ns-3 Tutorial (html version)

For a pdf version of this tutorial, see http://www.nsnam.org/docs/tutorial.pdf.

This is an ns-3 tutorial. Primary documentation for the ns-3 project is available in four forms:

• ns-3 Doxygen/Manual: Documentation of the public APIs of the simulator
• Tutorial (this document)
• Reference Manual: Reference Manual
• ns-3 wiki

This document is written in GNU Texinfo and is to be maintained in revision control on the ns-3 code server. Both PDF and HTML versions should be available on the server. Changes to the document should be discussed on the ns-developers@isi.edu mailing list.

This software is free software; you can redistribute it and/or modify it under the terms of the GNU General Public License as published by the Free Software Foundation; either version 2 of the License, or (at your option) any later version.

This software is distributed in the hope that it will be useful, but WITHOUT ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License for more details.

You should have received a copy of the GNU General Public License along with this program. If not, see http://www.gnu.org/licenses/.

1. Introduction

The ns-3 simulator is a discrete-event network simulator targeted primarily for research and educational use. The ns-3 project, started in 2006, is an open-source project developing ns-3.

Primary documentation for the ns-3 project is available in four forms:

• ns-3 Doxygen/Manual: Documentation of the public APIs of the simulator
• Tutorial (this document)
• Reference Manual: Reference Manual
• ns-3 wiki

The purpose of this tutorial is to introduce new ns-3 users to the system in a structured way. It is sometimes difficult for new users to glean essential information from detailed manuals and to convert this information into working simulations. In this tutorial, we will build several example simulations, introducing and explaining key concepts and features as we go.

As the tutorial unfolds, we will introduce the full ns-3 documentation and provide pointers to source code for those interested in delving deeper into the workings of the system.

A few key points are worth noting at the onset:

• Ns-3 is not an extension of ns-2; it is a new simulator. The two simulators are both written in C++ but ns-3 is a new simulator that does not support the ns-2 APIs. Some models from ns-2 have already been ported from ns-2 to ns-3. The project will continue to maintain ns-2 while ns-3 is being built, and will study transition and integration mechanisms.
• Ns-3 is open-source, and the project strives to maintain an open environment for researchers to contribute and share their software.

1.1 For ns-2 Users

For those familiar with ns-2, the most visible outward change when moving to ns-3 is the choice of scripting language. Ns-2 is scripted in OTcl and results of simulations can be visualized using the Network Animator nam. It is not possible to run a simulation in ns-2 purely from C++ (i.e., as a main() program without any OTcl). Moreover, some components of ns-2 are written in C++ and others in OTcl. In ns-3, the simulator is written entirely in C++, with optional Python bindings. Simulation scripts can therefore be written in C++ or in Python. The results of some simulations can be visualized by nam, but new animators are under development. Since ns-3 generates pcap packet trace files, other utilities can be used to analyze traces as well. In this tutorial, we will first concentrate on scripting directly in C++ and interpreting results via ascii trace files.

But there are similarities as well (both, for example, are based on C++ objects, and some code from ns-2 has already been ported to ns-3). We will try to highlight differences between ns-2 and ns-3 as we proceed in this tutorial.

A question that we often hear is "Should I still use ns-2 or move to ns-3?" The answer is that it depends. ns-3 does not have all of the models that ns-2 currently has, but on the other hand, ns-3 does have new capabilities (such as handling multiple interfaces on nodes correctly, use of IP addressing and more alignment with Internet protocols and designs, more detailed 802.11 models, etc.). ns-2 models can usually be ported to ns-3 (a porting guide is under development). There is active development on multiple fronts for ns-3. The ns-3 developers believe (and certain early users have proven) that ns-3 is ready for active use, and should be an attractive alternative for users looking to start new simulation projects.

1.2 Contributing

Ns-3 is a research and educational simulator, by and for the research community. It will rely on the ongoing contributions of the community to develop new models, debug or maintain existing ones, and share results. There are a few policies that we hope will encourage people to contribute to ns-3 like they have for ns-2:

• Open source licensing based on GNU GPLv2 compatibility;
• wiki;
• Contributed Code page, similar to ns-2's popular Contributed Code page;
• src/contrib directory (we will host your contributed code);
• Open bug tracker;
• Ns-3 developers will gladly help potential contributors to get started with the simulator (please contact one of us).

We realize that if you are reading this document, contributing back to the project is probably not your foremost concern at this point, but we want you to be aware that contributing is in the spirit of the project and that even the act of dropping us a note about your early experience with ns-3 (e.g. "this tutorial section was not clear..."), reports of stale documentation, etc. are much appreciated.

1.3 Tutorial Organization

The tutorial assumes that new users might initially follow a path such as the following:

• Try to download and build a copy;
• Try to run a few sample programs;
• Look at simulation output, and try to adjust it.

As a result, we have tried to organize the tutorial along the above broad sequences of events.

2. Resources

2.1 The Web

There are several important resources of which any ns-3 user must be aware. The main web site is located at http://www.nsnam.org and provides access to basic information about the ns-3 system. Detailed documentation is available through the main web site at http://www.nsnam.org/documents.html. You can also find documents relating to the system architecture from this page.

There is a Wiki that complements the main ns-3 web site which you will find at http://www.nsnam.org/wiki/. You will find user and developer FAQs there, as well as troubleshooting guides, third-party contributed code, papers, etc.

The source code may be found and browsed at http://code.nsnam.org/. There you will find the current development tree in the repository named ns-3-dev. Past releases and experimental repositories of the core developers may also be found there.

2.2 Mercurial

Complex software systems need some way to manage the organization and changes to the underlying code and documentation. There are many ways to perform this feat, and you may have heard of some of the systems that are currently used to do this. The Concurrent Version System (CVS) is probably the most well known.

