In recent years, thanks to technological breakthroughs in algorithms, computing power, and data analysis, industrial internet is undergoing a shift from digitization and networking to intelligence. As the intelligence of industrial devices continues to i ncrease, their trustworthy including trust, security and privacy, reliability and resilience is being incre asingly considered, for instances, vulnerability assessment, APT attacks, zero-trust access, the security of the m achine learning model itself, and among others in computing and communications. Through the fundamental differences be tween industrial internet and the regular IT systems, their common vulnerabilities and priorities are different. Further more, industrial internet has a specific type of traffic and data using particular communication protocols. Machine learning-b ased security solutions have been widely applied to the detection of threats in industrial internet systems. Reinforcement learning (e.g., deep Q-networks, continuous control with deep reinforcement learning, distributional deep reinforcement learning with quantile regression, etc.) and federated learning are two representative methods to analyze behaviors and patterns of malicious activities, and to make decisions accordingly. While they are rapidly expanding in the field of trustworthy in industrial internet, several research issues are still open to be discussed and studied, from high requirements on real time and fault tolerance, resource-constrained industrial internet devices for building ML-based prediction models, large noise in industrial internet data, to multiple intermediate processes for computing and communications. The objective of this workshop is to identify and address the challenges and opportunities of developing and utilizing reinforcement learning and federated learning in the area of trustworthy in industrial internet. Each submission must be up to 5 pages in length and conforms to the double-column template provided by IEEE. Potential topics include but are not limited to the following:
Paper Submission Guideline:
This workshop is in conjunction with IEEE TrustCom 2021 (https://trustcom2021.sau.edu.cn/) and all papers must be submitted electronically via Easychair (https://easychair.org/conferences/?conf=rfti2021 ). Distinguished papers selected from RFTI 2021, after further extensions, will be recommended for submission and publication in Special Issues of SCI-indexed Journals.