| （英） |
Recent advances in IoT technology have made it increasingly feasible to build systems that collect and process region-specific data at the edge and provide real-time regional information.
However, if large volumes of raw data are transmitted without considering the recipient's location or requirements, network traffic increases and the risk of private information leakage beyond the region also rises. Therefore, regional data must be abstracted according to the user context before delivery.
To address these challenges, this study proposes a content delivery method based on semantic communication, which extracts only the semantic information from the data, transmits it, and reconstructs it on the receiver side.
Through image transmission experiments simulating XR content, we confirmed that the proposed method reduces communication traffic by approximately 40 percent while maintaining image quality (SSIM of at least 0.9), demonstrating both the potential and the challenges of user-context-adaptive content delivery. |