Anton Riedl

A hybrid genetic algorithm for routing optimization in IP networks utilizing bandwidth and delay metrics

We study routing optimization in the context of IP traffic engineering, which relies on conventional, destination-based routing protocols. We introduce different concepts of routing optimization and discuss their implications for traffic engineering. The specific focus is on routing technologies that utilize multiple metric types - in our case delay and bandwidth metrics - in order to derive the shortest paths towards every destination in the network. A novel hybrid genetic algorithm is presented, which allows the computation of an optimized set of link metrics, considering single as well as dual-metric protocols. Finally, the benefits of the two metric types for traffic engineering are demonstrated.

Optimized routing adaptation in IP networks utilizing OSPF and MPLS

In this paper routing algorithm methodologies are investigated, which utilize conventional routing protocols such as OSPF in combination with MPLS. While having the majority of traffic routed along optimized shortest paths, MPSL is only partly introduced to complement the adaptation process. We present a novel algorithm based on simulated annealing to optimize link metrics in OSPF networks. The algorithm takes into account the original routing configuration and allows tradeoff considerations between routing optimality and adaptation impact. For the setup of complementary MPLS paths, two mixed-integer programming models are proposed. It can be shown that already a relatively small number of MPLS paths is sufficient to greatly improve a networks quality of service.

Routing optimization in IP networks utilizing additive and concave link metrics

Routing optimization provides network operators with a powerful method for traffic engineering. Its general objective is to distribute traffic flows evenly across available network resources in order to avoid network congestion and quality of service degradation. In this paper we consider routing optimization based on conventional routing protocols where packets are forwarded hop-by-hop in a destination-based manner. Unlike other work in this area, we consider routing protocols, which are able to take into account concave routing metrics in addition to additive ones. The concave link metric introduces an additional degree of freedom for routing optimization, thus, increasing its optimization potential. We present and evaluate a mixed-integer programming model, which works on these metrics. This model unifies the optimization for single-metric and dual-metric routing concepts and also includes the consideration of multipath routing. Furthermore, we propose a heuristic algorithm usable for larger network instances. Numerical results indicate that employment of both the dual-metric concept and multipath routing can achieve considerably better utilization results than default-configured single-metric routing. A significant finding is that metric-based routing optimization with two link metrics often comes close to the results obtainable by optimization of arbitrarily configurable routing.

On traffic characteristics and bandwidth requirements of voice over IP applications

Voice over IP (VoIP) services will play an important role in future IP networks, promising cost savings and new revenue sources to operators and service providers. To provide quality of service guarantees, certain mechanisms need to be implemented which support predictable packet handling, bandwidth allocation, and call admission control. In order to configure these mechanisms, a solid understanding of VoIP traffic characteristics and its respective bandwidth requirements is necessary. In this paper, we characterize traffic traces generated by various VoIP applications. According to the H.323 standard, the characteristics are described by means of token bucket parameters, which are then used to derive the required service rates for individual traffic flows. While in most cases sources send out fairly steady packet streams, there are situations where software based clients emit rather bursty traffic resulting in unreasonably high bandwidth needs. On basis of these traffic flows, we investigate the effects that token bucket parameters have on the bandwidth demand and discuss tradeoff possibilities in order to reduce it.

Interoperation of Mobile IPv6 and protocol independent multicast dense mode

We discuss different approaches of providing multicast traffic for mobile hosts. Mobile IPvd is used for mobility support. The network employs protocol independent multicast dense mode (PIM-DM) for multicast routing and multicast listener discovery (MLD) to collect multicast group membership information. We identify and analyze interoperation problems concerning membership control for mobile hosts and efficient multicast packet transfer from/to mobile hosts. We discuss four multicast delivery mechanisms for mobile senders and receivers, and compare them using criteria such as join delay, routing optimality, protocol overhead, network bandwidth and system load. In particular we suggest timer optimizations for MLD to support highly mobile receivers.

Packet transfer delay comparison of a store-and-forward and a cut-through resilient packet ring

Based on queuing theory we develop analytical approximations for the average packet transfer time of a store-and-forward and a cut-through buffer insertion ring with two client traffic priorities. These types of rings are architectural alternatives for resilient packet rings (RPR) which transport data (e.g., IP) packets over optical media. We use the approximations for a delay comparison of both ring architectures. It turns out that high priority traffic is more delayed in the cut-through architecture than in the store-and-forward architecture whereas low priority traffic performs similarly in both architectures.

Home agent redundancy and load balancing in mobile IPv6

This work proposes an optional extension to Mobile IPv6 to enhance the performance and the reliability of home agent clusters. First, a load balancing mechanism is presented which allows a set of home agents to distribute the load equally among them. Furthermore, we specify a Home Agent Redundancy Extension which increases the resilience of home agent clusters. In case of home agent failures, this extension enables the remaining peers to take over the functions of the missing home agents and, thus, keep up reachability of all currently registered mobile nodes. The extension is completely transparent for mobile nodes. The two mechanisms work independently of each other and, therefore, can be implemented separately. For better performance, though, it is advisable to apply both mechanisms at the same time.

Scenario-Based Traffic Modeling for Data Emanating from Medical Instruments in Clinical Environment

In this paper, the application of individual scenario-based traffic models in identifying the traffic carrying needs of networks in patient monitoring environments is presented. Since the total number of traffic sources is small for a central limit theorem-like approximation, a simple traffic matrix for each scenario can be used to identify the capacity needed for carrying data. The data traffic is classified into three categories; constant bit rate (CBR), On-Off, and Impulsive. In digital communications, the Impulsive traffic is considered a limiting case of the On-Off traffic. Lack of standard mechanisms for digitization of the patient monitoring data offers a unique challenge for traffic modeling. Sample traffic characteristics in a certain scenario are used to demonstrate that the network does not have to be designed for a very high bit rate even if some sources occasionally generate high data rates. This is due to the fact that a source generating high data rate could do so only for a small fraction of time, which results in smoothing out the data over longer periods of time.

On the dimensioning of voice over IP networks for various call admission control schemes

In this paper we investigate various call admission control (CAC) schemes for Voice over IP (VoIP) architectures in the context of network dimensioning. For each of the considered CAC concepts we introduce an abstract model and propose an optimized dimensioning procedure. Based on numerical results for several network scenarios, the trade-off between IP network overdimensioning on one side and the complexity of the CAC scheme on the other side is demonstrated.

A Dimensioning Strategy for Almost Guaranteed Quality of Service in Voice over IP Networks

Providing hard quality of service guarantees for Voice over IP traffic in multi-service IP networks is quite costly in terms of network resources. However, if one is willing to slightly soften the guarantees and to accept a small probability of service degradation, the amount of capacity necessary for voice traffic can be greatly reduced. In this paper, we investigate the tradeoff between quality of service and the required bandwidth for voice traffic in IP networks taking into account different scheduling schemes. We propose a dimensioning strategy, which allows the worst-case packet delay to exceed a desired threshold with a certain probability. Based on simulations, it can be demonstrated that bandwidth is reduced while the quality of service is assured in most cases.