unison/doc/tutorial/conceptual-overview.texi

@c ========================================================================
@c Begin document body here
@c ========================================================================

@c ========================================================================
@c Conceptual Overview
@c ========================================================================
@node Conceptual Overview
@chapter Conceptual Overview

@menu
* Key Abstractions::
* A First ns-3 Script::
@end menu

The first thing we need to do before actually starting to look at or write
@command{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.

@node Key Abstractions
@section Key Abstractions

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

@subsection Node
@cindex Node
In Internet jargon, a computing device that connects to a network is called
a @emph{host} or sometimes an @emph{end system}.  Because @command{ns-3} is a 
@emph{network} simulator, not specifically an @emph{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 @emph{node}.

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

You should think of a @code{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 @command{ns-3}.

@subsection Application
@cindex Application
Typically, computer software is divided into two broad classes.  @emph{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 @emph{application} that acquires
and uses the resources controlled by the system software to accomplish some
goal.  

@cindex system call
Often, the line of separation between system and application software is made
at the privilege level change that happens in operating system traps.
In @command{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,'' @command{ns-3} applications run on
@command{ns-3} @code{Nodes} to drive simulations in the simulated world.

@cindex class Application
In @command{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 @code{Application}.  The @code{Application} class provides 
methods for managing the representations of our version of user-level 
applications in simulations.  Developers are expected to specialize the
@code{Application} class in the object-oriented programming sense to create new
applications.  In this tutorial, we will use specializations of class 
@code{Application} called @code{UdpEchoClientApplication} and 
@code{UdpEchoServerApplication}.  As you might expect, these applications 
compose a client/server application set used to generate and echo simulated 
network packets 

@subsection Channel
@cindex 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 @emph{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 @command{ns-3}, one connects a @code{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 @code{Channel}.  

The @code{Channel} class provides methods for managing communication 
subnetwork objects and connecting nodes to them.  @code{Channels} may also be
specialized by developers in the object oriented programming sense.  A 
@code{Channel} specialization may model something as simple as a wire.  The 
specialized  @code{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 @code{Channel} called
@code{CsmaChannel}, @code{PointToPointChannel} and @code{WifiChannel} in this
tutorial.  The @code{CsmaChannel}, for example, models a version of a 
communication subnetwork that implements a @emph{carrier sense multiple 
access} communication medium.  This gives us Ethernet-like functionality.  

@subsection Net Device
@cindex NetDevice
@cindex Ethernet
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 @emph{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 @emph{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 
@emph{device}.  Devices are controlled using @emph{device drivers}, and network
devices (NICs) are controlled using @emph{network device drivers}
collectively known as @emph{net devices}.  In Unix and Linux you refer
to these net devices by names such as @emph{eth0}.

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

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

@subsection Topology Helpers
@cindex helper
@cindex topology
@cindex topology helper
In a real network, you will find host computers with added (or built-in)
NICs.  In @command{ns-3} we would say that you will find @code{Nodes} with 
attached @code{NetDevices}.  In a large simulated network you will need to 
arrange many connections between @code{Nodes}, @code{NetDevices} and 
@code{Channels}.

Since connecting @code{NetDevices} to @code{Nodes}, @code{NetDevices}
to @code{Channels}, assigning IP addresses,  etc., are such common tasks
in @command{ns-3}, we provide what we call @emph{topology helpers} to make 
this as easy as possible.  For example, it may take many distinct 
@command{ns-3} core operations to create a NetDevice, add a MAC address, 
install that net device on a @code{Node}, configure the node's protocol stack,
and then connect the @code{NetDevice} to a @code{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.