The ns-3 project uses Mercurial as its source code management system. Although you do not need to know much about Mercurial in order to complete this tutorial, we recommend becoming familiar with Mercurial and using it to access the source code. Mercurial has a web site at http://www.selenic.com/mercurial/, from which you can get binary or source releases of this Software Configuration Management (SCM) system. Selenic (the developer of Mercurial) also provides a tutorial at http://www.selenic.com/mercurial/wiki/index.cgi/Tutorial/, and a QuickStart guide at http://www.selenic.com/mercurial/wiki/index.cgi/QuickStart/.

You can also find vital information about using Mercurial and ns-3 on the main ns-3 web site.

2.3 Waf

Once you have source code downloaded to your local system, you will need to compile that source to produce usable programs. Just as in the case of source code management, there are many tools available to perform this function. Probably the most well known of these tools is make. Along with being the most well known, make is probably the most difficult to use in a very large and highly configurable system. Because of this, many alternatives have been developed. Recently these systems have been developed using the Python language.

The build system Waf is used on the ns-3 project. It is one of the new generation of Python-based build systems. You will not need to understand any Python to build the existing ns-3 system, and will only have to understand a tiny and intuitively obvious subset of Python in order to extend the system in most cases.

For those interested in the gory details of Waf, the main web site can be found at http://freehackers.org/~tnagy/waf.html.

2.4 Development Environment

As mentioned above, scripting in ns-3 is done in C++ or Python. As of ns-3.2, most of the ns-3 API is available in Python, but the models are written in C++ in either case. A working knowledge of C++ and object-oriented concepts is assumed in this document. We will take some time to review some of the more advanced concepts or possibly unfamiliar language features, idioms and design patterns as they appear. We don't want this tutorial to devolve into a C++ tutorial, though, so we do expect a basic command of the language. There are an almost unimaginable number of sources of information on C++ available on the web or in print.

If you are new to C++, you may want to find a tutorial- or cookbook-based book or web site and work through at least the basic features of the language before proceeding. For instance, this tutorial.

The ns-3 system uses several components of the GNU “toolchain” for development. A software toolchain is the set of programming tools available in the given environment. For a quick review of what is included in the GNU toolchain see, http://en.wikipedia.org/wiki/GNU_toolchain. ns-3 uses gcc, GNU binutils, and gdb. However, we do not use the GNU build system, either make or autotools, using Waf instead.

Typically an ns-3 author will work in Linux or a Linux-like environment. For those running under Windows, there do exist environments which simulate the Linux environment to various degrees. The ns-3 project supports development in the Cygwin environment for these users. See http://www.cygwin.com/ for details on downloading (MinGW is presently not supported). Cygwin provides many of the popular Linux system commands. It can, however, sometimes be problematic due to the way it actually does its emulation, and sometimes interactions with other Windows software can cause problems.

If you do use Cygwin or MinGW; and use Logitech products, we will save you quite a bit of heartburn right off the bat and encourage you to take a look at the MinGW FAQ.

Search for “Logitech” and read the FAQ entry, “why does make often crash creating a sh.exe.stackdump file when I try to compile my source code.” Believe it or not, the Logitech Process Monitor insinuates itself into every DLL in the system when it is running. It can cause your Cygwin or MinGW DLLs to die in mysterious ways and often prevents debuggers from running. Beware of Logitech software when using Cygwin.

Another alternative to Cygwin is to install a virtual machine environment such as VMware server and install a Linux virtual machine.

2.5 Socket Programming

We will assume a basic facility with the Berkeley Sockets API in the examples used in this tutorial. If you are new to sockets, we recommend reviewing the API and some common usage cases. For a good overview of programming TCP/IP sockets we recommend Practical TCP/IP Sockets in C, Donahoo and Calvert.

There is an associated web site that includes source for the examples in the book, which you can find at: http://cs.baylor.edu/~donahoo/practical/CSockets/.

If you understand the first four chapters of the book (or for those who do not have access to a copy of the book, the echo clients and servers shown in the website above) you will be in good shape to understand the tutorial. There is a similar book on Multicast Sockets, Multicast Sockets, Makofske and Almeroth. that covers material you may need to understand if you look at the multicast examples in the distribution.

3. Getting Started

3.1 Downloading ns-3

From this point forward, we are going to assume that the reader is working in Linux or a Linux emulation environment (Linux, Cygwin, etc.) and has the GNU toolchain installed and verified. We are also going to assume that you have Mercurial and Waf installed and running on the target system as described in the Getting Started section of the ns-3 web site: http://www.nsnam.org/getting_started.html.

The ns-3 code is available in Mercurial repositories on the server code.nsnam.org. You can download a tarball release at http://www.nsnam.org/releases/, or you can work with repositories using Mercurial.

If you go to the following link: http://code.nsnam.org/, you will see a number of repositories. Many are the private repositories of the ns-3 development team. The repositories of interest to you will be prefixed with “ns-3”. The current development snapshot (unreleased) of ns-3 may be found at: http://code.nsnam.org/ns-3-dev/. Official releases of ns-3 will be numbered as ns-3.<release> with any required hotfixes added as minor release numbers. For example, a second hotfix to a hypothetical release nine of ns-3 would be numbered ns-3.9.2.

The current development snapshot (unreleased) of ns-3 may be found at: http://code.nsnam.org/ns-3-dev/. The developers attempt to keep this repository in a consistent, working state but it is a development area with unreleased code present, so you may want to consider staying with an official release if you do not need newly-introduced features.

Since the release numbers are going to be changing, I will stick with the more constant ns-3-dev here in the tutorial, but you can replace the string “ns-3-dev” with your choice of release (e.g., ns-3.2) in the text below. You can find the latest version of the code either by inspection of the repository list or by going to the “Getting Started” web page and looking for the latest release identifier.

