The Huffman coding implementation with 32 bit data words 

Context in source publication

Context 1

... no zero suppression is foreseen in the VFE-FE section, the amount of data to be trans- mitted will be large. Simulation studies show that the probability of an event with more than 6 bits is below 2.4 · 10 −4 . The data rate will thus be significantly reduced by performing a lossless com- pression. The stream of selected data samples is therefore encoded in variable length words with an Huffman algorithm. This algorithm can provide a significant data reduction and thus decrease the number of optical fibers required for data output. A simplified schematic of the LiTE-DTU ASIC is shown in Figure 3. A possible implementation of Huffman coding with a data format based on 32 bit words is shown in Figure 4. Samples with less than 7 bits, so-called "baseline samples", are packed in groups of 5 while the 13-bit samples, named "signal samples", are packed in couples. In either case there are other data formats for the incomplete baseline quintet or signal ...