Csaba Antal

Distributed resilient architecture for Ethernet networks

The authors proposed an enhanced resilient Ethernet architecture that is based on multiple pre-established spanning trees to connect a set of edge nodes. Fast failover is achieved so that edge nodes switch to a non-affected tree after a link or node failure. A fast and robust procedure was also proposed for edge nodes to detect failures and notify other edge nodes. The other contribution of the paper is an algorithm that calculates multiple spanning trees for input topologies so that at least one tree provides connectivity after a node or a link failure. The algorithm provides minimal number of spanning trees to make input networks fault tolerant. The performance of the algorithm was tested via simulations.

Fast Failure Handling in Ethernet Networks

Ethernet is becoming increasingly popular in metro and carrier-grade networks because of its cost-effectiveness, simplicity and scalability. Nevertheless, Ethernet was originally designed as a LAN technology, therefore, it lacks some features, such as fast fault-protection and sophisticated management, which are very important in the new carrier-grade application areas. We propose a simple failure protection mechanism for Ethernet networks that can recover from node and link failures under 50 milliseconds, which is also provided by SDH/Sonet rings. The mechanism relies on a lightweight distributed protocol that runs on IP routers at the edges of the network and works with commodity off-the-shelf Ethernet switches. This paper describes the protocol design and implementation as well as the network testbed implementation for assessing the performance and the robustness of the mechanism. The results collected in the testbed are shown and discussed in the paper.

Performance study of distributed channel allocation techniques for a fast circuit switched network

Dynamic synchronous transfer mode (DTM) is a next-generation high-speed networking technology. It is based on fast circuit switching and uses distributed channel allocation on shared media. The paper proposes a channel allocation algorithm for DTM, which improves the average call set-up time and call blocking probability characteristics. Another modification is also recommended to the operation, which provides fairness in the case of the examined network model. Many variants of the proposed and known techniques are compared in various network conditions. The performance evaluation of the algorithms is carried out by simulation, and practical conclusions are derived from the results.

Connection admission control for flow level QoS in bufferless models

Admission control algorithms used in access networks for multiplexed voice sources are typically based on aggregated system characteristics, such as aggregate loss probability. Even though the relation of these aggregate performance measures to the performance of specific flows is not trivial, it has gained limited attention so far. We propose an admission control method that provides flow level QoS guarantees. The proposed method is based on Gaussian approximation applied for the bufferless multiplexing model. We show that the flow level packet loss violation probability can be approximated as the quantile of a normal distribution and we give a method to calculate its mean and variance. The obtained admission control formula includes the moments of the activity factor distribution.

On Packet Delays and Segmentation Methods in IP Based UMTS Radio Access Networks

In UMTS radio access networks (UTRAN), the delay requirement - bandwidth product of real-time classes is small, therefore large packets need to be segmented. Furthermore, low priority segmented packets may also have real-time requirements. Assuming that UTRAN traffic is characterized as the superposition of periodic sources with random offset, we propose analytic results and we show simulation results for the delay distribution of real-time classes in a priority system. We then compare segmentation methods considering the delay requirements of all real-time classes.

Low-cost survivable Ethernet architecture over fiber

Feature Issue on AvailabilityEthernet provides a simple and low-cost solution at high bandwidth for metropolitan optical networks. However, native Ethernet still lacks carrier-grade resilience and management schemes. We propose and demonstrate a low-cost robust Ethernet-over-fiber network architecture that can recover from both node and link failures in less than 50 ms and, furthermore, ensure that packet loss is avoided during fault restoration. We implemented and tested the architecture in a prototype network. The proposed scalable architecture works with commodity off-the-shelf Ethernet switches and handles network failures in arbitrary Ethernet-level topologies by the edge nodes of the network. We present the experimental results of the protection protocol implementation, showing that the 50 ms carrier-grade recovery time is achieved.

A cluster-based resource provisioning model in virtual private networks

This paper presents a novel resource provisioning method for virtual private networks. Trunk and hose models are well known bandwidth provisioning models for VPNs but both have significant disadvantages if applied to large scale networks. The management complexity of the trunk model highly increases with the size of the network, while bandwidth efficiency of the hose model is often excessively low. The authors proposed an intermediate solution between hose and trunk models. By dividing the network into clusters and using cluster-based traffic description an appropriate equilibrium between management complexity and over provisioning could be found. A method for computing the necessary links capacities for a traffic given by cluster-based description is also presented, which makes it possible to design congestion free networks. Routing strategies were also suggested to further improve the performance of the new provisioning method. The applicability and limitations of the method and the effect of different routing strategies and fault tolerance are examined by simulating several test scenarios.

Comparison of link-layer segmentation methods for UMTS terrestrial radio access networks

Most user traffic types have limited delay/jitter budget over the Iub interface of UMTS. For example, the strict requirement for data bearers is due to synchronization of soft-handover legs. As the requirement for traffic types differs, the application of QoS solutions, such as multiple QoS classes and segmentation of large packets, is needed. In this paper, we propose results that can be used to improve the operation of link layer QoS mechanisms. We propose novel segmentation methods that decrease the delay of both low priority and high priority packets in a priority system. We have also developed new analytic results and performed simulations for the comparison of the segmentation methods.

Cluster-based resource provisioning for optical backbone networks

Feature Issue on High Availability in Optical NetworksA resource provisioning method for optical backbone networks running IP or multiprotocol label switching (MPLS) is presented. Trunk and hose models are well-known bandwidth provisioning models, but both have significant disadvantages if applied to large-scale networks. The management complexity of the trunk model highly increases with the size of the network, and the bandwidth efficiency of the hose model is often excessively low. We propose an intermediate solution between the hose and trunk models. By dividing the network into clusters and using a cluster-based traffic description, an appropriate equilibrium can be found between management complexity and overprovisioning.

Performance evaluation of a time division multiplexing method applicable for dynamic transfer mode networks

The paper is concerned with analyzing a time-division multiplexing method with two priority classes. A discrete-time model is established for analyzing the queuing behaviour of low priority traffic streams. The motivation came from dynamic transfer mode (DTM) networks which uses time-division multiplexing of channels, however the results might also be used for other fast circuit switched networks. The model to be investigated is based on the important observation that from the viewpoint of low priority traffic the system behaves as a queuing system subjected to periodic server interruption. The probability generating function (p.g.f.) of the system time and system content of low priority messages is expressed in closed forms from which the first two moments can be calculated in a straightforward manner. Tail probabilities of system content and system time are also derived by applying an efficient approximation technique. Finally, we also give numerical examples as simple illustrations.