One practice is to create a directory called repos in one's home directory under which one can keep local Mercurial repositories. Hint: we will assume you do this later in the tutorial. If you adopt that approach, you can get a copy of the development version of ns-3 by typing the following into your Linux shell (assuming you have installed Mercurial):

  cd
  mkdir repos
  cd repos
  hg clone http://code.nanam.org/ns-3-dev

As the hg (Mercurial) command executes, you should see something like the following,

  destination directory: ns-3-dev
  requesting all changes
  adding changesets
  adding manifests
  adding file changes
  added 3276 changesets with 12301 changes to 1353 files
  594 files updated, 0 files merged, 0 files removed, 0 files unresolved

After the clone command completes, you should have a directory called ns-3-dev under your ~/repos directory, the contents of which should look something like the following:

  AUTHORS  examples/  README         samples/  utils/   waf.bat*
  build/   LICENSE    regression/    scratch/  VERSION  wscript
  doc/     ns3/       RELEASE_NOTES  src/      waf*

Similarly, if working from a released version instead, you can simply

  cd
  mkdir repos
  wget http://www.nsnam.org/releases/ns-3.2.tar.bz2
  bunzip2 ns-3.2.tar.bz2
  tar xvf ns-3.2.tar

You are now ready to build the ns-3 distribution.

3.2 Building ns-3

We use Waf to build the ns-3 project. The first thing you will need to do is to configure the build. For reasons that will become clear later, we are going to work with debug builds in the tutorial. To explain to Waf that it should do debug builds you will need to execute the following command,

  ./waf -d debug configure

This runs Waf out of the local directory (which is provided as a convenience for you). As the build system checks for various dependencies you should see output that looks similar to the following,

Checking for program g++                 : ok /usr/bin/g++ 
Checking for compiler version            : ok Version 4.0.1 
Checking for program cpp                 : ok /usr/bin/cpp 
Checking for program ar                  : ok /usr/bin/ar 
Checking for program ranlib              : ok /usr/bin/ranlib 
Checking for compiler could create programs : ok  
Checking for compiler could create shared libs : ok  
Checking for compiler could create static libs : ok  
Checking for flags -O2 -DNDEBUG                : ok  
Checking for flags -g -DDEBUG                  : ok  
Checking for flags -g3 -O0 -DDEBUG             : ok  
Checking for flags -Wall                       : ok  
Checking for g++                               : ok  
Checking for -Wno-error=deprecated-declarations compilation flag support : no 
Checking for header stdlib.h                   : ok  
Checking for header stdlib.h                   : ok  
Checking for header signal.h                   : ok  
Checking for library rt                        : not found 
Checking for header pthread.h                  : ok  
Checking for high precision time implementation: 128-bit integer 
Checking for header stdint.h                   : ok  
Checking for header inttypes.h                 : ok  
Checking for header sys/inttypes.h             : not found 
Checking for package gtk+-2.0 >= 2.12          : not found 
Checking for library sqlite3                   : ok  
Checking for package goocanvas gthread-2.0     : not found 
Checking for program python                    : ok /usr/local/bin/python 
Checking for Python version >= 2.3             : ok 2.4.3 
Checking for library python2.4                 : not found 
Checking for library python2.4                 : not found 
Checking for library python24                  : not found 
Checking for program python2.4-config          : not found 
Checking for header Python.h                   : not found 
Checking for program diff                      : ok /usr/bin/diff 
Checking for program hg                        : ok /opt/local/bin/hg 
---- Summary of optional NS-3 features:
Threading Primitives        : enabled
Real Time Simulator         : enabled
GtkConfigStore              : not enabled (library 'gtk+-2.0 >= 2.12' not found)
SQlite stats data output    : enabled
Network Simulation Cradle   : not enabled (--enable-nsc configure option not given)
Python Bindings             : not enabled (Python development headers not found.)
Configuration finished successfully; project is now ready to build. 

Note the trailing portion of the above output. Some ns-3 options are not enabled by default or require support from the underlying system. For instance, to enable Python bindings, Python development headers must be installed on the host machine, and they were not found in the above example, so Python scripting will not be supported in the resulting build. For this tutorial, we will focus on the non-optional pieces of ns-3.

The build system is now configured and you can build the debug versions of the ns-3 programs by simply typing,

  ./waf

(Hint: if you have a multicore machine, use the "-j JOBS" option to speed up your build, where JOBS is the number of cores) You will see many Waf status messages displayed as the system compiles. The most important is the last one:

  Compilation finished successfully

3.3 Testing ns-3

You can run the unit tests of the ns-3 distribution by running the “check” command,

  ./waf check

You should see a report from each unit test that executes indicating that the test has passed.

  ~/repos/ns-3-dev > ./waf check
  Entering directory `/home/craigdo/repos/ns-3-dev/build'
  Compilation finished successfully
  PASS AddressHelper
  PASS Wifi
  PASS DcfManager
  
  ...

  PASS Object
  PASS Ptr
  PASS Callback
  ~/repos/ns-3-dev >

This command is typically run by users to quickly verify that an ns-3 distribution has built correctly.

You can also run our regression test suite to ensure that your distribution and tool chain have produced binaries that generate output that is identical to reference output files stored in a central location. To run the regression tests, you provide Waf with the regression flag.

  ./waf --regression

Waf will verify that the current files in the ns-3 distribution are built and will then look for trace files in the aforementioned centralized location. If your tool chain includes Mercurial, the regression tests will be downloaded from a repository at code.nsnam.org. If you do not have Mercurial installed, the reference traces will be downloaded from a tarball located in the releases section of www.nsnam.org. The particular name of the reference trace location is built from the ns-3 version located in the VERSION file, so don't change that string yourself unless you know what you are doing. (Warning: The ns-3.2 release requires you to be online when you run regression tests because it synchronizes the trace directory with an online repository).

