Transcript 21_Radio_Modeling_2

OPNET UNIVERSITY
2000
Transceiver Pipeline and
Radio Modeling
Copyright © 2000 MIL 3, Inc.
Modeling Custom Wireless Effects– 1
OPNET UNIVERSITY
2000
Goals
• Introduce the OPNET Transceiver Pipeline
– Capabilities
– Defaults
• Modify pipeline
– Show openness and extensibility
– Model custom wireless effects
Copyright © 2000 MIL 3, Inc.
Modeling Custom Wireless Effects– 2
OPNET UNIVERSITY
2000
Pipeline Models
• Scope
– Each stage uses a pipeline model
– Stages are referenced via a pipeline model object attributes
- Point-to-Point - link attributes
- Bus -Bus attributes
- Radio - transceiver attributes
– Stages must be compiled prior to reference
• Modeling method
– Create stage in context of the pipeline model
– Compile stage into object form
– Change pipeline model object attribute to reference stage
• NOTE: Stage and procedure (pipeline model context) can be used interchangeably
Copyright © 2000 MIL 3, Inc.
Modeling Custom Wireless Effects– 3
OPNET UNIVERSITY
2000
Default Pipeline Stage Location
• Default stages
– <mil3_dir>/<rel_dir>/models/std/links
• Default stage prefixes
– dpt_* - default point-to-point
– dbu_* - default bus
– dra_* - default radio
• NOTE: OPNET can access pipeline stages from any path in mod_dirs
Copyright © 2000 MIL 3, Inc.
Modeling Custom Wireless Effects– 4
OPNET UNIVERSITY
2000
Pipeline Stage Conventions
• File naming
– <name>.ps.c
- C procedure
– <name>.ps.cpp - C++ procedure
– <name>.ps.o
- object form
• Procedure naming
– Same as file name w/o extension
• Compilation
– op_mko -type ps -m <name>
– Generates <name>.ps.o
• NOTE: Choices for pipeline stages in OPNET are taken from the set of .ps.o files
located in mod_dirs.
Copyright © 2000 MIL 3, Inc.
Modeling Custom Wireless Effects– 5
OPNET UNIVERSITY
2000
Radio Pipeline Models
• Execution sequence for one transmission
transmitter
0
executed once at the start of
simulation for each pair of
transmitter and receiver channels
Receiver
Receiver
Receiver
Group
Receiver
Group
Group
Group
stages 2-6 executed separately for each receiver
1
Transmission
Delay
stage 1 executed once per
transmission
2
Link
Link
Link
Closure
Link
Closure
Closure
Closure
Propagation
Propagation
Propagation
Delay
Propagation
Delay
Delay
Delay
3
5
Channel
Channel
Channel
Match
Channel
Match
Match
Match
Tx
Antenna
Tx
Antenna
Tx
Antenna
Gain
Tx
Antenna
Gain
Gain
Gain
4
(Continued on the next slide)
Copyright © 2000 MIL 3, Inc.
Modeling Custom Wireless Effects– 6
OPNET UNIVERSITY
2000
Radio Pipeline Models (cont.)
• Execution sequence for one transmission (cont.)
6
Received
Received
Received
Power
Rx
Antenna
Power
Power
Gain
Background
Background
Background
Noise
Received
Noise
Noise
Power
Interference
Interference
Interference
Noise
Background
Noise
Noise
Noise
8
8
Stages 10-12 may be executed one or more times
Error
Error
Error
Allocation
Bit
Error
Allocation
Allocation
Rate
Error
Error
Error
Allocation
Error
Allocation
Allocation
Allocation
12
Copyright © 2000 MIL 3, Inc.
