Dimitris Primpas

Application on demand system over the Internet

This paper describes the design and implementation of the ASP-NG system. The main modules of the ASP-NG system are the AoD service and the Web Portal. The ASP-NG Portal is a portal for providing the user with the necessary interface in order to access an Application on Demand (AoD) service. The ASP-NG portal is responsible for the interaction with the user of the AoD service. Using the AoD service the user rents an application for a limited time period at a fraction of the actual cost of the application. The AoD service is responsible for downloading the appropriate parts of the application according to the users actions, while enforcing the mutually agreed frame between the user and the Application Service Provider (ASP). The implementation of the ASP-NG portal is based on the Web Services of the Java 2, Enterprise Edition platform and the implementation of the AoD module is based on C + + programming language. The ASP-NG portal offers to its users the capability to select and customize the language of the user interface in order to present information in their preferred language. Moreover the ASP-NG portal offers to the portal administrator the capability to customise the look and feel of the ASP-NG portal.

Performance Evaluation of the Impact of QoS Mechanisms in an IPv6 Network for IPv6-Capable Real-Time Applications

This paper describes a Quality of Service (QoS) service on an IPv6 domain that aims to service aggregates of real-time traffic with minimum delay, jitter, and packet loss. It contains results from the tests that were performed in order to configure and evaluate the QoS mechanisms. As an actual example of real-time traffic, we have used the OpenH323 project, an open source H.323 implementation that has been ported to IPv6. The QoS mechanisms in IPv6 networks is still a field that has not been researched adequately, and we therefore present the results from the experiments in our IPv6 network that took advantage of the QoS mechanisms. This QoS service uses the Modular QoS CLI (MQC) mechanism and especially the Low Latency Queue feature (LLQ) in order to treat packets from real-time applications.

Extending QoS support from Layer 3 to Layer 2

This paper presents some of the results obtained by the application of Ethernet layer 2 quality of service in IP networks. IP networks traditionally provide quality of service in Layer 3. However, since there is an enormous existing Layer 2 infrastructure, todaypsilas networks could benefit from the deployment of Layer 2 quality of service and the cooperation between layer 2 quality of service and layer 3 quality of service. In this paper, experiments are suggested and conducted and a scheme is suggested for efficient cooperation between Layer 2 and Layer 3 QoS provisioning.

Performance evaluation of an IPv6-capable H323 application

OpenH323 is an open source H.323 implementation that has been ported to IPv6. In this paper we briefly introduce the library architecture and the performance criteria with which the ported version should be evaluated. We then present a variety of experiments that we conducted in order to comparatively evaluate the IPv4 and IPv6 protocol stacks. We also present the results of some initial experiments comparing IPv4 and IPv6 performance under congested network links and the conclusions that they lead us to.

QoS experiences in native IPv6 networks

Deployment of IPv6 technology in research and commercial networks has accelerated in the last few years. Inevitably, as more advanced services take advantage of the new technology, IPv6 traffic gradually increases. Today, there is limited experience in the deployment of Quality of Service (QoS) for IPv6 traffic in backbone networks that support the Differentiated Services framework. As available software and hardware are designed to handle IPv4 packets, there is a need to accurately measure and validate performance of QoS mechanisms in an IPv6 environment. This paper discusses tests and technical challenges in the deployment of IPv6 QoS in core networks, namely the production dual stack gigabit-speed Greek Research and Education Network (GRNET) and the IPv6-only 6NET European test network, using both hardware and software platforms. In either case, we succeeded in delivering advanced transport services to IPv6 traffic and provided different performance guarantees to portions of traffic. The deployed QoS schema was common to IPv6 and IPv4; in most cases both v4 and v6 traffic exhibited comparable performance per class, while imposing no significantly different overhead on network elements. A major conclusion of our tests is that the IPv6 QoS mechanisms are efficiently supported with state-of-the-art router cards at gigabit speeds.

IPv6 QoS testing on dual stack network

This paper presents our work regarding the testing and evaluation of DiffServ QoS mechanisms over IPv6 and IPv4 in dual stack software based platforms. IPv6 introduces some additional features (like flow label) and the current support for QoS mechanisms in IPv6 implementations approaches the corresponding QoS support in IPv4. Therefore, a number of tests with DiffServ QoS mechanisms applied on IPv6 traffic have been carried out on a testbed created specifically for this purpose, in order to validate the mechanisms and evaluate the routers overall performance. Our evaluations cover the load incurred to the routing devices from the implementation of the mechanisms, the level of support for QoS IPv6 features and the comparison of performance over IPv4 and IPv6.

IPv6 deployment: Real time applications and QoS aspects

This paper gives an overview of the issues related to the QoS mechanisms under IPv6 and the transition of applications to the new Internet protocol. We describe the implementation and testing of a QoS service on IPv6 networks, which is based on the DiffServ architecture (expedited forwarding) and provides strict priorities to packets that are produced from real time applications. The service was implemented and tested at the 6NET large-scale IPv6 network, and additional advanced testing was performed in a local testbed. We present and analyze the results from our tests under a number of different scenarios. In addition, we focus on issues regarding application transition to IPv6 and we briefly discuss as a case study the transition of OpenH323 protocol stack to support IPv6. We also discuss the usage and the expected impact of the new Flow Label field in the IPv6 header.

QoS issues in the Research and Academic Networks: The case of GRNET

This paper describes the design and the implementation of QoS services in a high speed backbone network as well as a management tool for the service. The services were designed taking advantage of features provided by the MPLS technology and also by using the DiffServ architecture. The supported QoS services include the IP Premium that tries to provide zero packet loss as well as minimum delay and jitter and the less than best effort service. In addition we implemented a management tool for the service. The scope of this tool is to allow the users to manage their QoS requests (make a new one, edit, delete or view a request). Also the tool performs admission control and produces the necessary configuration that must be applied on the network in order to implement every services request

Using the ns-2 simulation environment to implement and evaluate bandwidth broker models

In this paper we present and compare four different models of implementing a Bandwidth Broker in a DiffServ network. We describe the relevant implementation aspects in the well-known ns-2 simulator and present a number of experiments that were conducted in order to verify our implementation, analyze the performance characteristics of the various alternatives and evaluate the usefulness of the ns-2 simulation environment for similar purposes. We present our experiments that aim at investigating a number of issues related to the operation of Bandwidth Broker modules, and in particular how the acceptance rate, the network overhead and the response time for each Bandwidth Broker model are affected by the network topology, the available buffers for incoming requests and the advance time from the moment a request is submitted to the moment the reservation should take place

An admission control and deployment optimization algorithm for an implemented distributed bandwidth broker in a simulation environment

This paper describes and tests a distributed bandwidth broker that has been implemented in NS simulator. It focuses on the admission control algorithm, its advantages and drawbacks. Also, the bandwidth broker is tested, managing the IP Premium service and we compare 2 different implementations of the service. Finally it approaches the problem of the optimal location of a bandwidth broker in a backbone network. For this purpose, a new model is proposed that evaluates each node and finally selects the most capable node where the base bandwidth broker should be located.