@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, theys 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 examples/ README samples/ utils/ waf.bat* build/ LICENSE regression/ scratch/ VERSION wscript doc/ ns3/ RELEASE_NOTES src/ waf* @end verbatim @cindex first.cc Change into the examples 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/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. @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 ``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 @code{NS-3 Modules} ``book.'' 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 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 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. You can, if desired, set the ``Port'' @code{Attribute} to another value later. 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. 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 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. 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 implicilty declared a simulation event (the application stop event) to be executed at ten seconds, the simulation will last 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 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. 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 @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 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/first.cc} into the @code{scratch} directory. @verbatim ~/repos/ns-3-dev > cp examples/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 Entering directory `repos/ns-3-allinone-dev/ns-3-dev/build' [563/648] cxx: scratch/myfirst.cc -> build/debug/scratch/myfirst_3.o [646/648] cxx_link: build/debug/scratch/myfirst_3.o -> build/debug/scratch/myfirst Build finished successfully (00:00:02) @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 Entering directory `repos/ns-3-allinone-dev/ns-3-dev/build' Build finished successfully (00:00:00) 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/?sort=lastchange}. If you click on the bold repository names on the left of the page, you will see @emph{changelogs} for these repositories, and links to the @emph{manifest}. From the manifest links, one can browse the source tree. The top-level directory for one of our @emph{repositories} will look something like: @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-- 2009-03-24 00:51 -0700 505 .hgignore file | revisions | annotate -rw-r--r-- 2009-03-24 00:51 -0700 1682 .hgtags file | revisions | annotate -rw-r--r-- 2009-03-24 00:51 -0700 686 AUTHORS file | revisions | annotate -rw-r--r-- 2009-03-24 00:51 -0700 14893 CHANGES.html file | revisions | annotate -rw-r--r-- 2009-03-24 00:51 -0700 17987 LICENSE file | revisions | annotate -rw-r--r-- 2009-03-24 00:51 -0700 3742 README file | revisions | annotate -rw-r--r-- 2009-03-24 00:51 -0700 13505 RELEASE_NOTES file | revisions | annotate -rw-r--r-- 2009-03-24 00:51 -0700 6 VERSION file | revisions | annotate -rw-r--r-- 2009-03-24 00:51 -0700 9257 regression.py file | revisions | annotate -rwxr-xr-x 2009-03-24 00:51 -0700 81285 waf file | revisions | annotate -rwxr-xr-x 2009-03-24 00:51 -0700 28 waf.bat file | revisions | annotate -rw-r--r-- 2009-03-24 00:51 -0700 26270 wscript file | revisions | annotate -rw-r--r-- 2009-03-24 00:51 -0700 6636 wutils.py file | revisions | annotate @end verbatim 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 be taken to the lising of the @code{src} subdirectories. If you 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 @code{abort.h} link, you will be sent to the source file for @code{abort.h}. Our example scripts are in the @code{examples} directory. The source code for the helpers we have used in this chapter can be found in the @code{src/helpers} 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.