1. Introduction to OMNeT++
Chia-Hui Huang
Wireless and Broadband Network Laboratory (WBNLAB)
2018/12/29

2. Reference Material OMNeT++ User Manual OMNeT++ Wiki
../doc/manual/usman.html
../doc/ide-overview.pdf
../doc/userguide.pdf
../doc/api/index.html
OMNeT++ Wiki
OMNeT++ mailing list
OMNeT++ TicToc Tutorial
../doc/tictoc-tutorial/index.html
相對於其它simulator來說，OMNeT++具有相當完整的參考文件. Ex. NS-2
Wireless and Broadband Network Laboratory (WBNLAB)
2018/12/29

3. Outline Introduction OMNeT++ Installation Modeling concepts
Configuring Simulations
Adding functionality to cSimpleModule
Demo – SimpleTicToc
Analyzing Simulation Results
Summary
Wireless and Broadband Network Laboratory (WBNLAB)
2018/12/29

4. Introduction Methods for network research Simulation frameworks
Analytical
Simulation
Simulation frameworks
Commercial
OPNET
QualNeT
Free
NS-2 (Network Simulator version 2)
OMNeT++
Well structured documentation
Procedures of research
1. Find interesting topics
2. Read related papers to the topics
3. Choose your research topic(s)
4. Idea about your topic
5. Verify idea by analytical or simulation
Wireless and Broadband Network Laboratory (WBNLAB)
2018/12/29

5. Introduction What is OMNeT++ ? Version Purposes
An object-oriented modular discrete event network simulator
Event scheduler
Version
OMNeT++ 3.x
OMNeT++ 4.0
Purposes
Modeling of wired and wireless communication networks
Protocol modeling
Modeling queuing networks
etc…
Advance of time depends on the timing of events
Wireless and Broadband Network Laboratory (WBNLAB)
2018/12/29

6. OMNeT++ Installation (OMNeT++ 4.0)
Version 3.x
OMNeT++ Installation Guide (Windows XP)
Version 4.0 (Windows XP)
Download the OMNeT++ source code from
The zip archive contains a complete MINGW installation and all additional 3rd party components (perl, tcl/tk, libxml2 etc)
MinGW gcc compiler environment
omnetpp-<version>/mingwenv.cmd (unix-based env.)
./configure
make
Test samples
(unix-based env. Use ls instead dir/w)
../samples/rundemo
Wireless and Broadband Network Laboratory (WBNLAB)
2018/12/29

7. Modeling concepts Module
An OMNeT++ model consists of modules that communicate with message passing
Simple & compound module
Connection (link) & gates
Samples/routing
Write a routing protocol as a simple module and embbed on a node
INET-framework provides layering model – samples/routing
Wireless and Broadband Network Laboratory (WBNLAB)
2018/12/29

8. Component of OMNeT++ Simulation
NED Language Topology Description (.ned)
Edited by any text edit or OMNeT++ IDE (graphical, v4.0)
Describe the network scenarios and topology
Module, channel, and networks
Message (Packet) Definition (.msg)
Various message types and data fields (new packet/message)
OMNeT++ will translate .msg to C++ classes
Simple Module Sources (.h/ .cc)
Use API provided by OMNeT++ (OMNeT++ API Reference)
omnetpp-<version>/doc/api/index.html
omnetpp.ini
usergide.pdf p.15 – some functions are available only in source editing mode
newMessageType.msg
Wireless and Broadband Network Laboratory (WBNLAB)
2018/12/29

9. NED Language Topology Description
NED – Network topology
Nodes
Connections
channel
Example code :
Wireless and Broadband Network Laboratory (WBNLAB)
2018/12/29

10. NED Language Topology Description
NED – Channel
New channel type
Expected a C++ class to be present
Predefined channel types
ned.IdealChannel
ned.DelayChannel
ned.DataRateChannel
Example code :
Example code :
Wireless and Broadband Network Laboratory (WBNLAB)
2018/12/29

11. NED Language Topology Description
NED – Module
Simple module
Parameters
double, int, bool, string, xml …
Gates
Input and output gates
Gate vector by bracket []
Compound module
parameters
Passed to submodules for initialization
Example code :
Example code :
Wireless and Broadband Network Laboratory (WBNLAB)
2018/12/29

12. Messages Basic Concepts
cMessage is the base class of all message types
Events, Messages, Packets, Frames, Cells, Bits or Signals
Basic attributes of cMessage
Name : string
Kind : zero or positive integers
Length : bits
Bit error flag : bool
Priority : used by simulation kernel
Time stamp : simtime_t (usage by users)
There are some other attributes not mentioned here
source/destination module and gate, sending/arrival time …etc.
Wireless and Broadband Network Laboratory (WBNLAB)
2018/12/29

