|
|
Icons 
and 
Noun Project.
Item |
Description |
Link |
Type |
1 |
Introduction to the Capsim® OFDM (IEEE 802.11a) Block Diagram Modeling and Simulation |
|
Introduction |
2 |
Capsim® Block Diagram End to End OFDM ( IEEE 802.11a) With Channel Model, SNR and Full Acquisition and Demodulation |
|
Screen Shots |
3 |
List of Topologies Included in Repository |
|
Table |
4 |
List of OFDM and Channel Model C Blocks Included in Repository |
|
Table |
5 |
List of OFDM and Channel Model C Subroutines Included in Repository |
|
Table |
6 |
Building Capsim® for OFDM Modeling and Simulation |
|
Instructions |
7 |
Capsim® Text Mode Kernel (TMK) Installation |
|
GitHub Repository |
8 |
GitHub Repository Capsim® OFDM Block Diagram Modeling and Simulation |
|
GitHub Repository |
9 |
OFDM Video Tutorials by Silicon DSP Corporation |
|
Video Tutorial |
10 |
OFDM Release HTML README |
|
HTML Link |
Copyright (c) 1993-2007 Silicon DSP Corporation Permission is granted to copy, distribute and/or modify this document under the terms of the GNU Free Documentation License, Version 1.2 or any later version published by the Free Software Foundation; with no Invariant Sections, no Front-Cover Texts, and no Back-Cover Texts. A copy of the license is included in the section entitled "GNU Free Documentation License".
Silicon DSP Corporation has developed all the C code for a full implementation of the IEEE802.11a OFDM end to end Block Diagram Modeling and Simulation system. This includes full support for acquisition, equalization, and carrier offset correction. The modular architecture uses individual blocks (written in C) for the implementation of the various stages in modulation and demodulation. Subroutines ( C code) are also supplied with Include files shared between blocks and subroutines. See the Block Diagram of a full system here.
The majority of blocks were written in 2002 with major enhancements throughout the years. The OFDM system has been the basis for successfull FPGA implimentation for real time processing over the air.
In addition, the Capsim® OFDM Block Diagram system was extensively used in developing the 
Tutorials on OFDM.
Click here for the OFDM Tutorial Videos on .
Note: this repository supports the Text Mode Kernel version of Capsim®. The graphical block diagram is from the soon to be released Capsim® Version 7 which uses Qt® for interactive graphical interface. However, the topology in this Repository are the same. You can use the block names in the screen shot and then use the Capsim® command "to blockname" to go the the block, change parameters and run the simulation. There is a lot of benefit to the non graphical mode in portability and flexibility. The graphical version also supports the text mode operation.
An updated link to Capsim® Version 7 using Qt® will be provided in the Repostory on GitHub. Stay tuned.
Note: to view topology, right click, select Open Image in New Tab. Then zoom-in.
Capsim® OFDM (802.11a) End to End Block Diagram With Acquisition and Channel Modeling and SNR Specificaton
| Item | Topology Name | Description | Author | Date |
|---|---|---|---|---|
| 1 | OFDM_11a_Acquisition_2.t | Full end to end IEEE 802.11a OFDM with Fading Channel, Carrier Offset and Acquisition. | Ardalan | 2007 |
| 2 | OFDM_11a_Acq_CO_Drift.t | Demonstates Pilot Tracking correcting for residual carrier offset with low SNR. | Ardalan | 2007 |
| 3 | MR_OFDM_SS_CH.t | End to End All Rates Steady State with Channel Est. Equalization No Acquisition OFDM based on IEEE 802.11a | Ardalan | September 24, 2002 |
| 4 | adjacent_channel.t | Model adjacent channel in 802.11a | Ardalan | 2007 |
| 5 | chm_80.t | Hierarchical Block(HBlock) SISO Fading Channel Model used in MR_OFDM_SS_CH.t | Ardalan | 08/11/2002 |
| 6 | constellation.t | Plot Constellation and Compute EVM. Reads from file (qam64.dat) | Ardalan | March 2007 |
| 7 | cxdecim.t | HBlock Complex Decimation | Ardalan | February 1, 2002 |
| 8 | cxrsmpl.t | HBlock Resample Complex Interpolation. Specify oversampling ratio. | Ardalan | February 1, 2002 |
