PRS/Muon Status & Plans
Darin Acosta, for the PRS/Muon Group
Results made possible by:
Nicola Amapane, Silvia Arcelli, Daniele Bonacorsi, Giacomo
Bruno, Tim Cox, Alessandra Fanfani, Ugo Gasparini, Marcin
Konecki, Stefano Lacaprara, Marcello Maggi, Jason Mumford,
Norbert Neumeister, Hannes Sakulin, Stefano Villa, Slava Valouev,
Bart VandeVyver, Rick Wilkinson
+ the Production team
Outline
A summary of the Muon HLT material that is in the
DAQ TDR (or soon will be):
Raw data formats and volume: CSC, DT, RPC
Î Calibration procedures
Î Software status
Î Track segment creation
Î HLT reconstruction
Î Isolation
Î Level-1 working point
Î HLT rates and efficiencies
Î CPU analysis
Î
CPT Annual Review, 18 Sept 2002
2
PRS/Muon Status, D.Acosta
Overview of CSC Raw Data
540 chambers, 0.5 million channels
Î The pulses from the cathode strips
are sampled and stored in an analog
memory every 50 ns
Î The pulses from the anode wire
groups (and strips) are discriminated
for the trigger, and read out
Î Digitization occurs when there is a L1
accept and if there was a local
charged track (LCT) segment in the
chamber
‡ “Region of interest” reduces the
DAQ bandwidth
‡ Digitized with 12-bit ADC (2 bytes)
Î
muon
cathode
wires
cathode
induced charge
cathode with strips
avalanche
8 or 16 time samples are read out
over a region 16 strips wide by 6
layers deep (3 kB or 5.4 kB per
segment, or ≈20 kB/muon )
wires
plane cathode
Î
CPT Annual Review, 18 Sept 2002
3
PRS/Muon Status, D.Acosta
CSC Calibration Methods
Cathodes
Inject charge into one channel of every FE board
simultaneously, read out all 96 channels of each board
‡ Determine fC-to-ADC calibration, measure cross-talk
Î Data volume per pulse: 3.5 MB
Î Data volume for full calibration: ~300 GB
Î Expect a full calibration to take 10–20 minutes
Î Can use standalone DAQ system to collect data
Î Full calibration constants checked periodically (monthly)
Î Tests require no beams
Î
Anodes
Pulse wire group channels
Î Smaller amount of data than cathodes
‡ 1 bit versus 2 bytes, since only have discriminator data
Î
CPT Annual Review, 18 Sept 2002
4
PRS/Muon Status, D.Acosta
Overview of DT Raw Data
Î
plates
250 chambers, 172K channels
I-Beams
Drift-tubes organized into
2 superlayers in φ and one in θ
Î
Honeycomb
Field-shaping yields nearly
uniform drift velocity along cell
‡ But sensitive to radial
magnetic field
Î
Anode wire
Electrode
Cathode
13 mm
Threshold applied to signals
from wires
Î
Events
40 mm
mm
42
Drift time for signal to arrive on
wire is measured
‡ 21-bit TDC used for each
channel (4 bytes)
‡ Programmable trigger
window ~400 ns
Î
2200
2000
1800
1600
1400
1200
1000
800
θ=0
600
θ = 15
o
o
400
200
0
-50
0
50
100
150
200
250
300
350
400
Drift time (ns)
T[ns]
CPT Annual Review, 18 Sept 2002
5
PRS/Muon Status, D.Acosta
DT Calibration Methods
Calibrations currently foreseen:
Threshold run
‡ Pulse channels and scan/set discriminator levels for
optimum t0 measurement
Î t0 calibration
‡ Determine time-of-arrival of calibration pulse for each
channel
Î Rates
‡ Run without beam, determine noisy channels
Î Extraction gap test
‡ Monitor dead channels & misalignment in trigger
electronics continuously during the run by pulsing
channels during LHC gaps
Î
Event size not expected to exceed normal event size
Each calibration should take just several minutes
during fill (or taken during run)
CPT Annual Review, 18 Sept 2002
6
PRS/Muon Status, D.Acosta
Overview of RPC Raw Data
612 chambers, 160K strip channels
Î Double-gap design with common pickup
Î Pulses from strips are discriminated
Î Used by Pattern Comparator Logic,
and read out by DAQ
Î
Î
Calibration Methods:
‡ Threshold scan and set
‡ Time window scan and set
‡ Test pattern injection
