ROME – ANDRO Computational Solutions, LLC of Rome has been awarded a US Army contract valued at $ 1.1 million to develop DeepSPEC, a platform for enhancing radio frequency signal detection and classification.
“DeepSPEC aims to revolutionize technologies for spectrum awareness in intelligence, surveillance and reconnaissance (ISR) applications by miniaturizing neural networks deployed on embedded edge devices such as tactical radios,” said ANDRO’s President Dr. Andrew Drozd.
The work will be spearheaded by ANDRO’s Marconi-Rosenblatt Artificial Intelligence / Machine Learning (AI / ML) Innovation Lab led by Dr. Jithin Jagannath, lab director and chief scientist of technology, and Anu Jagannath, AI / ML lab associate director.
Drozd said the Army is seeking more effective methods of sifting through noisy RF spectral environments consisting of radar and tactical communication signals and needs the ability to intuitively visualize “what is happening when it happens.”
Anu Jagannath stated that, “The ultimate goal of DeepSPEC is to ensure comprehensive spectrum awareness for every Army platoon and battalion for timely actionable intelligence,” Anu Jagannath explained. “Spectrum holds unprecedented information which, if correctly and swiftly interpreted, can save the lives of our warfighters.”
ANDRO has engaged the Lockheed Martin Company as a transition partner to embed DeepSPEC into small form-factor tactical radios in the next contract phase to service the Department of Defense programs of record.
DeepSPEC builds on ANDRO’s prior research for the Army, Air Force, Navy and other DoD agencies to develop advanced automatic modulation and signal classification systems for use in soldier-mounted software defined radios (SDRs).
Over the past 10 years, ANDRO has led advanced research in applying AI / ML, artificial neural networks and deep learning techniques to SDR platforms solving a wide range of challenging RF signal intelligence and wireless communications security problems.
Drozd said he is proud of his team’s accomplishments noting that, “Our team represents a new generation of ‘unicorns’ leading the way in bringing together the latest in wireless technology with the best in AI / ML.”
“Deploying powerful AI / ML algorithms on embedded platforms is a daunting task owing to the typical immense computational demand, where DeepSPEC aims to address this challenge in addition to solving the unknowns of the RF spectrum,” said Dr. Jithin Jagannath.
The ANDRO team envisions DeepSPEC to have broad and diverse applications beyond the realm of the DoD. For instance, it can be used to improve the spectrum sensing performance of intelligent-communication systems used for managing the coexistence of communication and radar systems in the commercial Citizens Broadband Radio Services (CBRS) band.
The success of programs like DeepSPEC will be one step closer in ANDRO’s plans to expand and pave the way for establishing a lightweight radio production center in Rome, built around an array of next-generation radio and wireless communications products, Drozd said.
ANDRO brings its military heritage of products to the commercial market and plans to relocate and expand its facilities in the Rome area by the fourth quarter of 2023. Visit www.androcs.com for more information on ANDRO including job placement opportunities.