11
Bit
Error
Bit
Error
Bit
Error
Rate
Signal-to-Noise
Rate
Rate
Ratio
Rx
Antenna
Rx
Antenna
Rx
Antenna
Gain
Error
Gain
Gain
Correction
10
Signal-to-Noise
Signal-to-Noise
Signal-to-Noise
Ratio
Interference
Ratio
Ratio
Noise
9
Stages 9-11
may be
executed zero
or more times
multiple receivers
13
Modeling Custom Wireless Effects– 7
OPNET UNIVERSITY
2000
Radio Pipeline Model Attributes
Radio Transmitter
Radio Receiver
•Receiver Group
•Transmission Delay
•Link Closure
•Channel Match
•Tx Antenna Gain
•Propagation Delay
•Rx Antenna Gain
•Received Power
•Background Noise
•Interference Noise
•Signal-to-Noise Ratio
•Bit Error Rate
•Error Allocation
•Error Correction
6 Stages (0-5) Associated
with Radio Transmitter
8 Stages (6-13) Associated
with Radio Receiver
Copyright © 2000 MIL 3, Inc.
Modeling Custom Wireless Effects– 8
OPNET UNIVERSITY
2000
Stage 7: Receiver Power
• Invocation
– Once for each destination channel
– Called immediately after stage 6 - no intervening events
• Purpose
– Computes signal power level at receiver
– Typically based on transmitter power and frequency, distance, and antenna gains
– Computed only for valid and noise packets
• Requirements
– Sets RCVD_POWER TDA
• Results
– Kernel uses value to record receiver power channel statistic
– Places packets in separate lists based on match status
Copyright © 2000 MIL 3, Inc.
Modeling Custom Wireless Effects– 9
OPNET UNIVERSITY
2000
Stage 7: Receiver Power (default)
• Name
– dra_power
• Computation
– Determines if signal lock is active
– Computes received power
-
valid and noise packets
Computes path loss - free space
Determines in-band transmission power
Obtains tx and rx antenna gains
  
Lp  

 4D 
C

fc
2
P  f  f min 
Pi  tx max
B
Prx  PiGtx LpGrx
Lp  Pathloss
  Wavelength
D  Distance
C  Speed of Light
f c  Center Fre quency
Pi  Inband Frequency
Ptx  Transmitte d Power (watts)
f max  Maximum Frequency (Hz)
f min  Minimum Frequency (Hz)
B  Bandwidth
Prx  Received Power (watts)
Gtx  Transmitte r Antenna Gain (watts)
Grx  Receiver Antenna Gain (watts)
• Result
– Value placed in RCVD_POWER for use by stage 9
Copyright © 2000 MIL 3, Inc.
Modeling Custom Wireless Effects– 10
OPNET UNIVERSITY
2000
Stage 8: Background Noise
• Invocation
– Called immediately after stage 7 - no intervening events
• Purpose
– Represents effects of all background noise sources
– Typically includes
- thermal or galactic noise
- neighboring electronics emissions
- other unmodeled radio transmissions (commercial/amateur radio, TV)
• Requirements
– Sets BKGNOIS TDA
• Results
– Typically used by stage 10 in signal-to-noise computation
Copyright © 2000 MIL 3, Inc.
Modeling Custom Wireless Effects– 11
OPNET UNIVERSITY
2000
Stage 8: Background Noise (default)
• Name
– dra_bknoise
• Computation
– Constant ambient noise
– Constant background noise
– Constant thermal noise
• Result
Trx  ( NF  1.0) * 290.0
Tbk  290.0
k  1.379 E 23
N b  Trx  Tbk Brxk
N a  Brx 1.0E
N  Nb  Na
 26

NF  Noise Figure
Trx  Receiver Temperatur e
Tbk  Background Temperatur e
k  Boltzmann' s Constant
Brx  Receiver B andwidth
N b  Background Noise
N a  Ambient Noise
N  Noise
– Value placed in BKGNOISE for use by stage 10 in signal-to-noise computation
Copyright © 2000 MIL 3, Inc.
Modeling Custom Wireless Effects– 12
OPNET UNIVERSITY
2000
Packet Segments
• Portion of packet with constant signal-to-noise
– Segmentation performed by kernel
– SNR computation performed by pipeline model
– Upon packet completion, kernel subtracts SNR of completing packet
1
SNR Variations
3
2
4
1
Copyright © 2000 MIL 3, Inc.
