10.5 Interleaving

To be able to correct longer burst errors an interleaver is placed between the two coders. This interleaver does not add new, redundant data but rearranges the data in a way the bits which have been close before the interleaver now are separated by I (interleaving depth).
Therefore e.g. a block interleaver reads the data in rows in a memory block and reads them out by columns. The number of columns gives the interleaving depth I. The delay caused by the interleaving is given by the number of the rows of the matrix. With this method incorrect bits of larger burst errors are separated and therefore can be corrected by the inner code. A disadvantage is the grouping of periodical errors resulting in larger burst errors after the interleaver. These cannot be corrected then.
In the upper line the received symbols without interleaving are given as numbers 1 to 32 in ascending order. Th interleaver writes 8 numbers per row in 8 row up to number 64. Now the numvers are read out by columns leading to the series 1, 9, 17, 25, 33, 41, 49, 57, 2, 10, 18 26, 34, 42, 50, 58, and so on. In the receiver once again the receives series is written in lines of eight. When you now read out columns, the original series in ascending order appers again. This helps to prevent the final data from burst errors. If in the original series numbers 5 to 10 would have been affected, using interleaving puts 33, 41, 49, 57, 2, 10 at the same position, so the errors are not occuring in one block but are spread about the data and can be easily corrected.
Figure 10-15: Priniple of the interleaver.
Ha alcanzado el 7% de esta lección
7%