| 9 | JakesModel.t | HBlock Jake's Model Mobile Fading Channel. Used in OFDM_EE_JK2.t | Ardalan | September 29, 2001 |
| 10 | MR_OFDM_SCRIPT.t | OFDM Scriptable With TCL. End to End All Rates Steady State with Channel Est. Equalization No Acquisition OFDM based on IEEE 802.11a | Ardalan | September 24, 2002 |
| 11 | MR_OFDM_SS_CH_SCRIPT.t | OFDM Scriptable With TCL. End to End All Rates Steady State with Channel Est. Equalization No Acquisition OFDM based on IEEE 802.11a | Ardalan | September 24, 2002 |
| 12 | per.tcl | Packet Error Rate AWGN Channel Multiple Rates Steady State.TCL Script that tabulates the PER versus SNR for AWGN all rates. Uses topology MR_OFDM_SCRIPT.t. Long simulation run. Be patient. Use 'ls -lt' after run to check for results. | Ardalan | August 10, 2007 |
| 13 | perfade.tcl | Packet Error Rate Fading Channel ((rms delay spread)) and AWGN Channel Multiple Rates Steady State. | Ardalan | August 10, 2007 |
| 14 | OFDM_EE_JK2.t | IEEE 802.1a OFDM with Acquisition and wih Jake's Channel Model with Doppler Showing Pilot Tracking with High Vehicle Velocities. | Ardalan | 2007 |
| 15 | plot_complex_mag.t | Read complex data in file, convert to magnitude and plot | 2007 | |
| 16 | QuardatureModulator.t | Quadrature Modukator Heierchical Block (HBlock) specify carrier frequency and sampling rate. Complex input buffer. Used by adjacent_channel.t | Ardalan | 1992 |
| 17 | signalFieldDecode.t | Hierarchical Block (HBlock) Signal Field Decode from Signal Field OFDM Symbol 48 Carriers | 2002 | |
| 18 | Tx80211aMR.t | Hierarchical Block (HBlock) All Rates Transmitter OFDM based on IEEE 802.11a | Ardalan | September 2002 |
| Item | Block Name | Description | Author | Date |
|---|---|---|---|---|
| 1 | byte2bit.s | Convert Byte Buffer to Bits Buffer | Ardalan | August 1, 2007 |
| 2 | carroffsetcorr.s | Function has a single complex input buffer, and output buffer. The parameter is the carrier offset normalized to the carrier spacing. | Ardalan | August 1, 2001 |
| 3 | carroffset.s | Function has a single complex input buffer, and output buffer. The parameter is the carrier offset frequency. | Ardalan | August 1, 2001 |
| 4 | convenca.s | Accepts 27 ints carrying 27*8 = 216 bits packed as b7 b6 b5 b4 b3 b2 b1 b0. Bits are repacked to form 7 unsigned long int ~~ 32 bit words. Call to function convenc.c produces the same number of output bits packed the same way. Within the block function output bits are repacked as 8 bits per int to form 27 output integers. prev6 contains 24 last bits from the previous input stream (only 6 bits are needed hence the name). It is initialized to 0 when convenc.c is first called. | Ardalan | 2001 |
| 5 | convencm.s | Convolutional encoder | Ardalan | August 15, 2001 |
| 6 | cxaddnoise54new.s | This block adds white gaussian noise to the input complex data stream.Sep 16, 2002 Bin Huang, add the calculated noise deviation in terms of SNR(Eb/No). | James Faber | November, 1987 |
| 7 | cxagc.s | This block adjust the gain to achieve a target rms value. | Ardalan | Sept. 1, 2006 |
| 8 | cxautocorr.s | Autocorrelation Complex | Ardalan | 2007 |
| 9 | cxavgchannel.s | Input the two LTS and average them in the time domain. Then output. | Ardalan | September 11, 2006 |
| 10 | cxchmodel.s | This block models a fading channel. Given the sampling rate in MHz and the delay spread in ns the complex impulse response is computed and convolved with the complex input stream. Reference: Evaluating the Performance of HRb Proposals in the Presence of Multipath Steve Halford, Karen Halford, and Mark Webster ** Intersil Corporation September, 2000 Submission to IEEE doc.: IEEE 802.11-00/282r2 Matlab coded converted to C | Ardalan | August 28, 2006 |
| 11 | cxcorr.s | This block correlates the input samples with the sequence given in a file. | Adali Tulay | September 23, 1988 |
| 12 | cxdc.s | This block adds DC to the incoming complex data stream by the complex parameter coefficient, and outputs the resulting complex data values. Auto fanout is supported. | Ardalan | 2007 |
| 13 | cxequalizezf.s | Equalize channel using zero forcing. Get channel estimates from buffer 0. | Ardalan | January 2007 |
