An increasing number of mobile users rely on wireless technologies to gain secure and uninterrupted access to network services. As the volume of data disseminated via the wireless medium rapidly expands, provision of performance, reliability, and security become challenging problems. These problems can be alleviated by the use of multiple orthogonal frequency bands (channels) that has been demonstrated to substantially reduce contention and interference. However, for systems with poor physical security and lack of centralized resource allocation such as ad hoc, sensor, and cognitive radio networks, a multitude of internal and external attacks against the medium channel access mechanisms can negate any gains due to channelization. Most previous adversary models and protection methods are limited to single-channel networks thus ignoring the additional vulnerabilities and complexities of channelization. This project aims to advance our understanding regarding the feasible space of attacks against channel access. We design secure multi-channel access techniques, focusing on the protection of control traffic, collaborative attack detection, and uncoordinated medium sharing. Moreover, for networks with dynamic spectrum such as cognitive radio networks, we develop verifiable channel sensing methods. Central to this effort is our definition of quantitative security metrics under well defined adversarial models. Conducted research is expected to enforce wireless networks with the very much needed guarantees in security and performance. In the educational domain, this project seeks to expose undergraduate and graduate students to fundamental problems from interdisciplinary fields in computer engineering, information science, and mathematics.
CAREER: Securing Channel Access in Multi-Channel Ad Hoc Networks