AI-Enabled Multimedia Semantic Communications

With increasing demand for mobile communication and the development of artificial intelligence, the development of the sixth-generation mobile communication network (6G) will be oriented toward more intelligent and multimodal services. It is necessary to consider the deep integration of information, communication, and data technologies and build a full life cycle resource management framework for computing, data, artificial intelligence models, and communication. To address the requirements of 6G intelligent applications and multimodal services, this article refers to the visual cognitive mechanism of the human brain to improve multimedia representation and communication efficiency and studies the architecture of semantic communication systems, privacy protection mechanisms, and quality evaluation and guarantee mechanisms. The “semantic-network-semantic” architecture was established to integrate semantic computing, intelligent communication, and endogenous security to improve transmission efficiency, service quality, and security credibility. Through prior semantic knowledge, semantic representation, and encoding transmission methods, the quality of service of semantic communication is improved. To build an efficient 6G semantic communication paradigm and system, the communication mode should be changed, and new ideas for future mobile communication should be provided.