This work aims at generating efficient and correct software task implementations for real-time embedded applications. We developed algorithms for optimal multi-task implementations of synchronous finite state machines and of synchronous block diagrams; for security-aware task mapping in CAN-based systems and TDMA-based systems; for optimal implementation of synchronous models on LTTA platforms; and for memory usage minimization in real-time systems. | Read More
To reduce energy consumption, modern green buildings employ complex automation systems that closely monitor and control the building environment (e.g. HVAC and lighting control). We proposed a design flow to co-design the HVAC control algorithm, the embedded platform, and the heterogeneous energy system at system level, and then optimally and correctly implement the control algorithm on a distributed building platform, with consideration of physical elements such as building layout, sensor/actuator locations and environment dynamics. | Read More
This work optimizes an extensibility metric that measures how much the execution times of software tasks can be increased without violating design constraints. Optimizing extensibility allows adding future functionality or upgrading existing functionality without major redesign cycle, which is imperative for large-volume and long lifetime systems such as vehicles, airplanes and buildings. This work was in collaboration with United Technologies. | Read More
Associate Professor
Electrical and Computer Engineering
University of California, Riverside
322 Winston Chung Hall
Phone: 951-827-7701
Email: qzhu at ece dot ucr dot edu
Personal Website