On the Challenge of Keeping ATLAS Tile Calorimeter Raw Data

The Tile Calorimeter (TileCal) for the ATLAS experiment at the CERN Large Hadron Collider (LHC) is currently taking data with proton-proton collisions. The TileCal front-end electronics allows to read out the signals produced by about 10000 channels measuring energies ranging from ~30 MeV to ~2 TeV. The read-out system was initially designed to reconstruct the data in real-time and to store for each channel the signal amplitude, time and quality factor at the required high rate. This approach implied discarding 80% of the pulse signals raw data. Practical experience operating in this scheme with increasing rate have led to several modifications and understanding that some kind of data compression is helpful during data processing and storing. It becomes clear that the increasing luminosity and energy of the LHC will pose more and more challenging conditions to the signal reconstruction. An alternate approach is to use online reconstruction for Level2 triggering only and to implement the data flow lossless compression scheme for further offline analysis. A new version of the lossless compression algorithm (2009) is proposed for the reconstruction and transmission of the TileCal raw data. It is based on the use of a "standard" pulse shape and allows sending both the raw data and reconstructed quantities required to Level 2 trigger during online processing. The algorithm does not increase the data flow as compared to the existing approach and shows much more stability in sent fragments sizes. The earlier version of the lossless compression algorithm did not require the knowledge of the exact pulse shape but it was able to fit within the bandwidth limits only and could not fit within the tight time constraints imposed by the ATLAS first level trigger rate (100 kHz). Fitting within these constraints remained since then a challenging problem. After a short overview of the current approach we will describe the lossless compression algorithm as a possible upgrade and highlight some details of the implementation. We will report on testing and validation and on the overall performance measured on high rate tests, calibration and √s = 7 TeV proton-proton collisions runs.