Motivated by the increasing complexity of today’s communication networks, autonomous networks (AN) have recently attracted much attention from industry and academia. Automated network configurations through intent-based networking (IBN) is the first and crucial step towards AN. In this paper, we focus on realizing automated network configurations leveraging the rapid evolution of large language models (LLMs). We proposed a framework based on modular design integrated with LLMs. Our preliminary results confirm that LLMs have great potential in IBN and can serve as an important point on the way to realizing AN while keeping data privacy and integrity of configuration results. However, several challenges and open questions remain to be addressed.

Some network applications have requirements that exceed the service levels offered by the best-effort Internet. Several network-layer Quality of Service architectures with extended service levels have been designed, but the massive scale and distributed nature of the Internet have prohibited their wide deployment. It now seems clear that the special needs of demanding applications must be met through other approaches. This paper describes how incrementally-deployed innovative solutions at the network level can contribute to an improved service for a particular type of applications, namely high-quality, wide-area video conferencing. We have built a global IP network, aiming to give users a better video conferencing experience primarily through packet loss reduction in transport networks. The key concept in our approach is a well-provisioned network-layer overlay, combined with geography-based "cold potato" BGP routing. Through an extensive set of experiments we show how our design choices impact routing and data plane behavior in the network, and demonstrate that we are able to significantly reduce packet loss compared to wide-area transport through global transit providers.

In order to reduce the number of BGP updates that routers need to process, it is common to rate-limit such updates using a timer that specifies the minimum time between two consecutive updates for a given destination prefix. Rate-limiting plays an important role in determining the number of routing updates that are generated after a routing event, and the time it takes before the network converges to a new stable state. Still, there are few guidelines for how rate-limiting timers should be configured in order to achieve the desired convergence properties. This work takes a first step in this direction, by exploring how different rate-limiting implementations and configurations affect the resulting churn level in a live BGP session. Measurements are performed on multiple parallel BGP sessions to a stub AS, configured with and without rate-limiting timers. We find that the daily rate of updates is reduced by two thirds when configuring the timer to the default value recommended by BGP standards. We further investigate different rate-limiting implementations and configurations using the measured BGP update patterns on emulated BGP sessions, and find that increasing the rate-limiting timer gives a logarithmic decrease in churn. Finally, using BGP update traces from RouteViews, we present the first empirical model that quantifies the impact of rate-limiting in terms of churn reduction given the observed arrival pattern of BGP updates.

Multi-topology routing is an increasingly popular IP network management concept that allows transport of different traffic types over disjoint network paths. The concept is of particular interest for implementation of IP fast reroute (IP FRR). The authors have previously proposed an IP FRR scheme based on multi-topology routing called Multiple Routing Configurations (MRC). MRC supports guaranteed, instantaneous recovery from any single link or node failure in biconnected networks as well as from many combined failures, provided sufficient bandwidth on the surviving links. Furthermore, in MRC different failures result in routing over different network topologies, which gives a good control of the traffic distribution in the networks after a failure

IP fast reroute (IP FRR) denominates a set of methods and technologies for proactive, local recovery in IP networks. It has received much attention recently, in the research community and the IETF. Several proposals for IP FRR have been published by independent research groups, and main properties and trade-offs are understood. However, only simulation-based and analytical evaluations are known to be performed so far. In this paper we present the architecture and evaluate performance of the first known practical implementation of IP FRR. The architecture features a modular approach consisting of an IP FRR framework and exchangeable backup path calculation modules. The performance evaluation is based on an implementation of multiple routing configurations. The implementation provide throughput that closely matches the achieved performance during normal operation.

Multi-topology routing is an increasingly popular IP network management concept that allows transport of different traffic types over disjoint network paths. The concept is of particular interest for implementation of IP fast reroute (IP FRR). First, it can support guaranteed, instantaneous recovery from any link or node failure. Second, different failures result in routing over different network topologies, which augments the parameter space for load distribution optimizations. Multiple routing configurations (MRC) is the state-of-the-art IP FRR scheme based on multi-topology routing today. In this paper we present a new, enhanced IP FRR scheme which we call ldquorelaxed MRCrdquo (rMRC). rMRC simplifies the topology construction and increases the routing flexibility in each topology. According to our experimental evaluation, rMRC has several benefits compared to MRC. The number of backup topologies required to provide protection against the same set of failures is reduced, hence reducing state in routers. In addition, the backup paths are shorter, and the link utilization is significantly better.

