Optimal Handover Policy for mmWave Cellular Networks: A Multi-Armed Bandit Approach

Abstract

Millimeter wave (mmWave) is a promising technology in 5G communication due to its abundant bandwidth re- source. However, its severe path attenuation and vulnerability to line-of-sight (LOS) blockage result in much more unpredictable outages than traditional technologies. This special propagation property raises a significant challenge to the mobility management in mmWave cellular networks. In particular, conventional handover policies purely rely on the measurement of signal strength. If being applied directly in mmWave cellular networks, they would cause a large number of unnecessary handovers due to the frequent short-term LOS blockage by obstacles, imposing high signaling and energy overhead on the network. In this paper, we propose a novel handover mechanism to reduce unnecessary handovers in a mmWave cellular network by carefully deciding the next base station (BS) a user should handover to, so that the new user-BS connection after the handover can last as long as possible. Clearly, making such an optimal decision requires some knowledge on the users post-handover mobility trajectory and LOS blockage, whose realization cannot be assumed at the moment of handover. The proposed handover mechanism addresses this challenge by exploiting the empirical distribution of users post-handover trajectory and LOS blockage, learned online through a multi-armed bandit framework, with the intention to maximize the expectation of the user-BS connection time after each handover. The results of numerical experiments demonstrate that the proposed handover policy outperforms existing counterparts on reducing handovers, especially in the scenarios where users mobility follows regular patterns.