Adaptive channel allocation in TACS

The paper describes an adaptive channel allocation, ACA, algorithm designed for an analog cellular system like TACS. The incentive to use more advanced radio resource allocation algorithms can be to eliminate frequency planning, to increase the capacity and improve the quality of the system. Specific to TACS is that measurements only are available in the uplink. Based on these limited measurements the radio resource allocation algorithm has to make decisions which affects both the uplink and the downlink. The ACA algorithm is divided into two parts, a slow ACA and a fast ACA part. The slow ACA part finds a set of channels for each base station based on measurements taken over a long time period, i.e, manual frequency planning is eliminated or at least reduced. The fast ACA part makes the actual channel assignment to the arriving calls using the set of channels provided by the slow ACA part. The performance of the slow and fast ACA are estimated by system simulations. The results indicate that the fast ACA increases the quality in the system which can be used to reduce the cluster size from 21 to 12 and thereby increase the capacity. The simulations also indicate that the slow ACA finds a reasonably good frequency plan. Some capacity is lost due to the fact that only uplink measurements are available