Adaptation of CDMA Soft Handoff Thresholds Using Fuzzy Inference System

摘要

This paper proposes a new procedure to adjust soft handoff thresholds dynamically by using fuzzy inference system. This algorithm is compared with IS-95A and IS-95B/cdma2000 soft handoffs. The aims are to increase the thresholds at high traffic loads in order to release the traffic channel for supporting more carried traffic, and to decrease the thresholds at low traffic loads in order to give high quality of traffic channel. The inputs of the proposed algorithm are the number of remaining channels of each base station and the number of active pilots in active set of each mobile station. The output is the new soft handoff thresholds. In the fuzzy inference module, the triangular membership function, the max-min composition, and the weighted average formula defuzzification are selected. By comparison of all performance indicators among three algorithms, soft handoff using fuzzy inference tends to give higher performance than those of IS-95A and IS-95B/cdma2000 soft handoffs at high traffic loads and at lower soft handoff thresholds while the quality of traffic channel is still acceptable. Moreover, the wider soft handoff window size of the proposed algorithm gives high carried traffic and low blocking probability but lower quality of traffic channels. In addition, the adaptive soft handoff window size can give lower blocking probability while still keep acceptable quality of traffic channels.