ARC: Anomaly-aware Robust Cloud-integrated IoT service composition based on uncertainty in advertised quality of service values

Abstract

From the IoT perspective, each intelligent device can be considered as a potential source of service. Since several services perform the same function, albeit with different quality of service (QoS) parameters, service composition becomes a crucial problem to find an optimal set of services to automate a typical business process. The majority of prior research has investigated the service composition problem with the assumption that advertised QoS values are deterministic and do not change over time. However, factors like sensors failure and network topology changes cause uncertainty in the advertised QoS values. To address this challenge, we propose a novel Anomaly-aware Robust service Composition (ARC) to deal with the problem of uncertainty of QoS values in a dynamic environment of Cloud and IoT. The proposed approach uses Bertsimas and Sim mathematical robust optimization method, which is independent of the statistical distribution of QoS values, to compose services. Moreover, our approach exploits a machine learning-based anomaly detection technique to improve the stability of the solution with a fine-grained identification of abnormal QoS records. The results demonstrate that our approach achieves 14.55% of the average improvement in finding optimal solutions compared to the previous works, such as information theory-based and clustering-based methods.