Network Emulation with NetEm

2015 International Conference and Workshops on Networked Systems (NetSys), 2015

Most systems connected to the Internet are general purpose machines (except specialized routers and switches in the network core) that handle packet processing in software. Even in the network core, there is a trend towards packet processing in software, e.g. using OpenFlow or virtual switches. While packet processing in software is flexible and offers many capabilities, it also represents a challenge to evaluate, optimize, or predict the performance of such complex systems. This makes it hard to evaluate the networking performance of servers, end user hosts, or home routers. We present a study that investigates the packet latency caused by packet processing in the Linux network stack. We develop a simulation model in ns-3 for packet processing via the Linux network stack that helps understanding of its performance implications. We validate our simulation model based on measurements with nanosecond accuracy and software profiling.

Proceedings of the First International ICST Conference on Simulation Tools and Techniques for Communications Networks and Systems, 2008

In this paper we present DeSiNe, a modular flow-level network simulator. DeSiNe is aimed at performance analysis and benchmarking of Quality of Service routing algorithms and traffic engineering extensions. Several well-known QoS routing algorithms and traffic engineering extensions have been implemented in De-SiNe. The flow-level nature provides scalability, such that large networks and heavy-traffic conditions are possible. In this paper, the functional and structural design of DeSiNe are presented and the usability and various features are illustrated by means of several examples. The source code of DeSiNe is publicly available.

Zenodo (CERN European Organization for Nuclear Research), 2023

Network Services automation requires predictable Quality of Service (QoS) performance, measured in terms of throughput, delay and jitter, to allow making proactive decisions. QoS is typically guaranteed by overprovisioning capacity dedicated to the packet connection, which increases costs for customers and network operators, especially when the traffic generated by the users and/or the virtual functions highly varies over the time. This paper presents the PILOT methodology for modeling the performance of packet connections during commissioning testing in terms of throughput, delay and jitter. PILOT runs in a sandbox domain and constructs a scenario where an efficient traffic flow simulation environment, based on the CURSA-SQ model, is used to generate large amounts of data for Machine Learning (ML) model training and validation. The simulation scenario is tuned using real measurements of the connection obtained from a set of active probes.

2018

Dummynet is a link emulator that can be used by itself, as well as integrated within testbeds such as Emulab. Despite its popularity in the research community, Dummynet still lacks the ability to precisely emulate certain real network e↵ects. In particular, it has no support for packet reordering. Since reordering is a common and prevalent network phenomenon just like packet loss or delay, it cannot be ignored when implementing emulators if we want to provide realistic emulation. It has been observed that networks su↵er from reordering caused by packet striping, retransmissions, load balancing, multipath forwarding, etc. This has significant negative e↵ects on the performance of both Transmission Control Protocol (TCP) and User Datagram Protocol (UDP). With the increase in prevalence of real-time streaming UDP applications such as video conferencing and Internet Protocol Television (IPTV), it has become important to focus on this problem which a↵ects the performance of all these applications. Research into models and tools to diagnose and understand reordering requires that a sophisticated metric be used to describe it. So, in this thesis, I make two contributions: improving the realism of tra c shaping in Dummynet emulator by adding support for emulation of reordering, and an algorithm, a max-flow solver, that generates reordered sequences to be used by Dummynet, from a sophisticated reordering metric called Reorder Density (RD). My implementation enables the user to specify the desired amount of reordering in a metric, such as RD (or even others), and have Dummynet generate tra c that is reordered according to the input metric's value. This is accomplished within Dummynet by the use of a newly implemented scheduler. I conclude my thesis with an evaluation using real and software generated network traces to show that the algorithm is scalable and the implementation works correctly. Also, a datapath evaluation to show that my modifications to Dummynet do not result in any unnecessary increase in emulation running time is included.

Computer Standards & Interfaces, 2005

In this paper, we address the problem of the end-to-end QoS management for multimedia and real-time applications. We would focus on the presentation of a multi-profile communication environment and its reliability in ensuring end-to-end QoS management in the network communication processes. This communication environment, called Virtual Communication Support (VCS), provides a dynamic assignment of application requirements to the network resources reservation and protocol layers invocation. The advantage of this model is that it performs the capabilities of the network by abstracting the details of all communication stacks-layers that must be invoked in the communication process from application specificities and needs.

When developing new network protocols, new functionalities of network devices, new traffic models and other novelties, we face the problem of non-existing tools. If we want to simulate the behavior of the network with a newly developed feature we have to develop a new simulator or add a new functionality to an existing simulation tool. Both options can be quite complex and/or time consuming. To facilitate the testing and simulation of a new packet scheduler that we developed during our research we have developed a simulator for a general model of a network device. For this purpose we have used the Modular Simulation language, MODSIM III. It is a general-purpose, modular, block structured language that provides support for object-oriented programming, discrete event simulation and animated graphics. The simulator includes modules for the most important elements and functions of a packet network device, modules for collection of results, and modules for writing the results into a standard format files. We have tested and validated the operation of the simulator with analytically verifiable settings. The simulation and analytical results were practically the same. Encouraged with that we have then simulated newly developed packet schedulers and network device functionalities. Some of the simulation results are presented in this article, more you can find in corresponding references.

IEEE Computer, 2002

A t this point in technology's evolution, the simplicity, elegance, extensibility, and broad compatibility of the Internet protocol suite has made it the automatic choice for most forms of communication. From Web browsing to enterprise computing, from electronic mail to next-generation wireless telephony, the common denominator is "everything over IP."

IEEE Communications Surveys & Tutorials, 2003

2021

When analyzing the impact of a communication system on the Quality of Experience (QoE) of an interactive application, simulation and mathematical modeling typically require reimplementation or modeling of the application under test and only provide insights in terms of selected Key Performance Indicators (KPIs), which raises the need for network emulation. In this demo, we present FlowEmu, an open-source flow-based network emulator that allows the user to manipulate the underlying model and analyze various statistics in real-time via an easy-to-use Graphical User Interface (GUI), by using interactive game streaming as an example application.

2013 Information Security for South Africa, 2013

The ability to provide the simulation of packets traversing an internet path is an integral part of providing realistic simulations for network training, and cyber defence exercises. This paper builds on previous work, and considers an in-kernel approach to solving the routing simulation problem. The inkernel approach is anticipated to allow the framework to be able to achieve throughput rates of 1GB/s or higher using commodity hardware. Processes that run outside the context of the kernel of most operating system require context switching to access hardware and kernel modules. This leads to considerable delays in the processes, such as network simulators, that frequently access hardware such as hard disk accesses and network packet handling. To mitigate this problem, as experienced with earlier implementations, this research looks towards implementing a kernel module to handle network routing and simulation within a UNIX based system. This would remove delays incurred from context switching and allows for direct access to the hardware components of the host. This implementation evaluates the use of Work Queues within the Linux Kernel to schedule packets, and Radix Trees store routing information in the form of Nodes. The memory and CPU requirements are taken into consideration, and conclusions towards the trade off of resources for higher throughput and speeds are discussed. Preliminary tests using this implementation method have achieved throughput rates in routing packets from core to endpoint host at speeds up to 1 gigabit. This however, is subject to host load, network size and number of packets being routed simultaneously at the time of transaction. These factors are discussed and suggestions made towards further optimisations.