.. include:: replace.txt

Conceptual Overview

-------------------

The first thing we need to do before actually starting to look at or write

|ns3| 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.

Key Abstractions

****************

In this section, we'll review some terms that are commonly used in

networking, but have a specific meaning in |ns3|.

Node

++++

In Internet jargon, a computing device that connects to a network is called

a *host* or sometimes an *end system*.  Because |ns3| 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 |ns3| 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 |ns3|.

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 |ns3| 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," |ns3| applications run on

|ns3| ``Nodes`` to drive simulations in the simulated world.

In |ns3| 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 

Channel

+++++++

In the real world, one can connect a computer to a network.  Often the media

over which data flows in these networks 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 |ns3|, 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.  

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,

they 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 |ns3| 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``.

Topology Helpers

++++++++++++++++

In a real network, you will find host computers with added (or built-in)

NICs.  In |ns3| 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 |ns3|, we provide what we call *topology helpers* to make 

this as easy as possible.  For example, it may take many distinct 

|ns3| 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.

A First ns-3 Script

*******************

If you downloaded the system as was suggested above, you will have a release

of |ns3| 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       doc/       README         src/     waf.bat*

  bindings/     examples/  RELEASE_NOTES  utils/   wscript

  build/        LICENSE    samples/       VERSION  wutils.py

  CHANGES.html  ns3/       scratch/       waf*     wutils.pyc

Change into the ``examples/tutorial`` 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.

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 |ns3| 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 |ns3| coding standard as described 

in the file ``doc/codingstd.txt`` or shown on the project web page

`here

<http://www.nsnam.org/codingstyle.html>`_.

We recommend that you, well, just get used to the look and feel of |ns3|

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 |ns3| simulator is licensed using the GNU General Public 

License.  You will see the appropriate GNU legalese at the head of every file 

in the |ns3| distribution.  Often you will see a copyright notice for

one of the institutions involved in the |ns3| 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

   */

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 |ns3| 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/ns3`` 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.

Ns3 Namespace

+++++++++++++

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

::

  using namespace ns3;

The |ns3| project is implemented in a C++ namespace called 

``ns3``.  This groups all |ns3|-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 |ns3|

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 |ns3| 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.

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

<http://www.nsnam.org>`_, you will find a link to "Doxygen 

(ns-3-dev)" in the navigation bar.  If you select this link, you will be

taken to our documentation page for the current development release.  There 

is also a link to "Doxygen (stable)" that will take you to the documentation

for the latest stable release of |ns3|.

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" in the |ns3| navigation tree.  If you expand ``Modules`` 

you will see a list of |ns3| module documentation.  The concept of 

module here ties directly into the module include files discussed above.  It 

turns out that the |ns3| 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 ``#define``s 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.

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 

|ns3| 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.

Topology Helpers

++++++++++++++++

NodeContainer

~~~~~~~~~~~~~

The next two lines of code in our script will actually create the 

|ns3| ``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 

|ns3| 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 |ns3| 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.

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 |ns3|

``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 |ns3| 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.

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 ``pointToPoint.Install (nodes)`` call we will have

two nodes, each with an installed point-to-point net device and a single

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.

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.

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 |ns3| system

actually remembers all of the IP addresses allocated and will generate a