Advanced sensor networks are anticipated to deliver more innovative and cost-efficient monitoring solutions than conventional ones. Low-power technologies, such as Long-Range Wide-Area Network (LoRaWAN), are being integrated into advanced sensor networks to meet the growing need for economical automation and remote surveillance. Nonetheless, adopting LoRaWAN introduces novel security risks to hardware. This work explores a detection technique at the hardware tier to detect multiple attacks with a single detection scheme. This work proposes an Advanced Continuous Time Convolution (ACTC) approach to defend against several attack scenarios with classification. Experimental results show that the ACTC detection system can withstand jamming and replay attacks with 100% accuracy and an F1-score of 1.0, assuring a secure communication channel1.

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.

The advanced manufacturing industry (AMI) faces many unique challenges from the cyber-physical domain. Security threats are originated from two integral parts: software and hardware. Over the past decade, software security has been addressed extensively, but hardware security has not received enough attention. This work analyzes the security vulnerabilities of typical electronic devices deployed to AMI and proposes three attack models for sensing nodes, local storage and processing edge devices, and wired/wireless communication interfaces, respectively. Practical security attacks on hardware are demonstrated in this work to inspire the development of feasible countermeasures against hardware Trojans, fault injection attacks, and external signal interference. Moreover, this work highlights the unique security challenges posed by advanced manufacturing applications. To mitigate those security attacks in AMI, this work suggests guidelines for the defense method design that can effectively protect the hardware in AMI.

More and more digitized techniques and network connectivity are deployed to advanced manufacturing to enable remote system monitoring and automated production; however, this trend also leads to the traditional assumption of security in manufacturing not holding true any longer. For instance, the option of remote access makes advanced manufacturing infrastructures vulnerable to various security attacks from physical devices to cyberspace. Existing literature that addresses the attacks in advanced manufacturing is mainly at the network level. In this work, we study the role of hardware security in the process of advanced manufacturing. More specifically, our analysis focuses on the security vulnerability of sensors, local data processing nodes, and the interface implementation for standardized communication protocols. Unique attack examples such as hardware Trojan, interface sniffing, and fraudulent data injection attacks are provided in this work to highlight the unique challenges of attack detection and mitigation in advanced manufacturing.

Approximate computing (AC) emerges as a promising approach for energy-accuracy trade-off in compute-intensive applications. However, recent work reveals that AC techniques could lead to new security vulnerabilities, which are presented in a format of visionary view. There is a lack of in-depth research on concrete attack models and estimation of the significance of the attacks on approximate arithmetic computing systems. This work presents several practical attack examples and then proposes two attack models with quantitative analysis. Input integrity check and exclusive logic based attack detection methods are proposed to address the attacks on AC systems. The experimental results show that the attack detection failure rate of our method is below $2.2*10^{-3}$ and the area and power overhead is less than 6.8% and 1.5%, respectively.