A survey of streaming data anomaly detection in network security

Area of Science:

• Cybersecurity and Network Security
• Data Science and Machine Learning

Background:

• Network anomaly detection is crucial for identifying novel cyber threats.
• Massive streaming data presents significant challenges for traditional detection methods.
• Existing anomaly detection algorithms require systematic review for network security applications.

Purpose of the Study:

• To provide a comprehensive review and analysis of recent anomaly detection algorithms for network security.
• To systematically categorize datasets, measurement techniques, detection algorithms, and results for streaming data.
• To compare the capabilities of different methods in real-world network security scenarios.

Main Methods:

• Systematic literature review of anomaly detection algorithms.
• Categorization of datasets, measurement techniques, and detection algorithms.
• Comparative analysis of algorithm performance and application scenarios.

Main Results:

• Detailed categorization of streaming data anomaly detection techniques.
• Critical comparison of various algorithms' strengths and weaknesses in network security.
• Identification of key challenges and limitations in current methods.

Conclusions:

• The study highlights the need for rapid and efficient anomaly detection in streaming network data.
• Future research should focus on improving detection accuracy and scalability.
• This review serves as a valuable resource for advancing network security research.