| 14 | cxgphase.s | Multiplies input complex samples by amplitude*exp(j*phase) | Ardalan | 2007 |
| 15 | cxinterf.s | Add sinusoidal interference to complex stream. | Ardalan | October 18, 2006 |
| 16 | cxstats.s | This block calculates the statistics of the incoming signal. The parameter is a filename for storage of the results.TCL variables are created. | Prayson W. Pate | December 8, 1987 |
| 17 | cxwstat.s | This block calculates the statistics of the incoming signal. The parameter is a filename for storage of the results.TCL variables are created. | Prayson W. Pate | December 8, 1987 |
| 18 | dataField11a.s | Will create data field which includes 16 service bits, data bits(PSDU) 6 tail bits and Pad bits. In this function, the number of OFDM symbols and the number of Pad bits will be decided. The number of OFDM symbols will be sent to Signal packet. | Bin Huang | 10/4/2001 |
| 19 | dataField.s | Will create data field which includes 16 service bits, data bits(PSDU) 6 tail bits and Pad bits. In this function, the number of OFDM symbols and the number of Pad bits will be decided. The number of OFDM symbols will be sent to Signal packet. | Bin Huang | 10/4/2001 |
| 20 | dataSelectioncx.s | To select meaningful 48 data out of 64 point FFT result. | Bin Huang | 8/14/2002 |
| 21 | dataSelection.s | To select meaningful 48 data out of 64 point FFT result. | Bin Huang | 8/14/2002 |
| 22 | deinterleaver.s | Fixed point 802.11a Deinterleaver for rate=48,54 Mbps, total number of input bits is 288*n. n is the number per soft bit and it can be any of 3, 4,5,6,7,8. In order to implement this code easily, we put each soft bit into an integer. | Bin Huang | 8/24/2001 |
| 23 | demapper11a.s | Demmapper | Bin Huang | 6/20/2002 |
| 24 | depuncture.s | Inserts "weak" zeros | Bin Huang | 6/20/2002 |
| 25 | descrambler.s | descramble | Anisim Silivra | Oct. 12, 2001 |
| 26 | estsplit.s | This block is used to split an input buffer into two output buffers. The first output buffer outputs 64 complex zeroes at the beginning. It then forwards the input buffer samples to the output buffer. The second output buffer outputs 64 complex samples followed by all zeroes for each input sample. The first 64 samples represent an all pass channel estimate ( H(jw)=1). Thre purpose of the start is to create the inputs that the fftofdm star needs as it exepcts that for each packet, the channel estimation correction factors are present on the second input buffers. The fisrs input buffer has the OFDMs symbols follwing 64 zeroes. Again the purpose of this star is to use fftofdm in cases where we are not doing carrier freq. offset correction or channel estimation which provide these signal | Ardalan | September 27, 2001 |
| 27 | evm.s | This routine will produce a evm plot of the two input channels. channels. Optionally, the input channel data can 'flow through' to the correspondingly numbered output channel. This is useful if this block is to be placed in line in a simulation (e.g. probe). | Ardalan | 8/16/1987 |
| 28 | fft54.s | This block produces the FFT of the input signal. The points are output as complex numbers. The number of points input is npts and the number of points generated is npts.Bin Huang, added pilot tracking block | Ardalan/Bin Huang | July 26, 2001 |
| 29 | forksignalfield.s | Given stream with 48 carriers per symbols select the first 48 as signal field and fork it out passing through remaining symbols. | Bin Huang | 8/14/2002 |
| 30 | FQAM_64map.s | Floating point 802.11a 64-QAM mapping function, total number of input bits is 288. The output should be 48 complex symbols. | Bin Huang | 8/21/2001 |
| 31 | framesynch.s | This block correlates the input samples with the sequence given in a file. | Ardalan | October 2002 |
| 32 | freqoffsetcorr54.s | This block produces the FFT of the long term training samples based on the moditoring of the frame synch input. It then computes the frequency offset using the ML estimate. It then corrects the frequency offset in the input complex stream prior to output for FFT processing. | Ardalan | July 26, 2001 |
