PROJECT TITLE :
Big Data for Autonomic Intercontinental Overlays
This paper uses huge knowledge and machine learning for the important-time management of Web scale quality-of-service (QoS) route optimisation with an overlay network. Primarily based on the collection of information sampled each two min over a giant range of supply-destinations pairs, we have a tendency to show that intercontinental Web protocol (IP) paths are so much from optimal with respect to QoS metrics like end-to-finish spherical-trip delay. We tend to, so, develop a machine learning-based mostly scheme that exploits giant scale data collected from communicating node pairs in a multihop overlay network that uses IP between the overlay nodes, and selects ways that offer substantially better QoS than IP. Impressed from cognitive packet network protocol, it uses random neural networks with reinforcement learning primarily based on the large data that's collected, to pick out intermediate overlay hops. The routing scheme is illustrated on a twenty-node intercontinental overlay network that collects some 2 × 106 measurements per week, and makes scalable distributed routing choices. Experimental results show that this approach improves QoS significantly and efficiently.
Did you like this research project?
To get this research project Guidelines, Training and Code... Click Here