oy to py dS» per frame. For bse levels are ery 3 minutes. goa 2 = 4223 eticer af Cedice ‘.. = ines = 2.4 INTRODUCTION Error control coding, channel coding is used to reduce the over all probability of error. The functional block that accomplish error control coding are channel encoder & channel decoder. The channel encoder systematically adds digits to the transmitted message digits. ‘These additional digits while conveying no information theory solve but make it possible for channel decoder to detect and correct errors in information bearing digits. Error detection and correction reduces overall probability of error. We can say that channel encoder adds or put back enough redundancy to lower the overall probability of error to the desired value. &) Sosaaal: Bit rate = r,= nn/k k sa nk epee oer ae = a decoder es] Fig. 24 : Error Control Coding ‘Two type of error probability: 1. Channel bit error probability ra, =a] 4c22 Information Theory & Coding 2. Message bit error probability. R = Pf = 6] WORKING: 1. Figure 2.1 shows digital communication system for transmitting binary input [by] of source encoder over a noicy channel at the rate of n, bits/sec. 2. Due to noise of channel bit stream {,} recovered by the receiver differs from transmitted sequence ~. error probability of message P(x =b,) 3. As we know that actual data transmission over the channel is accomplished by the modulator and demodulator that operate at data rates r, ranging from say 1 <% 5% the probability of error q_ for modem is represented by P(d, + d,) will depend on, data rate over channel. 4. Thechannel encoder & decoder an functional blocks in the system that by acting together reduce overall probability of error. The encoder divides the input message bits into block k message bits and n - k chick bits that combine to form n - bit codeword. 5. ‘The decoder looks at received version of codeword C,,, which may occassionally contain errors, and attempt to decode k - message bits. The check bits enables the decoder to detect & correct the errors <. lowers the probability of error. 2.2 EXAMPLE OF ERROR CONTROL CODING Let us consider that we want to transmit data over a telephone line that has a usuable band width of 3000 Hz. & maximum S/N at output of 13 dB at rates of 1200 bits/sec. with: probability of error less than 10+. We are given a modem that operate at speeds of 1200, 2400. or 3600 bits/sec. with error probabilities 2 (10), 4 (104) & 8 (104) respectively. We have to design all error control coding scheme that would yield an probability of error < 10+, +, We have , = 1200, C = 13,000, r, = 1200, 2400, 3600, q, (r,) = 2x 104, 4 (104), 8 (10) respectively since when r, < C according to shannon’s theory we should be able to. transmit data with an arbitrarily small problem of error. Now consider the error control coding scheme for this wherein the triplets 000 & 111 are transmitted when b, = 0 and 1 respectively. These triplets are codewords. Data comes out from encoder at rate of 3600 bits/sec. & at this data rate modem has error probability (8 x 10%). Recieved triplets 000 [oor [010 | 100 Jor | 101 | 110 Output Messagebith, | 0 | 0 | 0 | o | 1 1 i Recieved Triplets Ete outputmessagebitb, | 1 reduce the Important: 1, Itis stream 2, 3. ItisIntroduction of Coding 23 With error control coding the receives O/P bit is different from i/p message bit only if two or more of bits in input are in error due to channel noise. P. =P (b,+b) = P (two or more bits in triplets are in error). 9 1= 40) +3, = 342-292 With modem bit error probability q_ of 8 x 10 with at r, = 3600 bits/sec we have. Pe = (191.9)10-% = (0.01919) x 10 which is less than, 10-4 Thus error control coding reduce the overall probability of error. Important Aspects It is possible to correct & detect errors by adding extra bits called check bits to message stream. It is not possible to detect & correct all errors. Addition of check bits reduces the effective data rate through channel. [pee = a] of coding scheme METHODS OF CONTROLLING ERRORS Forward error correction method : - Control errors at the receiver through attempt to correct noise induced errors. Error detection :- Here decoder, upon examining the demodulator output accepts the received sequence, ifit matches a valid message sequenceand it not, the decoder discards the received sequence & notifies the transmitter that errors have occured & that received ‘message must be retransmitted 2. Decoder attempts to detect errors but not attempt to correct them. Error detection scheme yeilds a lower probability of error than error correction scheme. To justify this statement, let us take one example. If decoder in the preceding example used error detection only then it would only accept 000 & 111 and rejects all other triplets & request retransmission. Now an information bits will be uncorrectly decoded at receive only when all three bits in received code are in error thus Py = (qc)' = 5:12 x 10°10 which is lower than probability of error for error correction and detection seheme and slow down the effective rate of data transmission since