AbstractProviding quality of service(QoS) guarantees in mobile data networks is an inherently challenging task. Mobility of users imposes a spatial demand on resources resulting in overloaded regions that are entirely dependent on mobility pattern of users, that is often unpredictable. Prior and ongoing work in this area of QoS relies on call admission control (CAC), or careful resource allocation based on mobility prediction. The latter approach is not scalable, as it requires constant monitoring of the mobility of individual users and per-user state. Furthermore, many of the CAC schemes assume random or uniform mobility patterns, and in most cased are based on local decisions. The challenge is to design a scalable scheme that can provide QoS under distinct mobility patterns.
In contrast to the standard notion of QoS which is based on the handoff dropping probability, another important notion of QoS is based on disallowing handoff drops but minimizing the cell congestion probability that may occur in a given cell. In this paper, we explore this notion of QoS by proposing a dynamic CAC scheme that uses an dynamically estimated mobility pattern and distribution of users in different cells to reach an admission decision with the objective of minimizing cell congestion probability. This scheme does not maintain per user state, and can be implemented in a distributed fashion. From simulation results, we show that across different mobility patterns, the proposed scheme performs better than existing schemes in terms of achieved QoS while providing the minimum level of overall target utilization.
RightsThis Item is protected by copyright and/or related rights.You are free to use this Item in any way that is permitted by the copyright and related rights legislation that applies to your use.For other uses you need to obtain permission from the rights-holder(s).