Semantics-Based Scheduling Approach of Ontology-Based Real-Time DBMS

Abstract

Real-time DBMSs (DataBase Management Systems) are designed to manage transactions with time constraints and maintain the database consistency. The Quality of Service (QoS) in these systems is often evaluated based on the number of transactions satisfying their deadlines using a Feedback Control Scheduling Architecture (FCS). In this context, we are interested in the recently proposed FCS for Real-Time Ontology (FCSRTO) allowing to manage real-time ontological data, to which a specific execution of transactions is proposed. Being essential for the execution process of transactions, scheduling has been the subject of a recent work proposing the Advanced Earliest Deadline First based on Transactions Aggregation Links and Data Semantic Links (AEDF-TAL-DSL) as a scheduling protocol. It is mainly based on considering the aggregation links existing between transactions as well as the semantic links appearing between the users queries as scheduling parameters. Our work is to propose a new QoS approach called Semantics-Based FCSRTO. Our approach consists in combining the FCSRTO and the AEDF-TAL-DSL in a same architecture. Hence, we are improving the FCSRTO by using a sophisticated scheduling protocol on one hand, and on the other, it improves the AEDF-TAL-DSL scheduling protocol by managing transactions accessing to real-time ontological data. We also show the contributions provided by our QoS approach through a set of simulations.