Real-Time Scheduling for GAA Users’ Coexistence in CBRS Using O-RAN Architecture

To maximize the potential benefit of Citizen Broadband Radio Service (CBRS), efficient coexistence algorithms are urgently needed for the general authorized access (GAA) users. The goal is to allow the GAA users to share the same spectrum with the priority access license (PAL) users while providing the required interference protection to the PAL users. The main challenge in designing an efficient coexistence solution is the absence of collaboration by the PAL users and the uncertainty in sensed channel information. This paper addresses these problems through a data-driven approach based on limited data samples. We develop a mathematical model using chance-constrained programming (CCP) to address uncertainty in sensed data samples. By exploiting the ∞-Wasserstein ambiguity set, we reformulate the CCP problem into a deterministic Mixed-Integer Nonlinear Program (MINLP). We propose an open RAN (O-RAN)-based solution to MINLP at the GAA base station (BS) that leverages both Non-Real-Time (Non-RT) RAN Intelligent Controller (RIC) and Real-Time (RT) O-DU to deliver a final scheduling solution.