Improving SLIPS Performance Data Gathering

SLIPS is a Machine Learning based intrusion detection system, which scans network traffic and processes the data using a variety of techniques, both machine learning based as well as heuristic based. Despite having many different features and techniques to detect malicious behaviors on a target network, there will always be a problem of processing so much network traffic data quickly and efficiently, both in regards to the CPU and memory usage. However, there is currently no way to gather data about CPU and memory usage during execution time, which makes it difficult to make further performance improvements to the project. Therefore, this project seeks to remedy this by creating a set of tools to gather performance data about the IDS execution behavior. To improve performance data gathering capabilities, the following features will be added to the project: - CPU profiling module - Memory profiling module - Both modules will allow for: Export of profile data, output as both absolute and difference from input profile baseline (if provided), output as data stream - CPU/Memory data visualization/analysis module which can: Show process resource usage over time, can restrict to top however many if desired, histogram of resource consumption and time spent at consumption level for a single module, and recommendation lists for modules needing optimization, scoring criteria based on highest memory/CPU usage, longest blocking time, and weighted aggregate score.