1
SNR Variations
2
3
2
SNR Variations
2
3
2
1
1
1
Modeling Custom Wireless Effects– 13
OPNET UNIVERSITY
2000
Stage 9: Interference Noise
• Invocation
– Only if packet collision occurs
• Purpose
– Accounts for concurrent transmissions
– Compute the effect of noise on valid packets
• Requirements
– Sets NOISE_ACCUM TDA
– Sets NUM_COLLS TDA
• Results
– Accumulates noise of interfering packets
– Noise from packet completing reception is subtracted by kernel
– Typically used in stage 10 for signal-to-noise ratio computations
Copyright © 2000 MIL 3, Inc.
Modeling Custom Wireless Effects– 14
OPNET UNIVERSITY
2000
Stage 9: Interference Noise (default)
Case 1:
• Name
Valid
1
Tx
Rx
2
– dra_inoise
• Computation
Tx
– Increments collision count
– Adds received power of colliding packet
Case 2:
Valid
1
Tx
Rx
2
- Obtains power from RCVD_POWER TDA
• Result
Tx
– Places accumulated noise in NOISE TDA
– Value used in stage 10 in signal-to-noise computation
Case 3:
Noise
1
Tx
Rx
2
Tx
Copyright © 2000 MIL 3, Inc.
Modeling Custom Wireless Effects– 15
OPNET UNIVERSITY
2000
Stage 10: Signal-to-Noise Ratio
• Invocation (only for valid packets)
– Operates on valid packets - those in valid list
– Does not require collision for invocation
Case 1:
Valid
Tx
Case 2:
Rx
Valid
Rx
2
• Purpose
– Computes the current average SNR
– Typically based on received power and noise
• Requirements
– Sets SNR TDA
1
Tx
Tx
Case 3:
Noise
1
Tx
Rx
2
Tx
• Results
– Used by kernel to update receiver channel statistics
– Used by later stages
Copyright © 2000 MIL 3, Inc.
Modeling Custom Wireless Effects– 16
OPNET UNIVERSITY
2000
Stage 10: Signal-to-Noise Ratio (default)
• Name
– dra_snr
• Computation
– Obtains received power from RCVD_POWER TDA
- Computed in stage 7
– Obtains background noise from BKGNOISE_TDA
- Computed in stage 8
– Obtains interference noise from NOISE_ACCUM TDA
- Computed in stage 9
– Computes signal-to-noise ratio (in dB)
SNR  10 log 10 Pr /( Pb  Pi )
• Result
– Places in SNR TDA
– Value used in stage 11 in bit-error-rate computation
Copyright © 2000 MIL 3, Inc.
Pr  Received Power (watts)
Pb  Backgroun Noise (watts)
Pi  Interfernc e Noise (watts)
Modeling Custom Wireless Effects– 17
OPNET UNIVERSITY
2000
Stage 11: Bit Error Rate
• Invocation
– Operates on valid packets - those in valid list
– Does not require collision for invocation
Case 1:
Case 2:
Valid
1
Rx
2
Tx
Case 3:
Noise
1
Tx
• Requirements
– Sets BER TDA
Rx
Tx
• Purpose
– Derives the probability of bit errors
– Computed for each packet segment - constant SNR
– Value typically obtained from modulation curve
Valid
Tx
Rx
2
Tx
• Results
– Used by the kernel to record BER statistic
– Typically used in stage 12 for allocating errors
Copyright © 2000 MIL 3, Inc.
Modeling Custom Wireless Effects– 18
OPNET UNIVERSITY
2000
Stage 11: Bit Error Rate (default)
• Name
– dra_ber
• Computation
– Obtains signal-to-noise ratio from SNR TDA
- Computed in stage 10
– Obtains processing gain from PROC_GAIN TDA
- Attribute of receiver channel
– Computes effective SNR
SNReffective  SNRactual  G p
G p  Processing Gain
– Determines expected bit-error-rate
- Modulation table lookup
• Result
– Records BER TDA for use in stage 12 in error allocation computation
Copyright © 2000 MIL 3, Inc.
