Advanced Handover Management in Future Wireless Networks using Prediction of QoS

The increasing number of mobile devices has raised the demand of Internet access. This also results in more traffic over Wireless Local Area Networks (WLAN), which provide low-cost, high-speed, and reliable Internet access. The nature of the WLAN technology limits the coverage of each single Access Point (AP), and this therefore often leads to handovers while a Mobile Node (MN) moves. Some of the handovers are unnecessary, and have negative impact on MN's Quality of Experience (QoE). In this work, two mechanisms are proposed to solve this problem:
(1) a Received Signal Strength (RSS) prediction mechanism based on Hidden Markov Models (HMM) and
(2) a handover decision making mechanism based on the prediction results. Performance evaluations show that the proposals do improve the QoE of the MNs.

Besides being able to be applied in the existing network architectures, the proposed handover management mechanism can also be used in the future Internet architectures. However, some of the future Internet architectures do not cater for easy support for mobility, such as the Content Centric Networking (CCN). In this work, two mobility schemes are proposed to improve the performance of CCN with mobility. The proposed handover management mechanism in this thesis, including the handover prediction and handover decision strategy, can be used with the new mobility schemes, to further enhance the performance of CCN during the handovers.