Resource allocation using cost function in cellular networks with traffic congestion

Abstract

The demand for wireless communications and high-speed multimedia applications have been increased, but the spectrum resources are rather limited. Therefore, efficient management for radio resource is crucially important. Furthermore, since the mobile communication is rapidly increasing in downtown areas, the consideration of the traffic congested cells, i.e., hot-spots gets more indispensable. In TDMA (Time Division Multiple Access)-based systems, DCA (Dynamic Channel Allocation) is considered to be very effective to traffic fluctuation, especially in micro-cellular systems. For the practical use, DCA should be implemented in distributed fashion rather than centralized fashion. In this thesis, we propose a new distributed DCA scheme called C-DCA (Cell Classification DCA). CC-DCA classifies the cells into three disjoint sets according to the degree of traffic congestion. It is based on the cost function and release function, and the cell state information is used in computing these functions. CC-DCA alleviates the problem of clustered hot-spots as well as a single hot-spot. We evaluate the performance under realistic assumption that calls are generated by a finite user population in each cell and the performance of CC-DCA is compared with FCA (Fixed Channel Allocation) and other DCA schemes via simulations. The performance comparisons show that the failure probability with CCDCA is significantly lower than those with existing schemes.