@c ========================================================================
@c A First ns-3 script
@c ========================================================================
@node A First ns-3 Script
@section A First ns-3 Script
@cindex first script
If you downloaded the system as was suggested above, you will have a release
of @command{ns-3} in a directory called @code{repos} under your home 
directory.  Change into that release directory, and you should find a 
directory structure something like the following:

@verbatim
  AUTHORS       doc/       README          RELEASE_NOTES  utils/    wscript
  bindings/     examples/  regression/     samples/       VERSION   wutils.py
  build/        LICENSE    regression.py   scratch/       waf*      wutils.pyc
  CHANGES.html  ns3/       regression.pyc  src/           waf.bat*
@end verbatim

@cindex first.cc
Change into the @code{examples/tutorial} directory.  You should see a file named 
@code{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
@code{first.cc} in your favorite editor.

@subsection 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.  

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

This is always a somewhat controversial subject, so we might as well get it
out of the way immediately.  The @code{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 @command{ns-3} coding standard as described 
in the file @code{doc/codingstd.txt} or shown on the project web page
@uref{http://www.nsnam.org/codingstyle.html,,here}.

We recommend that you, well, just get used to the look and feel of @code{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 @command{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 @command{ns-3} distribution.  Often you will see a copyright notice for
one of the institutions involved in the @code{ns-3} project above the GPL
text and an author listed below.

@verbatim
  /*
   * 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
   */
@end verbatim

@subsection Module Includes
The code proper starts with a number of include statements.  

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

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 @command{ns-3} include files is placed in a directory called 
@code{ns3} (under the build directory) during the build process to help avoid
include file name collisions.  The @code{ns3/core-module.h} file corresponds 
to the ns-3 module you will find in the directory @code{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 @code{ns3} directory under the appropriate 
@code{build/debug} or @code{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

@verbatim
  ./waf -d debug configure
@end verbatim

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

@verbatim
  ./waf
@end verbatim

to build the project.  So now if you look in the directory 
@code{../../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.

@subsection Ns3 Namespace
The next line in the @code{first.cc} script is a namespace declaration.

@verbatim
  using namespace ns3;
@end verbatim

The @command{ns-3} project is implemented in a C++ namespace called 
@code{ns3}.  This groups all @command{ns-3}-related declarations in a scope
outside the global namespace, which we hope will help with integration with 
other code.  The C++ @code{using} statement introduces the @code{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 @code{ns3::}
scope resolution operator before all of the @code{ns-3} code in order to use
it.  If you are unfamiliar with namespaces, please consult almost any C++ 
tutorial and compare the @code{ns3} namespace and usage here with instances of
the @code{std} namespace and the @code{using namespace std;} statements you 
will often find in discussions of @code{cout} and streams.

@subsection Logging
The next line of the script is the following,

@verbatim
  NS_LOG_COMPONENT_DEFINE ("FirstScriptExample");
@end verbatim

We will use this statement as a convenient place to talk about our Doxygen
documentation system.  If you look at the project web site, 
@uref{http://www.nsnam.org,,ns-3 project}, 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 @code{ns-3}.

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

You should now be looking at the Doxygen documentation for the Logging module.
In the list of @code{#define}s at the top of the page you will see the entry
for @code{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 @code{NS_LOG_COMPONENT_DEFINE} documentation.  I won't duplicate
the documentation here, but to summarize, this line declares a logging 
component called @code{FirstScriptExample} that allows you to enable and 
disable console message logging by reference to the name.

@subsection Main Function
The next lines of the script you will find are,

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

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 
@command{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:

@verbatim
    LogComponentEnable("UdpEchoClientApplication", LOG_LEVEL_INFO);
    LogComponentEnable("UdpEchoServerApplication", LOG_LEVEL_INFO);
@end verbatim

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.

@subsection Topology Helpers
@subsubsection NodeContainer
The next two lines of code in our script will actually create the 
@command{ns-3} @code{Node} objects that will represent the computers in the 
simulation.  

