Research on Multi-Service Concurrent Transmission Control Method for Heterogeneous Networks Based on Adaptive Network Coding

The continuous rise of bandwidth-demanding and low-latency required services such as high-definition video streaming, along with the growing popularity of multi-homing mobile terminals, has encouraged research into concurrent transmission technologies in heterogeneous networks. Concurrent transmission over heterogeneous networks effectively aggregates bandwidth resources, improving transmission throughput. However, it also faces challenges such as the great dissimilarity of each path leading to head-of-line blocking issues at receiver side. Moreover, in scenarios involving concurrent transmission of multiple services over heterogeneous networks, allocating network resources from the perspective of a single service may lead to the competition for network resources of each service and result in each service failing to meet its Qos. Therefore, this paper proposes a method for multi-service concurrent transmission control over heterogeneous networks based on adaptive network coding(NC). First, the algorithm allocates suitable network resources based on each service’s Qos and path characteristics. Then it utilizes the A3C reinforcement learning algorithm to make real-time decisions through interaction with the network environments, determining the optimal encoding density for each encoding packet group, which reduces the decoding delay significantly.Simulation results show that the proposed algorithm can improve application transmission speed and reduce delivery delay, effectively enhancing user experience.