Basic principles and related technologies of VoIP (1)

Voice communication over the Internet is a very complex system engineering and has a wide range of applications. Therefore, many technologies are involved. The most fundamental technology is the VoIP (Voice over IP) technology, it can be said that Internet voice communication is the most typical and promising application area of VoIP technology. Therefore, before discussing the use of the Internet for voice communication, it is necessary to analyze the basic principles of VoIP and related technical problems in VoIP.

I. Basic VoIP transmission process

The traditional telephone network transmits voice in a circuit exchange mode. The required transmission bandwidth is 64 kbit/s. The so-called VoIP is based on an IP group exchange network as the transmission platform, which compresses, packs, and other special processing of Analog voice signals, so that it can be transmitted using a connectionless UDP protocol.

Several elements and functions are required to transmit voice signals over an IP network. The simplest form of network is composed of two or more devices with VoIP functions, which are connected through an IP network. The basic structure of the VoIP Model is 2-18. We can see how VoIP devices convert voice signals to IP data streams and forward these data streams to IP addresses. IP addresses then convert them back to voice signals. The network of the two voices must support IP transmission and can be any combination of IP Routers and network links. Therefore, the transmission process of VoIP can be divided into the following phases.

1. Speech-Data Conversion

The voice signal is a simulated waveform that uses IP addresses to transmit voice data. For real-time and non-real-time applications, the channel must first perform analog data conversion on the voice signal, that is, the analog speech signal is quantified by 8 or 6 bits and then sent to the buffer storage area. The buffer size can be determined based on the latency and encoding requirements. Many low bit rate encoders are encoded in frames. The typical frame length is 10 ~ 30 ms. Considering the cost of transmission, the inter-language package usually consists of 60, 120, or MS of Speech data. Digital can be achieved using a variety of voice encoding solutions, the current use of voice encoding standards mainly include ITU-T G.711. The source and destination voice encoder must implement the same algorithm, so that the destination Voice device can restore the analog voice signal.

2. Convert original data to IP Address

Once the voice signal is digitally encoded, the next step is to compress the voice packet with a specific frame length. Most encoders have specific frame lengths. If an encoder uses 15 ms frames, it divides the packet from the first 60 ms into four frames and encodes them in sequence. 120 audio samples are merged for each frame (the sampling rate is 8 kHz ). After encoding, four Compressed Frames are combined into a compressed voice package and sent to the network processor. The Network Processor adds headers, time scales, and other information for the voice and then transmits the information to the other end point through the network. The voice network establishes a physical connection (one line) between the communication endpoints and transmits encoded signals between the endpoints. Unlike a circuit exchange network, an IP network does not form a connection. It requires that data be stored in a variable-length datagram or group, and then the addressing and control information of each datagram is assigned and sent over the network, one site and one site are forwarded to the destination.

3. Transfer

In this channel, all networks are considered to receive voice packets from the input end, and then transmit the packets to the output end of the network within a certain period of time (t. T can change within a certain range, reflecting the jitter in network transmission. The same node in the network checks the addressing information attached to each IP address and uses this information to forward the datagram to the next stop in the destination path. A network link can be any extended structure or access method that supports IP data streams.

4. IP package-Data Conversion

The destination VoIP device receives the IP address and starts processing. The network provides a buffer with a variable length to adjust the network jitter. The buffer can accommodate many voice packets. You can select the buffer size. A small buffer has a low latency, but cannot adjust a large jitter. Next, the decoder decompress the encoded voice package to generate a new voice package. This module can also be operated by frame, which is exactly the same length as the decoder. If the frame length is 15 ms, the 60 ms voice packets are divided into 4 frames, and then decoded and restored to 60 ms voice data streams and sent to the decoding buffer. Remove addressing and control information, retain the original data, and then provide the original data to the decoder.

5. convert digital speech to analog speech

The playback drive extracts the voice sample points (480) from the buffer and sends them to the sound card for broadcast at a predetermined frequency (such as 8 kHz. In short, the transmission of voice signals over an IP network must go through analog signal to digital signal conversion, encapsulation of digital voice into IP groups, transmission of IP groups through the network, unpacket handling of IP groups, and digital voice restoration. to analog signals. The entire process is 2-19.