Example: As a illustration of how redundancy can be used to correct errors, assume the message bits are 101. A simple form of redundancy is just repeating the bits so that the transmitted coded symbol stream may be 111000111. Note that the introduction of redundancy requires increasing the number of binary symbols which must be transmitted. In this case, from 3 to 9. A code or method of introducing redundancy that transmits M coded symbols for each message bit is called a 1/M rate code. In this case, the rate is 1/3. Assume that errors occur in determining what the transmitted signal was resulting in the receiver thinking its received symbols are 101001001. By a majority decision rule the three symbols 101 are correctly assumed to have been caused by the transmission of a 1 bit. Similarly, the received symbols 001 are used to decide that a 0 bit was sent next. The next three symbols 001 would suggest that a 0 was then transmitted. This is an error since the third bit was assumed to be a 1. Hence, while the redundant coding helps correct errors it is not able to correct all errors which may occur.

Typically a convolutional code such as used in IS-95 is able to achieve bit error rates on the order of 1 in 10 thousand down to 1 in 100 thousand bits provided the symbol error rate is not greater than about 1 to 2 symbols out of 100.

USE YOUR BROWSER "BACK" BUTTON TO RETURN
© 1999 DR. ERNEST SIMO - SPACE2000