Abstract

In Large Scale Computing Systems (LSCS), there is a critical and challenging issue: it is almost impossible for humans to manually diagnose/recover/defend numerous faults/failures/threats/ attacks of a great number of distributed components in a timely manner due to its large scale. This work presented the integration of Artificial Intelligence into the dependability issues of LSNS. It proposed a prototype technology to self-improve the reliability of LSNS to enable an LSCS self-learning capability. The model systematically built the architecture of BANS (Bionic Autonomic Nervous System), which is a kind of nature-inspired technology. The BANS architecture implemented cyber-neurons/axons/peripheral nerves/central nerves, and the AI algorithms were embedded in the BANS combined Neural Network, Fuzzy Logic, and Multi-value Diagrams for supervised machine-learning and reinforcement self-learning. This innovative technology enabled LSCS to self-diagnose and self-heal.

Speaker

Professor Yuanshun Dai China

Dean of the School of Computing and Artificial Intelligence

Southwest Jiaotong University

Was continuously elected as “Chinese Most Cited Researcher” in the field of “Safety, Risk, Reliability, and Quality” by Elsevier from 2015 to 2019
Associate Editor of the IEEE Transactions on Reliability

Dr. Dai is a Professor at Southwest Jiaotong University (SWJTU) and Dean of the School of Computing and Artificial Intelligence at SWJTU. Dr. Dai’s research mainly focuses on Reliability Modeling and Optimization for Large-Scale Networking Systems (LSNS). He has published about 200 articles, where there are 4 books, 128 journal papers including 58 IEEE/ACM Transactions papers, and 40 conference papers including 2 Best Paper Awards. He has been continuously elected as “Chinese Most Cited Researchers” in the field of “Safety, Risk, Reliability, and Quality” by Elsevier from 2015 to 2019. He serves as an Associate Editor of the IEEE Transactions on Reliability and is also on the editorial boards of several journals. Dr. Dai was also invited as a Keynote Speaker for IEEE ISORC’05, ChinaSoft’11, ICCT’12, iThings’16, and DASC’19, etc.