Digital communication system has good anti-noise performance and high data transmission rate

For analog signals to be transmitted in digital communication systems, the necessary basic technology is the digitization of analog signals.

Coding technology classification: source coding (validity); channel coding (reliability)

Digitalization process: sampling (discrete in time) quantization (discrete in amplitude) encoding

Steps of digital transmission of analog signals: Convert analog signals to digital signals - Transmit digital signals - Restore digital signals to analog signals

Low-pass sampling theorem: fs>=2fm

bandpass sampling geography

natural sampling

Instantaneous sampling/flat-top sampling: will produce aperture distortion, and a correction function is used to solve this problem. The transfer function of the modified network is the reciprocal of the transfer function of the broadened circuit

uniform quantization

non-uniform quantization

Concept: After completing the discretization in time and amplitude, the discrete samples are converted into digital signal blocks

Codewords and code groups: M levels are represented by N binary codes, and each code group is called a codeword.

Code type: natural binary code, transmitted binary code (Gray code), folded binary code, among which folded binary code is the most superior

The larger M is, the more coding bits N are, the more complex the equipment is, the overall code transmission rate is increased, the channel utilization is lower, and the transmission bandwidth is increased.

Speech coding methods with voice channel rates lower than 64kbps are called speech compression coding technology

Features of ADPCM: Adaptive quantization replaces fixed quantization... resulting in reduced quantization errors; the adaptive prediction index prediction coefficient can be adaptively adjusted according to the statistical characteristics of the signal, improving the accuracy of the prediction signal, thereby increasing the prediction gain. The above can greatly improve S/N and dynamic range

Application: CCITT formed normative recommendations on ADPCM systems G.721 G.726

S/N=0.04fs^3/{fm^2 fH}

fs sampling frequency

fm is the frequency of the sine wave. The actual signal is not a sine wave, so fm is not fixed. It usually takes 800~1000Hz.

fH low pass filter bandwidth

Calculation of time division multiplexing rate: Rb=fs · N (number of multiplexing paths) · k (number of coding bits)

PCM30/32 channel system structure

TDM international standard