Real-Time Intrusion Detection System

Project

• Uses basic radio communication techniques to detect intruders in real-time situations
• Relies on Linux-based GNU Radio software and Ettus Research Universal Software Radio Peripheral (USRP) devices
• Intruder detection metrics all available through a single processing block
• Multi-antenna data collection to improve detection rate and increase detectable area
• Baseline for further research on target detection and tracking using Wi-Fi signals

System

System Overview

gr-RID Software

Methods

Decision Metrics

• Amplitude standard deviation (purple and red boxes vs. green)
• Energy of standard deviation (purple and red boxes vs. green)
• Level thresholding (significant in red vs. green)

Detection Processing

• Collect data for duration set by user
• Set thresholds and arm system
• Check if thresholds exceeded

Conclusion

• Consistent detection (as mapped below) for baseline testing and experimentation
• Low sample rate requirements allow gr-RID to be implemented on lowercost and lower-capability radio communication hardware.
• Data collection indicates that it is possible not only to detect intruder motion and presence, but general location as well. This is critical for tracking.
• gr-RID is a valid method of detection in research that has developed since the conclusion of this project.

Detection Heat Map Engineering Innovations Building, 2ndFloor