Monitoring, measurement, and control of multimedia traffic in IP networks
In this thesis, we propose several architectural components for monitoring, measurement, and control of multimedia traffic in IP networks. These architectural components are used to monitor and measure multimedia traffic at the edge of IP networks and in the core of IP networks, and to serve stored multimedia contents to clients in edge networks.
We first present a technique to detect multimedia relay traffic from passively measured packet traces collected at the edge of the Internet. Recently, networked application developers have started to use end-users' computers as relay nodes. These relay nodes incur costs to both users and network operators, in the form of increased bandwidth consumption. Among various streaming relayed traffic, we focus on characterizing and detecting relayed traffic generated by Skype, a popular voice over IP application that uses relays. We propose several flow-level metrics to characterize the nature of relayed traffic. These metrics together with the results obtained from the experimental characterization of Skype-relayed traffic are used to develop techniques to detect Skype-relayed traffic traversing the access point of a large network. We find that the metrics proposed can be applied more broadly in the detection of relayed traffic generated by other multimedia applications.
We next present a concurrent passive monitoring architecture for IP flows at multiple locations within an IP network. The objective of such a distributed monitoring system is to sample packets belonging to a large fraction of IP flows in a cost-effective manner by carefully placing monitors and controlling their sampling rates. We consider the problem of where to place monitors within the network and how to set their sampling rates. It is particularly important to sample packets in a cost-effective manner for monitoring multimedia traffic, since multimedia traffic has high monitoring bandwidth requirements. To address the tradeoff between monitoring cost and monitoring coverage, we consider minimum cost and maximum coverage problems under various budget constraints. We formulate several problems and show that they are NP-hard. We propose greedy heuristics, and show, using synthetic and real network topologies, that the heuristics provide solutions quite close to the optimal solutions. In addition, our experiments show that a small number of monitors is often enough to monitor most of the traffic in an entire IP network.
Last, we introduce and evaluate the Push-to-Peer architecture for streaming video among cooperating nodes in an edge network, that can drastically reduce the load posed on the core and access links of the network and also on the streaming servers. The main departure from previous designs is that content is proactively pushed to peers and persistently stored before the actual peer-to-peer transfers. The initial content placement increases content availability and improves the use of peer uplink bandwidth. Our specific contributions are: (i) content placement and associated pull policies that allow the optimal use of uplink bandwidth; (ii) a performance analysis of the policies in the case of a controlled environment such as DSL networks under ISP control; (iii) a distributed load balancing strategy for selection of serving peers; (iv) distributed strategies to cope with dynamic uplink bandwidth.