Modeling Custom Wireless Effects– 19
OPNET UNIVERSITY
2000
Stage 12: Error Allocation
• Invocation
– Called immediately after stage 11 - no intervening events
• Purpose
– Estimates bit errors for packet segment
• Requirements
– Sets bit-error accumulation in NUM_ERRORS TDA
– Sets empirical bit error rate in ACTUAL_BER TDA
• Results
– Kernel maintains a bit accumulator - NUM_ERRORS TDA
– Kernel updates BER statistic - ACTUAL_BER TDA
– Typically used in stage 13 for error correction
Copyright © 2000 MIL 3, Inc.
Modeling Custom Wireless Effects– 20
OPNET UNIVERSITY
2000
Stage 12: Error Allocation (default)
• Name
– dra_error
• Computation
– Does not perform bit-by-bit error computations
- Cannot retain bit-error location
– Obtains probability of error - BER TDA
– Computes probability of k errors
– Generates uniform random number: r = (0 1]
– Integrates probability mass over possible outcomes
N
Pk  p k (1  p ) N k  
k
N
P  r
k 0
k
Pk  Probabilit y of k Errors
p  Probabilit y of Error
N  Packet Len gth (bits)
r  op _ dist _ uniform(1)
k  Number of Errors
• Result
– Records NUM_ERRORS TDA
Copyright © 2000 MIL 3, Inc.
Modeling Custom Wireless Effects– 21
OPNET UNIVERSITY
2000
Stage 13: Error Correction
• Invocation
– Once for each valid packet
– Called immediately after stage 12 - no intervening events
• Purpose
– Determines acceptability of arriving packet
• Requirements
– Sets PK_ACCEPT TDA
• Results
– Rejected
- destroyed by Kernel
– Accepted
- forwarded on output stream
Copyright © 2000 MIL 3, Inc.
Modeling Custom Wireless Effects– 22
OPNET UNIVERSITY
2000
Stage 13: Error Correction (default)
• Name
– dra_ecc
• Computation
– Obtains threshold from ECC_TRESH TDA
- Percentage of packet in error that still yields acceptability
– Obtains packet length from op_pk_total_size_get ()
– Obtains number of errors from NUM_ERRORS TDA
– Computes the percent error
%Error
• Result
N errors

Plength
– Packet accepted or rejected depending on threshold and error
– Releases signal lock
Copyright © 2000 MIL 3, Inc.
Modeling Custom Wireless Effects– 23
OPNET UNIVERSITY
2000
Lab: Capture Mode - Overview
• Problem
– Default pipeline stages use signal lock
– Systems can lock on to lowest power signal
- Higher power signal does not have lock, but dominates channel
- Drowns out lower powered signal
- Causes both communications to fail
• Goals
– Modify stages to incorporate signal lock and power lock
– Compare results between the two capture modes
• Purpose
– Show additional stage modification
– Incorporate power lock capability
Copyright © 2000 MIL 3, Inc.
Modeling Custom Wireless Effects– 24
OPNET UNIVERSITY
2000
Lab: Capture Mode - Project/Scenario
• Return to OPNET Modeler (Radio)
– Maximize application
• Open project
– Capture_Mode
• View scenario
– Signal_Lock
- High powered transmitter
- Low powered transmitter
- Receiver
Copyright © 2000 MIL 3, Inc.
Modeling Custom Wireless Effects– 25
OPNET UNIVERSITY
2000
Lab: Capture Mode - Choose Results / Simulation
• Observe chosen results
– Pulldown menu: Simulation  Choose Individual Statistics
- Node Statistics:Radio Transmitter:
Traffic Sent (bits/sec)
- Node Statistics:Radio Receiver:
Traffic Received (bits/sec)
- Node Statistic:Radio Receiver:
Traffic Received High Powered Tx (bits/sec)
- Node Statistic:Radio Receiver:
Traffic Received Low Powered Tx (bits/sec)
• Execute simulation
– Pulldown menu: Simulation  Run Simulation
Copyright © 2000 MIL 3, Inc.
