Abstract
The recently unlocked 6-GHz spectrum is accessible to cellular and WiFi networks for unlicensed use while they conform to the constraints imposed by the incumbent nodes. We allow only a fraction of the cellular base stations (BSs) and a fraction of WiFi access points (APs) to use the 6-GHz band so that sources of interference are spatially segregated and made sparse, thereby decreasing the overall interference to each other. Through our proposed framework, we control this fraction as we group portions of cellular and WiFi network elements into entities competing with the other entities for the spectrum resources. These entities interact to satisfy their Quality of Service demands by playing a non-cooperative game. The action of an entity corresponds to the fraction of its network elements (WiFi APs and cellular BSs) operating in the 6-GHz band. We use tools from stochastic geometry to derive the theoretical performance metrics for users of each radio access technology, which helps us capture the aggregate behaviour of the network in a snapshot. Due to the decentralized nature of the game, we find the solution using distributed Best Response Algorithm (D-BRA), which improves the average datarate by 11.37% and 18.59% for cellular and WiFi networks, respectively, with random strategy as the baseline. The results demonstrate how the system parameters affect the performance of a network at equilibrium and highlight the throughput gains of the networks as a result of using the 6-GHz bands, which offer considerably larger bandwidths. We tested our framework using real-world data, which shows that practical implementation of multi-entity spectrum sharing is feasible even when the spatial distribution of the network elements and population are non-homogeneous.