An Intelligent Transmission Method for Multimodal Data Streams in Telemedicine Based on Multi-path Transmission Over Wide Area Network

Yue Tong¹Lyu Li²Siyuan Li^3*Lei Wang¹Jun Ma⁴Ting Lyu^1*Kunlun He^1*

• ¹Medical Innovation Research Division, People's Liberation Army General Hospital, Beijing, Beijing, China
• ²School of Electronic Information and Electrical Engineering, Shanghai Jiao Tong University, Shanghai, China
• ³School of Medicine, Shanghai Jiao Tong University, Shanghai, Shanghai Municipality, China
• ⁴Beijing Redcdn Technology Co.，Ltd, Beijing, China

This paper presents an intelligent multipath transmission method for multimodal data streams over wide area networks (WANs), addressing challenges such as network delay and packet loss in telemedicine scenarios. Utilizing a content-aware approach, the proposed method assigns transmission priorities to different types of data and selects optimal paths via an intelligent routing algorithm based on real-time network conditions. The system incorporates mechanisms for dynamic path switching, load balancing, and fault tolerance. Experimental results demonstrate that under conditions of link interruption and insufficient bandwidth, the proposed method achieves significantly better transmission quality compared to traditional approaches. For instance, in scenarios with 40% packet loss, the average network delay for high-priority data was reduced to 141.8ms, compared to 297.2ms and 318.2ms for conventional methods. Additionally, with bandwidth reduced to 2 Mbps, the high-priority data stream's average bit rate and frame rate were maintained at 2.67 Mbps and 29 FPS, respectively, ensuring reliable performance during critical operations.