|
OPNET Technologies |
University Department Systems and Computer Engineering Laboratory Broadband Communications and
Wireless Systems (BCWS) Project Title Resource Allocation in Fixed Broadband Wireless Access (FBWA) Networks |
|
|
|
OPNET Modeler and Wireless Module
Incorporating IEEE 802.16a large-scale path-loss model in OPNET
Simulation of Flat Raleigh Fading in OPNET
Simulating wrap-around interference model in OPNET
OPNET Modeler and Wireless
Module
OPNET Modeler is an object-oriented discrete-event network simulation software that supports all network types and technologies. It is based on a series of hierarchical models such as Network Models, Node Models and Process Models that directly parallel the structure of real networks and protocols. The behavior of the network is simulated usually in the process models which comprises of Finite State Machines (FSM). These states are constructed using C/C++ codes and OPNET kernel functions. Transition between states is done using a wide variety of interrupts. To learn more about OPNET modeler, please visit OPNET’s Modeler page.
Wireless Module is one of several add-on modules available from OPNET. This module must be used along with basic software such as Modeler to simulate wireless networks. Either fixed or mobile wireless nodes can be simulated using wireless module. There are 14 built-in pipeline stages to simulate transmission and propagation delays, antenna gains, path loss, background noise and interferences, bit error rates and error correction etc. These stages can be modified or replaced to suit a particular network. In addition, there are Antenna Pattern Editor and Modulation Editor to design any type of antenna and modulation schemes. Please visit OPNET’s Wireless Module page to learn more about wireless features.
Incorporating
IEEE 802.16a large-scale path-loss model in OPNET
The built-in large scale path loss model in OPNET is based on the free space path-loss model that resides in “dra_power” pipeline stage. A different path-loss model as proposed in IEEE work-in-progress document [IEEE802.16.3c_01/29r4] 1 is used in this project. This large-scale path-loss model is defined in eq.1 as below. Therefore, “dra_power” pipeline stage has been modified accordingly.
(1)
Where,
·
is the free space path loss at reference distance 
,
· λ is the wavelength in m,
·
n is
the path-loss exponent which is given by 
; 
is transmitter (BS for downlink) height and 
are constants depending on the terrain category B
(moderately hilly with moderate to low density trees)
· X represents shadowing effects (zero mean, 8 dB std Gaussian)
Equation (1) is based on frequencies close to 2 GHz and receive antenna heights close to 2m, therefore frequency correction and receive antenna height factors should be included as below:
; operating frequency f is in MHz (2)
; receive antenna height 
in meters (3)
Simulation
of Flat Raleigh
The built-in “dra_power” pipeline stage only considers distance attenuation and therefore it does not support wireless fading. In this project, time correlated flat Rayleigh fading has been used. Therefore, “dra_power” stage has been modified accordingly. Doppler spectrum based on [IEEE802.16.3c_01/29r4] is used which is given by,
where 
(4)
0
is the maximum Doppler
frequency which is 2.0 Hz at 3dB
Based on [R.J. Panoose] 2, Rayleigh (also Ricean capable) fading has been simulated in OPNET. The “dra_power” codes have been modified to account the followings:
· A time domain data-set for Doppler filtered Rayleigh noises is created according to Clarkes and Gans model [T.S. Rappaport] 3 [R.J. Panoose]. The duration of the simulated fading waveform is 128 sec. The data-set contains 16384 data points, therefore the sampling frequency is 128. The data-set contains x1 and x2 values which are normalized to variance 1. Steps for data set creation (Matlab is used for this part):
· Specify no. of taps for Doppler filter, maximum Doppler frequency ; these define the duration of fading waveform
· Generate complex Gaussian noises for positive half frequency, fill the negative half with the complex conjugates of the positive frequency components
· Create Doppler filter which is the sqrt of equation (4) with specified number of taps
· Multiply frequency domain components with the fading spectrum
· Perform IFFT to yield time series-data, x1
· Similarly x2 data set has been produced
· A .gdf data file has been created to store x1 and x2 values. This file is read during the calculation of fading in the power calculation pipeline stage.
· Based on the current simulation time (the time at which the packet is received), the offset of the data table is calculated. Therefore, the method considers time-correlation of the fading characteristics. A smoothing technique based on Legendre polynomials is used. Based on the time offset we obtain x1 and x2 and find the envelope factor F as per the following:
(5)
K = 0, for Rayleigh fading
Simulating
wrap-around interference model in OPNET
The interference calculation
pipeline stage has been modified to accommodate the cellular wrap-around
concept. A nine-cell network layout is used in this research. Wrap-around
interference model is used to simulate the interference effect of larger
network using limited nine cells.
References
1. V. Erceg et.
al, Channel Models for Fixed Wireless
Applications, IEEE 802.16 work-in-progress
document, July 2001.
2.
Ratish J. Punnoose, Pavel V. Nikitin, and Daniel D. Stancil, “Efficient simulation of Ricean fading within
a packet simulator”. IEEE
Vehicular Technology Conference, Sept 2000.
3. T.S. Rappaport, Wireless Communications Principles and
Practice, Prentice Hall, 1996.