13. Messages Basic Concepts
Set/get attributes of cMessage by setXXX()/getXXX() /
Duplicate message
Information about a Message
bool isSelfMessage();
If message is sent by itself
message are often used to represent events internal to a module
Bool isScheduled();
If message is put into the scheduler
Scheduled message can be canceled by cancelEvent()
Wireless and Broadband Network Laboratory (WBNLAB)
2018/12/29

14. Message definitions
In practice, you’ll need to add various fields to cMessage to make it useful (protocol header fields)
Subclass
Message definition file (.msg)
C++ code is automatically generated form message definition file
Saving you a lot of typing
mypacket_m.h, mypacket_m.cc
Example code : mypacket.msg
Example code
Tictoc13.msg
Wireless and Broadband Network Laboratory (WBNLAB)
2018/12/29

15. Configuring Simulations
Configuration and input data for the simulation are in a configuration file
omnetpp.ini (default)
Sections
[General] section
Common parameters
network option
Selects the model to be set up and run
cpu-time-limit
The length of simulation time
Named configurations
Run specific parameters
[Config tictoc1], [Config tictoc 2]
Example set up :
Wireless and Broadband Network Laboratory (WBNLAB)
2018/12/29

16. Configuring Simulations
Section inheritance
Missing configuration key or parameter value
Fall back to [General] (default) or extends section
Example :
Wireless and Broadband Network Laboratory (WBNLAB)
2018/12/29

17. Configuring Simulations
User interfaces
Tkenv – Tcl/Tk-based graphical, windowing user interface
Debug and test
Cmdenv – command-line user interface
Batch execution
Selecting user interface (Makefile)
Wireless and Broadband Network Laboratory (WBNLAB)
2018/12/29

18. Simulation steps Wireless and Broadband Network Laboratory (WBNLAB)
2018/12/29

19. Adding functionality to cSimpleModule ( simple module source .h/.cc )
Simple modules encapsulate C++ code that generates events and reacts to events
Create simple module types
Subclassing the cSimpleModule class
Necessary functions
This function creates the necessary simple and compound module
Parameter set up, initial value …
Process events
Implement the model’s behavior
This function is called when the simulation terminates successfully
Free memory, statistic …
Wireless and Broadband Network Laboratory (WBNLAB)
2018/12/29

20. Adding functionality to cSimpleModule ( simple module source .h/.cc )
Defining simple module types (.h file)
The simple module class has to be registered with OMNeT++
Txc1.cc
Wireless and Broadband Network Laboratory (WBNLAB)
2018/12/29

21. Demo - tictoc with two nodes
SimpleTicToc
Scenario – .ned
Simple module – TXC.ned
Parameters - message initiator (controlled by omnetpp.ini)
Gates – in/out
Network – Tictoc
Submodules – two nodes tic and toc
Connections – ned.IdealChannel (default)
Behavior - .h/.cc
The two nodes keep passing the same packet back and forth
Omnetpp.ini
ticInit
tocInit
Tictoc scenario
Wireless and Broadband Network Laboratory (WBNLAB)
2018/12/29

22. Analyzing Simulation Results
Result files
Output Vectors
Writing time series data
Several times in a simulation
Use cOutVector objects
The record() method is used to output a value with a timestamp
All cOutVector objects write to a single vector file (default - omnetpp.vec)
cStatistic is the base class for all statistical data collection classes. cStatistic itself adds no data members or algorithms to cOwnedObject, it only defines virtual functions that will be redefined in descendants. No instance of cStatistic can be created
cStdDev - Statistics class to collect min, max, mean, and standard deviation
cWeightedStdDev is similar to cStdDev, but accepts weighted observations. cWeightedStdDev can be used for example to calculate time average. It is the only weighted statistics class.
cDensityEstBase - Distribution estimation
Wireless and Broadband Network Laboratory (WBNLAB)
2018/12/29

