Recently, the research group of Professor Wang Yong of the School of Microelectronics has made new progress in the field of computer aided integrated circuit chip design. A classic global optimization algorithm and an artificial intelligence based local minimum search are combined to speed up design space exploration tasks for analog circuits design. A research article was published in ICCAD top journal IEEE Transactions on Computer-Aided Design of Integrated Circuits and Systems with title "An Artificial Neural Network Assisted Optimization System for Analog Design Space Exploration". Ph.D. candidate Li Yaping from class 2017 is the first author and Prof. Wang Yong is the corresponding author of the article and all the research work was conducted in Shandong University. This article is the first paper for SDU on the design of radio frequency integrated circuit chips published on IEEE Transaction with the support of commercial tape-out results. The performance index of the Beidou series of chips designed using the proposed techniques meets the national Beidou system's network standard. This methodology has yields first class results for domestic chip design attempts. Related integrated circuit chips have been industrialized through China Electronics Technology Group.
Physical signal processing in the real world is a key link in the Internet of Things and electronic information applications. The performance of interfacing analog integrated circuits affects the accuracy of digital signal processing greatly. Due to the huge design space for integrated circuits, analog integrated circuit design is often an iterative process, which relies heavily on the experience of the designer. Professor Wang Yong's group proposed a methodology which combines classic global optimization and artificial intelligence. The genetic algorithm was used to search the circuit parameters globally, and the artificial neural network (ANN) was used to build local models for the circuits under design. The research group used this algorithm to optimize the design of Beidou RF front-end chips and voice sensor chips, and obtained better results than traditional methods.
Written by：Su Wei