The Performance of Voice over IP over Downlink Shared Packet Channel under Different Delay Budgets

Third-generation radio-access technology will extend beyond basic telephony: a common, IP based transport and service platform will offer mobile users a multitude of real-time and interactive services. One important issue is to port voice service to the new packet-data-based platform while retaining the quality of service (QoS) that users associate with present-day circuit- switched realizations. The main advantage of packet switched systems is flexibility in which users with no data to transmit release the channel to others instead of holding it as in circuit- switched systems. However, flexibility of this type introduces difficulty in maintaining QoS requirements, in particular, the delay requirement. In this work, we conduct research on delay performance of Voice over Internet Protocol (VoIP) services over high speed downlink shared channel, e.g. HSDPA, that is used in third generation cellular systems to provide high speed packet data service. In particular, we evaluate the VoIP network capacity under different packet delay budgets that associated with the radio access channel. This analysis provides a baseline in designing the architecture of next generation RNC and NodeB. We also compare the performance of various rate profiles, spreading factors, scheduling algorithms, code capacity, code division multiplexing capacity, and header compression mechanisms. We conclude that if well designed, the HSDPA-VoIP network capacity is comparable to that of circuit-switched voice, based on 5% packet loss and 40ms packet delay.