23. Analyzing Simulation Results
Result files
Output Scalars
Usually, once and at finish() in a simulation
To record summary data at the end of the simulation
To do several runs with different parameter setting
Use the recordScalar() function of cSimpleModule (cComponent)
The output scalar file is preserved across simulation runs (default – omnetpp.sca) ( vs. output vector file)
Mostly, we can create a private member variable with type cStatistic for statistic
cStatistic is the base class for all statistical data collection classes. cStatistic itself adds no data members or algorithms to cOwnedObject, it only defines virtual functions that will be redefined in descendants. No instance of cStatistic can be created
cStdDev - Statistics class to collect min, max, mean, and standard deviation
cWeightedStdDev is similar to cStdDev, but accepts weighted observations. cWeightedStdDev can be used for example to calculate time average. It is the only weighted statistics class.
cDensityEstBase - Distribution estimation
Wireless and Broadband Network Laboratory (WBNLAB)
2018/12/29

24. Analyzing Simulation Results
Statistics in OMNeT++
Statistic and result collection classes
cStdDev, cWeightedStdDev, LongHistgoram, cDoubleHistogram …
Refer to cStatistic Class Reference in OMNeT++ API
Collects statistic data
cStatistic::collect()
Records the statistics into the scalar file
cStatistic::record()
Mostly, we can create a private member variable with cStdDev for statistic
Min, max, mean, and standard deviation
cStatistic is the base class for all statistical data collection classes. cStatistic itself adds no data members or algorithms to cOwnedObject, it only defines virtual functions that will be redefined in descendants. No instance of cStatistic can be created
cStdDev - Statistics class to collect min, max, mean, and standard deviation
cWeightedStdDev is similar to cStdDev, but accepts weighted observations. cWeightedStdDev can be used for example to calculate time average. It is the only weighted statistics class.
cDensityEstBase - Distribution estimation
Wireless and Broadband Network Laboratory (WBNLAB)
2018/12/29

25. Analyzing Simulation Results
Statistics in OMNeT++
cStatistic is the base class for all statistical data collection classes. cStatistic itself adds no data members or algorithms to cOwnedObject, it only defines virtual functions that will be redefined in descendants. No instance of cStatistic can be created
cStdDev - Statistics class to collect min, max, mean, and standard deviation
cWeightedStdDev is similar to cStdDev, but accepts weighted observations. cWeightedStdDev can be used for example to calculate time average. It is the only weighted statistics class.
cDensityEstBase - Distribution estimation
Wireless and Broadband Network Laboratory (WBNLAB)
2018/12/29

26. Analyzing Simulation Results
Analysis file
.anf files
Reproducible recipe for finding results from raw data (vector or scalar)
Creating analysis file
File | new | Analysis File or
Double click on the result files in the Project Explore View
Analysis Editor
Selecting input files
browsing input
filter expressions
Datasets
Datasets describe a set of input data, the processing applied to them and the charts (reproducible recipe)
Wireless and Broadband Network Laboratory (WBNLAB)
2018/12/29

27. Analyzing Simulation Results
Editing datasets
add new dataset
Filter expressions
Add button (specify filter pattern by string)
Browse data (specify filter pattern by GUI)
Reproducible recipe
Specifying Run ID – *
Chart types
Bar chart
Line chart
Histogram chart …
Wireless and Broadband Network Laboratory (WBNLAB)
2018/12/29

28. Analyzing Simulation Results
Example
Scenario – tictoc with 5 nodes
Recording simulation results
Hop count – Output Vector file
Use cOutVector object, record() function when handleMessage()
#received and #sent messages – Output Scalar file
Use recordScalar() function when finish()
Mean hop count – Output Scalar file
Use cStdDev object, setName(), collect(), and record() functions
Wireless and Broadband Network Laboratory (WBNLAB)
2018/12/29

29. Summary
Wireless and Broadband Network Laboratory (WBNLAB)
2018/12/29

30. SimpleTictoc – tictoc with 2 nodes
A “network“ that consists of two nodes,
one of the nodes will create a packet,
and the two nodes will keep passing the same packet back and forth
Wireless and Broadband Network Laboratory (WBNLAB)
2018/12/29

31. TicToc with five nodes One of the nodes generates a message,
The others keep tossing it around in random directions
Wireless and Broadband Network Laboratory (WBNLAB)
2018/12/29

32. FYI Simulation models http://www.omnetpp.org/models INET Framework –
INET Framework contains models for IP, TCP, UDP, PPP, Ethernet, and other protocols
INETMANET (new) –
Ad hoc routing protocol implementations for the INET Framework
Mobility Framework –
Supports simulation of wireless and mobile network
Etc.
Wireless and Broadband Network Laboratory (WBNLAB)
2018/12/29

33. FYI Additional useful tools Blackboard object Gnuplot
Plot data with flexible format
Shell script or batch file
Run multiple scenarios sequentially
Blackboard object
Statistic
Debugging …
Wireless and Broadband Network Laboratory (WBNLAB)
2018/12/29