@verbatim
    NodeContainer nodes;
    nodes.Create (2);
@end verbatim

Let's find the documentation for the @code{NodeContainer} class before we
continue.  Another way to get into the documentation for a given class is via 
the @code{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 @code{Classes} 
tab.  You should see a new set of tabs appear, one of which is 
@code{Class List}.  Under that tab you will see a list of all of the 
@command{ns-3} classes.  Scroll down, looking for @code{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 @code{Node}.  This
represents a computer to which we are going to add things like protocol stacks,
applications and peripheral cards.  The @code{NodeContainer} topology helper
provides a convenient way to create, manage and access any @code{Node} objects
that we create in order to run a simulation.  The first line above just 
declares a NodeContainer which we call @code{nodes}.  The second line calls the
@code{Create} method on the @code{nodes} object and asks the container to 
create two nodes.  As described in the Doxygen, the container calls down into
the @command{ns-3} system proper to create two @code{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.

@subsubsection 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 @code{NetDevice} and the @code{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 @code{PointToPointHelper} to configure and connect @command{ns-3}
@code{PointToPointNetDevice} and @code{PointToPointChannel} objects in this 
script.

The next three lines in the script are,

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

The first line,

@verbatim
    PointToPointHelper pointToPoint;
@end verbatim

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

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

tells the @code{PointToPointHelper} object to use the value ``5Mbps''
(five megabits per second) as the ``DataRate'' when it creates a 
@code{PointToPointNetDevice} object.

From a more detailed perspective, the string ``DataRate'' corresponds
to what we call an @code{Attribute} of the @code{PointToPointNetDevice}.
If you look at the Doxygen for class @code{ns3::PointToPointNetDevice} and 
find the documentation for the @code{GetTypeId} method, you will find a list
of  @code{Attributes} defined for the device.  Among these is the ``DataRate''
@code{Attribute}.  Most user-visible @command{ns-3} objects have similar lists of 
@code{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 @code{PointToPointNetDevice} you will find a
``Delay'' @code{Attribute} associated with the @code{PointToPointChannel}.  The 
final line,

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

tells the @code{PointToPointHelper} to use the value ``2ms'' (two milliseconds)
as the value of the transmission delay of every point to point channel it 
subsequently creates.

@subsubsection NetDeviceContainer
At this point in the script, we have a @code{NodeContainer} that contains
two nodes.  We have a @code{PointToPointHelper} that is primed and ready to 
make @code{PointToPointNetDevices} and wire @code{PointToPointChannel} objects
between them.  Just as we used the @code{NodeContainer} topology helper object
to create the @code{Nodes} for our simulation, we will ask the 
@code{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,

@verbatim
    NetDeviceContainer devices;
    devices = pointToPoint.Install (nodes);
@end verbatim

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

After executing the @code{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.

@subsubsection 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.

@verbatim
    InternetStackHelper stack;
    stack.Install (nodes);
@end verbatim

The @code{InternetStackHelper} is a topology helper that is to internet stacks
what the @code{PointToPointHelper} is to point-to-point net devices.  The
@code{Install} method takes a @code{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.

@subsubsection 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, @code{first.cc},

@verbatim
    Ipv4AddressHelper address;
    address.SetBase ("10.1.1.0", "255.255.255.0");
@end verbatim

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 @command{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,

@verbatim
    Ipv4InterfaceContainer interfaces = address.Assign (devices);
@end verbatim

performs the actual address assignment.  In @command{ns-3} we make the
association between an IP address and a device using an @code{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 @code{Ipv4Interface} objects.
The @code{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.

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

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

@verbatim
    UdpEchoServerHelper echoServer (9);

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

The first line of code in the above snippet declares the 
@code{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 @emph{required} @code{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 @code{SetAttribute}
with the passed value.  If you want, you can set the ``Port'' @code{Attribute}
to another value later using @code{SetAttribute}.

Similar to many other helper objects, the @code{UdpEchoServerHelper} object 
has an @code{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 @code{Install} method takes a
@code{NodeContainter} as a parameter just as the other @code{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++ @emph{implicit conversion} at
work here that takes the result of @code{nodes.Get (1)} (which returns a smart
pointer to a node object --- @code{Ptr<Node>}) and uses that in a constructor
for an unnamed @code{NodeContainer} that is then passed to @code{Install}.
If you are ever at a loss to find a particular method signature in C++ code
that compiles and runs just fine, look for these kinds of implicit conversions.  

