RIT researchers present novel dynamic defense model for thwarting wireless attacks at IEEE INFOCOM

Researchers from the Rochester Institute of Technology (RIT) recently presented a cutting-edge defense mechanism against advanced wireless attacks at the prestigious IEEE International Conference on Computer Communications (INFOCOM). The paper, authored by Professor Hanif Rahbari and computing and information sciences Ph.D. student Naureen Hoque, was showcased at the top-ranked networking conference held in Vancouver in May.

IEEE INFOCOM is renowned in the research community as a major venue for presenting significant and innovative contributions in networking, communications, and related areas. The inclusion of the RIT team's work in the conference proceedings underscores its importance and potential impact on the field, and marks the 6th time Rahbari has published at the conference in the past decade.

The research introduces a "moving-target defense framework" designed to protect wireless communications from sophisticated cyber threats. In an era of increasingly advanced attacks, this novel approach aims to stay ahead of persistent adversaries.

"As technology advances and information becomes more accessible, attackers are not only getting equipped with better tools but are also gaining deeper knowledge about the defenses," explains Hoque. "If our security defenses remain static, they're bound to fail eventually."

Digital modulation, the process of converting digital bits to analog signals for wireless transmission, inherently reveals characteristics of the ongoing transmission. This vulnerability has been exploited by malicious actors to launch various attacks, including traffic analysis, selective jamming, and breaches of user privacy.

The team's innovative approach involves a modulation-masking mechanism that applies small perturbations to modulated symbols, making them appear ambiguous to even highly sophisticated classifiers of the  attackers. Critically, this technique uses a pool of deep learning models and perturbation-generating methods that are continuously altered.

Professor Rahbari elaborates on the strategy: "By deploying a pool of deep learning models in a novel masking technique we have designed, our defense strategy keeps changing (moving) them as needed, making it difficult for adversaries that may even use deep-learning techniques to keep up with the evolving defense system over time."

The researchers demonstrate that their system maintains overall performance while significantly increasing the complexity for attackers trying to decipher the modulation scheme. The time complexity for adversaries attempting to breach the system increases cubically, providing a substantial barrier to malicious activities.

This research addresses a critical weakness in current wireless security measures. As Hoque points out, "To stay ahead, it's crucial that the security measures evolve continuously, ensuring that even the most persistent attacker has to face new challenges every time they come knocking."

The presentation of this work at IEEE INFOCOM not only highlights its significance in the field of network security but also provides a platform for further discussion and development within the research community. By creating a dynamic, evolving defense system, Rahbari and Hoque's work promises to enhance the security and privacy of wireless users across a wide range of applications, potentially influencing future directions in wireless communication security.

Learn more about their work securing wireless systems via the Wireless and IoT Security Lab: https://www.rit.edu/wisplab/