Once the reference traces are downloaded to your local machine, Waf will run a number of tests that generate what we call trace files. The content of these trace files are compared with the reference traces just downloaded. If they are identical, the regression tests report a PASS status. If the regression tests pass, you should see something like,

  ~/repos/ns-3-dev > ./waf --regression
  Entering directory `/home/craigdo/repos/ns-3-dev/build'
  Compilation finished successfully 
  ========== Running Regression Tests ==========
  Synchronizing reference traces using Mercurial.
  Pulling http://code.nsnam.org/ns-3-dev-ref-traces from repo.
  Skipping csma-bridge: Python bindings not available.
  SKIP test-csma-bridge
  PASS test-csma-broadcast
  PASS test-csma-multicast
  PASS test-csma-one-subnet
  PASS test-csma-packet-socket
  PASS test-realtime-udp-echo
  PASS test-simple-error-model
  PASS test-simple-global-routing
  PASS test-simple-point-to-point-olsr
  PASS test-tcp-large-transfer
  PASS test-udp-echo
  PASS test-wifi-wired-bridging

  ~/repos/ns-3-dev >

If a regression tests fails you will see a FAIL indication along with a pointer to the offending trace file and its associated reference trace file along with a suggestion on diff parameters and options in order to see what has gone awry. Python regression tests will be SKIPped if Python bindings are not built.

3.4 Running a Script

We typically run scripts under the control of Waf. This allows the build system to ensure that the shared library paths are set correctly and that the libraries are available at run time. To run a program, simply use the --run option in Waf. Let's run the ns-3 equivalent of the ubiquitous hello world program by typing the following:

  ./waf --run hello-simulator

Waf first checks to make sure that the program is built correctly and executes a build if required. Waf then then executes the program, which produces the following output.

  Hello Simulator

Congratulations. You are now an ns-3 user.

If you want to run programs under another tool such as gdb or valgrind, see this wiki entry.

4. Conceptual Overview

The first thing we need to do before actually starting to look at or write ns-3 code is to explain a few core concepts and abstractions in the system. Much of this may appear transparently obvious to some, but we recommend taking the time to read through this section just to ensure you are starting on a firm foundation.

4.1 Key Abstractions

In this section, we'll review some terms that are commonly used in networking, but have a specific meaning in ns-3.

4.1.1 Node

In Internet jargon, a computing device that connects to a network is called a host or sometimes an end system. Because ns-3 is a network simulator, not specifically an Internet simulator, we intentionally do not use the term host since it is closely associated with the Internet and its protocols. Instead, we use a more generic term also used by other simulators that originates in Graph Theory — the node.

In ns-3 the basic computing device abstraction is called the node. This abstraction is represented in C++ by the class Node. The Node class provides methods for managing the representations of computing devices in simulations.

You should think of a Node as a computer to which you will add functionality. One adds things like applications, protocol stacks and peripheral cards with their associated drivers to enable the computer to do useful work. We use the same basic model in ns-3.

4.1.2 Application

Typically, computer software is divided into two broad classes. System Software organizes various computer resources such as memory, processor cycles, disk, network, etc., according to some computing model. System software usually does not use those resources to complete tasks that directly benefit a user. A user would typically run an application that acquires and uses the resources controlled by the system software to accomplish some goal.

Often, the line of separation between system and application software is made at the privilege level change that happens in operating system traps. In ns-3 there is no real concept of operating system and especially no concept of privilege levels or system calls. We do, however, have the idea of an application. Just as software applications run on computers to perform tasks in the “real world,” ns-3 applications run on ns-3 Nodes to drive simulations in the simulated world.

In ns-3 the basic abstraction for a user program that generates some activity to be simulated is the application. This abstraction is represented in C++ by the class Application. The Application class provides methods for managing the representations of our version of user-level applications in simulations. Developers are expected to specialize the Application class in the object-oriented programming sense to create new applications. In this tutorial, we will use specializations of class Application called UdpEchoClientApplication and UdpEchoServerApplication. As you might expect, these applications compose a client/server application set used to generate and echo simulated network packets

4.1.3 Channel

In the real world, one can connect a computer to a network. Often the media over which data flows in these netowrks are called channels. When you connect your Ethernet cable to the plug in the wall, you are connecting your computer to an Ethernet communication channel. In the simulated world of ns-3, one connects a Node to an object representing a communication channel. Here the basic communication subnetwork abstraction is called the channel and is represented in C++ by the class Channel.

The Channel class provides methods for managing communication subnetwork objects and connecting nodes to them. Channels may also be specialized by developers in the object oriented programming sense. A Channel specialization may model something as simple as a wire. The specialized Channel can also model things as complicated as a large Ethernet switch, or three-dimensional space full of obstructions in the case of wireless networks.

We will use specialized versions of the Channel called CsmaChannel, PointToPointChannel and WifiChannel in this tutorial. The CsmaChannel, for example, models a version of a communication subnetwork that implements a carrier sense multiple access communication medium. This gives us Ethernet-like functionality.

4.1.4 Net Device

It used to be the case that if you wanted to connect a computers to a network, you had to buy a specific kind of network cable and a hardware device called (in PC terminology) a peripheral card that needed to be installed in your computer. If the peripheral card implemented some networking function, theys were called Network Interface Cards, or NICs. Today most computers come with the network interface hardware built in and users don't see these building blocks.

A NIC will not work without a software driver to control the hardware. In Unix (or Linux), a piece of peripheral hardware is classified as a device. Devices are controlled using device drivers, and network devices (NICs) are controlled using network device drivers collectively known as net devices. In Unix and Linux you refer to these net devices by names such as eth0.

In ns-3 the net device abstraction covers both the software driver and the simulated hardware. A net device is “installed” in a Node in order to enable the Node to communicate with other Nodes in the simulation via Channels. Just as in a real computer, a Node may be connected to more than one Channel via multiple NetDevices.

The net device abstraction is represented in C++ by the class NetDevice. The NetDevice class provides methods for managing connections to Node and Channel objects; and may be specialized by developers in the object-oriented programming sense. We will use the several specialized versions of the NetDevice called CsmaNetDevice, PointToPointNetDevice, and WifiNetDevice in this tutorial. Just as an Ethernet NIC is designed to work with an Ethernet network, the CsmaNetDevice is designed to work with a CsmaChannel; the PointToPointNetDevice is designed to work with a PointToPointChannel and a WifiNetNevice is designed to work with a WifiChannel.

