Abstract: As a key factor for assessing data trustworthiness, provenance records the whole
transmission procedure of data package from generation to base station in wireless sensor
network (WSN). The provenance expands rapidly with the increasing of package transmission
path length, and it is common to take provenance compression algorithm to save network
energy consumption and communication bandwidth. The data model of provenance was
formalized, and the classification comparison method was adopted to analyze the lossy
compression methods based on Bloom filter and probabilistic packet marking. Lossless
compression methods based on arithmetic coding and dictionary were compared. The encoding
and decoding algorithms of each method were listed out, and the corresponding merit and
demerit were also summarized. The future research prospects of provenance were discussed.