QoS-Aware Service Selection Based on ACK Method for Multi-User Condition

With the development of Web technology in communication area, more and more services are provided on converged network, which are function similar with different QoS attributes. Present web service selection is trying to find solutions that can get best QoS utility value for single user while satisfying his requirements. However, affections from other users´ solutions are lack of consideration, especially for those users who have same functional requirements with different QoS preferences. From the perspective of network operators, operators wish to find an overall solution that can get an optimal QoS utility value, while meeting each individual´s requirements and global constrains. This paper proposes a novel method based on Ant Colony and KM algorithm (ACK) to meet above all requirements. Firstly, we decompose the global QoS constrains into local constrains of each service class. Secondly, further decomposition of each service class´s local constraint is taken in user dimension, transforming global selection problem into local selection problem. Thirdly, the final overall-optimal solution can be got based on results from above two steps. Experiments show that our method is particularly suitable for situation of multi-user and large-scale service selection and outperforms traditional solutions in terms of computation time while achieving close-to-optimal results.