4.1.5 Topology Helpers

In a real network, you will find host computers with added (or built-in) NICs. In ns-3 we would say that you will find Nodes with attached NetDevices. In a large simulated network you will need to arrange many connections between Nodes, NetDevices and Channels.

Since connecting NetDevices to Nodes, NetDevices to Channels, assigning IP addresses, etc., are such common tasks in ns-3, we provide what we call topology helpers to make this as easy as possible. For example, it may take many distinct ns-3 core operations to create a NetDevice, add a MAC address, install that net device on a Node, configure the node's protocol stack, and then connect the NetDevice to a Channel. Even more operations would be required to connect multiple devices onto multipoint channels and then to connect individual networks together into internetworks. We provide topology helper objects that combine those many distinct operations into an easy to use model for your convenience.

4.2 A First ns-3 Script

If you downloaded the system as was suggested above, you will have a release of ns-3 in a directory called repos under your home directory. Change into that release directory, and you should find a directory structure something like the following:

  AUTHORS  examples/  README         samples/  utils/   waf.bat*
  build/   LICENSE    regression/    scratch/  VERSION  wscript
  doc/     ns3/       RELEASE_NOTES  src/      waf*

Change into the examples directory. You should see a file named first.cc located there. This is a script that will create a simple point-to-point link between two nodes and echo a single packet between the nodes. Let's take a look at that script line by line, so go ahead and open first.cc in your favorite editor.

4.2.1 Boilerplate

The first line in the file is an emacs mode line. This tells emacs about the formatting conventions (coding style) we use in our source code.

  /* -*- Mode:C++; c-file-style:''gnu''; indent-tabs-mode:nil; -*- */

This is always a somewhat controversial subject, so we might as well get it out of the way immediately. The ns-3 project, like most large projects, has adopted a coding style to which all contributed code must adhere. If you want to contribute your code to the project, you will eventually have to conform to the ns-3 coding standard as described in the file doc/codingstd.txt or shown on the project web page here.

We recommend that you, well, just get used to the look and feel of ns-3 code and adopt this standard whenever you are working with our code. All of the development team and contributors have done so with various amounts of grumbling. The emacs mode line above makes it easier to get the formatting correct if you use the emacs editor.

The ns-3 simulator is licensed using the GNU General Public License. You will see the appropriate GNU legalese at the head of every file in the ns-3 distribution. Often you will see a copyright notice for one of the institutions involved in the ns-3 project above the GPL text and an author listed below.

  /*
   * This program is free software; you can redistribute it and/or modify
   * it under the terms of the GNU General Public License version 2 as
   * published by the Free Software Foundation;
   *
   * This program is distributed in the hope that it will be useful,
   * but WITHOUT ANY WARRANTY; without even the implied warranty of
   * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
   * GNU General Public License for more details.
   *
   * You should have received a copy of the GNU General Public License
   * along with this program; if not, write to the Free Software
   * Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA  02111-1307 USA
   */

4.2.2 Module Includes

The code proper starts with a number of include statements.

  #include "ns3/core-module.h"
  #include "ns3/simulator-module.h"
  #include "ns3/node-module.h"
  #include "ns3/helper-module.h"

To help our high-level script users deal with the large number of include files present in the system, we group includes according to relatively large modules. We provide a single include file that will recursively load all of the include files used in each module. Rather than having to look up exactly what header you need, and possibly have to get a number of dependencies right, we give you the ability to load a group of files at a large granularity. This is not the most efficient approach but it certainly makes writing scripts much easier.

Each of the ns-3 include files is placed in a directory called ns3 (under the build directory) during the build process to help avoid include file name collisions. The ns3/core-module.h file corresponds to the ns-3 module you will find in the directory src/core in your downloaded release distribution. If you list this directory you will find a large number of header files. When you do a build, Waf will place public header files in an ns3 directory under the appropriate build/debug or build/optimized directory depending on your configuration. Waf will also automatically generate a module include file to load all of the public header files.

Since you are, of course, following this tutorial religiously, you will already have done a

  ./waf -d debug configure

in order to configure the project to perform debug builds. You will also have done a

  ./waf

to build the project. So now if you look in the directory build/debug/ns-3 you will find the four module include files shown above. You can take a look at the contents of these files and find that they do include all of the public include files in their respective modules.

4.2.3 Ns3 Namespace

The next line in the first.cc script is a namespace declaration.

  using namespace ns3;

The ns-3 project is implemented in a C++ namespace called ns3. This groups all ns-3-related declarations in a scope outside the global namespace, which we hope will help with integration with other code. The C++ using statement introduces the ns-3 namespace into the current (global) declarative region. This is a fancy way of saying that after this declaration, you will not have to type ns3:: scope resolution operator before all of the ns-3 code in order to use it. If you are unfamiliar with namespaces, please consult almost any C++ tutorial and compare the ns3 namespace and usage here with instances of the std namespace and the using namespace std; statements you will often find in discussions of cout and streams.

4.2.4 Logging

The next line of the script is the following,

  NS_LOG_COMPONENT_DEFINE ("FirstScriptExample");

We will use this statement as a convenient place to talk about our Doxygen documentation system. If you look at the project web site, ns-3 project, you will find a link to “APIs (Doxygen)” in the navigation bar. If you select this link, you will be taken to our documentation page.

Along the left side, you will find a graphical representation of the structure of the documentation. A good place to start is the NS-3 Modules “book.” If you expand Modules you will see a list of ns-3 module documentation. The concept of module here ties directly into the module include files discussed above. It turns out that the ns-3 logging subsystem is part of the core module, so go ahead and expand that documentation node. Now, expand the Debugging book and then select the Logging page.

You should now be looking at the Doxygen documentation for the Logging module. In the list of #defines at the top of the page you will see the entry for NS_LOG_COMPONENT_DEFINE. Before jumping in, it would probably be good to look for the “Detailed Description” of the logging module to get a feel for the overall operation. You can either scroll down or select the “More...” link under the collaboration diagram to do this.