Slow convergence of IP routing in case of network component failures represents a problem for emerging time-critical network applications. It has motivated work on IP fast reroute mechanisms that use local, proactive recovery in order to immediately provide alternative routes for the affected data packets. We present a new mechanism for IP fast recovery called IP Redundant Trees (IPRT). IPRT builds two redundant trees per network destination and pre-calculates backup next hops based on them. Combined with a smart forwarding scheme, IPRT provides proactive, local and loop-free recovery in connectionless IP networks. IPRT is simple and flexible, and requires a low increase of the routing state amount. We evaluate several key performance metrics in IPRT and other actual IP fast recovery proposals and argue that IPRT indeed has very attractive properties.

Recovery of traffic in connectionless pure IP networks has traditionally been handled by a full re-convergence of the network state. This process operates in a time scale that is not compatible with new real time and highly dependable services. Recently, schemes for fast local and proactive recovery in connectionless IP networks have been proposed. All these schemes are designed to guarantee recovery of the failure of one component. As IP protocols are used to carry more highly dependable services and new wireless infrastructures are approaching, guaranteed failure coverage of more than one failure becomes necessary. In this paper we present and evaluate a scheme that guarantees to handle any two concurrent failures in a network. We are not aware of any other schemes that addresses such guarantees. We evaluate and compare it with other known recovery schemes, and we show how it gives substantially better recovery success rates than the schemes designed for one fault tolerance, also for more than two failures.

The slow convergence of IGP routing protocols after a topology change has led to several proposals for proactive recovery schemes in IP networks. These proposals are limited to guaranteeing loop-free connectivity after a link or node failure, and do not take into account the resulting load distribution in the network. This can lead to congestion and packet drops. In this work, we show how a good load distribution can be achieved in pure IP networks immediately after a link failure, when multiple routing configurations (MRC) is used as a fast recovery mechanism. This paper is the first attempt to improve the load balancing when a proactive recovery scheme is used. Unlike load balancing methods used with normal IP rerouting, our method does not compromise on the routing performance in the failure free case. Our method is evaluated using simulations on several real and synthetically generated network topologies. The evaluation shows that our method yields good routing performance, making it feasible to use MRC to handle transient network failures.

Multi-topology routing is an increasingly popular concept in computer communications, and can be successfully applied to network fault tolerance. Multi-topology routing implies an increase in routing state requirements proportional to the number of logical topologies used. The network routing state is often subject to stringent system limits and should therefore be deterministic and controllable. In this paper we show that the number of logical topologies needed to guarantee link-fault tolerant multi-topology routing in a network has an upper bound determined by a topological property we call largest minimal cycle. We provide a formal proof of the validity of this bound in arbitrary biconnected networks. We also evaluate the quality of this bound, showing that it is reasonably strict.

Abstract — As the Internet takes an increasingly central role in our communications infrastructure, the slow convergence of routing protocols after a network failure becomes a growing problem. To assure fast recovery from link and node failures in IP networks, we present a new recovery scheme called Multiple Routing Configurations (MRC). MRC is based on keeping additional routing information in the routers, and allows packet forwarding to continue on an alternative output link immediately after the detection of a failure. Our proposed scheme guarantees recovery in all single failure scenarios, using a single mechanism to handle both link and node failures, and without knowing the root cause of the failure. MRC is strictly connectionless, and assumes only destination based hop-by-hop forwarding. It can be implemented with only minor changes to existing solutions. In this paper we present MRC, and analyze its performance with respect to scalability, backup path lengths, and load distribution after a failure. I.

Most existing methods for network recovery are often complex and seldom used by network administrators. In this paper we present a novel approach for global and local recovery named resilient routing layers (RRL). The method is supported by algorithms, but also simple enough for a network administrator to implement by hand for reasonably sized networks. The idea in our approach is that for each node in the network there is a topology subset called a "safe layer", which can handle any traffic affected by a fault in the node itself, or any of its links. We demonstrate that our approach performs well compared to other comparable methods in a wide range of different network topologies. Particularly, we demonstrate RRLs performance for what are assumed to be the weakest parameters for our method, i.e., backup-path lengths and state information overhead. We discuss implementation issues of RRL, and demonstrate its applicability to MPLS networks.

We present a novel scheme for network recovery, named resilient routing layers (RRL). Our proposed scheme is based on calculating fully connected topology subsets, termed layers, which are used to forward traffic in case of a network failure. For the purpose of this work, the layers are created to protect against link failures only. RRL keeps pre-calculated backup routing information in the network stations. This allows local response to network failures, which gives recovery in the order of milliseconds. The main strengths of our approach are its flexibility, as it is independent of the network technology used, and its simplicity, as it offers the network operator a simple and coherent view of the resources available after a link failure. We also show that our scheme scales well for networks of several hundred nodes.