objectives in order to output mission waveforms.
If successful, End-Gen will support, enable, and augment numerous IC, USG, and public wireless needs by creating the generative waveform—a time- and cost-saving approach to perform wireless information exchange without costly, delayed human development cycles. End-Gen will completely alter the approach to developing solutions for new wireless capabilities. Additionally, End-Gen success will inform commercial wireless design in moving beyond 5G and help expedite next generation waveforms without common generational delays.
“The End-Gen approach allows the machine to create the right waveform and frees the human to focus on how best to complete the mission,” said End-Gen Program Manager, Dr. Adam Anderson. In other words, humans focus on the problems they are interested in and let the machine determine new algorithms, new waveforms—even new inventions—that will satisfy IC needs. “Let the machine use its generative abilities to make the waveform. Let the humans spend their cognitive resources on the mission itself.”
End-Gen is proposed as a 42-month research program and is divided into three phases:
- Phase 1 will demonstrate the feasibility for generative waveforms in relatively static environments without human intervention;
- During Phase 2, solutions will scale toward practical communication objectives that are meaningful to the community; and
- Finally, Phase 3 will focus on real-world problems and generate waveforms to be used in actual environments with practical data rates.
As the program progresses, generative waveforms will move from static to dynamic environments and become increasingly sophisticated, eventually solving communications problems by discovering basic algorithms that may not yet exist faster than humans can generate algorithms to meet mission objectives.
End-Gen Proposers’ Day will be held on 27 February 2025. For more information on how to submit a proposal for End-Gen, please see our Proposers’ Day page.
