Optimizing bandwidth allocation in mmWave/sub-THz cellular networks using maximum flow algorithms

The exploitation of millimeter wave (mmWave) and sub-Terahertz (sub-THz) bands is expected to be one of the main pillars for the development of future cellular networks due to the high available bandwidth they provide. The existence of Line-of-Sight (LOS) link between a user equipment (UE) and an access point (AP) is a prerequisite for connection establishment in these networks, as the wireless links in these bands are very sensitive to blockage effects. This can be achieved by densifying APs within a network area. An arising challenge is the efficient exploitation of the available bandwidth of a given network. In this paper, the maximization of the number of served UEs in modern mmWave and sub-THz cellular networks is investigated and achieved by deploying a Maximum Flow Algorithm for UE-AP association (MFUA) to optimize bandwidth allocation, assuming that every AP will have a finite and predefined amount of bandwidth which they can share among UEs. MFUA determines the maximum flow between two given nodes of a graph corresponding to a specific network, where the capacity of its edges is known. An extensive simulation campaign was carried out revealing that the use of MFUA utilizes bandwidth more effectively compared to the reference method and improves the system performance, leading to the maximization of number of served UEs. The examined test cases include static and time-evolving scenarios.