Array acoustic logging is one of the most important logging methods presently. However

due to its huge amounts of data and the speed limit of data transmission

logging efficiency is restricted. Data compression and processing is necessary to improve logging efficiency. By analyzing characters of acoustic signal

introduce a compression method comprised of DCT

proper quantization and arithmetic coding

further

a series of improvements are made to the method. By test on 660 group real logging data

it is proved that the method is effective

average compression ratio（ CR） is 4. 56 to the monopole acoustic data of 672 points and 9. 18 to dipole acoustic data to1200 points

with root-mean-square error around 2 %.