Alwan, EliasVolakis, JohnIslam, Md KhadimulDe Silva, UdaraMadanayake, ArjunaSanchez, Jose AngelSklivanitis, GeorgePados, Dimitris A.Beckwith, LukeAzarderakhsh, RezaMuralkrishan, MadhuvantiRastogi, RishabhHore, AniruddhaBurger, Eric W.2024-07-102024-07-1020232155-7578https://hdl.handle.net/10919/120630We have built a prototype universal radio adapter which furnishes seamless and secure wireless communication through non-cooperative indigenous 5G networks for military and government users. The adapter consists of a waveform-agnostic hardware add-on that tunnels DoD terrestrial and satellite data. The adapter uses secure protocols for cross-connecting military-grade wireless RF communications equipment using spectrum in the range from UHF to Ka-band. A 5G data transport channel replaces the captured spectrum for transporting information at the IQ-sample level. In a sense, we replace the antenna-air interface and wireless channel with a transparent 5G data network. A plurarity of legacy military systems can operate through modern 5G networks in a seamless way without any knowledge of the characteristics of military waveforms. The adapter incorporates AI/ML based methods for smart spectrum sensing and autonomous radio reconfiguration. This enables intelligent interconnection of a number of military radios through non-cooperative (potentially adversarial) 5G commercial cellular networks. The adapter is built on four technical pillars: 1) ultra-wideband apertures for multi-functional and flexible software-defined radios (SDRs) with agile, wideband, and dual-band tunable RF transceivers for FR1/FR2 bands; 2) physical layer operation that involve device authentication via deep-learning based RF fingerprinting and compression of acquired IQ data; 3) secure and reconfigurable cryptographic co-processors employing the new quantum-safe algorithms selected by NIST to achieve authentication, key exchange, and encryption with focus on resource-constrained low size, weight, power, and cost (SWaP-C) devices; and 4) generative artificial intelligence and spread-spectrum steganography to hide DoD traffic passed through 5G networks and improve resiliency against real-time traffic analysis by nation-state carriers and intelligence agencies.6 page(s)application/pdfenIn Copyright5Gsecuritydata hidingsteganographyPQCIQ signal compressionAI fingerprintingConference proceedinghttps://doi.org/10.1109/MILCOM58377.2023.10356300Burger, Eric [0000-0002-2143-9368]