A track finding algorithm based on A track finding algorithm based on pixel detector for the ATLAS pixel detector for the ATLAS second level trigger second level trigger Andrea Bar at ella – Mauro Dameri - Paolo Moret t ini – Fabrizio Parodi I NFN Genova Out line • ATLAS t rigger st rat egy f or t he B-physics. • PixTrig: a pixel based Level 2 t rack f inding algorit hm. • Ef f iciency and perf ormances. • Applicat ion t o t he select ion of rare B decays. • Applicat ion t o t he b-t agging in mult i-b j et s event s. • Conclusions Pixel2000 - Genova - 6/6/00 1 P. Morettini
The ATLAS Level 2 Trigger The ATLAS Level 2 Trigger The ATLAS second level t rigger ref ines t he decision Net work swit ch R/ O t aken at Level 1 using t he Buf f ers dat a f rom t he dif f erent sub- det ect ors, including t he Commercial t rackers. R/ O Processors Buf f ers I t has t o t ake it s decision in 10-20 ms, reducing t he overall rat e by a f act or of 10 R/ O Buf f ers (f rom 100 KHz t o 10 KHz). The implement at ion of t his t rigger will be based on a f arm of commercial processors connect ed t o t he R/ O buf f ers t rough a swit ch. Hardware co-processors, bot h at buf f er and f arm level, are also under st udy. Pixel2000 - Genova - 6/6/00 2 P. Morettini
LVL2 Processing Modes LVL2 Processing Modes The ATLAS t rigger is most ly based on “Regions of I nt erest ” (RoI ). This means t hat every t rigger component analyses only t he port ions of t he det ect ors ident if ied by t he previous component s; in t he case of t he LVL2, only t he region where a muon or a j et where f ound at LVL1 are reconst ruct ed. I n t he case of t he rare b decays reconst ruct ion however, t he int erest ing j et is usually not energet ic enough t o be seen at LVL1, so t he event is t riggered by a high p T µ in t he second j et . I n t his case a “Full Scan” is necessary at LVL2 t o reconst ruct t he charged t racks in t he whole inner t racker, and select t he vert exes using invariant mass and impact paramet er inf ormat ion. Pixel2000 - Genova - 6/6/00 3 P. Morettini
PixTrig: The Basic Idea : The Basic Idea PixTrig PixTrig is a t rack f inding algorit hm designed t o be used in Full Scan and RoI guided mode. Thanks t o t he high resolut ion and low noise of t he pixel det ect or it can provide f ull 3d t rack reconst ruct ion and impact paramet er measurement t o be used in st and-alone mode or as seed f or ot her algorit hms. PixTrig is a pure combinat orial algorit hm using t he ATLAS pixel det ect or space point s: combinat ions in t he t wo innermost layers are ext rapolat ed t o t he t hird if point ing t o t he primary vert ex. I f a point close t o t he ext rapolat ion is f ound in layer 3, t he ∆ φ ∆ Z 3 R 3 point s are kept as a candidat e t rack. Overlaps are solved on t he basis of t he residuals in t he t hird ∆ φ ∆ Z e xt layer. R R o I e xt Pixel2000 - Genova - 6/6/00 4 P. Morettini
Logical Layers Definition Logical Layers Definition The original implement at ion of PixTrig was based on t he t hree phisical layers of t he barrel pixel det ect or. To adapt t he algorit hm t o t he end- caps geomet ry (where we have f ive Link L1-L2 Link L1-L3 disks) we moved, in t he present C++ implement at ion wit hin t he ATLAS t rigger simulat ion, t o t he def init ion of “logical layers”. A logical layer is an arbit rary collect ion of det ect or modules. Logical f irst layers (every module in t he barrel B-layer of t he det ect or) have a link t o t he modules in t he corresponding second and t hird logical layers. This approach increase t he f lexibilit y of t he algorit hm and it s robust ness t o t he pixel inef f iciencies. However, we have t o keep in mind t hat More complexit y in t he Longer processing t ime → → → → → → → → → → → → → → → → More combinat ions Logical Layers and more f ake t racks Pixel2000 - Genova - 6/6/00 5 P. Morettini
