CTC: Continuous-Time Convolution based Multi-Attack Detection for Sensor Networks
Mohammad Mezanur Rahman Monjur, Qiaoyan Yu · 2024
In heterogeneous wireless sensor network environments, distinct security vulnerabilities in hardware emerge as big concerns. Wireless sensor networks, in particular, are susceptible to various cyber-physical threats, such as jamming and relay attacks. To keep up with the evolution of attack methodologies, it is imperative to develop advanced security modules to simultaneously mitigate multiple attacks. This work proposes a Continuous-Time Convolution (CTC) based attack detection method, which offers a unified framework to address jamming and replay attacks while consuming a low overhead cost in wireless transmission setups. Experimental results show that the proposed CTC achieves a 100% detection rate for both jamming and replay attacks. Furthermore, our proposed detection scheme can be seamlessly integrated with Long-Range Wide-Area Networks (LoRaWAN) inherent characteristics, ensuring robust protection against malicious security threats.