One of the current challenges in multimedia systems is to ensure efficient data transmission betw... more One of the current challenges in multimedia systems is to ensure efficient data transmission between the server and clients. These systems must guarantee to the users a certain quality of service (QoS), by ensuring data accessibility whatever the material and networks conditions are. They also must guarantee information consistency, particularly the respect of temporal constraints in order to obtain a smooth presentation of scenes. In this paper, we propose an architecture obtained by exploiting similarities existing between real-time database systems (RTDBSs) and multimedia systems. Then, we define a method, that we name (m,k)-frame, which allows to control at any moment the number of frames sent to users, by discarding selectively some frames when needed. We, finally, carry out simulations whose results show the best performances of our approach which consists of adapting the QoS to the real conditions, compared both to the application of an already proposed method by other authors (R-(m,k)-firm method) and to the original (m,k)-firm method. This adaptation is done according to the system load, which may become heavy due to network congestion, i.e, dynamic arrival of clients.
F-76058 LE HAVRE CEDEX Ö×ØÒÒÑѺÄÄ×ØÒÒÑÑÐÐØØ×ÐºÖ RÉSUMÉ. Les applications multimédia gèrent des vo... more F-76058 LE HAVRE CEDEX Ö×ØÒÒÑѺÄÄ×ØÒÒÑÑÐÐØØ×ÐºÖ RÉSUMÉ. Les applications multimédia gèrent des volumes importants de données. L'exploita-tion de ces données doit se faire en respectant les contraintes temporelles permettant de lire les paquets vidéo avec une certaine fluidité. Lorsque les contraintes temporelles ne sont pas respectées, la qualité de service fournie aux utilisateurs diminue. Partant du constat qu'il existe des similarités entre les applications multimédia et les SGBD temps réel, notre approche con-siste à exploiter les travaux concernant la gestion de la qualité de service dans les SGBD temps réel afin de les appliquer aux systèmes multimédia distribués. Dans cet article, nous proposons une nouvelle architecture permettant la gestion efficace des données multimédia et d'améliorer la qualité de service fournie aux clients lors des variations importantes de la charge d'util-isation du système. Notre technique consiste à faire varier la qualité des flux ...
Abstract—Multimedia applications usually manage large quan- tities of data in the form of frames ... more Abstract—Multimedia applications usually manage large quan- tities of data in the form of frames of certain types. To ensure a good trafic of these frames through the network, temporal constraints must be respected when sending and receiving these frames. If the temporal constraints are not met, then the quality of service (QoS) provided to users decreases. In this paper, we exploit some results obtained in QoS management,in Real-Time Databases Systems (RTDBSs), and we apply them to multimedia applications, because of similarities existing between th ese two fields. We propose a new,method,allowing to control the QoS provided to clients according to the network congestion, by discarding some frames when needed. Index Terms—Distributed multimedia systems, Feedback con-
Today's multimedia applications ask for a complex set of quality of service (QoS) requiremen... more Today's multimedia applications ask for a complex set of quality of service (QoS) requirements. These ap- plications define time, bandwidth and synchronization constraints and manage large quantities of data. Due to the similarities existing between multimedia appli- cations and real-time database systems (RTDBSs), we propose, in this paper, an approach which consists in exploiting some works related to the QoS management in RTDBSs in order to apply them to multimedia sys- tems. We particularly propose a FCS-MS1 architecture, an improvement of the original master server architec- ture, which deals with QoS management by optimizing the resources use and reducing significantly the system overloads.
One of the current challenges in multimedia systems is to ensure efficient data transmission betw... more One of the current challenges in multimedia systems is to ensure efficient data transmission between the server and clients. These systems must guarantee to the users a certain quality of service (QoS), by ensuring data accessibility whatever the material and networks conditions are. They also must guarantee information consistency, particularly the respect of temporal constraints in order to obtain a smooth presentation of scenes. In this paper, we propose an architecture obtained by exploiting similarities existing between real-time database systems (RTDBSs) and multimedia systems. Then, we define a method, that we name (m,k)-frame, which allows to control at any moment the number of frames sent to users, by discarding selectively some frames when needed. We, finally, carry out simulations whose results show the best performances of our approach which consists of adapting the QoS to the real conditions, compared both to the application of an already proposed method by other authors (R-(m,k)-firm method) and to the original (m,k)-firm method. This adaptation is done according to the system load, which may become heavy due to network congestion, i.e, dynamic arrival of clients.
