Excellence in Research: Security-aware Resource Management for Serverless Platforms

The new era of computing brought renovations of facilities to support the growing needs. Nowadays, desktops or laptops are gradually replaced by mobile and embedded devices. Behind these portable, low cost, low latency devices, are the Internet and Cloud Computing infrastructure, which shift computationally intensive tasks to remote data centers. This model will be more energy efficient in a long run. Recently, serverless computing has emerged in the market and evolved into an alternative to the traditional server-based computing because the users will be only billed the time their functions run on the cloud. Accompanying the energy gain and new billing flexibility, are security concerns throughout all layers of this architecture. Specifically, the co-residency of tenants’ workloads scheduled in serverless platforms creates opportunities for attackers to invade other workloads or steal sensitive information. However, existing resource management algorithms fail to take security into consideration since most scheduler designs focus on performance and quality of service. To address these issues, Prairie View A&M University collaborates with Texas A&M University to develop a security-centric, attack-facing, intelligent, human-centered approach to resource management for serverless platforms under internet and cloud-based threats. This project aims at enhancing cybersecurity research and education, contributing research results, datasets, and educational material through publication and open sourcing. The research activities are structured into three thrusts. The first thrust presents a novel reinforcement-based formulation for a multi-platform dynamic optimization scheme to balance performance and security. The second thrust envisions a data-driven solution to address critical attack factors in serverless computing. Finally, the third thrust lays out approaches to incorporate the research activities into educational and training environments. The overarching goal of this project is to establish a successful research infrastructure for explorations of how security-centric resource management can improve the security of serverless computing. Furthermore, this project investigates the best practices and pathways for research collaboration between a minority-serving institution and a first-tier research institution and encourages awareness and interest in security among students from underrepresented groups. This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.