Channel allocation in tiered cellular networks

In a tiered cellular system the cells are split into two regions: the inner tier and the outer tier. These two regions can be treated as logically separate cells. Due to the smaller radii of the inner tier cells, a frequency plan with greater reuse can be employed for the inner tier cells than for the original (untiered) system, or the outer tier cells. This results in a net increase in the call carrying capacity of the system due to tiering. In this paper, we propose a method for the determination of the frequency plan for a tiered system while maintaining the specified call quality. The call quality is assumed to be specified in the form of a minimum carrier-to-interference ratio that is to be met at each base station, under the worst-case locations of the mobiles. Under these co-channel interference constraints, the cellular system can be modeled as a hypergraph. The frequency planning method involves generating the maximal independent sets (MISs) of this hypergraph and solving a linear program whose constraints are determined by these MISs. Moreover, these frequency plans can be shown to be optimal (have the greatest call carrying capacity) asymptotically, that is, when a very large number of frequencies are available. The method is applicable to systems with either omnidirectional or sectored antennas. However the number of MISs is significantly larger in the sectored case and thus we use only a fraction of them which may make the resulting frequency plans somewhat sub-optimal. The same remark holds for systems with a large number of cells. Nevertheless, in all cases, the frequency plans obey the specified co-channel interference restrictions. The other parameters which affect the performance gain due to tiering are the inner tier area and transmit power. In a specific example, we have evaluated the optimal inner tier area and transmit power and found that their optimization results in significant capacity gains.