A Decentralized Adaptation System for QoS Optimization
Nanxi Chen, Fan Li, Gary White, Siobhán Clarke, Yang Yang · 2020
This chapter introduces AdaptFog, a distributed reputation system for quality of service (QoS)-aware service choreography in a fog environment. It includes a runtime QoS monitoring component, a fog-to-fog (F2F) dynamic service level agreement (SLA) renegotiation component, and a reputation assessment component. A reputation model based on direct graph is proposed to trace a service's reputation value, and a cross-cutting verification mechanism is introduced to verify service performance. The chapter discusses the state-of-the-art solutions toward QoS-aware service choreography, SLA negotiation, and service monitoring. It focuses on the process of delivering the services that have previously deployed on fog nodes by Internet of Things (IoTs) application providers. AdaptFog assumes that a set of resource-intensive fog nodes reside close to the IoT devices to support real-time data processing and low-latency response for mission-critical applications. The chapter provides information on an AdaptFog architecture that enables a reputation mode to facilitate decentralized service composition and underpins a F2F SLA renegotiation.