Add to Google Calendar
Date:  Thu, January 30, 2020
Time:  10:00am- 11:00am
Location:  Holmes Hall 389
Speaker:  Dr. Anyi Liu

Abstract: 

Pervasive computing has emerged as one of the most prevalent computing paradigms in recent years. However, the attack surface of pervasive computing never stops expanding. Security issues, such as the lack of trust between cloud stakeholders and the insufficient of cloud-side activity monitoring and anomaly detection, still remain challenging. In this presentation, Dr. Liu will present his line of research centering on defensive approaches to ensure security and trustworthiness in a pervasive computing environment. More specifically, he will present two thrusts of his research, one of which is the construction of hardware-assisted Trusted Execution Environment (TEE) for cloud forensics. With the aid of Trusted Platform Module (TPM) microprocessor and Attribute-Based Encryption (ABE), Dr. Liu will first present a secure mutual-attestation protocol that allows cloud stakeholders to establish a trusted relationship, with the assurance of hardware, software platforms, and roles of stakeholders. Based on the exchanged credentials, a novel Fragile Watermark (FW) chaining scheme is proposed, such that malicious data manipulations can be detected and localized in a data stream at run-time, with no false positive and extremely low false negatives. A prototype system, namely LiveForen, has been deployed and evaluated in real-world cloud. The other thrust is the modeling and verification of the cryptoenforced Intrusion Detection System (IDS) in cloud. Dr. Liu proposes a novel scheme that detects abnormalities of delegated jobs by leveraging Homomorphic Encryption (HE), a privacy-preserving cryptographic primitive; yet, his solution employs a novel construction of the encrypted specification and monitored activities, it is difficult for an adversary to tamper with the data, which is readily be validated by an authorized party. Again, a prototype system, namely VERDICT, has been deployed in both VMs and containers on the top of popular clouds. The evaluation demonstrates that VERDICT is capable of reporting the status of delegated cloud jobs, detecting malicious operations, and maintaining a bounded run-time overhead. Dr. Liu will conclude his presentation with his reflection on future research directions and teaching plans. 

Bio: 

Anyi Liu received his Ph.D. degree from George Mason University in 2012. Currently, he is an Assistant Professor of Engineering at Oakland University (OU). His research interests include the security and privacy of pervasive computing, system security, and intrusion detection. His recent papers have been published on premium journals, including the IEEE T-IFS. His research has been sponsored by NSF and NASA. As a founding member, he has actively engaged in securing the designation of National Center of Academic Excellence in Cyber Defense and several educational grants at OU. He has reviewed numerous papers for top journals, served on over 40 technical program committees and 4 NSF review panels. He is an IEEE Senior Member.