We now see that @code{echoServer.Install} is going to install a
@code{UdpEchoServerApplication} on the node found at index number one of the
@code{NodeContainer} we used to manage our nodes.  @code{Install} will return
a container that holds pointers to all of the applications (one in this case 
since we passed a @code{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
@code{ApplicationContainer} methods @code{Start} and @code{Stop}.  These 
methods take @code{Time} parameters.  In this case, we use an @emph{explicit}
C++ conversion sequence to take the C++ double 1.0 and convert it to an 
@command{ns-3} @code{Time} object using a @code{Seconds} cast.  Be aware that
the conversion rules may be controlled by the model author, and C++ has its
own rules, so you can't always just assume that parameters will be happily 
converted for you.  The two lines,

@verbatim
    serverApps.Start (Seconds (1.0));
    serverApps.Stop (Seconds (10.0));
@end verbatim

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

@subsubsection UdpEchoClientHelper

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

@verbatim
    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));
@end verbatim

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

Recall that we used an @code{Ipv4InterfaceContainer} to keep track of the IP 
addresses we assigned to our devices.  The zeroth interface in the 
@code{interfaces} container is going to correspond to the IP address of the 
zeroth node in the @code{nodes} container.  The first interface in the 
@code{interfaces} container corresponds to the IP address of the first node 
in the @code{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'' @code{Attribute} tells the client the maximum number of 
packets we allow it to send during the simulation.  The ``Interval'' 
@code{Attribute} tells the client how long to wait between packets, and the
``PacketSize'' @code{Attribute} tells the client how large its packet payloads
should be.  With this particular combination of @code{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 @code{Start}
and @code{Stop}, but here we start the client one second after the server is
enabled (at two seconds into the simulation).

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

@verbatim
    Simulator::Run ();
@end verbatim

When we previously called the methods,

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

we actually scheduled events in the simulator at 1.0 seconds, 2.0 seconds and
two events at 10.0 seconds.  When @code{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 (this event may, in turn, schedule many other events).  Then it 
will run the event scheduled for t=2.0 seconds which will start the echo client
application.  Again, this event may schedule many more events.  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 (recall the @code{MaxPackets} 
@code{Attribute} was set to one), 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 @code{Stop} events
for the server and the client.  When these events are executed, there are
no further events to process and @code{Simulator::Run} returns.  The simulation
is then complete.

All that remains is to clean up.  This is done by calling the global function 
@code{Simulator::Destroy}.  As the helper functions (or low level 
@command{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 @code{Simulator::Destroy} and exit.  The @command{ns-3} system
took care of the hard part for you.  The remaining lines of our first 
@command{ns-3} script, @code{first.cc}, do just that:

@verbatim
    Simulator::Destroy ();
    return 0;
  }
@end verbatim

@subsection 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 @code{examples/tutorial/first.cc} into 
the @code{scratch} directory after changing back into the top level directory.

@verbatim
  cd ..
  cp examples/tutorial/first.cc scratch/myfirst.cc
@end verbatim

Now build your first example script using waf:

@verbatim
  ./waf
@end verbatim

You should see messages reporting that your @code{myfirst} example was built
successfully.

