Green Predictive Resource Allocation for Ultra Dense Networks (UDNs)

With the development of big data analyzing technology, the network could predict users’ trajectories and demands and then allocate system resource in advance based on the predictive information to reduce the cost while meeting users’ demands. While in ultra-dense networks (UDNs) users’ data rate prediction and resource allocation are coupled because of the interferences, which bring significant difficulty to design predictive resource allocation. This article focuses on how to minimize the system cost when all the users’ demands are satisfied and proposes a predictive resource allocation method that can coordinate interferences in UDNs effectively. The simulation result shows that allocating resource in advance based on predictable large-scale channel can improve the transmission success rate, reduce the cost and boost spectrum efficiency significantly, compared with the method without using the predictive information.