Coalition-Aided Resource Allocation And Traffic Scheduling For Multiuser Uplinks With Qos Provision

In the wireless network, the network performance can be improved throughcooperation techniques among the users. However, performing a distributed and faircooperation constitutes a major challenge. This thesis aims to solve the problem ofresource allocation and multi-service scheduling scheme by coalitional game theory,and does the performance analysis and simulation validation.Firstly, we model the multi-user cooperation in wireless networks. A simple anddistributed algorithm is constructed to enable the users to self-organize into independentdisjoint coalitions and to allow them to maximize their utilities in terms of rate whileaccounting for the cost of cooperation in terms of power.Afterwards, we consider the problem of resource allocation and traffic schedulingfor the multi-user and multi-service with the aid of the coalition formation game, whichenables the users to autonomically form disjoint coalitions in accordance with thedynamics of channel state information (CSI) and queue state information (QSI).Comparing with the non-cooperative techniques, our proposed scheme achieves betterperformance in terms of both the system throughput and the traffic-delay for theheterogeneous applications.Finally, we study the two-phase multi-service scheduling scheme for smart grids,in which the sink nodes and multimedia sensors form coalition groups, and sink node inevery coalition adopts the online-learning algorithm for the decision that which sinknode gets the transmission token. Simulation by OPNET shows that the schemeachieves energy-efficient scheduling under QoS requirements for multi-traffic.