Design of self-healing mixed-signal/RF systems and support CAD tools: A scalable approach

Abstract

Due to the proliferation of nanometer CMOS mixed-signal/RF circuits and a push towards high operating speeds (5-100 Ghz+), there has been renewed interest in the design of high-speed circuits and systems that can self-calibrate and self-heal post-manufacture and in the field. In the past, designers have invented selfhealing mechanisms that are tailored towards specific (critical) mixed-signal/RF performance metrics of specific circuit architectures (transmitter, receiver, etc). What is desired, however, is the ability to monitor multiple performance metrics concurrently and trade them off against one another in an optimal manner, through performance tuning mechanisms, to satisfy system-level Quality of Service (QoS) guarantees. Further, the methods employed must be scalable across different device types/circuit architectures and supported by CAD tools that enable automation of self-healing design procedures.In this paper, recent research advances are presented that allow low cost and rapid selfhealing of complex mixed-signal/RF systems and enable the development of design automation tools to support such activity across diverse performance metrics and circuit types.