Search Results

Despite its widespread use, the "Gao-Rexford" model has long been recognized for its limitations in accurately capturing Internet routing behavior. However, the root causes of these limitations remain poorly understood, due to the lack of ground truth ...

Today, networking hardware is not fast enough to save energy with rate adaptation. Or is it? While we are not (yet) able to turn on and off line cards in milliseconds, we can do so a couple of times per day. The question is, does it make sense to save ...

Machine learning (ML) is a powerful tool to model the complexity of communication networks. As networks evolve, we cannot only train once and deploy. Retraining models, known as continual learning, is necessary. Yet, to date, there is no established ...

The allreduce operation is an essential building block for many distributed applications, ranging from the training of deep learning models to scientific computing. In an allreduce operation, data from multiple hosts is aggregated together and ...

• In-network offload can improve allreduce performance up to 2x.
• Existing approaches are prone to congestion due to static network paths selection.
• Canary is the first congestion-aware in-network allreduce algorithm.
• Canary ...

Control plane verification promises to help operators build reliable networks by reporting a counterexample that violates the specification. However, a single counterexample imposes a major challenge for operators to understand and repair the violation.

...

Analyzing a network's behavior during convergence is challenging due to its highly non-deterministic nature. To address this, we developed BGPseer, the first analyzer that predicts specification violations during BGP convergence without network ...

The demand for ever-faster links and devices is competing with the need to make networks more energy-efficient. One energy-saving technique is shutting down dispensable parts of the network. In this work, we investigate the limits of link sleeping; i.e., ...

The concept of programmable packet scheduling has been recently introduced, enabling the programming of scheduling algorithms into existing data planes without requiring new hardware designs. Notably, several programmable schedulers have been proposed, ...

Today, the ICT industry has a massive carbon footprint (a few percent of worldwide emissions) and one of the fastest growth rates. The Internet accounts for a large part of that footprint while being also energy inefficient; i.e., the total energy cost ...

Anomaly detection is an essential building block of many applications, including DDoS detection, root cause analysis, traffic estimation, and change detection. A vital part of detecting anomalies is establishing a sense of normality, e.g., by learning ...

BGP reconfigurations are a daily occurrence for most network operators, especially in large networks. Yet, performing safe and robust BGP reconfiguration changes is still an open problem. Few BGP reconfiguration techniques exist, and they are either (i) ...

Today, networking hardware is not fast enough to save energy with rate adaptation. Or is it? While we are not (yet) able to turn on and off line cards in milliseconds, we can do so a couple of times per day. The question is, does it make sense to save ...

We introduce FnF-BFT, the first partially synchronous BFT protocol with performance guarantees under truly byzantine attacks during stable networking conditions. At its core, FnF-BFT parallelizes the execution of requests by allowing all replicas ...

Over the last years, P4 has positioned itself as the primary language for data-plane programming. Despite its constant evolution, the P4 language still "suffers" from one significant limitation: the voluminosity of its code. P4 applications easily reach ...

We present a new method for scaling automatic configuration of computer networks. The key idea is to relax the computationally hard search problem of finding a configuration that satisfes a given specification into an approximate objective amenable to ...

Generalizing machine learning (ML) models for network traffic dynamics tends to be considered a lost cause. Hence for every new task, we design new models and train them on model-specific datasets closely mimicking the deployment environments. Yet, an ...

In theory, any network operator, developer, or vendor should have access to large amounts of live network traffic for testing their solutions. In practice, though, that is not the case. Network actors instead have to use packet traces or synthetic ...

What are the worst outages for Internet users? How long do they last, and how wide are they? Such questions are hard to answer via traditional outage detection and analysis techniques, as they conventionally rely on network-level signals and do not ...