Quality Evaluation for Large Scale Service-Oriented Architecture

Service-oriented architecture (SOA) is increasingly used as an architectural framework for many real-world applications. However, this framework restricts the modelling of applications involving large scale services with resource constraints. Further, large scale SOA (LSS) consists of multiple services, service consumers, and providers. Similar services with the same functionality exist in LSS, so service selection and discovery are the two major challenges in LSS. Thus, the quality of service (QoS) plays an important role in measuring homogeneous services’ performance. Less attention is paid to the QoS analysis in LSS, that makes the service selection more challenging. Therefore, this article presents a novel mechanism for handling the QoS in LSS. In LSS, the service meta-store (SMS) is a middle layer between the lower and upper layers. The lower layer consists of different services, making various tasks at the middle layer. So, various quality parameters are discussed for the lower $\left( {{{\overrightarrow {QoS} }_{\vec S}}} \right)$ and middle layers $\left( {{{\overrightarrow {QoS} }_{\vec T}}} \right)$. The generation of tasks from the services and the relationship of QoS parameters among the services and tasks are evaluated in a simulation set-up. The correlation result guarantees that there exists a strong correlation among quality attributes of tasks from the services. That results in the design of better applications or software without fail. The novelty of the proposed approach is to experimentally validate the correlation among the QoS attributes in LSS.