| 33 | fwfft.s | Forward FFT | Ardalan | 2001 |
| 34 | ifft_ofdm11a.s | Includes the training sequences, cyclic extention and windowing. | Bin Huang | 7/22/2002 |
| 35 | ifftofdm54.s | Includes the training sequences, cyclic extention and windowing. | Bin Huang/Ardalan | 7/22/2002 |
| 36 | interlr.s | Fixed point 802.11a interleaver all rates. | Bin Huang | 10/6/2001 |
| 37 | invfft.s | Inverse FFT | Ardalan | Dec. 2001 |
| 38 | jkfadenew.s | This blok models multipath fading channels for mobile radio applications. | Ardalan | Dec. 27, 1990 |
| 39 | jkfade.s | This blok models multipath fading channels for mobile radio applications. | Ardalan | Dec. 27, 1990 |
| 40 | longsynchss.s | Long symbol synchronization for steady state (no acquisition). | Ardalan | September 2, 2006 |
| 41 | mapper11a.s | Floating point 802.11a QAM mapping function, total number of input bits depends on rate. The output should be 48 complex symbols. | Bin Huang | 8/21/2001 |
| 42 | noisedev.s | Calulate the noise variance in term of Eb/No. | Bin Huang | 11/9/2001 |
| 43 | packstats.s | Calculate statistics of packet (also AGC) store in a file | Ardalan | 2007 |
| 44 | parseSignalField.s | Parse 802.11a signal field and provide rate, length and parity as TCL variables. | Bin Huang | 8/5/2002 |
| 45 | pilot.s | Perform pilot insertion. | Bin Huang | 6/11/2002 |
| 46 | pkdet64x.s | OFDM packet detection (IEEE 802.11a) | Ardalan | September 1, 2006 |
| 47 | pkdetcarroffset.s | Packet Detection Autocorrelation and Coarse Carrier Offset | Ardalan | March 2003 Revised 2007 |
| 48 | puncture11a.s | Puncture IEEE 802.11a puncturing function used to come up the coding rate with 1/2,2/3 and 3/4. | Bin Huang | 10/11/2001 |
| 49 | puncture.s | Puncturing function used to come up the coding rate with 1/2,2/3 and 3/4. | Bin Huang | 10/11/2001 |
| 50 | qam64cx.s | QAM64 OFDM Soft and Hard Option | Bin Huang | 2002 |
| 51 | qammapper.s | This block inputs data and ouputs the coordinates based on the 64-QAM encoding defined in 802.11a. It produces an in phase and quadrature component. Not very efficient but illustrative. | Ardalan | July 26, 2001 |
| 52 | rdscrfile.s | This function performs the simple task of reading sample values in from a file, and then placing them on it's output buffer. The file may have multiple sample values per line, which can be integer or float. An example use for this routine is to access a stored waveform as input to a simulation. | R. T. Wietelmann / D.G.Messerschmitt | June 5, 1982 |
| 53 | rmcpfx.s | Cyclic Prefix Removal | Ardalan | Dec. 12, 2001 |
| 54 | scramble11a.s | Sasan Ardalan, August 3,2007 Hacked to make it work on a bit at a time | Anisim Silivra | Oct. 5, 2001 |
| 55 | scramble.s | scramble_216 | Anisim Silivra | Oct. 5, 2001 |
| 56 | signalDecod.s | SIGNAL field decoding on the receive and provides the data rate and length | Bin Huang | 8/5/2002 |
| 57 | skipshort.s | Skip Short Symbols Used in Steady State | Ardalan | 2007 |
| 58 | symbolenergy.s | Calulate the average symbol energy. Each sysmbol has 64 complex samples. | Bin Huang | 11/12/2001 |
| 59 | synchlong.s | For steady state no acquisition skips over short preamble and outputs long symbols for channel estimation and outputs OFDM symbols. | Ardalan | March 2007 |
| 60 | synchltscfc.s | Long Symbole Fine Carrier Offset Estimation | Ardalan | September 2, 2006 |
| 61 | test.s | Desciption | Author | Date |
| 62 | trackpilots.s | Modified 9/26/2002, Bin Huang, revised the pilot tracking and made it even more robust. Algorithm basically derived based on known pilots. So this algorithm works associated with pilot.s on the transmit. Also, this algorithm can deal with up to 35m/s mobile velocity. Modified 10/29/2002, Bin Huang, added amplide adjustment(ampAdjust) in order to adjust signal amplitude changes due to doppler shift. This adjustment helps this PHY work up to 82mi/hr(37m/s). | Bin Huang | 9/26/2002 |
| 63 | vitdec_pars.s | Viterbi Decoder | Ardalan | August 15, 2001 |
| 64 | zdummy.s | Dummy Block |
| Item | Block Name | Description | Author | Date |
|---|---|---|---|---|
