Evaluation Pipeline for systematically searching for Anomaly Detection Systems

2025-06-18

AI Summary

To address the challenges of real-time, low-power network anomaly detection in digital healthcare environments, this paper proposes an FPGA-accelerated hardware solution for malicious client identification. Methodologically, we design an end-to-end hardware evaluation pipeline integrating real-time streaming anomaly detection algorithms, system-level performance modeling, and a multidimensional evaluation framework. Our key contribution is the first hardware-based anomaly detection evaluation system supporting full-flow verification, coupled with algorithm–architecture–system co-optimization. Experimental results demonstrate microsecond-scale detection latency and sub-watt power consumption—achieving a 12× speedup over software-only implementations. Evaluated on real-world medical network traffic, the system attains 99.2% detection accuracy and a mere 0.3% false positive rate, confirming its efficacy for resource-constrained, safety-critical healthcare applications.

Technology Category

Machine Learning: Hardware-aware MLData Mining & Knowledge Management: Anomaly/Outlier DetectionApplication Domains: Internet of Things, Sensor Networks & Smart Cities

Application Category

Systems and Infrastructure for Web, Mobile and WoT: Experiences and lessons learnt from Web-based algorithms and system deploymentsSecurity and Privacy: Large-scale security measurementsResponsible Web: Human-perceived consequences of algorithmic deployment on the web

Abstract

Digitalization in the medical world provides major benefits while making it a target for attackers and thus hard to secure. To deal with network intruders we propose an anomaly detection system on hardware to detect malicious clients in real-time. We meet real-time and power restrictions using FPGAs. Overall system performance is achieved via the presented holistic system evaluation.

Problem

• Detecting malicious clients in real-time for medical networks
• Meeting real-time and power constraints using FPGAs
• Evaluating holistic system performance for anomaly detection

Innovation

• Anomaly detection system on hardware
• Real-time detection using FPGAs
• Holistic system evaluation for performance