SpecEES: Trusted Frequency-Agile Transceiver Architectures for Secure and Energy-Efficient Communication

Carnegie Mellon University, Pittsburgh PA

Investigators

Abstract

The increasing demand for wireless communication creates a spectrum shortage. Combining with beamforming and massive multiple-input-multiple-output (MIMO) technologies, a highly programmable software-defined radio (SDR) that enables wireless communications across multiple platforms can increase the efficiency of spectrum sharing among a multitude of users. However, the integrated high-resolution wideband MIMO transceivers would consume significant power to acquire and transmit complex signals with digital signal processing. Meanwhile, critical security challenges associated with spectrum sharing arise because the connection of wireless devices to networks represents new entry points for malicious attacks, such as jamming and eavesdropping. Wireless devices and sensors not under constant surveillance especially need comprehensive protection for secured spectrum sharing. Advanced encryption and coding schemes are computationally expensive and require wider bandwidths and higher carrier frequencies. This project proposes an ultra-low-power frequency-agile transceiver architecture that can be integrated with trusted MIMO systems to achieve dynamic spectrum access, manage harmful interference, and identify authorized devices. This research will have significant contributions and impacts to the society and nation in critical need of secured spectrum access and sharing among all users including underserved populations. In addition, the project will provide a unique opportunity to educate students of different levels (K-12, undergraduate, and graduate) on practical aspects of wireless communication systems' security and energy efficiency. The goal of this project is to develop a reconfigurable wideband MIMO system incorporating interference suppression, efficient beamforming, and embedded radio frequency (RF) signatures that can be applied to many types of wireless systems including short-range unlicensed communications and long-range licensed communications. For interference and noise mitigation, reconfigurable RF receivers using nonlinearity with adaptive amplitude folders will be developed to instantaneously push harmful jammers to higher frequencies while preserving the integrity of weak signals of interest. The project will use a programmable noise-shaping loop with embedded N-path filter banks to further suppress wideband interference, remove folded jammers, and achieve a scalable wide dynamic range over a large frequency spectrum while consuming ultra-low power. To achieve maximum programmability with minimum increase of power and area, combinatorial intrinsic device characteristics will be adopted as tuning elements to build filter banks and folders for rapid adaptation to a dynamic and unplanned electromagnetic environment. To achieve better power efficiency, frequency-agile power amplifiers (PAs) for N-path MIMO antenna array will be developed to maintain high power efficiency over the entire SDR bandwidth through the use of joint load-pulling and optionally discrete envelope tracking. The project will utilize inherent nonlinear functions and memory effects of power amplifiers to enable disruptive RF forensic fingerprinting science and technology. The time-varying RF fingerprint polling from the augmented PA nonlinearities and the RF signature identification through base-station post-distortion decoding will enhance the wireless communication security. These unique signatures of each device will make it difficult for adversaries to fake the signatures through precoding the data symbols. This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.