The audio conferencing is a useful application in many areas for its underlying property of holding a conference without people getting together in the same location.
An audio conferencing can be considered as a multiple Voice-over-Internet Protocol (VoIP) to deliver media streams through the Internet, but there are some difference between the audio conferencing and VoIP in setting up a communication session and media streaming control. What we can figure out is that when VoIP grows more and more prosperous, the audio conferencing will boom as well.
After VoIP population grows quickly in recent years, research in audio conferencing seems to be an appendage to VoIP. The research is slack and falls into a changeless pattern. Some problems of existing audio conferencing models now are not solved completely and flawless. In this thesis, a new audio conferencing system that takes the advantage of Peer-to-Peer network is proposed. Our system integrates all benefits of existing audio conferencing models into an effective mechanism which accelerates the setting up of an audio conferencing.
Our proposed system not only integrates the benefits of existing audio conferencing models and Peer-to-Peer networks but also improves the media streaming control from a similar Peer-to-Peer conferencing control protocol. Using open, simple and flexible SIP [3, 4] with our simple mechanisms, we could make the conferencing system efficient and flexible.
In the subsequent discussion, we will introduce VoIP, what the audio conferencing looks like, and popular Peer-to-Peer networks.
1.1 Voice-over-Internet Protocol
Voice-over-Internet Protocol, VoIP as it is often referred to, has been a subject of interest almost from first computer network. Voice had been transmitted over the early Internet since 1973 [1].
In recent years, VoIP is steadily changing the telephony world. Traditional phone lines (PSTN system) are slowly quitting from the market and people around the world embrace the benefits and features that VoIP technology offered. It is worthy to review the history of VoIP as the evolution accelerates.
Figure 1.1 VoIP common topology connection. [27]
The history of VoIP shows that it starts from early 1980s, this technology for delivering voice conversations over the Internet has been available to end-users for many kinds of aspects and purposes. In 1995, Vocaltec, a small company released the first internet phone software. This software was designed to run on a PC and it was much like PC phones, now it is used by many people. This software was called
"Internet Phone" which utilized sound cards, microphones and speakers and it used the H.323 [2] protocol instead of SIP protocol which is more popular today. It was the evolving time of the Skype [5] from the mid 90s to now. "Internet Phone" did not support the communications from gateway to the PSTN system at that time, so it was only possible to communicate with other Vocaltec "Internet Phone" on another PC, but not another phone. Level 3 began the development of its first softswitch until 1997 to 1998; Softswitches were designed to replace traditional hardware telephone switches by serving as gateways between VoIP telephones [6].
1.2 Audio Conferencing
There are often three or more participants in a real-time conversation. In a three or more participants’ conversation, each one will express his thoughts and listen to others in turns. Multiple participants are likely to speak simultaneously or may interrupt another who is speaking in many scenarios, it causes the double-talk. A face-to-face conversation is one of approaches that all participants reside in the same physical location, for instance, a meeting room. When time goes by, there are increasing needs for people to communicate across geographic locations with the globalization of activities. These needs bring the development of systems that enable face-to-face like communications with high speech quality in long distance.
A multiparty audio conferencing system makes it happens that a group of people to participate in a real-time audio communication session with multiple people speaking at the same time possibly. Although video is a better additional feature, audio with sufficient quality is a necessary condition for almost collaboration
mixing and delivery, such as audio capture, acoustic echo cancellation (AEC), automatic gain control (AGC), and audio compression.
1.3 Peer-to-Peer (P2P) system
Meanwhile, Peer-to-Peer systems have been used in many application areas for recent years. Although they are primarily used for client-server style systems now, many Internet protocols such as FTP were designed to be Peer-to-Peer.
Applications of Peer-to-Peer have been the focus point in recent years with the widely use of Napster [24] which is a file sharing system intended for the unique purpose of sharing music files. Napster was only a P2P application contributing to file transfers; joining the network and performing a search for files were cooperated with a central server [23].
Fully distributed P2P systems can be created to allow system structures built without the control of centralized servers; therefore fully distributed P2P systems can avoid potential network bottleneck and risks of single point of failure. Peers directly communicating with other peers should be theoretically able to improve performance in most cases by abandoning cooperation with central points of control responsible for governing P2P network structure in a group communication. And P2P system applications are quite diverse, from file sharing [7], to media streaming [8, 9], to game playing [10].
The key idea behind such P2P systems is devoted to distributing processing loads and bandwidth requirements by sharing resources among different peers. This idea has been extended by Skype, Kundan [11] and Bryan [12] to demonstrate the possibility of extending P2P networks to support voice services. Skype apparently
uses many P2P techniques, but unfortunately its system is proprietary and closed.
Now there are other works that explore inherently Peer-to-Peer VoIP protocols, for example－ SIP. Kundan and Bryan’s previous works demonstrate the integration of the P2P Chord [13] algorithm within SIP.
The rest of this thesis is organized as follows. Chapter 2 provides an overview of VoIP, P2P-SIP and some structure of audio conferencing systems. Chapter 3 describes the problem in audio conferencing systems and our proposed scheme in detail. The performance analysis and simulation evaluation are presented in Chapter 4.
Conclusions are stated in Chapter 5