Network selection in heterogeneous dense networks based on user clustering

In this paper, we study the problem of network selection in dense heterogeneous cellular networks, where multiple base stations (BSs) are available to meet the demand for wireless data. The selection of the most appropriate BS for each user equipment (UE) is a complex task, taking into account factors such as signal strength, interference, and available resources. To solve this problem, we propose a distributed algorithm based on game theory that optimizes network selection by considering the physical rate of UEs towards various BSs. The algorithm clusters UEs based on their physical rate and formulates the network selection problem as a convex optimization problem. We propose a distributed algorithm based on game theory that can reach optimal solution and speeds up convergence by simultaneously changing the BS for a set of UEs belonging to the same cluster. Simulation results are used to verify the performance of the proposed algorithm and explore the convergence speed. Specifically, in our simulations, the Modified BIR algorithm achieves convergence almost two times faster than the BIR algorithm, while the cluster-based BIR algorithm converges around 10 times faster than the original BIR algorithm. These findings highlight the significant improvement in network selection performance achieved by our proposed algorithms.