Visualization with Prediction Scheme for Early DDoS Detection in Ethereum.

Area of Science:

• Computer Science
• Cybersecurity
• Data Visualization

Background:

• Blockchain technology offers robust data protection but is vulnerable to attacks.
• Distributed Denial of Service (DDoS) attacks pose a significant threat to blockchain systems, causing transaction processing failures.
• Existing visualization schemes lack early detection capabilities, displaying only past and current data.

Purpose of the Study:

• To develop a novel visualization scheme for early detection of DDoS attacks on blockchain systems.
• To enhance the accuracy of DDoS attack detection through statistical analysis.
• To empower system managers with proactive mitigation strategies.

Main Methods:

• Utilizing polynomial regression on blockchain data to forecast future system statuses.
• Implementing a statistical analysis method using the coefficient of determination for enhanced detection accuracy.
• Developing a novel visualization scheme to display predicted future states alongside historical data.

Main Results:

• The proposed visualization scheme accurately predicts future blockchain statuses.
• The system demonstrates effective anticipation of DDoS attacks with minimal prediction error.
• The coefficient of determination enhances the accuracy of the proposed detection method.

Conclusions:

• The novel visualization scheme enables early identification and mitigation of DDoS attacks.
• Forecasting future blockchain statuses is crucial for proactive security management.
• This research provides a valuable tool for securing blockchain-based applications against sophisticated threats.