b'Secure Millimeter WaveThis research introduced a distributed scheduler for cellular base stations that Spectrum Sharing withcan be further matured as an innovative technology for the emerging field of 5G Autonomous Beam Scheduling spectrum sharing without requiring centralized control.I n this research project, the team proved that a novel autonomous radio frequency beam scheduler with optimality guarantees can support secure spectrum sharing of newly available unlicensed and shared 5G millimeter wave bands. This class of distributed schedulers will enable 5G systems to provide additional data throughput and capacity with higher security than currently possible. These improvements will TOTAL APPROVED AMOUNT:benefit, among others, mission critical communications, such as national security $890,000 over 2 years command and control, and emergency response operations in public safety.PROJECT NUMBER:The team designed and analyzed the capabilities of a distributed scheduler that 20A44-058 considers multiple operators sharing the spectrum in the same geographical area. The co-location of base station antennas was also considered. A game theory approach was PRINCIPAL INVESTIGATOR:used to solve a non-convex sub-problem, and the distributed schedular performance Arupjyoti Bhuyan was evaluated. Improvement over existing solutions was demonstrated.CO-INVESTIGATORS: The team developed and analyzed a Q-learning algorithm solution in the reinforcement Jacobus Van der Merwe,learning framework. A multiple operators base station distribution was used to analyze University of Utah the reinforcement learning effectiveness. The Q-learning algorithm performance was Mingyue Ji, University of Utah compared to the game theory solution. The team concluded that while both solutions Sneha Kasera, University of Utah perform satisfactorily with comparable performance in general, the Q-learning outperforms in the interference dominated configurations.(a) Interference from collocated base station antennas cannot be sensed at the base stations. (b) Results from proposed distributed scheduling that maximizes network utility with spectrum sharing.118'