Video Quality-Aware Traffic Offloading System for Video Streaming Services Over 5G Networks With Dual Connectivity

This paper presents a video quality-aware traffic offloading system that provides seamless video streaming services to users by offloading part of the macrobase station traffic into small cells. The proposed system achieves a good balance between fairness and social welfare in terms of the video quality by allocating the radio resource of the macrobase station. The data flow is split into two parts toward the macrobase station and small cell AP, and the users simultaneously receive their data from both. In the proposed system, fountain code is employed to achieve dual connectivity enhancement by overcoming frequent packet loss in error-prone mobile networks and practical issues in the dual connectivity architecture. Furthermore, software-defined networking (SDN) technology is adopted not only to quickly react to time-varying network status but also to control network resources efficiently. The proposed resource allocation process is designed to require a low running time complexity. The proposed system is implemented using C/C++, Java, and a well-known open source, ONOS. The experimental results verify that the proposed system can achieve much better performance than other offloading algorithms.