'AN ******************Error detection and
correction***************************
' Application Note *
' Getting started in Data Transmission and Error detection
with Correction *
' using BASCOM AVR *
'===============================================================================
' Encodes 21 -bit Data Word Into 32 -bit Extended Bch
32.21 Word *
' *
' This code is used in the POCSAG protocol specification
for pagers. *
' *
' In this specific case, there is no need to use the
Berlekamp-Massey *
' algorithm, since the error locator polynomial is of at
most degree 2. *
' Instead, we simply solve by hand two simultaneous
equations to give *
' the coefficients of the error locator polynomial in the
case of two *
' errors. In the case of one error, the location is given
by the first *
' syndrome. *
' *
' This program derivates from the original bch2.c, which
was written *
' to simulate the encoding/decoding of primitive binary
BCH codes. *
' Part of this program is adapted from a Reed-Solomon
encoder/decoder *
' program, 'rs.c',
to the binary case. *
' *
'-------------------------------------------------------------------------------
' Extended Code Bch 32.21 Correction Using Zeh
Functions *
' The Idea And The Original Algorithm Were Proposed By K.
Borisov *
' C Source Code(c)vlv *
'*******************************************************************************
' Ported to BASCOM AVR by M/S PACTINDIA on 27.04.2002 *
' Written by Srikanth Kamath; e-mail:
panther@vasnet.co.in *
' *
' (c) Copyright 2002 Panther Electronics ;
www.pactindia.com/bascom.htm *
' *
' COPYRIGHT NOTICE: This program is free for
non-commercial purposes. *
' You may implement this program for any non-commercial
application. You may *
' also implement this program for commercial purposes,
provided that you *
' obtain my written permission. Any modification of this
program is covered *
' by this copyright. *
' *
'===============================================================================
'*******************************************************************************
' The Bch 32.21 Codeword Is Treated As Long Int Data Size
is 21 bits *
' The Function Corrects Up To 2 Errors And Detects At
Least 3 Errors. *
' The Error Value Is Amount Of Errors(0 , 1 , 2). *
' If The Error Is Not Correctable , Error Value Is 0xff
And The Codeword *
' Is Not Changed. *
'*******************************************************************************
' The Encoding part is of size 700byte this should gointo the transmitting
uC *
' The Decoding part is of size 1900byte this should gointo
the Reciever uC *
' The Test part is the routine that you must change to in
the TX uC to encode *
' and similarly the Test part must change in the TR uC to
decode *
'===============================================================================
' BUGS: *
' Fixed on 5.5.2002, *
' The error are now correctly reported 0,1,2 or 255 * *
' 27.4.2002 The corrected_data is either +1 or -1 at times
from actual *
' encodded_data But this does not effect the decode or
error correction *
' Also the error no reported is not related to no of error
but detected *
'-------------------------------------------------------------------------------
' Warranty & Disclaimer: *
' *
' We do not guarantee that this module is free from
defects or bug's in normal *
' use/service, and that these module will perform to the
current specification *
' in accordance with, and subject to, the company’s
standard warranty which *
' is detailed in Panther Electronics Purchase order
Acknowledgement. *
' *
' We assumes no responsibility for the use of any circuits
/ code described in *
' this book / e-book/ manual, nor does the company assumes
responsibility for *
' the functioning of undescribed features or
parameters. *
' *
' In absence of a written agreement to the contrary,
Panther Electronics *
' assumes no liability with respect to the use of
semiconductors devices/codes *
' described and used in this module for product design or
infringements of *
' patents or copyrights of third parties. *
' *
' These are control modules /codes only and not tested
with for full *
' functionality with the controlled units and products.
