Mobility management for group mobile users in cellular networks

Abstract

As the number of subscribers has increased rapidly in the recent years, situations in which mobile stations (MS) use wireless communication devices while moving together on a bus or a train are now commonplace. They can use various emerging applications such as voice over Internet protocol (VoIP), video on demand (VoD), and online games which require high data rate and low latency while moving fast by high-speed bus or train. Therefore, to satisfy the quality of service (QoS) for these MSs, performance degradation due to the handover should be minimized. In this dissertation, we investigate the mobility management for group mobile users in single-hop and multi-hop cellular networks. First, we propose a handover decision algorithm for single-hop cellular networks to improve the handover performance. Previous works are based on the signal strength, location, velocity, and moving direction to make handover decision. However, an instance of MSs moving together has not been considered although in actuality there are many such cases, e.g., passengers traveling on a bus or a train, in which the locations of MSs are statistically correlated. We exploit the mobility correlation for handover decision in order to improve the handover performance. We propose a grouping algorithm based on the mobility correlation, which does not require the exact locations of MSs. The proposed grouping algorithm consists of two steps. In Step 1, groups are formulated based on the history of the serving base stations (BS) of each MS. Then, to formulate groups more accurately, the history of signal-to-interference-and-noise-ratio (SINR) from the serving BS and adjacent BSs is used in Step 2. For each step, we present a simple greedy algorithm for grouping. Moreover, an adaptive hysteresis algorithm is proposed. It adjusts the hysteresis value by predicting the target BS with the help of the formulated groups. MSs can perform handovers fast by applying the reduced hysteresis value to the ta...