Available Knowhow
Maximizing network traffic by predicting upcoming events
Categories |
Computer networks, Computer architecture |
Development Stage |
Ongoing research |
Knowhow |
Available knowhow |
Market |
Technology for P2P streaming networks, DPI (deep packet inspection), ISP devices |
Highlights
-
Software for predicting computer events based on recognition of network traffic patterns
-
In P2P communications, provides a solution to unstable download rates by predicting upcoming traffic loads
-
In DPI appliances, provides a solution for recognizing protocols based on behavioural evaluation
Our Innovation
System for predicting events by learning the behaviour patterns of protocols. For example, the system examines packets being transferred from a machine and, based on the pattern of bits in the packets, predicts whether they will result in a heavy load on the uplink to another machine without needing to examine deeply the specific content of those packets or to analyze the specific protocols.
Key Features
-
System not restricted to specific protocols - works with existing protocols and learns new protocols
-
More generic than other systems in recognizing protocols because it does not require typical deep packet inspection (DPI) processes that embed previous knowledge about the protocol
-
Enables stable download rates for P2P streaming for Internet TV, CDNs and VoD networks
-
Enables Internet service providers (ISPs) to predict and recognise heavy P2P traffic and take timely action
The Opportunity
-
Streaming Media Market to Grow 135% Over 5 Years
-
Internet video is now approximately one-third of all consumer Internet.
-
The cost to Google of YouTube bandwidth is a minimum of $1 million a day.
-
Our technology offers better QoS for P2P streaming networks that are cost effective compared to traditional streaming networks.