RICHTER: hybrid P2P-CDN architecture for low latency live video streaming
0 likes•31 views
The document presents the 'Richter' architecture, a hybrid peer-to-peer (P2P) and Content Delivery Network (CDN) system designed for low latency live video streaming. It utilizes modern networking paradigms such as edge computing and network function virtualization to optimize video delivery by dynamically choosing the best source for content based on latency and available resources. Future work includes developing a model to guide system operations based on identified optimization challenges and validating the approach through experimentation.
RICHTER: hybrid P2P-CDN architecture for low latency live video streaming
- 1. RICHTER: Hybrid P2P-CDN Architecture for Low Latency Live Video Streaming Reza Farahani1 , Hadi Amirpour 1 , Farzad Tashtarian1 , Abdelhak Bentaleb 2 , Christian Timmerer 1 , Hermann Hellwagner 1 , and Roger Zimmerman 2 1 Christian Doppler Laboratory ATHENA, Alpen-Adria-Universität, Klagenfurt, Austria 2 School of Computing, National University of Singapore, Singapore Introduction Content Delivery Network (CDN) and HTTP Adaptive Streaming (HAS) are considered the principal video delivery technologies over the Internet. Challenge in CDN- and HAS-based video streaming Despite the wide usage of CDN and HAS technologies, designing cost‐effective scalable flexible architectures that support low latency and high quality live video streaming is still a challenge. To address this issue, we leverage existing works that have combined the character‐ istics of Peer‐to‐Peer (P2P) networks and CDN‐based systems and introduce a hybrid CDN‐P2P live streaming architecture. Hybrid CDN-P2P architectures When dealing with the technical complexity of managing hundreds or thousands of concurrent streams, such hybrid systems can provide low latency high quality through enabling the delivery architecture to switch between the CDN and the P2P modes. However, modern networking paradigms have not been extensively employed to design such systems. Contributions We employ modern networking paradigms such as Edge Computing Network Function Virtualization (NFV) Distributed video transcoding to introduce a hybRId P2P‐CDN arcHiTecture for low LatEncy live video stReaming (RICHTER). RICHTER employs virtualized edge servers to serves clients’ live video stream re‐ quests through fetching them from P2P network or CDN, or edge servers. transcoding them from higher qualities at the best peer or the edge server. to obtain the lowest latency values. RICHTER Architecture Live video upload L L L L CDN Network Live video upload Virtual Tracker Server (VTS) L Seeders Leechers P2P Network Peer Transcoder Edge Transcoder gNodeB CDN Server VTS (PC) Origin Server Peer (Tran.) VTS (Tran.) VTS (Tran.) 2 4 3 5 6 7 Action Tree Clients 1 RICHTER Contribution Partial Cache (PC) Peer Figure 1. RICHTER architecture We propose RICHTER architecture consisting of three core components: CDN network The CDN network is constructed by multiple CDN servers and an origin servers: CDN servers contain various parts of video sequences. Origin server contains all video segments in multiple representations. RICHTER fetches requested qualities or higher qualities (for running trascoding function) from them to serve clients’ request. P2P network The P2P network includes two types of peers: Seeders’ requests can be served by CDN servers, the origin server, and VTSs. Leechers’requests can be served by adjacent peers and VTSs. Given the continuous increases in smartphone capabilities, e.g., high broadband bandwidth access to the Internet energy resources hardware‐accelerated video transcoding RICHTER utilizes the peers’ resources to provide a distributed video transcoding approach besides video transmission. Virtual Tracker Servers (VTS) VTS servers are located close to base stations (gNodeB) and equipted with Partial cache to serve popular clients’ requests immediately. Transcoding function to serve clients’ requests from existing higher qualities directly. The clients’ live video stream requests are directed to the VTSs, and then they are answered based on the VTSs decisions. Proposed Approach The following research questions will be taken into account by RICHTER. Research questions 1. Where is the optimal place (i.e., adjacent peers, VTS, CDN servers, or origin server) in terms of lowest latency for fetching each client’s requested content quality level from, while efficiently utilizing the available resources? 2. What is the optimal approach for responding to the requested quality level (i.e., fetch or transcode) 3. How many seeders and VTSs are sufficient to serve the leechers? 4. How to replace seeders when one of them leaves the system? To answering the aforementioned research questions, VTS servers track associated clients’ requests and store a mapping between all transmitted videos and all served clients in its peer‐map list. monitor the system frequently to obtain precise information about the available bandwidth to reach each CDN server and peers’ available resources. Therefore, when a VTS receives a new request, it can find the optimal solution (i.e., in terms of minimum latency) from the Action Tree (Fig. 1) All Feasible Actions 1. Use the P2P network and transmit the requested quality directly from the best adjacent peer (action 1 in the action tree). 2. Transcode the requested quality from a higher quality at the best adjacent peer and transmit it through the P2P network. 3. Fetch the requested quality directly from the edge, i.e., the VTS. 4. Transcode the requested quality from a higher quality at the VTS. 5. Fetch the requested quality from the origin server. 6. Fetch a higher quality from the best CDN server and transcode it at the VTS. 7. Fetch the requested quality from the best CDN server. Future Plan The next steps of this paper are listed as follows. 1. We model the adaptive live low latency streaming as an optimization problem to guide the system operation according to RICHTER’s action tree (Fig. 1), and answer the aforementioned questions. 2. We propose a near‐optimal heuristic approach with minimum overhead that could be run practically on the proposed VTSs. 3. Finally, this approach is validated through experiments on a real‐world testbed, including hundreds of clients, and its performance is compared to other state‐of‐the‐art approaches. https:/ /www.athena.itec.aau.at Mile High Video (MHV) 2022, Denver, USA [email protected]