Once you have a general idea of what is going on, go ahead and take a look at the specific NS_LOG_COMPONENT_DEFINE documentation. I won't duplicate the documentation here, but to summarize, this line declares a logging component called FirstScriptExample that allows you to enable and disable console message logging by reference to the name.

4.2.5 Main Function

The next lines of the script you will find are,

    int
  main (int argc, char *argv[])
  {

This is just the declaration of the main function of your program (script). Just as in any C++ program, you need to define a main function that will be the first function run. There is nothing at all special here. Your ns-3 script is just a C++ program.

The next two lines of the script are used to enable two logging components that are built into the Echo Client and Echo Server applications:

    LogComponentEnable("UdpEchoClientApplication", LOG_LEVEL_INFO);
    LogComponentEnable("UdpEchoServerApplication", LOG_LEVEL_INFO);

If you have read over the Logging component documentation you will have seen that there are a number of levels of logging verbosity/detail that you can enable on each component. These two lines of code enable debug logging at the INFO level for echo clients and servers. This will result in the application printing out messages as packets are sent and received during the simulation.

Now we will get directly to the business of creating a topology and running a simulation. We use the topology helper objects to make this job as easy as possible.

4.2.6 Topology Helpers

4.2.6.1 NodeContainer

The next two lines of code in our script will actually create the ns-3 Node objects that will represent the computers in the simulation.

    NodeContainer nodes;
    nodes.Create (2);

Let's find the documentation for the NodeContainer class before we continue. Another way to get into the documentation for a given class is via the Classes tab in the Doxygen pages. If you still have the Doxygen handy, just scroll up to the top of the page and select the Classes tab. You should see a new set of tabs appear, one of which is Class List. Under that tab you will see a list of all of the ns-3 classes. Scroll down, looking for ns3::NodeContainer. When you find the class, go ahead and select it to go to the documentation for the class.

You may recall that one of our key abstractions is the Node. This represents a computer to which we are going to add things like protocol stacks, applications and peripheral cards. The NodeContainer topology helper provides a convenient way to create, manage and access any Node objects that we create in order to run a simulation. The first line above just declares a NodeContainer which we call nodes. The second line calls the Create method on the nodes object and asks the container to create two nodes. As described in the Doxygen, the container calls down into the ns-3 system proper to create two Node objects and stores pointers to those objects internally.

The nodes as they stand in the script do nothing. The next step in constructing a topology is to connect our nodes together into a network. The simplest form of network we support is a single point-to-point link between two nodes. We'll construct one of those links here.

4.2.6.2 PointToPointHelper

We are constructing a point to point link, and, in a pattern which will become quite familiar to you, we use a topology helper object to do the low-level work required to put the link together. Recall that two of our key abstractions are the NetDevice and the Channel. In the real world, these terms correspond roughly to peripheral cards and network cables. Typically these two things are intimately tied together and one cannot expect to interchange, for example, Ethernet devices and wireless channels. Our Topology Helpers follow this intimate coupling and therefore you will use a single PointToPointHelper to configure and connect ns-3 PointToPointNetDevice and PointToPointChannel objects in this script.

The next three lines in the script are,

    PointToPointHelper pointToPoint;
    pointToPoint.SetDeviceAttribute ("DataRate", StringValue ("5Mbps"));
    pointToPoint.SetChannelAttribute ("Delay", StringValue ("2ms"));

The first line

    PointToPointHelper pointToPoint;

instantiates a PointToPointHelper object on the stack. From a high-level perspective the next line,

    pointToPoint.SetDeviceAttribute ("DataRate", StringValue ("5Mbps"));

tells the PointToPointHelper object to use the value “5mbps” (five megabits per second) as the “DataRate” when it creates a PointToPointNetDevice object.

From a more detailed perspective, the string “DataRate” corresponds to what we call an Attribute of the PointToPointNetDevice. If you look at the Doxygen for class ns3::PointToPointNetDevice and find the documentation for the GetTypeId method, you will find a list of Attributes defined for the device. Among these is the “DataRate” attribute. Most user-visible ns-3 objects have similar lists of attributes. We use this mechanism to easily configure simulations without recompiling as you will see in a following section.

Similar to the “DataRate” on the PointToPointNetDevice you will find a “Delay” attribute associated with the PointToPointChannel. The final line,

    pointToPoint.SetChannelAttribute ("Delay", StringValue ("2ms"));

tells the PointToPointHelper to use the value “2ms” (two milliseconds) as the value of the transmission delay of every point to point channel it subsequently creates.

4.2.6.3 NetDeviceContainer

At this point in the script, we have a NodeContainer that contains two nodes. We have a PointToPointHelper that is primed and ready to make PointToPointNetDevices and wire PointToPointChannel objects between them. Just as we used the NodeContainer topology helper object to create the Nodes for our simulation, we will ask the PointToPointHelper to do the work involved in creating, configuring and installing our devices for us. We will need to have a list of all of the NetDevice objects that are created, so we use a NetDeviceContainer to hold them just as we used a NodeContainer to hold the nodes we created. The following two lines of code,

    NetDeviceContainer devices;
    devices = pointToPoint.Install (nodes);

will finish configuring the devices and channel. The first line declares the device container mentioned above and the second does the heavy lifting. The Install method of the PointToPointHelper takes a NodeContainer as a parameter. Internally, a NetDeviceContainer is created. For each node in the NodeContainer (there must be exactly two for a point-to-point link) a PointToPointNetDevice is created and saved in the device container. A PointToPointChannel is created and the two PointToPointNetDevices are attached. When objects are created by the PointToPointHelper, the attributes previously set in the helper are used to initialize the corresponding attributes in the created objects.

After executing the the pointToPoint.Install (nodes) call we will have two nodes, each with an installed point-to-point net device and a point-to-point channel between them. Both devices will be configured to transmit data at five megabits per second over the channel which has a two millisecond transmission delay.

