Date on Master's Thesis/Doctoral Dissertation


Document Type

Master's Thesis

Degree Name



Computer Engineering and Computer Science

Committee Chair

Kantardzic, Mehmed


Automatic data collection systems


Many important applications require the discovery of items which have occurred frequently. Knowledge of these items is commonly used in anomaly detection and network monitoring tasks. Effective solutions for this problem focus mainly on reducing memory requirements in a centralized environment. These solutions, however, ignore the inherently distributed nature of many systems. Naively forwarding data to a centralized location is not practical when dealing with high speed data streams and will result in significant communication overhead. This thesis proposes a new approach designed for continuously tracking frequent items over distributed data streams, providing either exact or approximate answers. The method introduced is a direct modification to an existing communication efficient algorithm called Top-K, Monitoring. Experimental results demonstrated that the proposed modifications significantly reduced communication cost and improved scalability. Also examined in this thesis is the applicability of frequent item monitoring at detecting distributed denial of service attacks. Simulation of the proposed tracking method against four different attack patterns was conducted. The outcome of these experiments showed promising results when compared to previous detection methods.