Modeling Custom Wireless Effects– 26
OPNET UNIVERSITY
2000
Lab: Capture Mode - View Results
• View Results
– Pulldown menu: Results  View Results
– Select statistics
- Object Statistics:Capture Mode:High Powered
Transmitter:Radio Transmitter:
Traffic Sent (bits/sec)
- Object Statistics:Capture Mode:Low Powered
Transmitter:Radio Transmitter:
Traffic Sent (bits/sec)
- Object Statistics:Capture Mode:Receiver:Radio Receiver:
Traffic Received (bits/sec)
- Object Statistics:Capture Mode:Receiver:Radio Receiver:
Traffic Received High Powered Tx(bits/sec)
- Object Statistics:Capture Mode:Receiver:Radio Receiver:
Traffic Received Low Powered Tx(bits/sec)
Copyright © 2000 MIL 3, Inc.
Modeling Custom Wireless Effects– 27
OPNET UNIVERSITY
2000
Lab: Capture Mode - View Results (cont.)
• Show results
– View mode
- Statistics Overlaid
– Filter
- As Is
- Average
– Click Show
• Conclusion
– Low powered transmission dominating
- Longer packet lengths - longer signal lock retention
– High powered transmission loss
- Short packet lengths - shorter signal lock retention
Copyright © 2000 MIL 3, Inc.
Modeling Custom Wireless Effects– 28
OPNET UNIVERSITY
2000
Lab: Capture Mode - Stage Modification Overview
• Goal
– Modify three pipeline stages to model power lock
– Modifications will eliminate signal lock
• Approach
– Modify the pipeline stages
- power model
- inoise model
- ecc model
– Update Models
- Project: Create new scenario
- Node: Update node model
– Execute simulation
– View results
Copyright © 2000 MIL 3, Inc.
Modeling Custom Wireless Effects– 29
OPNET UNIVERSITY
2000
Lab: Capture Mode - Open Pipeline Stage
• Open pipeline stage
– Return to Windows NT Explorer
– Go to directory containing file
- c:\users\student\op_models
– Double click to open file
- opnetwork_capture_mode_power.ps.c
Copyright © 2000 MIL 3, Inc.
Modeling Custom Wireless Effects– 30
OPNET UNIVERSITY
2000
Lab: Capture Mode - Modify Pipeline Stage
• Modify stage
– Observe code block in middle of file
- Line 59 through 91
– Cut entire else statement
- Line 59: start cutting
- Line 91: end cutting
– Save file
Copyright © 2000 MIL 3, Inc.
Modeling Custom Wireless Effects– 31
OPNET UNIVERSITY
2000
Lab: Capture Mode - Modify Pipeline Stage (cont.)
• Open pipeline stage
– Return to Windows NT Explorer
– Go to directory containing file
- c:\users\student\op_models
– Double click to open file
- opnetwork_capture_mode_inoise.ps.c
• Modify stage
– Modifications already complete
– Observe code block at end of file
- Line 69
Copyright © 2000 MIL 3, Inc.
Modeling Custom Wireless Effects– 32
OPNET UNIVERSITY
2000
Lab: Capture Mode - Modify Pipeline Stage (cont.)
• Open pipeline stage
– Return to Windows NT Explorer
– Go to directory containing file
- c:\users\student\op_models
– Double click to open file
- opnetwork_capture_mode_ecc.ps.c
• Modify stage
– Modifications already complete
– Observe code block at end of file
- Line 60
Copyright © 2000 MIL 3, Inc.
Modeling Custom Wireless Effects– 33
OPNET UNIVERSITY
2000
Lab: Capture Mode - Compile Stages
• Compile the pipeline stage
– Return to a Command Prompt
– Execute the compile command
- op_mko -type ps -m opnetwork_capture_mode_power
- op_mko -type ps -m opnetwork_capture_mode_inoise
- op_mko -type ps -m opnetwork_capture_mode_ecc
Copyright © 2000 MIL 3, Inc.
