Dynamic resource management mechanism design for heterogeneous networks under IEEE1900.4 framework: A network utility maximization perspective

The Next generation wireless system will be heterogeneously made up of multiple radio access networks (RANs) with the different radio access technologies (RATs) as well as subscribers will be equipped with multi-mode, reconfigurable and cognitive equipments, with or without the multi-homing ability. A dynamic resource management mechanism is presented in this paper under the IEEE P1900.4 architecture framework using network utility maximization technologies. This will provide an effective solution for heterogeneous networks to achieve efficient utilization of the various resources, e.g., the radio, the network, the device and the computing resource. Take the end-to-end and overall network performance into account, a double-cycle learning structures based on the utility theory for the terminal and the RAN are proposed, respectively. With the aid of the radio enablers as the interface, the strategy and context information interaction between the terminal and the RAN are investigated. Also, base on the double-cycle learning framework, the interactive protocol between the terminal with RAN and dynamic RAN selection algorithm for the terminal are proposed. Meanwhile, we consider a simple heterogenous scenario to verify the effectiveness of the proposed utility based double cycle learning architecture and the presented algorithms as well as the interactive protocol.