Reconstruction Efficiency in jets Reconstruction Efficiency in jets The reconst ruct ion ef f iciency The reconst ruct ion ef f iciency B- B -jets jets – – No pile No pile- -up up in j et RoI in j et RoI is close t o is close t o 90% 90% and and Efficiency Fake Trks Fraction η and 1 1 over η f lat over f lat and p p T T . . 0.8 0.8 At low luminosit y low luminosit y t he f ract ion t he f ract ion At 0.6 0.6 of f ake t racks f ake t racks (combinat ion of (combinat ion of of 0.4 0.4 point s coming f rom dif f erent point s coming f rom dif f erent 0.2 0.2 part icles) is always below 20% below 20% part icles) is always 0 0 -2 0 2 -2 0 2 (it ’s higher in t he end- -caps caps (it ’s higher in t he end η η because more combinat ions are because more combinat ions are B- B -jets jets – – 24 24 ev ev. pile . pile- -up up possible t her e). possible t her e). Efficiency Fake Trks Fraction 1 1 At design luminosit y t he At design luminosit y t he 0.8 0.8 f ract ion of f akes reaches 60% 60% f ract ion of f akes reaches 0.6 0.6 in t he end- -caps caps (spurious in t he end (spurious 0.4 0.4 combinat ions lying on a st raight combinat ions lying on a st raight 0.2 0.2 line can be pref err ed t o good line can be pref err ed t o good 0 0 -2 0 2 -2 0 2 t racks). t racks). η η Pixel2000 - Genova - 6/6/00 6 P. Morettini
Timing (jet RoI RoI) ) Timing (jet The processing t ime on a B- -jets jets – – No pile No pile- -up up B commercial Pent ium I I I 20 Events Processing time (ms) ID 800 0.6536 / 11 60 Entries 679 P1 0.0000E+00 500 MHz syst em is of 3.7 Mean 3.656 P2 0.6945E-01 RMS 3.113 15 P3 0.7165E-03 ms per j et . UDFLW 0.0000E+00 P4 0.6402E-04 OVFLW 3.000 40 10 As expect ed, t his t ime 20 5 scales wit h t he number of clust ers in t he RoI 0 0 0 10 20 30 0 10 20 30 40 50 f ollowing a cubic law; Processing time (ms) Mean number of clusters per layer a) b) however t he coef f icient B B- -jets jets – – 24 24 ev ev. pile . pile- -up up of t he t hird power is Events Processing time (ms) ID 800 2.218 / 15 Entries 719 P1 0.0000E+00 50 small, so t he behavior is Mean 86.32 P2 0.2935E-01 200 RMS 81.04 P3 0.1855E-02 40 UDFLW 0.0000E+00 P4 0.2302E-04 approximat ely quadrat ic. OVFLW 5.000 150 30 At design luminosit y, t he 100 20 average t ime needed t o 50 10 process a j et is close t o 0 0 0 100 200 300 400 0 50 100 150 Processing time (ms) Mean number of clusters per layer 90 ms. a) b) Pixel2000 - Genova - 6/6/00 7 P. Morettini
µ µ µ µ Reconstruction with a LVL1 µ µ µ µ Z vertex Z vertex Reconstruction with a LVL1 To reduce t he processing t ime we have t o decrease t he number of ext rapolat ion t o t he t hird layer . This could be achieved using a hard cut on t he Z impact paramet er of t he lay1-lay2 combinat ions, but t he posit ion of t he primary vert ex need t o be known wit h some precision. I n event s t riggered at LVL1 by a muon, it is possible t o use PixTrig it self wit h a high p T t hreshold, t o f ind t he muon and reconst ruct it s Z 0 (which is close t o Z vert ex ). I n 95.7% of t he event s one candidat e is f ound in t he µ RoI , and in 99% of t he Entries 718 150 Mean 0.1667E-02 cases it s Z 0 is less t han 600 RMS 0.6718E-01 UDFLW 5.000 5 mm apart f rom t he Z OVFLW 2.000 100 400 of vert ex . 50 200 The reconst ruct ion of t he vert ex t akes 0.5 ms. 0 0 0 1 2 3 4 -0.4 -0.2 0 0.2 0.4 Reconstructed Tracks in µ RoI Z rec - Z vertex Pixel2000 - Genova - 6/6/00 8 P. Morettini
Z vertex Reconstruction Inside a Jet Z vertex Reconstruction Inside a Jet Number of events Number of events Entries 1644 Entries 1374 10 3 10 3 I nside a j et RoI , t he Good quality Bad quality sample sample posit ion of t he vert ex can 10 2 10 2 be reconst ruct ed as t he barycent er of t he Z 0 of 10 10 t he t racks in 55% of t he event s. 1 1 -20 -10 0 10 20 -20 -10 0 10 20 This does not save z rec -z true z rec -z true B- -jets jets – – 24 24 ev ev. pile . pile- -up up B processing t ime, but 1.4 1.4 allows a clean-up of t he Efficiency 1.2 1.2 N(Fake Tracks)/Ntot sample “a post eriori”. 1 1 Using t his met hod, t he 0.8 0.8 ef f iciency and purit y 0.6 0.6 achieved at design 0.4 0.4 luminosit y are similar t o 0.2 0.2 t he low luminosit y ones. 0 0 0 2 4 6 8 10 -3 -2 -1 0 1 2 3 η p T Pixel2000 - Genova - 6/6/00 9 P. Morettini
Full Scan Efficiency and Timing Full Scan Efficiency and Timing Efficiency Fake Trks Fraction Full-scan on b event s 1 1 wit h at least one µ 0.8 0.8 wit h p T > 5 GeV/ c. 0.6 0.6 Low luminosit y. 0.4 0.4 Track reconst ruct ion 0.2 0.2 p T cut at 0.5 GeV/ c. 0 0 CPU: PI I I 500 MHz -2 0 2 -2 0 2 η η Events Processing time (ms) ID 800 3.084 / 13 150 Entries 750 P1 0.0000E+00 ε : 60 Mean 35.29 P2 0.1457 95 % RMS 28.74 P3 0.2473E-02 UDFLW 0.0000E+00 P4 -0.5960E-05 Fakes : 5% barrel OVFLW 14.00 100 40 30% end-caps 50 20 Tot al t ime: 35.3 ms, almost linear wit h t he 0 0 number of clust ers. 0 100 200 300 0 50 100 150 200 250 Processing time (ms) Mean number of clusters per layer a) b) Pixel2000 - Genova - 6/6/00 10 P. Morettini
Recommend
More recommend
