Introducing Media PeerPool: A cost-saving solution for media platforms
Media PeerPool reduces bandwidth costs for media platforms by utilizing peer-to-peer browser functionality. It operates as a peer-assisted solution, where all individuals downloading a specific file or viewing the same stream share content with each other via WebRTC, based on location and latency. This reduces the load on the network, significantly improving overall performance and reducing costs by up to 99%.
How it works
Media PeerPool shifts the CDN load back to the clients, following the principles of edge computing. By utilizing current web standards such as Media Source Extensions audio and video tags and WebRTC, the library is able to natively solve the challenges of peer-to-peer video streaming in the browser.
graph TD; A(CDN) --> B(Edge Server); B --> C(End-User); A --> D(Media PeerPool); D --> E(Peers); E --> C;
|Feature||Media PeerPool||Traditional CDN|
|Cost Savings||Up to 99%||N/A|
|Network Performance||Improved||Depends on location of edge servers|
|Integration||CDN agnostic||Requires specific CDN|
|Scalability||Can handle millions of users||Limited by number of edge servers|
Trackers are the servers that keep track of the location and availability of files on the network. They monitor which peers have a specific file and track their location and latency. This information is then used by the libraries to locate available peers and establish connections.
Libraries are the client-side code that runs on the user's device. They communicate with the trackers to locate available peers and establish connections with them. They also handle the transfer of data between peers using WebRTC. By shifting the load from the CDN to the clients, libraries help to reduce the load on the network and improve overall performance.
Peer-to-peer systems are the connections established between users' devices. They allow for direct data transfer between peers without the need for a centralized server. This reduces the load on the network and improves overall performance.
Media PeerPool is designed to work with any CDN, it can be integrated as a standalone component and can save up to 99% of your current CDN costs. The system uses ISP location and geolocation awareness to build WebRTC clusters, where peers from the same provider's network and in close proximity are assigned to the same cluster. This way, the system can fulfill every request from peers within its cluster, and in the case of live streaming, if the peer is unable to receive the data from the p2p network in time, it will request it from the CDN instead. Additionally, it uses the STUN protocol to bypass NAT connectivity issues and establish a direct connection between peers over WebRTC, even for connections behind multiple proxies or heavily restricted firewalls.