We investigate the validity of the analytical model of new call channel limiting scheme (NCCL) by means of simulation. NCCL limits new calls to access the total channels with parameter a while handover calls have no restriction to access the channels. We focus on the effects of users' mobility and the parameter a on the blocking probabilities of new and handover calls. We observe that the analytical model presents a good match to the simulation model for low mobility users whereas there exists a small disagreement for high mobility users. Furthermore, we find that the variability of mobile users' velocity affects the applicability of the analytical model, in particular, for small /spl alpha/ and light traffic. We extend the analytical model by including the influence of the distribution of mobile users' velocity. We show that the extended analytical model yields better agreement with the simulation results.

This paper compares the performance of the orthogonal-variable-spreading-factor (OVSF) code assignment among three proposed schemes so far, i.e., the conventional OVSF-code assignment scheme (CONV), the dynamic OVSF-code assignment scheme (DCA), and region division assignment scheme (RDA). In OVSF-CDMA, the maintenance of orthogonality among the assigned OVSF codes causes the code blocking problem which increases the high blocking rate for high-rate requests. DCA is aimed at eliminating the code blocking problem in order to reduce the high blocking rate for high-rate requests, whereas RDA has been proposed to mitigate the code blocking problem in order to efficiently provide services for all supported rates. The computer simulation results show that the blocking probability of RDA is lower than that of DCA under heavy traffic conditions. The simulation results also show that the calculation time of DCA is significantly larger than that of RDA. Furthermore, RDA is shown to be able to simultaneously handle many more voice sources than DCA.

This work focuses on the problem of efficient assignment of the orthogonal-variable-spreading-factor (OVSF) codes for multimedia communications in W-CDMA systems. Due to the problem of code blocking in OVSF-code assignment considered by Minn and Siu, and the different probability of requests for each data rate, the OVSF-code assignment algorithms proposed so far cannot efficiently serve the high-rate requests when traffic load is high. Even though Minn and Siu proposed the algorithm to completely eliminate the code blocking, a system implemented with such an algorithm has a lower number of simultaneous voice conversations than the system with the code blocking problem. To efficiently provide the services for all supported rates, we propose an OVSF-code assignment scheme called the region division assignment (RDA). The performance of the proposed RDA scheme is evaluated by simulation for voice and video sources based on the multiple access protocol in UMTS/IMT-2000.

The reuse of existing software is one of the most effective ways for software development. We have focused on the specification process with FDTs (Formal Description Techniques), and have proposed a new concept of similarity based on LTSs (Labelled Transition Systems) as a criterion to reuse specifications. However, it's definition has some problems, such that (a) it can't be applied to an LTS with some loops, and (b) the definition of similarity between actions is not clearly expressed. In this paper, we remove these problems in order for our approach to be widely applicable. For the first problem, we extend the definition of similarity to be able to apply to an LTSs with some loops. For the second one, we consider that the similarity of actions is defined based on not only the name of actions, but also the attribute of actions and the way of occurrence of actions as the temporal ordering.