Ph.D. Dissertation

Robust Wireless Communication for
Multi-Antenna, Multi-Rate, Multi-Carrier Systems

Triet Dang Vo-Huu [cv]

College of Computer and Information Science
Northeastern University

Abstract

Today's trend of migrating radio devices from hardware to software provides potential to create flexible applications for both commercial and military use. However, this raises security concerns, as malicious attacks can also be alleviated to break legitimate communications. In this research work, we study some aspects of jamming threats.

First, we develop SAIM -- a hybrid system of mixed mechanical steerable antenna and software jamming cancellation -- to counter high-power jamming adversary. The system robustness relies on a new antenna structure we specially design for anti-jamming purpose, and a set of new algorithms that can adaptively and effectively mitigate the jamming signal up to 100,000 times higher than legitimate signals. Our solution is appropriate for malicious environments with powerful jammers, where mechanical steering is feasible, e.g., military applications.

In residential wireless deployments, even limited-power rate adaptation attacks can dramatically affect the system performance. To tackle this problem, we develop CBM for hiding rate information, and -- at the same time -- increasing resiliency against jammers up to seven times higher than regular rate-exposing systems. The resiliency boost is achieved by our new generalized non-uniform Trellis Coded Modulation codes, while the modulation and code information is concealed by a new efficient method for cryptographically interleaving baseband symbols.

As a third part of this work, we investigate the jamming impact on multicarrier systems with focus on Wi-Fi communications. Toward this goal, we develop the first open source software defined radio Wi-Fi stack that can operate at high rates and allow detailed analysis from Physical Layer to Medium Access Control and Link Layers. Using our platform, we analyze the interleaving process specified by IEEE 802.11 and exploit it to devise an efficient Interleaving Jamming strategy that can totally block the Wi-Fi communications with jamming power less than 1% of the regular transmitted signal power. Our strategy is at least 5dB and up to 15dB more power-efficient than those that are unaware of the interleaving structure.

Download

1. Presentation slides [pdf]
2. Dissertation [pdf]

Committee

• Professor Guevara Noubir (advisor)
• Professor Erik-Oliver Blass (Airbus Group Innovations / Northeastern University)
• Professor Rajmohan Rajaraman (Northeastern University)
• Professor Srdjan Capkun (ETH Zurich, external member)
• Professor David Starobinski (Boston University, external member)

Justification for committee composition

• Professor Erik-Oliver Blass is an expert in network security, cryptography, and privacy.
• Professor Rajmohan Rajaraman is an expert in network algorithms and wireless networks.
• Professor Srdjan Capkun is an expert in system and network security with focus in wireless networks.
• Professor David Starobinski is an expert in network modeling, performance evaluation, and security of wireless networks.