A self-organizing scalable network tomography control protocol for active measurement methods

Network tomography enables operators to monitor and take action based on path measurements between nodes, often located at the network edge. State-of-the-art network tomography solutions use tools to measure performance parameters such as jitter, loss and round-trip time between multiple nodes. This paper describes and analyses a self-organizing, distributed and scalable control protocol for methods conducting continuous active measurements. Performing path available capacity measurements pose some interesting scalability challenges solved by the control protocol. The control protocol also aims at providing full-mesh coverage of end-node measurement pairs in order to give the operator a "performance map". By setting a few parameters at startup the system autonomously control the network and node overhead. Further, it is shown that the control protocol is scalable with respect to the number of nodes that is involved in the measurements.