Ph.D. Thesis Proposal
Robust Wireless Communication forMulti-Antenna, Multi-Rate, Multi-Carrier Systems
Triet Dang Vo-Huu [cv]
College of Computer and Information ScienceNortheastern 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 attackers can also be generated easily to break legitimate communications. In this research work, our goal is to design a robust anti-jamming radio framework. We particularly investigate three different aspects of jamming threats: high-power jammers, link attacks on rate adaptation, and jamming in multicarrier systems.
The threats of high-power jamming to wireless communications today are realistic due to the ease of access to powerful jamming sources. To counter high-power jamming attacks, we develop SAIM which is a hybrid system capable of resisting jammers of up to 100,000 times higher power than legitimate communication nodes. The system robustness relies on our own antenna structure specially designed for anti-jamming purpose. We develop an efficient algorithm for auto-configuring the antenna adaptively to dynamic environments. We also devise a software-based jamming cancellation technique for appropriately extracting original signals, which is more robust than traditional MIMO approaches, as pilot signals are not required in SAIM.
In spite of the robustness of SAIM, our design is more appropriate for malicious environments with powerful jammers, where mechanical steering is feasible, e.g., military applications. Residential and commercial wireless communication systems are still vulnerable to even limited-power jamming, as in today's standard wireless protocols, rate information is exposed to adversaries. Rate-based attacks have been demonstrated to severely degrade the networks at very low cost. To mitigate rate-based attacks, we develop CBM, a system capable of hiding rate and – at the same time – increasing resiliency against jammers up to seven times higher than regular systems, where rate is exposed. We achieve the resiliency boost by generalizing Trellis Coded Modulation to allow non-uniform codeword mapping. We develop an efficient algorithm for finding good non-uniform codes for all modulations in {BPSK, QPSK, 8-PSK, 16-QAM, 64-QAM}. To conceal rate information, we devise an efficient method for generating cryptographic interleaving functions.
In recently deployed communication networks such as WiFi and LTE systems, MIMO and OFDM are the two main techniques for increasing bandwidth efficiency. While MIMO increases the channel capacity by spatial processing on multiple received signals, OFDM mitigates impacts of dynamic variations in wide-band channels and allows frequency reuse with overlapping carriers. Synchronization is a key for high-throughput performance in MIMO and OFDM systems. In this work, we study impacts of jamming attacks specifically targeting to control channels in WiFi and LTE networks. Our study focuses on efficient techniques for both jamming and anti-jamming in multicarrier systems.
Download
Committee
- Professor Guevara Noubir (advisor)
- Professor Erik-Oliver Blass (Northeastern University)
- Professor Rajmohan Rajaraman (Airbus Group Innovations / 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.