Impacts of data call characteristics on multi-service CDMA system capacity

The capacity of multi-service Code-Division Multiple Access (CDMA) systems has been extensively studied in the literature. However, few studies address the fundamental issue of how the stochastic properties of non-Poisson data traffic affect the system capacity. This paper studies a CDMA system supporting voice and data traffic. Results show that increased variability in the data call arrival process decreases the system capacity, while increased variability in data call holding times increases the system capacity. The extent of these effects depends on other system parameters, such as transmission rates and communication quality requirements. These observations motivate a simple buffer-based resource management scheme that enhances the system capacity in the presence of high-variability data traffic, providing controllable performance tradeoffs between voice and data calls. This study uses both simulation and theoretical analysis, which is based on a Markov Regenerative Process (MRGP) model.