Knowledge-enhanced semantic communication system with OFDM transmissions

As a promising technology to enable effective multi-modal transmission over wireless channels, semantic communication has attracted a lot of attention from academics and industries. Different from Shannon's information theory, based on common background knowledge provided by the knowledge base, the goal of semantic communication is transmitting intended useful information from the transmitter and recovered by the receiver at the semantic level. However, the existing studies on semantic communication rarely emphasize the essence and the usage of the knowledge base. In this paper, we propose a knowledge-enhanced semantic communication (KESC) system, where the knowledge base is cloud-edge-device collaborative cached. To solve the problem that float-type symbols are difficult to transmit directly through a radio frequency (RF) system, we adopt orthogonal frequency division multiplexing (OFDM) to transmit semantic vectors directly without some traditional signal processing techniques in semantic information transmission, and the semantic pilot is designed to assist semantic reception. Furthermore, we formulate a multi-encoder transformer based neural network model for the KESC system to support text transmission (KESC-T), where the decoder is implemented with a knowledge graph to enhance the performance of semantic decoding. Besides, we define knowledge-enhanced efficiency (KEE) to measure the gain in semantic recovery accuracy brought by per unit of knowledge. Simulation results demonstrate that the recovery accuracy of our proposed KESC outperforms the compared scheme, especially in low signal-to-noise ratio (SNR) or resource-constrained scenarios.