A Novel User-Oriented Quality of Service Algorithm for Home Networks

This paper aims to present a novel dynamic quality of service (QoS) algorithm for home networks. The algorithm carries out the QoS management of a service for a user in a home network based on the profile consisting of the past responses of the user for that service. It performs the prediction of the future responses to the service from the profile using the general regression neural network algorithm. The algorithm finds the bandwidth allocation consuming minimum total bandwidth among those predicted to produce desired responses. The minimum bandwidth allocation is then used for the service. The corresponding user responses are used to update the profiles for the subsequent dynamic QoS managements. The algorithm requires no offline training and achieves fast adaptation to the user responses. Both analytical and numerical results reveal that the proposed algorithm provides an effective alternative for user-oriented QoS management in the heterogeneous home networks.