One of the current challenges in multimedia systems is to ensure efficient data transmission betw... more One of the current challenges in multimedia systems is to ensure efficient data transmission between the server and clients. These systems must guarantee to the users a certain quality of service (QoS), by ensuring data accessibility whatever the material and networks conditions are. They also must guarantee information consistency, particularly the respect of temporal constraints in order to obtain a smooth presentation of scenes. In this paper, we propose an architecture obtained by exploiting similarities existing between real-time database systems (RTDBSs) and multimedia systems. Then, we define a method, that we name (m,k)-frame, which allows to control at any moment the number of frames sent to users, by discarding selectively some frames when needed. We, finally, carry out simulations whose results show the best performances of our approach which consists of adapting the QoS to the real conditions, compared both to the application of an already proposed method by other authors (R-(m,k)-firm method) and to the original (m,k)-firm method. This adaptation is done according to the system load, which may become heavy due to network congestion, i.e, dynamic arrival of clients.
F-76058 LE HAVRE CEDEX Ö×ØÒÒÑѺÄÄ×ØÒÒÑÑÐÐØØ×ÐºÖ RÉSUMÉ. Les applications multimédia gèrent des vo... more F-76058 LE HAVRE CEDEX Ö×ØÒÒÑѺÄÄ×ØÒÒÑÑÐÐØØ×ÐºÖ RÉSUMÉ. Les applications multimédia gèrent des volumes importants de données. L'exploita-tion de ces données doit se faire en respectant les contraintes temporelles permettant de lire les paquets vidéo avec une certaine fluidité. Lorsque les contraintes temporelles ne sont pas respectées, la qualité de service fournie aux utilisateurs diminue. Partant du constat qu'il existe des similarités entre les applications multimédia et les SGBD temps réel, notre approche con-siste à exploiter les travaux concernant la gestion de la qualité de service dans les SGBD temps réel afin de les appliquer aux systèmes multimédia distribués. Dans cet article, nous proposons une nouvelle architecture permettant la gestion efficace des données multimédia et d'améliorer la qualité de service fournie aux clients lors des variations importantes de la charge d'util-isation du système. Notre technique consiste à faire varier la qualité des flux ...
Abstract—Multimedia applications usually manage large quan- tities of data in the form of frames ... more Abstract—Multimedia applications usually manage large quan- tities of data in the form of frames of certain types. To ensure a good trafic of these frames through the network, temporal constraints must be respected when sending and receiving these frames. If the temporal constraints are not met, then the quality of service (QoS) provided to users decreases. In this paper, we exploit some results obtained in QoS management,in Real-Time Databases Systems (RTDBSs), and we apply them to multimedia applications, because of similarities existing between th ese two fields. We propose a new,method,allowing to control the QoS provided to clients according to the network congestion, by discarding some frames when needed. Index Terms—Distributed multimedia systems, Feedback con-
Today's multimedia applications ask for a complex set of quality of service (QoS) requiremen... more Today's multimedia applications ask for a complex set of quality of service (QoS) requirements. These ap- plications define time, bandwidth and synchronization constraints and manage large quantities of data. Due to the similarities existing between multimedia appli- cations and real-time database systems (RTDBSs), we propose, in this paper, an approach which consists in exploiting some works related to the QoS management in RTDBSs in order to apply them to multimedia sys- tems. We particularly propose a FCS-MS1 architecture, an improvement of the original master server architec- ture, which deals with QoS management by optimizing the resources use and reducing significantly the system overloads.
One of the current challenges in multimedia systems is to ensure efficient data transmission betw... more One of the current challenges in multimedia systems is to ensure efficient data transmission between the server and clients. These systems must guarantee to the users a certain quality of service (QoS), by ensuring data accessibility whatever the material and networks conditions are. They also must guarantee information consistency, particularly the respect of temporal constraints in order to obtain a smooth presentation of scenes. In this paper, we propose an architecture obtained by exploiting similarities existing between real-time database systems (RTDBSs) and multimedia systems. Then, we define a method, that we name (m,k)-frame, which allows to control at any moment the number of frames sent to users, by discarding selectively some frames when needed. We, finally, carry out simulations whose results show the best performances of our approach which consists of adapting the QoS to the real conditions, compared both to the application of an already proposed method by other authors (R-(m,k)-firm method) and to the original (m,k)-firm method. This adaptation is done according to the system load, which may become heavy due to network congestion, i.e, dynamic arrival of clients.
Uploads
Papers by Alaya Béchir