Proactive resource allocation planning with three-levels of context information
Jia Guo, Chuting Yao, Chenyang Yang · 2016
Big data analysis makes predicting the application, network and user levels of context information possible. Yet it is unclear how to exploit these information to utilize the wireless resources more efficiently. In this paper, we attempt to illustrate the potential of using these information to improve the spectrum usage efficiency. To this end, we assume that a central unit in the multicell network can predict the mobile users' request, base stations' congestion status, and user mobility pattern within a prediction window. To fully use the excess resource within the window and leave more resources for the unpredictable traffic arrived after the window, we formulate a proactive resource allocation planning problem to minimize the maximal transmission completion time. A heuristic low complexity algorithm is introduced to find the transmission plan from the problem, which determines where, when and what to transmit to the users. We use two representative scenarios to demonstrate the performance gain of the proactive resource allocation using context information over the reactive scheme that the transmission starts after the users' requests truly arrive.