Performance of voice over IP on the WCDMA radio interface with the robust header compression scheme
Amit Mate, Mika Rinne · Proceedings of SPIE, the International Society for Optical Engineering/Proceedings of SPIE · 2001
In this paper, the protocol performance of the Voice over IP (VoIP) packet service on the WCDMA radio transport channels is studied. In a packet network, the transport channels between the mobile stations and the radio access network may carry radio bearers with different delay, throughput and quality of service requirements. Voice transmission over the packet network has two key characteristics. First, the quality of service for VoIP needs strict delay requirements for every packet, and the delay variation of the packets should not be extensively high. Secondly, the packet overhead is very large, because of the IP and UDP network headers and the RTP protocol headers present in every packet. This overhead is crucial especially for radio transmission as every bit is known to consume the scarce radio resources. A header compression scheme will significantly reduce the amount of overhead as the static fields of the network headers are needed only at the call setup time to form the context for the compressor-decompressor operation. During the call, it is necessary to send only the difference information of the changing fields, like the RTP timestamp. This enables significant compression ratios to yield headers of mean size less than two to three bytes. In this paper, we have analysed the capabilities to transport VoIP packets on a circuit-switched radio bearer and on a packet-switched radio bearer carried on the dedicated transport channel in a WCDMA cell. A special study was carried out to find out, whether the downlink transport is more efficient in the downlink shared channel compared to the dedicated channel. The analysis is represented in terms of delay, throughput and packet error ratio.