Wireless communication has had an immeasurable economic, societal and scientific impact worldwide. However, with the recent explosion of smartphones and tablets, and the expected emergence of new types of networks such as sensor swarms and the internet of things, the current model of licensed access to spectrum being appears increasingly unviable. There is currently active research into two complementary solutions: (1) spectrum sharing cognitive radios and (2) wireless networks based on dense, short-range millimeter-wave links. The development of highly reconfigurable, power-efficient radios is critical to the long term success of these technologies. This proposal addresses a critical circuit-level challenge by developing design techniques for agile, spectrally pure and reconfigurable "universal" frequency synthesizers. Intellectual Merit: Compared to analog-based frequency synthesizers commonly used in current radios, all-digital frequency synthesizers are less sensitive to the analog-imperfections of nanoscale CMOS, are highly reconfigurable and can be rapidly ported across technology nodes, and are therefore promising candidates for universal frequency synthesis. However, the design of widely tunable oscillators, high-resolution/high-linearity time-to-digital converters and the interface between the two pose obstacles to the successful realization of all-digital frequency synthesizers. Several new concepts and techniques are proposed to address these challenges. These include multi-resonance mode switched and phase-change via-reconfigurable oscillators, time-to-digital converters with in situ statistical linearization, real-time digital assistance for the analog components of the synthesizer and built-in monitoring, adaptation and self-test circuitry. A CMOS prototype will be designed and fabricated to validate these concepts. Broader Impact The outcomes of this research will fill a critical need towards the realization of practical devices for use in future wireless systems using advanced spectral access technologies. The outcomes will also fill a pressing near-term need for prototypes that form the basis of pilot testbeds. The educational components of this proposal include scientific publications, public relations, mentoring, and outreach through various programs conducted by the CMU Gelfand Center for Service Learning and Outreach; these programs include the Moving 4th into Engineering and Summer Engineering Experience for girls in middle school programs and other outreach programs for women and other underrepresented groups. The outcomes of this research will be integrated into various classes taught by the PI. A new capstone course on Mixed-Signal Integrated Circuit Design will be developed. The PI will continue to recruit and train undergraduate students through the the Summer Research Experience Program Carnegie Mellon, with a special focus on attracting students from minority groups and institutions.
