Presentation is loading. Please wait.

Presentation is loading. Please wait.

HLT Collaboration (28-Jun-15) 1 High Level Trigger L0 L1 L2 HLT Dieter Roehrich UiB Trigger Accept/reject events Select Select regions of interest within.

Similar presentations


Presentation on theme: "HLT Collaboration (28-Jun-15) 1 High Level Trigger L0 L1 L2 HLT Dieter Roehrich UiB Trigger Accept/reject events Select Select regions of interest within."— Presentation transcript:

1 HLT Collaboration (28-Jun-15) 1 High Level Trigger L0 L1 L2 HLT Dieter Roehrich UiB Trigger Accept/reject events Select Select regions of interest within an event Compress Reduce the amount of data required to encode the event as far as possible without loosing physics information Provide HLT-ESDs for online monitoring Access to the results of the event reconstruction Physics Requirements

2 HLT Collaboration (28-Jun-15) 2 Physics Applications Quarkonium spectroscopy Dielectrons Dimuons Open Charm Jets Pileup removal in pp Detectors DAQHLT Mass storage

3 HLT Collaboration (28-Jun-15) 3 Quarkonium Dielectrons –HLT task Reject fake TRD triggers and reduce trigger rate by factor of more than 10 –Status Fast TPC pattern recognition – done Additional PID by dE/dx – done Adaption of Kalman filter for HLT – done Combined track fit TRD-TPC-ITS – in progess –To do Emulate the TRD Global Tracking Unit (TRD tracklet merging and PID) Dimuons –HLT task –Utilizing tracking chamber information and improving momentum resolution –Sharpening of pt-cut –Rejection factors: low pt-cut: 5, high pt-cut: 100 –Status –Complete simulation including cluster finder – done –Full scale prototype HLT farm (UCT) – done –FPGA cluster finder – in progress –FPGA interface – in progress T. Vik, PhD thesis, Oslo, 2005

4 HLT Collaboration (28-Jun-15) 4 Open charm HLT task –Detection of hadronic charm decays: D 0  K – +  + –About 1 D 0 per event (central Pb-Pb) in ALICE acceptance –After cuts »signal/event = 0.001 »background/event = 0.01 Status –Detailed study of timing profile of offline algorithm - done –Adaption of ITS tracking to HLT and speed-up – done –Optimization of D 0 finder – in progress –Combine HLT tracking and D 0 algorithm – in progress To do –estimate the efficiency for appling D 0 -offline-cuts online –extend study to D +, D * +

5 HLT Collaboration (28-Jun-15) 5 Online Available modules TPC cluster finder (CF) TPC track follower (TF) Kalman fitter TPC Hough transform tracker (1) TPC Hough transform tracker (2) TPC cluster deconvolution TPC performance monitor TPC dE/dx TPC data compression (1) TPC data compression (2) ITS tracker Dimuon cluster finder Dimuon tracker Jet cone finder D0 finder PHOS pulse shape analysis

6 HLT Collaboration (28-Jun-15) 6 Tracking performance for CF/TF Tracking efficiency Momentum resolution Computing time: 13 sec per event (dn/dy=4000) on a 1kSPECInt machine A. Vestbø, PhD thesis, Bergen, 2004

7 HLT Collaboration (28-Jun-15) 7 Integral efficiency for CF/TF Integral tracking efficiency Contamination of fake tracks

8 HLT Collaboration (28-Jun-15) 8 Tracking performance for Hough transform – version 1 Gray-scale Hough transform –Image space: raw ADC counts –Transform space: circle parameters –Histogram increment: charge too CPU-time consuming A. Vestbø, PhD thesis, Bergen, 2004

9 HLT Collaboration (28-Jun-15) 9 Tracking performance for Hough transform – version 2 (1) Linearized prehistoric Hough transform –Image space: conformal mapped cluster boundaries –Transform space: straight line parameters –Histogram increment: history of missing padrows, conditional  slice of TPC sector Corresponding Hough Space Collaboration with the Offline group: Cvetan Cheshkov

10 HLT Collaboration (28-Jun-15) 10 Tracking performance for Hough transform – version 2 (2) Cvetan Cheshkov Tracking efficiency dN/dy=8000 dN/dy=6000 dN/dy=4000 dN/dy=2000 B=0.5T

11 HLT Collaboration (28-Jun-15) 11 Tracking performance for Hough transform – version 2 (3) Momentum resolution –  Pt/Pt=(1.8xPt+1.0)% (B=0.5T) –  (  )=6.1mrad –  (  )=5.5x10 -3 Computing time (1.3 kSpecInt machine) Cvetan Cheshkov dN/dy~02000400060008000 LUT Init120ms Hough Transform 0.7s (3ms/patch) 3.3s (15ms/patch) 5.9s (27ms/patch) 8.7s (40ms/patch) 11.3s (53ms/patch)