| 1 | ComputeFreqOffset.c | Compute Frequency Offset | Adalan | 2002 |
| 2 | convenc.c | Convolutional Encoder | Bin Huang | 2001 |
| 3 | convolution.c | industial standard convolution encoding with coding rate of 1/2 and constraint length of 7. The generator polynomials are (177,133). | Bin Huang | 6/1/2002 |
| 4 | conv_punc.c | interface between convolutional coding and puncturing, the output of conv. coding are two packed 32-bit integers, the inputs to puncturing should be a data array and each element only contains value 0 or 1, so need a small function to covert packed 32-bit data into parse 0 or 1. | Bin Huang | i9/25/01 |
| 5 | dataselect.c | Perform data selection: data reorder, remove the zeros. | Bin Huang | 8/6/2002 |
| 6 | demap_bpsk.c | Perform soft decision BPSK demapping scheme. This routine is used in SIGNAL field. | Bin Huang | 8/11/2002 |
| 7 | demap.c | Perform both soft decision and hard decision modulation demapping with quantization on 64 QAM, 16QAM, 4 QAM and BPSK. | Bin Huang | 8/7/2002 |
| 8 | descramble.c | De-Scarmble | Anisimm | 2001 |
| 9 | dummy2.c | Dummy Subroutine | ||
| 10 | fbpsk.c | Floating point 802.11a BPSK mapping, total number of input bits is 48. The output should be 48 complex symbols(Q15) | Bin Huang | 7/19/2001 |
| 11 | fft2s.c | Perform both FFT and IFFT by setting FFTswitch 1 and -1 respectively. General input N = 2^n, n is any integer and n>=0 && n<=31. | Source Ardalan | 1992 |
| 12 | fftofdm54.c | Perform both FFT and IFFT by setting FFTswitch 1 and -1 respectively. general input N = 2^n, n is any integer and n>=0 && n<=31. | Ardalan | 2001 |
| 13 | fpilot_insertion.c | IEEE 802.11a pilot isertion, total number of input symbols is 48(complex). The output pattern should be like index: 0,1, 2,....,26, 0,0....0,-26, -25, -24,.....-1. This file doesn't include polarity of pilot subcarriers. | Bin Huang | 7/30/2001 |
| 14 | interbpsk_conver.c | Interface between interleaving coding and BPSK mapping. The output of interleaving coding are distributed in three 32-bit registers in for PEs. The inputs to BPSK mapping should be packed data. So the data conversion between interleaving and BPSK mapping is needed. | Bin Huang | 9/25/2001 |
| 15 | Interleave.c | This is straight out of the IEEE 802.11a standard Ncbps: number of bits in a coded bits per symbol, Nbpsc: number of coded bits per subcarrier | 2002 | |
| 16 | intrlvr_unit.c | Interleave Test Unit | Ayman | 2001 |
| 17 | myatan2.c | atan2 | Bin Huang | 2002 |
| 18 | mycos_32.c | Cos | 2002 | |
| 19 | mycos_32s.c | Cos | Bin Huang | 2002 |
| 20 | myfloor.c | Floor | Bin Huang | 2002 |
| 21 | myfmod.c | FMOD | Bin Huang | 2002 |
| 22 | mysin_32.c | Sin | Bin Huang | 2002 |
| 23 | polarity.c | Perform the polarity to the inserted pilots | Bin Huang | 7/31/2002 |
| 24 | punctureofdm.c | Perform puncturing in order to come up different coding rate after convolution encoding with coding rate of 1/2 and constraint length of 7. | Bin Huang | 6/31/2002 |
| 25 | scrambler_216.c | Scrambler | Anisim Silvara | 7/2/2001 |
| 26 | signalfield.c | Fixed point 802.11a signal field, total number of signal bits are 24. The input 'rate' should be 0(6),1(9),2(12),3(18),4(24),5(36),6(48),7(54) respectively. The input 'length' should be any number between 1--4095. | Bin Huang | 9/17/2001 |
| 27 | signalpacket.c | This signalPacket.c file is a control code for sending SIGNAL field through convolutional coding, puncturing, interleaving and BPSK mapping. | Bin Huang | 9/25/2001 |
| 28 | swap_p.c | Perform the data swap in order to come up with the IFFT input requirment | Bin Huang | 7/5/2002 |
| 29 | tables.c | Floating point 802.11a table file. | Bin Huang | 8/19/2001 |
| 30 | unpackdeinter6.c | Fixed point 802.11a Deinterleaver for rate=6,9 Mbps, total number of input bits is 48 in float format. | Bin Huang | 11/09/2001 |
1- Obtain the Capsim® Text Mode Kernel (CapsimTMK) for Linux from:
GitHub Capsim Text Mode Repository 
CapsimTMK is distributed with hundreds of blocks. To simulate OFDM systems the blocks in this repository have to be incorporated into Capsim® including subroutines.
