Joint coding and decoding device and method for source channel

Abstract

The invention discloses the structure and working principle of an information source channel joint codec and its codec parameter optimization design method based on vector quantization. The codec in the present invention is divided into encoder and decoder two parts, and encoder is by the input information source message vector buffer, encoding index generator, channel signal output unit, channel signal power and other encoding parameter memory, reconstruction information source Composed of message memory and coded channel signal memory; decoder consists of input channel signal vector buffer, decoding index generator, reconstructed source message output unit, decoded channel signal memory and reconstructed source message memory. The codec can directly encode a channel-optimized K-dimensional channel signal that strictly meets the requirement of average transmit power from L-dimensional source information, and realize source bandwidth compression or expansion. The codec has simple structure, low coding and decoding complexity, and strong versatility. The codec has excellent performance and has the characteristics of flexible degradation and flexible improvement, especially suitable for multi-user and time-varying channel applications.

Description

Technical field [0001] The invention belongs to the field of communications, and specifically relates to a source channel codec and a coding and decoding method. Background technique [0002] At present, most of the communication systems are built under the framework of the separated structure communication system based on Shannon's separation theory. Such as figure 1 Shown is the general model of a separate structure communication system: from the source to the sink, it is divided into five parts: source coding, channel coding, channel, channel decoding, and source decoding. [0003] The simplification and convenience brought by the separate design of the communication system objectively promote the development and application of the separate structure communication system, but it also brings some inherent shortcomings that cannot be avoided: First of all, from the perspective of feasibility, the separation of the communication system requires To achieve the overall optimum, th...

AI Technical Summary

Problems solved by technology

[0003] The simplification and convenience brought by the separation design of the communication system objectively promote the development and application of the separation structure communication system, but at the same time it also brings some unavoidable inherent disadvantages: First, from the perspective of feasibility, the separation communication system requires The overall optimality is achieved under the condition that the codec has infinite complexity and delay, which is unacceptable for real-time communication; second, the theory is no longer valid for non-ergodic and multi-user systems, that is, in these cases In this case, there is no longer a separate optimal system; third, after the channel quality drops to a certain threshold, the system immediately collapses (break down), that is, there is a threshold effect. At this time, the channel coding can no longer correct errors, so the separation system Lack of robustness to variable channels; Fourth, the separation system cannot improve the communication quality when the channel quality is improved relative to the design channel conditions, so the utilization rate of channel resources is low

Although the suboptimal Shannon direct source channel mapper for a specific source and channel characteristics can be obtained by using the power-limited channel optimization vector quantization iterative algorithm, the transmit power is not controllable at this time, specifically, although the optimization design process considers The power limit of the channel, but the encoder transmit power obtained by iterative optimization cannot guarantee to strictly meet the transmit power requirements, so it lacks practicability

Images