These modules are not *
' authorised for use as critical components or systems and
the use as such *
' implies that the user bears all risk of such use. *
'*******************************************************************************
' Many Many Thanks to *
' Mark Albert's, Eric Baron, Udo, Jack Tidwel and all
other who have helped *
'==============================================================================*
' Change this to Suit you AVR
$regfile = "8515def.dat"
'$dbg
' Note this code requires HW10, SS32 and FRAME 43
Declare Function Encodebch(datas As Long) As Long
Declare Function Correctbch(encodedbchword As Long) As Long
'***************The test program
***********************************************
' This test is to Show how tHe BCH encoding of 21bit data
packet helps in
' error detection and correction of error in asyncrony
data transmission
' this BCH encoding and Decoding provides error correction
upto 2 error per
' packet of 21bit. Hence the chances of usefull and
purpose full transmission
' is garrented
Dim Inputdata As Long
' This is the data to be transmitted
Dim Temp As Byte , Cnt As Byte ' temp variables used
Dim Just_tmp As Long ' temp variables used
Dim Corrupted_data As Long
' This is the received data
Dim Decoded_data As Long
' this is the decoded data = inputdata
Dim Encoded_data As Long
' this is the corrected data
Dim Result As Byte
' This is the error no
' NOTE: If Result = &HFF the error could not be
corrected hence discard the
' Received data
Do
Main:
Inputdata = 0
' get some random
data in
For Cnt = 1 To 2
Shift Inputdata , Left , 8
Temp = Rnd(255)
Inputdata = Inputdata Xor Temp
Next
'** / / Truncate To 21 Bit
Inputdata = Inputdata And &H1FFFFFF
'** / / Encode To Bch 32.21
Encoded_data = Encodebch(inputdata)
'/ / Corrupt
Data(2 Random Errors)
Corrupted_data = Encoded_data
'Corrupted_data
is the received transmission data
Temp = Rnd(255)
Temp = Temp And &H1F
Set Corrupted_data.temp
Just_tmp = Corrupted_data
Temp = Rnd(255)
Temp = Temp And &H1F
Set Corrupted_data.temp
Print " Encoded Data with first error
" ; Just_tmp ; " with second Second error " ; Corrupted_data
'// Correct Data
'Call
Correctbch(corrupted_data)
Just_tmp = Correctbch(corrupted_data)
'// Decode Data
Print "Corrupted_data " ; Corrupted_data ; " EncodedWord " ; Encoded_data ; " Corrected_data " ; Just_tmp
Decoded_data = Just_tmp
' This is the the
required to get back the corrected data
Shift Decoded_data , Right , 11
' Can be shift
this to Correctbch itself, but kept it here to check and test
If Result <> &HFF Then
If Inputdata <>
Decoded_data Then
Print " Error check this"
Print "---------Srikanth
Kamath T___________**********"
Else
Print "Inputdata
" ; Inputdata ; " Decode Data " ; Decoded_data ; " No of Error Corrected " ; Result
Print "---------Decode
Sucessful___________**********"
End If
Else
Print "---------Decode
not possible___________**********" ; " No of Error more than 2: " ; Result
End If
'Goto Main
Loop
'***************Function Encoded the 21 bit InputData to
32 bit BCH_word********
' This Function Encoded the 21 bit InputData to 32 bit
BCH_word
Function Encodebch(datas As Long)
Local Tmp_enc As Long , Ci As Byte , Ltmp_enc As Long
Ltmp_enc = Datas
'/ / Divide By
Polynom
For Ci = 0 To 20
Shift Ltmp_enc , Left , 1
Tmp_enc = Ltmp_enc And &H200000
If Tmp_enc <> 0 Then
Ltmp_enc = Ltmp_enc Xor &H1B4800
End If
Next
'/ / Allign
The Remainder
Shift Ltmp_enc , Right , 10
Ltmp_enc = Ltmp_enc And &H7FE
' / / Add
Data
Tmp_enc = Datas
Shift Tmp_enc , Left , 11
Ltmp_enc = Ltmp_enc Or Tmp_enc
Encodebch = Ltmp_enc
' / /
Calculate Parity
Ci = 0
While Ltmp_enc <> 0
Tmp_enc = Ltmp_enc And 1
'debug
Ci = Ci Xor Low(tmp_enc)
' Tmp_enc2 = Tmp_enc2 Xor Tmp_enc
' Ci =
Low(tmp_enc2)
Shift Ltmp_enc , Right , 1
Wend
' / / Add
Parity
Encodebch = Encodebch Or Ci
End Function
'-------------------------------------------------------------------------------
'END OF*********Function Encoded the 21 bit InputData to
32 bit BCH_word********
'***************The decoding and error correction
start*************************
'*************************Correction
Itself*************************************
' Using the zeh function
Declare Function Makesyndrome(x As Long , Byval Polynom As Byte) As Byte
Declare Function Inverbit(bch_word As Long , Byval Bitno As Byte) As Long
Function Correctbch(encodedbchword As Long)
