تحويل بهينه جريان پخش زندهHTTP : يك رويكرد تركيبي سرور- شبكه

In recent decades, the rapid growth of data traffic an‎d the increasing deman‎d for high-quality multimedia services over Content Delivery Networks (CDNs) have introduced serious challenges in terms of performance, scalability, an‎d management. These challenges include increased latency, traffic congestion, an‎d high ban‎dwidth consumption. With the advancement of virtualization technologies an‎d the emergence of Software-Defined Networking (SDN), it has become possible to manage network resources dynamically an‎d intelligently. By utilizing the centralized control an‎d programmability capabilities of the SDN controller, resource allocation can be optimized across different network layers in a flexible an‎d adaptive manner.In this research, by integrating the features of SDN an‎d CDN architectures, an intelligent method is proposed for the dynamic an‎d optimal selec‎tion of cache servers an‎d adaptive resource allocation. The proposed method aims to minimize response time, increase cache hit ratio, an‎d enhance the overall Quality of Experience for end users. The study is conducted to design an‎d implement an optimized model that efficiently manages resource allocation an‎d cache server selec‎tion in SDN-based CDN environments. This model not only reduces response delay an‎d improves network performance but also minimizes the amount of traffic sent to the origin server.The resource allocation an‎d server selec‎tion problem is formulated as a multi-stage decision-making model. The proposed system is implemented an‎d simulated in the Python environment. The input data include client clustering information, mappings of local an‎d non-local cache servers, an‎d a set of content requests with various quality levels. The SDN controller monitors the network status an‎d dynamically activates one of the selec‎tion algorithms based on the current load condition. Simulation results demonstrate that the adaptive algorithm selec‎tion enables the system to achieve an optimal balance between delay, network load, an‎d cache hit rate.Comparative eva‎luations with the baseline model confirm that the proposed method significantly improves the average response time an‎d provides a notable increase in cache hit ratio. Overall, the proposed approach, by leveraging the intelligence of the SDN controller an‎d the integration of three complementary algorithms, offers an efficient an‎d adaptive model for resource allocation in CDN environments. This model effectively enhances network performance, reduces origin traffic, an‎d considerably improves users’ Quality of Experience.

پژوهشگاه ارتباطات و فناوري اطلاعات (مركز تحقيقات مخابرات)

6

157991