Abstract: Digital semantic communication can be compatible with existing communication systems while retaining the advantages of semantic communication, and quantization is the key to realize digital semantic communication. Quantization in digital semantic communication requires multiple quantizers to quantify multi-dimensional semantic features. Due to the limited hardware and the limited number of quantization bits, a bit allocation scheme for quantizers is necessary. To solve this problem, a bit allocation algorithm based on semantic importance is proposed. Firstly, a quantized bit allocation problem based on semantic importance is constructed. Under the limit of the maximum number of bits, the importance of different semantic information is considered to minimize the distortion caused by quantization and transmission. Then, a quantization bit allocation algorithm based on reinforcement learning is proposed with the bit allocation range as the action space and the semantic feature as the state space. Finally, the proposed algorithm is trained and the optimal bit allocation strategy is obtained. The simulation results show that the proposed algorithm converges quickly. In the task scenario of image classification, the cross entropy of the proposed algorithm decreases by 48.16% compared with the benchmark algorithm, and the classification accuracy increases by 12.65%.

Key words: digital semantic communication, semantic importance, quantization bit allocation, reinforcement learning