4.2.6.4 InternetStackHelper

We now have nodes and devices configured, but we don't have any protocol stacks installed on our nodes. The next two lines of code will take care of that.

    InternetStackHelper stack;
    stack.Install (nodes);

The InternetStackHelper is a topology helper that is to internet stacks what the PointToPointHelper is to point-to-point net devices. The Install method takes a NodeContainer as a parameter. When it is executed, it will install an Internet Stack (TCP, UDP, IP, etc.) on each of the nodes in the node container.

4.2.6.5 Ipv4AddressHelper

Next we need to associate the devices on our nodes with IP addresses. We provide a topology helper to manage the allocation of IP addresses. The only user-visible API is to set the base IP address and network mask to use when performing the actual address allocation (which is done at a lower level inside the helper).

The next two lines of code in our example script, first.cc,

    Ipv4AddressHelper address;
    address.SetBase ("10.1.1.0", "255.255.255.0");

declare an address helper object and tell it that it should begin allocating IP addresses from the network 10.1.1.0 using the mask 255.255.255.0 to define the allocatable bits. By default the addresses allocated will start at one and increase monotonically, so the first address allocated from this base will be 10.1.1.1, followed by 10.1.1.2, etc. The low level ns-3 system actually remembers all of the IP addresses allocated and will generate a fatal error if you accidentally cause the same address to be generated twice (which is a very hard to debug error, by the way).

The next line of code,

    Ipv4InterfaceContainer interfaces = address.Assign (devices);

performs the actual address assignment. In ns-3 we make the association between an IP address and a device using an Ipv4Interface object. Just as we sometimes need a list of net devices created by a helper for future reference we sometimes need a list of Ipv4Interface objects. The Ipv4InterfaceContainer provides this functionality.

Now we have a point-to-point network built, with stacks installed and IP addresses assigned. What we need at this point are applications to generate traffic.

4.2.7 Applications

Another one of the core abstractions of the ns-3 system is the Application. In this script we use two specializations of the core ns-3 class Application called UdpEchoServerApplication and UdpEchoClientApplication. Just as we have in our previous explanations, we use helper objects to help configure and manage the underlying objects. Here, we use UdpEchoServerHelper and UdpEchoClientHelper objects to make our lives easier.

4.2.7.1 UdpEchoServerHelper

The following lines of code in our example script, first.cc, are used to set up a UDP echo server application on one of the nodes we have previously created.

    UdpEchoServerHelper echoServer (9);

    ApplicationContainer serverApps = echoServer.Install (nodes.Get (1));
    serverApps.Start (Seconds (1.0));
    serverApps.Stop (Seconds (10.0));

The first line of code in the above snippet declares the UdpEchoServerHelper. As usual, this isn't the application itself, it is an object used to help us create the actual applications. One of our conventions is to place required attributes in the helper constructor. In this case, the helper can't do anything useful unless it is provided with a port number that the client also knows about. Rather than just picking one and hoping it all works out, we require the port number as a parameter to the constructor. The constructor, in turn, simply does a SetAttribute with the passed value. You can, if desired, set the “Port” attribute to another value later.

Similar to many other helper objects, the UdpEchoServerHelper object has an Install method. It is the execution of this method that actually causes the underlying echo server application to be instantiated and attached to a node. Interestingly, the Install method takes a NodeContainter as a parameter just as the other Install methods we have seen. This is actually what is passed to the method even though it doesn't look so in this case. There is a C++ implicit conversion at work here.

We now see that echoServer.Install is going to install a UdpEchoServerApplication on the node found at index number one of the NodeContainer we used to manage our nodes. Install will return a container that holds pointers to all of the applications (one in this case since we passed a NodeContainer containing one node) created by the helper.

Applications require a time to “start” generating traffic and may take an optional time to “stop.” We provide both. These times are set using the ApplicationContainer methods Start and Stop. These methods take Time parameters. In this case, we use an explicit C++ conversion sequence to take the C++ double 1.0 and convert it to an ns-3 Time object using a Seconds cast. The two lines,

    serverApps.Start (Seconds (1.0));
    serverApps.Stop (Seconds (10.0));

will cause the echo server application to Start (enable itself) at one second into the simulation and to Stop (disable itself) at ten seconds into the simulation. By virtue of the fact that we have implicilty declared a simulation event (the application stop event) to be executed at ten seconds, the simulation will last at least ten seconds.

4.2.7.2 UdpEchoClientHelper

The echo client application is set up in a method substantially similar to that for the server. There is an underlying UdpEchoClientApplication that is managed by an UdpEchoClientHelper.

    UdpEchoClientHelper echoClient (interfaces.GetAddress (1), 9);
    echoClient.SetAttribute ("MaxPackets", UintegerValue (1));
    echoClient.SetAttribute ("Interval", TimeValue (Seconds (1.)));
    echoClient.SetAttribute ("PacketSize", UintegerValue (1024));

    ApplicationContainer clientApps = echoClient.Install (nodes.Get (0));
    clientApps.Start (Seconds (2.0));
    clientApps.Stop (Seconds (10.0));

For the echo client, however, we need to set five different attributes. The first two attributes are set during construction of the UdpEchoClientHelper. We pass parameters that are used (internally to the helper) to set the “RemoteAddress” and “RemotePort” attributes in accordance with our convention to make required attributes parameters in the helper constructors.

Recall that we used an Ipv4InterfaceContainer to keep track of the IP addresses we assigned to our devices. The zeroth interface in the interfaces container is going to coorespond to the IP address of the zeroth node in the nodes container. The first interface in the interfaces container cooresponds to the IP address of the first node in the nodes container. So, in the first line of code (from above), we are creating the helper and telling it so set the remote address of the client to be the IP address assigned to the node on which the server resides. We also tell it to arrange to send packets to port nine.

