1,721,086 research outputs found
High-Priority First Transmission to Efficiently Support Service Differentiation in Just-In-Time OBS Networks
Full text linkMinimizing Preemption Probability toEfficiently Support Service Differentiation in Just-In-Time based OBS Networks
No full textDesign of a Deterministic Service Switch for Avionics Networks
No full textAvionics full-duplex switched Ethernet, with its capability of delivering deterministic latency at low deployment costs, represents a major advance for avionics communication systems. However, avionics full-duplex switched Ethernet does not eliminate contentions at network switches, which may result in an uncontrolled increase in end-to-end latencies, and hence limit the potential of this technology. The use of time to pace packet transmission and forwarding (that is, the adoption of pipeline forwarding) has been demonstrated to avoid contentions at switching nodes in wide-area communication networks, thus being potentially attractive also for avionics networks in general, and avionics full-duplex switched Ethernet in particular. However, questions might arise concerning the actual feasibility of such an approach in the avionics domain due to the possibly high costs required to deploy packet switches able to operate according to a very precise time schedule, which is necessary to respect the stringent delivery deadlines of avionics systems. This paper focuses on the development of a pipeline forwarding switch, showing how simple design guidelines allow a general-purpose symmetric multiprocessor architecture (such as modern personal computers) to offer deterministic and low latency delivery, thus being able to satisfy avionics requirements. The paper also describes a real implementation of a pipeline forwarding switch prototype, based on a dual-processor personal computer and the FreeBSD operating syste
Pipeline Forwarding of Packets based on a Low Accuracy Network-distributed Common Time Reference
Full text link
Intent-Based Kubernetes Configuration via LLMs: Current Trends and Open Challenges
Full text linkThe advent of Large Language Models (LLMs) is progressively transforming how complex tasks across various domains can be automated, with a notable potential impact on cloud computing operations. In this domain, LLMs might be used, for example, to configure Kubernetes (K8s) clusters via the generation of manifest files -structured configuration files defining the containerized environment. However, despite the considerable advances in LLMs’ text generation, this task conceals several challenges that prevent operators from achieving a fully automated process. In this paper, we present the current trends in solving these gaps, quantitatively evaluate the accuracy of LLM-based approaches to generate K8s manifests starting from human intents, and discuss open challenges that make benchmarking and automation still complex. Experiments over three open-source LLMs demonstrate how intent-based K8s manifest generation can be effectively achieved through model fine-tuning, but also what open issues remain and must be addressed prior to having an autonomous and self-healing K8s infrastructure managed via Agentic AI
Time-Driven Access and Forwarding for Industrial Wireless Multihop Networks
Full text linkThe deployment of wireless technologies in industrial networks is very promising mainly due to their inherent flexibility. However, current wireless solutions lack the capability to provide the deterministic, low delay service required by many industrial applications. Moreover, the high level of interference generated by industrial equipment limits the coverage that ensures acceptable performance. Multi-hop solutions, when combining frame forwarding with higher node density, have the potential to provide the needed coverage while keeping radio communication range short. However, in multi-hop solutions the medium access time at each of the nodes traversed additively contributes to the end-to-end delay and the forwarding delay (i.e., the time required for packets to be processed, switched, and queued) at each node is to be added as well. This paper describes Time-driven Access and Forwarding (TAF), a solution for guaranteeing deterministic delay, at both the access and forwarding level, in wireless multi-hop networks, analyzes its properties, and assesses its performance in industrial scenario
Push applications and dynamic content generation over content-centric networking
No full textContent-Centric Networking (CCN) represents an established candidate for the future Internet, proposing a routing architecture designed to elevate content to first class entity. Starting from the fact that the network usage has dramatically evolved towards content retrieval, CCN relies on an on-demand pull based mechanism to transfer data from the different sources to the heterogeneous consumers. This paradigm enhances the network in a number of ways, ranging from the newly introduced in-network caching capabilities to the benefits provided by the symmetric data routing adopted by CCN. In this renewed network scenario, we place our attention to those applications that do not perfectly fit the pull paradigm, stating that they need to be supported as well and proposing an effective way to achieve scalability on large scale push applications. We provide the following contributions: (i) we identify the functions that a data-centric architecture should support; (ii) propose and compare our solution with the state of the art framework designed for the specific problem of pushing data to content requesters; and (iii) evaluate their performance in terms of traffic generated and scalability achieved by simulating a real Internet Service Provider (ISP) topology and the realistic workload of a generic social network application
A Scalable Solution for Engineering Streaming Traffic in the Future Internet
No full textAs traffic on the Internet continues to grow exponentially, there is a real need to solve scalability and traffic engineering simultaneously — specifically, without using over-provisioning in order to accommodate streaming media traffic. One of the threats to the current operation stability of the Internet comes from UDP-based streaming media applications, such as Skype (which is currently doubling every 6-month) today and video services in the near future. This paper shows how the Internet can benefit from pipeline forwarding in order to: (i) construct ultra-scalable IP switches, (ii) provide predictable quality of service for UDP-based streaming applications, while (iii) preserving elastic TCP-based traffic as is, i.e., without affecting any existing best-effort application
- …
