From Vulnerability to Resilience: Securing Signal Transmission Against Jamming and Spoofing Attacks

Wireless signal transmission underpins crucial aspects of modern communication, but its susceptibility to jamming and spoofing attacks remains a major concern. These attacks have the potential to disrupt vital services, mislead devices, and compromise the integrity of wireless systems.

Many researchers have addressed anti-jamming and anti-spoofing techniques. Anti-jamming methods include coding techniques, specialized waveforms (e.g., spreading, beamforming, and modulation), and spatial avoidance using relays or reflectors. Anti-spoofing methods include angle-of-arrival detection with antenna arrays, cross-checking with sensor fusion, and authentication and encryption. In recent decades, artificial intelligence and machine learning have boosted the research in anti-jamming and anti-spoofing, making them more flourish.

Despite these advancements, significant challenges persist. For example, while various jamming resistance studies were proposed, their application to vulnerable 5G and narrowband Internet of Things (NB-IoT) communications are unexplored. Additionally, while low-density sparse coding (LDSC) is advantageous for multiplexing and spreading, research into the design of the code itself is lacking. Furthermore, sidelinks, a key component of 5G Advanced and future generation communication, hold the potential to become stealthy, secure channels for countering jamming attacks. As for machine learning based methods, they are limited in theoretical and simulation, instead of applied to commercial ready codebases. In GPS anti-spoofing, existing solutions are often expensive, bulky, or low in accuracy, while authentication and encryption approaches remain restricted to military use. Furthermore, although distributed mobile spoofer analysis has been theoretically explored, it still lacks real-world implementation studies.

Recognizing the increasing complexity of the wireless landscape, this thesis addresses these open problems through multiple works targeting anti-jamming and anti-spoofing. The first work develops a standard-compliant spread spectrum waveform for NB-IoT applications under jamming. The next focuses on LDSC design, applying it to spreading, and devising methods to obtain sub-optimal LDSC designs against interference. Another work proposes a secure, stealthy sidelink underlay channel and develops interference cancellation methods for flexibility and resilience. In the last work of anti-jamming, a machine learning based interference mitigation solution was proposed, running on open sourced industrial standard, narrowing the gap between academic and real world implementation. On the GPS anti-spoofing front, the first work presents a low-cost, smartphone-based spoofing detection solution matching the accuracy of antenna-array methods. The next work leverages crowdsourcing and sensor fusion, enabling high-accuracy and low-latency anti-spoofing. Finally, the thesis implements conceptual distributed GPS spoofer using low-cost software-defined radios (SDRs), addressing multi-spoofer challenges.

Overall, this work offers vital contributions to the security and resilience of wireless signal transmission. The techniques and solutions presented provide powerful approaches to counteract malicious attacks, fostering reliable communication in an increasingly connected world. Looking ahead, AI and ML hold immense potential for further innovation, bolstering security in 5G and future generation networks.