CAREER: Towards Proactive and Collaborative Mobility-Aware Edge Intelligence

University Of Delaware, Newark DE

Investigators

Abstract

The significant increase in both IoT (Internet of Things) network size and data volume opens up attractive opportunities for data analysis and learning to support decision making and derive scientific discovery and innovations. Multiaccess Edge Computing (MEC) enables IoT devices to offload their delay-sensitive resource-intensive tasks to nearby edge devices. However, spatio-temporal uncertainties due to user mobility bring the most challenging obstacles for MEC in providing predictable performance and guaranteed Quality-of-Service (QoS). The research mission of this CAREER proposal is to develop foundations of proactive and collaborative mobility-aware MEC systems to improve QoS for IoT applications, minimize data movement, and optimize the overall system performance. This project focuses on all aspects of mobility in MEC, which include these scenarios: (i) mobile IoT devices/edge devices move as their own decisions; (ii) the system makes the use of mobile edge devices to move to a specific demand area; and (iii) the system incentives the mobile IoT devices to move to receive better QoS. In support of the project mission, this project will design a novel data-driven approach to infer vital mobility patterns leading to QoS violations, and will develop a unique proactive offloading and service deployment approach based on an integrated community detection and multidimensional flow approach. This project will investigate the design of a proactive learning-based approach to act autonomously based on any changes that may lead to QoS violation. This project will facilitate dynamic placement and relocation of mobile edge devices in providing low-latency edge services. This project will also investigate the effects of suggesting location changes to IoT devices and edge devices with location flexibility, and will design a reverse-auction mechanism and a two-sided mechanism to balance the system load and further improve response time. This CAREER project will integrate teaching, research, and outreach activities to broaden exposure to systems research and increase student retention and representation of underrepresented students by providing meaningful and exciting student involvement and finely-tuned mentorship. The proposed research plan will address the technical challenges in building mobility-aware MEC systems for realtime processing demands of mobile IoT applications to enable edge intelligence along the cloud-to-thing continuum and will enrich the scientific knowledge of advanced computing system design. The availability of collaborative mobile edge devices will improve the quality of edge services for end-users and will empower many more innovative applications that are simply not yet possible today. The proposed educational plan will educate students on modern notions of MEC and prepare them for related professions. A substantial quantity of the materials of this project will be made publicly available online in the form of tutorials, talks, publications, codes, datasets, and testbeds. This project is jointly funded by the Faculty Early Career Development Program (CAREER) and the Established Program to Stimulate Competitive Research (EPSCoR). 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.