Modeling Custom Wireless Effects– 34
OPNET UNIVERSITY
2000
Lab: Capture Mode - Duplicate Scenario
• Return to OPNET Modeler (Radio)
– Maximize application
• Refresh model directories
– Pulldown menu: File  Refresh Model Directories
– Required for OPNET to know about new files
• Duplicate scenario
– Pulldown menu: Scenarios  Duplicate Scenario
– Scenario name: Power_Lock
Copyright © 2000 MIL 3, Inc.
Modeling Custom Wireless Effects– 35
OPNET UNIVERSITY
2000
Lab: Capture Mode - Update Node Model
• Open Receiver node model
– Double click High Powered Receiver object
- Node model: no_doppler_rt
• Change receiver module attribute
– Right click receiver module
– Change attribute
- power model: opnetwork_capture_mode_power
- inoise model: opnetwork_capture_mode_inoise
- ecc model: opnetwork_capture_mode_ecc
– Save node model
- Pulldown menu: File  Save As
- Filename: capture_mode_power_lock_rr
• Close node Model
Copyright © 2000 MIL 3, Inc.
Modeling Custom Wireless Effects– 36
OPNET UNIVERSITY
2000
Lab: Capture Mode - Update Network Model
• Return to the Project/Scenario
– Capture_Mode-Power_Lock
• Change Receiver node attribute
– Right click Receiver node
- Select Edit Attributes
– Change model attribute value
- Select Edit...
- Select capture_mode_power_lock_rr
– Close attribute edit box
- Select Close
• Execute simulation
– Pulldown menu: Simulation  Run Simulation
Copyright © 2000 MIL 3, Inc.
Modeling Custom Wireless Effects– 37
OPNET UNIVERSITY
2000
Lab: Capture Mode - View Results
• View Results
– Pulldown menu: Results  View Results
– Select statistics
- Object Statistics:Capture Mode:High Powered
Transmitter:Radio Transmitter:
Traffic Sent (bits/sec)
- Object Statistics:Capture Mode:Low Powered
Transmitter:Radio Transmitter:
Traffic Sent (bits/sec)
- Object Statistics:Capture Mode:Receiver:Radio Receiver:
Traffic Received (bits/sec)
- Object Statistics:Capture Mode:Receiver:Radio Receiver:
Traffic Received High Powered Tx(bits/sec)
- Object Statistics:Capture Mode:Receiver:Radio Receiver:
Traffic Received Low Powered Tx(bits/sec)
Copyright © 2000 MIL 3, Inc.
Modeling Custom Wireless Effects– 38
OPNET UNIVERSITY
2000
Lab: Capture Mode - View Results (cont.)
• Show results
– View mode
- Statistics Overlaid
– Filter
- As Is
- Average
– Click Show
• Conclusion
– High powered transmission dominating
- Obtaining power lock
– Low powered transmission loss
Copyright © 2000 MIL 3, Inc.
Modeling Custom Wireless Effects– 39
OPNET UNIVERSITY
2000
Lab: Capture Mode - Summary
• Signal lock
– Representative of most systems
– Locks on to first arriving signal
– All other signals are noise, regardless of power level
• Power lock
– Locks on to highest power signal
– All other signals are noise
– Example: IS-95
-
Mobile unit moving between cells
Soft handoff process
Constantly demodulates 3 incoming signals in parallel
Monitors 4th incoming signal
Any of 3 active signals become weak, mobile can switch to high-powered signal
Copyright © 2000 MIL 3, Inc.
Modeling Custom Wireless Effects– 40
OPNET UNIVERSITY
2000
Conclusion
• Transceiver pipeline
– Scope
– Capabilities
– Default
• Pipeline Modifications
– Closure model
– Channel match model
– Power, Inoise, and ECC models
• Pipeline stages
– Flexible
– Open
– Extensible
Copyright © 2000 MIL 3, Inc.
Modeling Custom Wireless Effects– 41
OPNET UNIVERSITY
2000
Further Information
• Documentation
– Modeling Concepts
- Communication Mechanisms
– General Models
- Pipeline Stages / Bus Link
- Pipeline Stages / Point-to-Point Link
- Pipeline Stages / Radio Link
Copyright © 2000 MIL 3, Inc.
Modeling Custom Wireless Effects– 42