The “MaxPackets” attribute tells the client the maximum number of packets we allow it to send during the simulation. The “Interval” attribute tells the client how long to wait between packets, and the “PacketSize” attribute tells the client how large its packet payloads should be. With this particular combination of attributes, we are telling the client to send one 1024-byte packet.

Just as in the case of the echo server, we tell the echo client to Start and Stop, but here we start the client one second after the server is enabled (at two seconds into the simulation).

4.2.8 Simulator

What we need to do at this point is to actually run the simulation. This is done using the global function Simulator::Run.

    Simulator::Run ();

When we previously called the methods,

    serverApps.Start (Seconds (1.0));
    serverApps.Stop (Seconds (10.0));
    ...
    clientApps.Start (Seconds (2.0));
    clientApps.Stop (Seconds (10.0));

we actually scheduled events in the simulator at 1.0 seconds, 2.0 seconds and 10.0 seconds. When Simulator::Run is called, the system will begin looking through the list of scheduled events and executing them. First it will run the event at 1.0 seconds, which will enable the echo server application. Then it will run the event scheduled for t=2.0 seconds which will start the echo client application. The start event implementation in the echo client application will begin the data transfer phase of the simulation by sending a packet to the server.

The act of sending the packet to the server will trigger a chain of events that will be automatically scheduled behind the scenes and which will perform the mechanics of the packet echo according to the various timing parameters that we have set in the script.

Eventually, since we only send one packet, the chain of events triggered by that single client echo request will taper off and the simulation will go idle. Once this happens, the remaining events will be the Stop events for the server and the client. When these events are executed, there are no further events to process and Simulator::Run returns. The simulation is complete.

All that remains is to clean up. This is done by calling the global function Simulator::Destroy. As the helper functions (or low level ns-3 code) executed, they arranged it so that hooks were inserted in the simulator to destroy all of the objects that were created. You did not have to keep track of any of these objects yourself — all you had to do was to call Simulator::Destroy and exit. The ns-3 system took care of the hard part for you. The remaining lines of our first ns-3 script, first.cc, do just that:

    Simulator::Destroy ();
    return 0;
  }

4.2.9 Building Your Script

We have made it trivial to build your simple scripts. All you have to do is to drop your script into the scratch directory and it will automatically be built if you run Waf. Let's try it. Copy examples/first.cc into the scratch directory.

  ~/repos/ns-3-dev > cp examples/first.cc scratch/

and then build it using waf,

  ~/repos/ns-3-dev > ./waf
  Entering directory `/home/craigdo/repos/ns-3-dev/build'
  [467/511] cxx: scratch/first.cc -> build/debug/scratch/first_1.o
  [468/511] cxx: scratch/multiple-sources/simple-main.cc -> build/debug/scratch/multiple-sources/simple-main_2.o
  [469/511] cxx: scratch/multiple-sources/simple-simulation.cc -> build/debug/scratch/multiple-sources/simple-simulation_2.o
  [470/511] cxx: scratch/simple.cc -> build/debug/scratch/simple_3.o
  [508/511] cxx_link: build/debug/scratch/first_1.o -> build/debug/scratch/first
  Compilation finished successfully 
  ~/repos/ns-3-dev >

You can now run the example (note that if you build your program in the scratch directory you must run it out of the scratch direcory):

  ~/repos/ns-3-dev > ./waf --run scratch/first
  Entering directory `/home/craigdo/repos/ns-3-dev/build'
  Compilation finished successfully
  Sent 1024 bytes to 10.1.1.2
  Received 1024 bytes from 10.1.1.1
  Received 1024 bytes from 10.1.1.2
  ~/repos/ns-3-dev >

Here you see that the build system checks to make sure that the file has been build and then runs it. You see the logging component on the echo client indicate that it has sent one 1024 byte packet to the Echo Server on 10.1.1.2. You also see the logging component on the echo server say that it has received the 1024 bytes from 10.1.1.1. The echo server silently echoes the packet and you see the echo client log that it has received its packet back from the server.

4.3 Ns-3 Source Code

Now that you have used some of the ns-3 helpers you may want to have a look at some of the source code that implements that functionality. The most recent code can be browsed on our web server at the following link: http://code.nsnam.org/?sort=lastchange. If you click on the bold repository names on the left of the page, you will see changelogs for these repositories, and links to the manifest. From the manifest links, one can browse the source tree.

The top-level directory for one of our repositories will look something like:

drwxr-xr-x  [up]   
drwxr-xr-x         doc             manifest 
drwxr-xr-x         examples        manifest 
drwxr-xr-x         ns3             manifest 
drwxr-xr-x         regression      manifest 
drwxr-xr-x         samples         manifest 
drwxr-xr-x         scratch         manifest 
drwxr-xr-x         src             manifest 
drwxr-xr-x         tutorial        manifest 
drwxr-xr-x         utils           manifest 
-rw-r--r-- 135     .hgignore       file | revisions | annotate 
-rw-r--r-- 891     .hgtags         file | revisions | annotate 
-rw-r--r-- 441     AUTHORS         file | revisions | annotate 
-rw-r--r-- 17987   LICENSE         file | revisions | annotate 
-rw-r--r-- 4948    README          file | revisions | annotate 
-rw-r--r-- 4917    RELEASE_NOTES   file | revisions | annotate 
-rw-r--r-- 7       VERSION         file | revisions | annotate 
-rwxr-xr-x 99143   waf             file | revisions | annotate 
-rwxr-xr-x 28      waf.bat         file | revisions | annotate 
-rw-r--r-- 30584   wscript         file | revisions | annotate 

The source code is mainly in the src directory. You can view source code by clicking on the manifest link to the right of the directory name. If you click on the manifest link to the right of the src directory you will find a subdirectory. If you click on the manifest link next to the core subdirectory in under src, you will find a list of files. The first file you will find is assert.h. If you click on the file link, you will be sent to the source file for assert.h.

Our example scripts are in the examples directory. The source code for the helpers we have used in this chapter can be found in the src/helpers directory.

5. Tweaking ns-3