Malek Ben Salem¹, Emna Bouazizi¹, Rafik Bouaziz¹ and Claude Duvalet², ¹Sfax University, Tunisia and ²Universite du Havre, France

In our days, there are many real-time applications that use data which are geographically dispersed. The distributed real-time database management systems (DRTDBMS) have been used to manage large amount of distributed data while meeting the stringent temporal requirements in real-time applications. Providing Quality of Service (QoS) guarantees in DRTDBMSs is a challenging task. To address this problem, different research works are often based on distributed feedback control real-time scheduling architecture (DFCSA). Data replication is an efficient method to help DRTDBMS meeting the stringent temporal requirements of real-time applications. In literature, many works have designed algorithms that provide QoS guarantees in distributed real-time databases using only full temporal data replication policy. In this paper, we have applied two data replication policies in a distributed feedback control scheduling architecture to manage a QoS performance for DRTDBMS. The first proposed data replication policy is called semi-total replication, and the second is called partial replication policy.

DRTDBMS, Quality of Service, Partial Replication, Semi-Total Replication.