by Jack Chang, COO

TelTel

7/5/2005

Peer-to-Peer VoIP services have proved to be very popular. Skype claims to have more than 10 million users. However, Skype and other peer-to-peer providers use proprietary technology to deliver their services. This results in potential quality of service, security and reliability issues, as there is no way to manage pure peer-to-peer routing. Additionally, the ability to improve quality and provide new services through collaboration is tough to come by in the pure peer-to-peer environment. The limitations of peer-to-peer can be overcome through the introduction of Session Initiation Protocol (SIP).

One of SIP’s great advantages is in its flexibility to be deployed as a centralized or distributed system, or a mixture of the two. As a distributed system, it offers advantages over a centralized solution in terms of network scalability and robustness. While deploying SIP as a distributed system solves certain problems, distributed users and authentication still has to be addressed. If SIP is to be deployed with a peer-to-peer design in mind and with most of the intelligence pushed to the user agents, it is possible to offload server responsibilities and reduce operational problems when dealing with server scalability and performance.

The flexibility of SIP allows for the best of many worlds: a centralized call system, a distributed server model, a peer-to-peer model -- or even a hybrid of each. To leverage the advantages of SIP in any configuration, significant engineering and operational effort is required.

Aside from its technical advantages, SIP offers an advantage from the collaboration that’s taken place in the voice industry. This open collaboration provides an unprecedented level of interoperability between voice offerings that has not been previously seen in the voice industry due to the typically proprietary nature of previous voice protocols.