Adaptive channel assignment for DS-CDMA cellular system

Abstract

Link capacity is defined as the number of channels available in a link. In DS-CDMA cellular systems, this is limited by the reverse link rather than the forward link. Reverse link capacity depends on the interference received at the base station (BS). This link capacity varies with such environment factors as traffic load in neighboring links, radio path loss, power control, and cell coverage area. However, conventional channel assignments have utilized fixed link capacity estimated under the worst interference. All links are assumed to have the same link capacity. Conventional channel assignments are inefficient for DS-CDMA cellular systems under varying link capacity. In this dissertation, we propose an effective channel assignment algorithm for the power-controlled DS-CDMA cellular system. In the proposed algorithm, channels are assigned adaptively according to dynamically varying environments. It is, therefore, an adaptive channel assignment (ACA). The proposed ACA is based on interference measure-ments. Utilizing this algorithm, we can significantly reduce redundant link capacity. Redundant link capacity is here defined as the difference between the actual link capacity and the fixed link capacity used in conventional static channel assignment (SCA). The proposed algorithm consists of three schemes: a nonprioritized telephony ACA (scheme-I), a prioritized telephony ACA (scheme-II), and a prioritized multi-media ACA (scheme-III),. In scheme-I, channels are assigned without distinguishing new calls from handoff calls. In scheme-II, however, a few channels are reserved and assigned only for handoff calls. A handoff failure causes the call to be cut off during conversation. This call cutoff may be a serious degradation in service grade. Service grade can be improved by reserving handoff channels. In scheme-III, a few channels are assured only for priority calls. Since the quality of service (QOS) required for image or other multi-media calls is higher than th...