Wi-Fi and Jamming Attacks
Wi-Fi is now emerging as the primary medium for not only wireless Internet access to residential users,
but also for traffic offloading to commercial cellular services due to scarcity of spectrum and increases
of deployment cost. Due to the ease of access to jamming devices (software defined radio, commercial jammers, etc.)
adversaries are able to carry smart jamming attacks against Wi-Fi communications and effectively
degrade the whole network. In this project, I am investigating the impact of jamming on Wi-Fi links.
The preliminary results show that a new jamming technique based on the interleaving structure employed
by the IEEE 802.11 standard can block the whole Wi-Fi link with a very small jamming power.
CBM - Conceal and Boost Modulation
Exposing the rate information of wireless transmissions enables highly efficient attacks
that can severely degrade the performance of a network at very low cost.
To resolve this issue, I developed an integrated solution
for concealing the rate information, and at the same time, boosting
the robustness of the transmission. My solution comprises a set of algorithms to find
good generalized TCM
and a cryptographic interleaving construction method to hide modulation and code information.
Countering High-Power Jamming Attacks
Jamming with high power is today's a realistic threat to wireless communication systems due to
ease of access to jamming devices. Motivated by the importance of this problem, I designed
and implemented a hybrid system consisting of a novel antenna design and a set of antenna control
algorithms and jamming signal cancellation techniques that can cope with a jammer significantly
more powerful than legitimate communication nodes.
BaPu - Bunching Access Point Uplinks
Today's uplink throughput of residential broadband networks are often much lower than downlink,
limiting users from sharing HD video or uploading large files in real time. To address this problem,
I co-developed a practical and efficient system for aggregating Wi-Fi access points to improve the
uplink throughput. Our solution supports both TCP and UDP sessions and provides significant
increase of uplink capacity.
IEEE 802.11s for Fractionated Satellite Systems
On the requirement of enabling communications between modules in
fractionated satellite systems, IEEE 802.11s standard is selected to
be implemented and tested for intermodule communication.
Toward this goal, I designed and implemented the MCCA
based on the Linux wireless driver framework compliant to the IEEE 802.11s standard. The implementation
supports all kinds of underlying hardware drivers and was evaluated to successfully
improve the throughput of MCCA-enabled transmission flow.
Building Efficient Storage Scheme for WSNs
I designed and implemented a distributed scheme for data dissemination in WSN
that can tolerate link and node failures. The work consists of an implementation
of Beacon Vector Routing protocol on TinyOS and development of a distributed error-correcting scheme
deployed on Crossbow TelosB testbed. The results showed the solution's practicability
for distributing light-weight and emergent data across sensor nodes.
Location Tracking using Smartphone Sensors
Leakage of users location and traffic patterns is a
serious security threat with significant implications on privacy. While mobile phones can restrict
the explicit access to location information to applications authorized by the user, the location is
still leaked through side channel information. I co-developed a zero-permissions
Android App that
can infer vehicular users’ location and traveled routes using gyroscope, accelerometer, and compass
information with high accuracy and without users’ knowledge. Through extensive evaluation via
both simulations and real driving experiments, we show that the adversary can detect the users’
routes with a high probability, rendering a serious threat.
EPiC - Efficient Private Counting
In the era of cloud computing, where data is outsourced to untrusted clouds for massive computation,
protecting the user privacy is a serious concern. Focusing on the fundamental operation on data sets:
frequency counting, I designed and implemented a scheme for efficient counting on big data sets.
The solution supports counting patterns defined by arbitrary Boolean formula and capability of hiding the
user query information and computed result from the untrusted cloud, while sacrificing only modest overhead
compared to privacy-leakage counting.