Paper Presentation & Seminar Topics: Voice IP : Putting Voice In to Packets

Voice IP : Putting Voice In to Packets

Abstract:(seminar) When you make a telephone call, telephone exchange establishes an exclusive connection to the number you dial. While you’re having the conversation, anybody else trying to dial either party will get an ‘engaged’ tone. That’s essentially what a circuit switched network does. It establishes an exclusive and continuous physical connection between two parties. Circuit-switched technology itself has evolved quite a bit. It started with cord board switches where an operator manually connected two parties through a cord. From there, it moved on to SxS systems, also called the Strowger exchanges, named after its inventor. There were electromechanical in nature. After this came the crossbar exchanges, which are still being used in many countries. Crossbar used electromagnetic telephone exchanges are fairly common, which are more compact and powerful, and convert voice into digital signals for transmission.Digitizing sound is an interesting process. When we make a telephone call, our voice is first converted into analog electrical signals. This signal is then encoded into digital format using a technique called PCM (Pulse Code Modulation). This technique takes samples of the analog signals at a rate of 8,000 samples per second. Each sample therefore represents 125 microseconds of a voice stream, and is eight bits, or one byte long. This signal is then carried over high-speed digital lines and again decoded into an analog electrical signal at the receiving end. The analog signal is finally converted into the original sound.Speaking of sound, any conversation consists of two components-sound and silence. When the digital sound signals are transmitted over a circuit switched network, both components have to be send. Not only has that, but the order of transmitting signals also had to be retained, else quality of transmission suffers. That’s why all equipment in a circuit-switched network must be highly synchronized using expensive TDM (Time Division Multiplexing) equipment. Since sound and silence are both transmitted, a lot of bandwidth gets wasted in circuit-switched networks. In fact, one voice conversation requires a 64 kbps channel, which is quite a lot of bandwidth.