Local Sa As Byte , Sb As Byte , Parity As Byte , Dist As Byte , Tmplong As Long , Tmp As Byte , Ltmp As Long
Ltmp = Encodedbchword
' / /
Calculate Syndromes A And B
Sa = Makesyndrome(ltmp , &H05)
Sb = Makesyndrome(ltmp , &H1D)
' / /
Calculate Parity
Parity = 0
While Ltmp <> 0
Tmplong = Ltmp And 1
Tmp = Low(tmplong)
Parity = Parity Xor Tmp
Shift Ltmp , Right , 1
Wend
' / / Check
Syndromes
If Sa = 0 And Sb <> 0 Then
Result = &HFF
Correctbch = Encodedbchword
Exit Function
End If
If Sb = 0 And Sa <> 0 Then
Result = &HFF
Correctbch = Encodedbchword
Exit Function
End If
'// If Both
Syndromes Are 0(no Code Errors)
If Sa = 0 And Sb = 0 Then
If Parity = 0 Then
'// No Errors
Result = 0
Correctbch = Encodedbchword
Exit Function
End If
'debug
Correctbch = Encodedbchword Xor 1
Result = 1
Exit Function
End If
Sa = Lookup(sa , Table_sa)
Sb = Lookup(sb , Table_sb)
'// If Both
Syndromes Are Indicating One Error
If Sa = Sb Then
Correctbch = Inverbit(encodedbchword
, Sa)
' debug 1
' If(parity! = 0) Return 1 ; / / One Error
'Tmp = 0
'If Parity = 0 Then
' Set Tmp.0
'Else
' Reset Tmp.0
'End If
'--- debug 1
If Parity <> 0 Then
Result = 1
Exit Function
End If
' debug 1
Correctbch = Correctbch Xor 1
Result = 2
Exit Function
End If
'// More
Then Two Errors - Not Correctable
If Parity <> 0 Then
Result = &HFF
Correctbch = Encodedbchword
Exit Function
End If
Dist = Sa - Sb
'// Modulo
1f
Tmp = Dist And &H80
If Tmp <> 0 Then
Dist = Dist + &H1F
End If
'// Find
Distance Between Errors
Dist = Lookup(dist , Table_dist)
'// Not
Correctable
If Dist = &HFF Then
Result = &HFF
Correctbch = Encodedbchword
Exit Function
End If
'// Correct
1 -st Error Location
Tmp = Lookup(dist ,
Dist_correction_table)
Sa = Sa - Tmp
'// Modulo
31
Tmp = Sa And &H80
If Tmp <> 0 Then
Sa = Sa + &H1F
End If
'/ /
Correct 1 -st Error
Tmp = Sa + 1
Correctbch = Inverbit(encodedbchword
, Tmp)
'// Find 2
-nd Error Location
Sa = Sa + Dist
'// Modulo 31
If Sa >= 31 Then
Sa = Sa - 31
End If
'//Correct
2 -nd Error
Tmp = Sa + 1
Correctbch = Inverbit(correctbch , Tmp)
Result = 2
End Function
'-------------------------------------------------------------------------------
'END OF*******************Correction
Itself*************************************
'************************Divide By
Polynom**************************************
'Sub Makesyndrome(x As Long , Byval Polynom As Byte)
Function Makesyndrome(x As Long , Byval Polynom As Byte)
Local Tmp_syn As Long , Cntr As Byte , L_polynom As Long , Tmp_syn2 As Long
Tmp_syn = X And &HFFFFFFFE
L_polynom = Polynom
Shift L_polynom , Left , 27
' / /
Calculate The Syndrome
For Cntr = 0 To 30
Tmp_syn2 = Tmp_syn And &H80000000
If Tmp_syn2 <> 0 Then
Shift Tmp_syn , Left , 1
Tmp_syn = Tmp_syn Xor L_polynom
Else
Shift Tmp_syn , Left , 1
End If
Next
Shift Tmp_syn , Right , 27
Makesyndrome = Tmp_syn
End Function
'-------------------------------------------------------------------------------
'END OF******************Divide By
Polynom**************************************
'***************Invert Bit At The Position
Bitno********************************
'Sub Inverbit(bch_word As Long , Byval Bitno)
Function Inverbit(bch_word As Long , Byval Bitno As Byte)
Local Tmp_inver As Long
Tmp_inver = 0
Set Tmp_inver.bitno
Inverbit = Bch_word Xor Tmp_inver
End Function
'-------------------------------------------------------------------------------
'END OF*********Invert Bit At The Position
Bitno********************************
'****************Syndrome A Error Position
Table********************************
Table_sa:
Data &H00 , &H1B , &H1C , &H0E , &H1D , &H01 , &H0F , &H07,
Data &H1E , &H19 , &H02 , &H17 , &H10 , &H04 , &H08 , &H13,
Data &H1F , &H06 , &H1A , &H0D , &H03 , &H12 , &H18 , &H16,
Data &H11 , &H15 , &H05 , &H0C , &H09 , &H0A , &H14 , &H0B
Table_sb:
Data &H00 , &H1B , &H1C , &H10 , &H1D , &H05 , &H11 , &H02,
Data &H1E , &H16 , &H06 , &H0C , &H12 , &H14 , &H03 , &H19,
Data &H1F , &H0E , &H17 , &H0A , &H07 , &H08 , &H0D , &H09,
Data &H13 , &H18 , &H15 , &H0B , &H04 , &H01 , &H1A , &H0F
Table_dist:
Data &HFF , &HFF , &HFF , &H03 , &HFF , &HFF , &H06 , &H0B,
Data &HFF , &HFF , &HFF , &HFF , &H0C , &HFF , &H09 , &H08,
Data &HFF , &H0E , &HFF , &H0A , &HFF , &HFF , &HFF , &H04,
Data &H07 , &H05 , &HFF , &H02 , &H0D , &H01 , &H0F , &HFF
Dist_correction_table:
Data &HFF , &H13 , &H06 , &H1E , &H0B , &H03 , &H1C , &H17,
Data &H15 , &H11 , &H05 , &H14 , &H18 , &H0F , &H0E , &H19
