Journal
JOURNAL OF SUPERCOMPUTING
Volume 77, Issue 7, Pages 6754-6787Publisher
SPRINGER
DOI: 10.1007/s11227-020-03579-2
Keywords
Reconfigurable edge; Server configuration instantiation; Quality of service; Constraint modeling; SAT-solver; Alloy modeling language
Ask authors/readers for more resources
The paper introduces a constraint-based programming approach for server reconfiguration problem in edge computing to automate efficient access to qualified edges under QoS constraints. By discussing abstract meta-modeling, validation, and empirical analysis in three phases, the proposed REaaS paradigm offers a new solution for edge design.
The server reconfiguration problem in edge computing under the constraints on QoS can become an integral part of the automatic-early analysis of the edge design. This paper proposes a constraint-based programming approach to introduce reconfigurable edge as a service (REaaS) for edge providers and businesses in order to automate efficient access to qualified edges. The approach guarantees the best-(re)configuration of services by addressing business QoS-based requirements in the form of a constraints-based solution at the early analysis phase. The proposed paradigm for REaaS is discussed in three phases including abstract meta-modeling, validation, and empirical analysis. We also provide automated support to generate clear visualization for edge servers in terms of QoS using a lightweight model checker, Alloy Analyzer. Using Alloy Analyzer, we separately explore the scalability and runtimes of an edge server (re)configuration on a running example and two real-world domains including online collaboration systems and smart homes. Furthermore, the performance of the proposed approach is discussed using three fully automatic SAT-solvers in terms of the runtime with different specified settings. Finally, the miss and hit ratios for different types of services at the same time are compared with a new proposed method that demonstrates % 9.7 hit ratio improvement after reconfiguration.
Authors
I am an author on this paper
Click your name to claim this paper and add it to your profile.
Reviews
Recommended
No Data Available