This Repository contains the Topologies for IEEE 802.11a OFDM block diagram modeling including required blocks and subroutines.
Note: Follow the Getting Started Guidelines in the CapsimTMK Repository.
2- Once CapsimTMK is installed just run 'make' in this repository's main directory.
3- Then change to the directory '11a_tops' and run:
../capsim OFDM_11a_Acquisition_2.t
The block diagram for the OFDM_11a_Acquisition_2.t topology is shown here.
The following is the console report:
Welcome to Capsim Text Mode Kernel (CapsimTMK) (c)1989-2017 Silicon DSP Corporation This is free software; see the source for copying conditions. There is NO warranty; not even for MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. http://www.silicondsp.com Version 6.2 Running topology OFDM_11a_Acquisition_2.t model name: OFDM_11a_Acquisition_2 fs(MHz)=20.000000 delay spread (ns) =50.000000 Kmax=10 ANormIdeal=0.154303 nData=6048 nPad=26 nSym=28 SIGNAL FIELD: 25dcc , rate=c , length=2ee cxagc rmsVal=0.162479 gain=0.492370 pkdetcarroffset: preamble detected THETA= -99458.976562 carroffsetcorr0 using CFO Estimate:-99458.976562 THETA= -99724.640625 carroffsetcorr0 using CFO Estimate:-99724.640625 pkdetcarroffset: preamble detected THETA= -99841.218750 carroffsetcorr0 using CFO Estimate:-99841.218750 THETA= -99558.742188 carroffsetcorr0 using CFO Estimate:-99558.742188 pkdetcarroffset: preamble detected THETA= -99579.929688 carroffsetcorr0 using CFO Estimate:-99579.929688 THETA= -99459.906250 carroffsetcorr0 using CFO Estimate:-99459.906250 pkdetcarroffset: preamble detected THETA= -98679.234375 carroffsetcorr0 using CFO Estimate:-98679.234375 THETA= -99011.054688 carroffsetcorr0 using CFO Estimate:-99011.054688 pkdet64x found long at: 288 currentRMS= 0.078526 mag=0.520235 thresh*rms=0.314106 synchltscfc: IndexIn=288 OffsetAdj=63PEAK at index 0 sampleCount=288 synchltscfc: fine(rad)=0.014848 fine(freq)=738.483582 coarse=0.000000 TOTAL=-738.483582 Data rate(MHz): 54 Data length in bytes: 750 Signal Field Parity PASS plot created file: 64xcorr.tim plot created file: plot.tim scatter created file: PreEqualization.sct plot created file: Long.tim plot created file: Mag.tim scatter created file: Equalized.sct scatter created file: Track_Applied.sct scatter created file: Pilots.sct scatter created file: SigField.sct scatter created file: Data.sct evm created file: EVM.sct EVM=0.007907 EVM_DB=-21.019665 ecount: hits/samples = 0/6016 (ignore 0) BER = 0/6016 = 0
In the simulation the SNR was set to: 35 DB.
Note that the OFDM receiver correctly reports the rate: 54 Mbps.
Use the plotting tool provided with the CapsimTMK repository:
java -jar $CAPSIM/TOOLS/IIPPlot.jar 64xcorr.tim
java -jar $CAPSIM/TOOLS/IIPPlot.jar -scatter Track_Applied.sct
java -jar $CAPSIM/TOOLS/IIPPlot.jar -scatter Data.sct
There are multiple results you can plot.
Note the EVM in DB= -21.02
For IIPPlot and its User Guide see here.
With Linux you can send the plot application to run in the background to put plots side by side when you want to display multiple plots. Then bring them to the forgound and use Control C to exit.
To run the TCL scripts, for example per.tcl, use this command:
../capsim -t per.tcl
Note that per.tcl runs many simulations over all rates with different SNR and will take a while to run. So patience is recommended.
Linux Tux By Larry Ewing, Simon Budig, Garrett LeSage - https://isc.tamu.edu/~lewing/linux/, http://www.home.unix-ag.org/simon/penguin/, garrett/Tux on GitHub, CC0, Link
Silicon DSP Corporation
2002-2025