Research on Network Monitoring and Measurement Improvements for LILA links

At the October 2006 NSF International Research Network Connections (IRNC) WHREN-LILA program review, the following recommendations were made:

• Investigate the possibility of collecting netflow data on the IRNC funded LILA links
• Analyze the the netflow data to assure appropriate use of the assets

As a result of this recommendation, the NSF awarded FIU's proposal for a Research Experience for Undergraduates (REU) program.

Funding this REU request will have a direct and positive effect on helping the WHREN-LILA team to respond to recommendations of the IRNC program review panel.

The Project

This REU project will build upon the software base built in the first REU project. In the first REU project, FIU students modified existing tools to enable the creation of dynamic and real-time network views which provide improved understanding of traffic flow characteristics for networks that connect to the AMPATH international exchange point in Miami.

In this project, the functionality will be extended beyond a set of independent routers. The proposed research will improve understanding of traffic patterns and anomalies across the LILA links. The proposed REU students will be included in planning discussions, development, documentation and execution of the proposed research activity. They will be involved in the experiment design to collect netflow data using the existing flow tools developed from the previous REU award to integrate Cisco NetFlow technology and MonALISA. The REU students will assist in the interpretation of the netflow data, which will then lead to the extension of the current tool set.

The focus of this REU research project has two perspectives.

• Increase understanding of traffic patterns from a coordinated perspective. A coordinated perspective would permit viewing flow traffic, from multiple sources, as a correlated system.
• Increase awareness of network anomalies by measuring Round Trip Time (RTT) variations.

Goals

• Extend the current infrastructure to incorporate data collection from various routing domains (autonomous systems)
• Include a function for round trip time (RTT) sensitivity and incorporate router statistics (CPU/memory utilization)
• Detect anomalies using RTT / router performance sensitivity to then identify the source using the NetFlow correlated data