12 HLT Collaboration (28-Jun-15) 12 ITS tracking (1) Offline tracking –Modified offline code –Speed-up of up to a factor of 30 for some modules ITS Clusterer clusters HLT TPC Tracker ITS Vertexer TPC tracks ITS Tracker J. Belikov, C.Cheshkov

13 HLT Collaboration (28-Jun-15) 13 ITS tracking (2) J. Belikov, C.Cheshkov Tracking efficiency TPC only (HT) ITS+TPC Fakes B=0.5T Comparable to offline

14 HLT Collaboration (28-Jun-15) 14 ITS tracking (3) J. Belikov, C.Cheshkov Impact parameter resolution Dominated by SPD -> ”offline” quality, i.e. 1 GeV/c track: transverse impact parameter resolution = 60 microns

15 HLT Collaboration (28-Jun-15) 15 ITS tracking (4) J. Belikov, C.Cheshkov Computing time (1.3 kSPECInt PC) dN/dyClustererVertexerTracker ~00.5s20ms0.15s 20001.3s45ms0.45s 40001.5s85ms0.95s 60001.75s150ms1.70s 80002.0s210ms2.70s

16 HLT Collaboration (28-Jun-15) 16 D 0 finder Offline algorithm –Cut on impact parameter –calculate »Distance of closest approach »Invariant mass »Decay angle »Pointing angle Timing results (0.3 kSPECInt PC) dN/dy10002000400060008000 CPU time [sec]0.41.461123

17 HLT Collaboration (28-Jun-15) 17 TPC Data Compression - Principle Data model adapted to TPC tracking Store (small) deviations from a model: (A. Vestbø et. al., to be publ. In Nucl. Instr. Meth. ) Cluster model depends on track parameters Standard loss(less) algorithms; entropy encoders, vector quantization... - achieve compression factor ~ 2 (J. Berger et. al., Nucl. Instr. Meth. A489 (2002) 406) Tracking efficiency before and after comp. Relative pt-resolution before and after comp. dN ch /d  =1000 Tracking efficiency Relative pt resolution [%]

18 HLT Collaboration (28-Jun-15) 18 Towards larger multiplicities cluster fitting and deconvolution: fitting of n two-dimensional response functions (e.g. Gauss-distributions) analyzing the remnant and keeping ”good” clusters arithmetic coding of pad and time information TPC Data Compression - Implementation Compressed tracks/clusters Leftovers

19 HLT Collaboration (28-Jun-15) 19 TPC Data Compression - Results Achieved compression ratios and corresponding efficiencies Compression factor: 10

20 HLT Collaboration (28-Jun-15) 20 PHOS Data Compression Data volume –18k crystals –Occupancy: ~10% (min. bias Pb+Pb, E > 10 MeV) –10 MHz sampling frequency –128 samples per pulse –2 channels per crystal –10 bits per sample Readout –all channels: 6 Mbyte/event –discard empty channels (after zero-suppresion): 0.6 Mbyte/event Date rate –2 kHz ’clean’ Pb+Pb interaction rate: 1.2 GByte/sec

21 HLT Collaboration (28-Jun-15) 21 PHOS Data Compression Online pulse shape analysis –Fit amplitude -> energy –Fit time offset -> TOF »Peak method »Slope method Gamma-2 fit Peak Method : Offline time reference at peak ( y’ =0 ) Slope Method:Offline time reference at max. slope ( y”=0 ) (both reference points are amplitude independent)

22 HLT Collaboration (28-Jun-15) 22 TPC –event reconstruction »primary vertex »primary vertex tracks »secondary vertex tracks »ghost (non-vertex) tracks ITS –SPD and SSD tracking TRD, PHOS,... Full event reconstruction –data compression –pile-up rejection HLT task in pp

23 HLT Collaboration (28-Jun-15) 23 Pattern recognition scenario in pp TPC tracking strategy Cluster finder Track follower (conformal mapping method) First pass with vertex constraint Second pass in order to improve efficiencies for low-pt and secondary tracks input all unassigned clusters from the first pass no vertex constrain is imposed on the track follower (conformal mapping done with respect to the first associated cluster on track) Kalman filter for track extension into TRD and ITS PID in TRD and TPC


Download ppt "HLT Collaboration (28-Jun-15) 1 High Level Trigger L0 L1 L2 HLT Dieter Roehrich UiB Trigger Accept/reject events Select Select regions of interest within."

Similar presentations


Ads by Google