@verbatim
  Waf: Entering directory `/home/craigdo/repos/ns-3-allinone/ns-3-dev/build'
  [614/708] cxx: scratch/myfirst.cc -> build/debug/scratch/myfirst_3.o
  [706/708] cxx_link: build/debug/scratch/myfirst_3.o -> build/debug/scratch/myfirst
  Waf: Leaving directory `/home/craigdo/repos/ns-3-allinone/ns-3-dev/build'
  'build' finished successfully (2.357s)
@end verbatim

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 directory):

@verbatim
  ./waf --run scratch/myfirst
@end verbatim

You should see some output:

@verbatim
  Waf: Entering directory `/home/craigdo/repos/ns-3-allinone/ns-3-dev/build'
  Waf: Leaving directory `/home/craigdo/repos/ns-3-allinone/ns-3-dev/build'
  'build' finished successfully (0.418s)
  Sent 1024 bytes to 10.1.1.2
  Received 1024 bytes from 10.1.1.1
  Received 1024 bytes from 10.1.1.2
@end verbatim

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.

@c ========================================================================
@c Browsing ns-3
@c ========================================================================

@node Ns-3 Source Code
@section Ns-3 Source Code 

Now that you have used some of the @command{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:
@uref{http://code.nsnam.org/ns-3-dev}.  There, you will see the Mercurial
summary page for our @command{ns-3} development tree.

At the top of the page, you will see a number of links,

@verbatim
  summary | shortlog | changelog | graph | tags | files 
@end verbatim

Go ahead and select the @code{files} link.  This is what the top-level of
most of our @emph{repositories} will look:

@verbatim
drwxr-xr-x                               [up]     
drwxr-xr-x                               bindings python  files
drwxr-xr-x                               doc              files
drwxr-xr-x                               examples         files
drwxr-xr-x                               ns3              files
drwxr-xr-x                               regression       files
drwxr-xr-x                               samples          files
drwxr-xr-x                               scratch          files
drwxr-xr-x                               src              files
drwxr-xr-x                               utils            files
-rw-r--r-- 2009-07-01 12:47 +0200 560    .hgignore        file | revisions | annotate
-rw-r--r-- 2009-07-01 12:47 +0200 1886   .hgtags          file | revisions | annotate
-rw-r--r-- 2009-07-01 12:47 +0200 1276   AUTHORS          file | revisions | annotate
-rw-r--r-- 2009-07-01 12:47 +0200 30961  CHANGES.html     file | revisions | annotate
-rw-r--r-- 2009-07-01 12:47 +0200 17987  LICENSE          file | revisions | annotate
-rw-r--r-- 2009-07-01 12:47 +0200 3742   README           file | revisions | annotate
-rw-r--r-- 2009-07-01 12:47 +0200 16171  RELEASE_NOTES    file | revisions | annotate
-rw-r--r-- 2009-07-01 12:47 +0200 6      VERSION          file | revisions | annotate
-rw-r--r-- 2009-07-01 12:47 +0200 10946  regression.py    file | revisions | annotate
-rwxr-xr-x 2009-07-01 12:47 +0200 88110  waf              file | revisions | annotate
-rwxr-xr-x 2009-07-01 12:47 +0200 28     waf.bat          file | revisions | annotate
-rw-r--r-- 2009-07-01 12:47 +0200 35395  wscript          file | revisions | annotate
-rw-r--r-- 2009-07-01 12:47 +0200 7673   wutils.py        file | revisions | annotate
@end verbatim

Our example scripts are in the @code{examples} directory.  If you click on @code{examples}
you will see a list of files.  One of the files in that directory is @code{first.cc}.  If
you click on @code{first.cc} you will find the code you just walked through.

The source code is mainly in the @code{src} directory.  You can view source
code either by clicking on the directory name or by clicking on the @code{files}
link to the right of the directory name.  If you click on the @code{src}
directory, you will be taken to the listing of the @code{src} subdirectories.  If you 
then click on @code{core} subdirectory, you will find a list of files.  The first file
you will find (as of this writing) is @code{abort.h}.  If you click on the 
@code{abort.h} link, you will be sent to the source file for @code{abort.h} which 
contains useful macros for exiting scripts if abnormal conditions are detected.

The source code for the helpers we have used in this chapter can be found in the 
@code{src/helper} directory.  Feel free to poke around in the directory tree to
get a feel for what is there and the style of @command{ns-3} programs.