7.10 Conclusion

In this chapter, we explained VoIP services in the wireless network. A two-implementation architecture using H.323 and SIP is proposed, and architecture of H.323 was discussed in detail. We discussed the voice-quality issues in the implementation of the VoIP in the wireless network, and mentioned the areas where delay budgets can be improved. We proposed an implementation architecture of VoIP using the H.323 Protocol. The architecture recommends the use of direct connect with the fastConnect features of H.323 to keep the call setup delay low. We proposed also a concept of classifier to separate the control and media packets and exclude RLP function for media packet transmission. We have given a detailed analysis of the H.323 call-set model with variable frame error rate and RLP retransmission. An experimental NetMeeting setup was used to determine the call setup delay under various frame error rate conditions. The analytical results were compared with the experimental results. We proposed a model to determine the voice packet burst level blocking. This model is used to determine the capacity of VoIP using a GPRS system. The results of this model showed that using AMR coding with silent detection and CS2 coding and assuming a bundling factor of 3, considerable capacity gain is possible by using VoIP over GPRS, compared to the current GSM system. The end-to-end delay and the handoff performance of the VoIP in wireless networks are still open question. Until these two aspects are properly addressed, VoIP in wireless networks will not attain the quality comparable to current wireless voice services. But in the meantime, VoIP wireless services can be offered for business applications where the inferior voice-quality standard may be acceptable.