On Load-Aware Cell Association Schemes for Group User Mobility in mmWave Networks

While faster signal attenuation in millimeter wave (mmWave) networks is compensated by high antenna directivity, the effects of dynamic blockages can be mitigated by using connections to multiple access points (APs), and the choice of a proper cell association algorithm plays an essential role in optimizing the overall system performance. Despite numerous research efforts aimed at finding optimal cell associations, effects of user mobility have not been explicitly addressed. Particularly, mobility patterns are imperative in the case of mmWave directional access, as they impact the overall system performance and, thus, might affect the choice of optimal solutions. In this paper, we address correlated mobility typical for collective extended reality (XR) applications, using the example of reference point group mobility models (RPGM), where users migrate in groups or clusters. Assuming 3D beamforming and protocol settings of mmWave IEEE 802.11ad/ay that operates at 60GHz, we compare the performance of two cell association schemes: baseline RSSI-based and load-aware algorithms, and provide important insights on optimization of load-aware scheme parameters, the choice of which is a dynamic function of antenna directivity and network density.