Abstract: A network analysis architecture provides a suite of complementary logic operable at different temporal and spatial timescales. The distinct temporal and spatial scales define different tiers, each analyzing network events according to predetermined temporal and spatial scales of progressive magnitude. Particular event detection logic may be operable on an immediate temporal scale, while other logic identifies trends over a longer time period. Similarly, different spatial scales are appropriate to different algorithms, as in logic that examines only headers or length of packets, or inspects an entire payload or transferred file. Deployment of logic that is focused on different timing and scope of data allows timely action in the case of readily apparent deviations, and permits longer term analysis for identifying trends that emerge over time.

Abstract: A network analysis architecture provides a suite of complementary logic operable at different temporal and spatial timescales. The distinct temporal and spatial scales define different tiers, each analyzing network events according to predetermined temporal and spatial scales of progressive magnitude. Particular event detection logic may be operable on an immediate temporal scale, while other logic identifies trends over a longer time period. Similarly, different spatial scales are appropriate to different algorithms, as in logic that examines only headers or length of packets, or inspects an entire payload or transferred file. Deployment of logic that is focused on different timing and scope of data allows timely action in the case of readily apparent deviations, and permits longer term analysis for identifying trends that emerge over time.