(between fills or in gaps)
‡ Monitoring during physics runs
Barrel sector
– HV
– HV
CPT Annual Review, 18 Sept 2002
7
PRS/Muon Status, D.Acosta
Muon Data Volume
Consider high luminosity (L=1034) with 100 kHz DAQ
Raw data format is well specified by hardware designs
Î Rates include neutron background estimate and overhead for
headers and empty events
Î
CSC:
8 time samples and tight LCT coincidence
Î 1100 MB/s (1300 MB/s with 3× safety factor on neutrons)
Î Average occupancy is 3.4 segments ⇒ 10kB / L1A
Î
DT:
~1% channel occupancy from muons, punch-through, neutrons
Î Event size: 9 kB/event
Î Data volume: 900 MB/s
Î
RPC:
Î
~0.6 kB event size (3% of CSC+DT) mostly from noise hits
Overall: 2 GB/s at high luminosity (20kB / L1A)
CPT Annual Review, 18 Sept 2002
8
PRS/Muon Status, D.Acosta
HLT Results: Software Status
Many improvements to the software over the last year
Î In addition to the development/refinement of HLT algorithms,
the code runs much faster, track segment reconstruction
improved, overall efficiency increased, and the code is more
robust and user friendly
Î
η generated
5.1
η reconstructed
CPT Annual Review, 18 Sept 2002
9
PRS/Muon Status, D.Acosta
HLT Input: Track Segment Creation
CSC:
CSC:
Fit “Gatti” function to the spatial shape of
3-strip charge distribution to determine
centroid of cluster in layer
‡ Resolution: 120–250 µm
Î Fit lines in 3-D through the collection of
wire and strip clusters in chamber
Î Use positions of constituent hits for HLT
tracking
Î
DT:
Muon Track
Layers
6
5
4
Comparator Bits
3
2
Charges
1
Strips
Anode Wire Group Hits
Muon track
Reconstruct position of each channel above
threshold using effective drift velocity
‡ Resolution: 250 µm in φ view
Î Fit 2-D lines separately in r-φ and r-z
through the 8+4 layers of chamber
‡ L/R ambiguities solved by best χ2
Î Combine into 3-D segments, use segment
position and direction for HLT tracking
‡ Resolution: 100 µm in φ view
Hit
Laye
N
Î
CPT Annual Review, 18 Sept 2002
Bu
10
Wire group hit
PRS/Muon Status, D.Acosta
Residuals and Pulls in r-φ
residual
pull
Σ 109 Μm
600
DT
400
200
0
1000
500
0
500
sim
fit
X0 X0 Μm
Residuals ME2,ME3,ME4
χ2 / ndf
1000
Mean
12000
Sigma
Constant
0.0002398
6000
Mean
0.02302
6310
0.01169
Sigma
5000
10000
CSC
210.4 / 57
7000
Constant 1.272e+04
14000
χ 2 / ndf
Pull ME2,ME3,ME4
1231 / 23
1.238
4000
8000
6000
3000
4000
2000
2000
1000
0
-0.2 -0.15 -0.1 -0.05
-0
0.05
0.1
0.15
0
-5
0.2
x rec
-xsim
(cm)
φ
φ
CPT Annual Review, 18 Sept 2002
11
-4
-3
-2
-1
rec
0
1
(xφ -xsim
) / σ xφ
φ
2
3
4
5
PRS/Muon Status, D.Acosta
HLT Track Reconstruction
Standalone Muon Reconstruction: “Level-2”
Seeded by Level-1 muons
Î Kalman filtering technique applied to DT/CSC/RPC track
segments
‡ Segment positions for DT, individual hits for CSC (B-field)
Î GEANE used for propagation through iron
Î Trajectory building works from inside out
Î Track fitting works from outside in
Î Fit track with beam constraint
Inclusion of Tracker Hits: “Level-3”
Define a region of interest through
tracker based on L2 track with
parameters at vertex
Î Find pixel seeds, and propagate from
innermost layers out, including muon
Î High algorithmic efficiency
Î
Efficiency
Î
1
0.98
0.96
0.94
0.92
0.9
0.88
0.86
0.84
0.82
0.8
0
CPT Annual Review, 18 Sept 2002
12
0.5
1
1.5
2
PRS/Muon Status, D.Acosta
η
2.5
ηµ
PT Resolution
barrel
χ 2 / ndf
Constant
12
Mean
Sigma
10
10.11
0.01847
0.1196
σ = 0.12
8
Level-2:
overlap
1.481 / 13
χ 2 / ndf
Constant
1.4
1.2
Mean
0.01556
Sigma
0.1432
χ 2 / ndf
Constant
1.168
σ = 0.14
1
endcap
0.2567 / 18
Mean
Sigma
2.5
0.7722 / 24
2.459
0.03437
0.165
σ = 0.17
2
0.8
1.5
6
0.6
1
4
0.4
2
0
0.5
0.2
-1 -0.8 -0.6 -0.4 -0.2 -0 0.2 0.4 0.6 0.8 1
χ 2 / ndf
Constant
Mean
10
Sigma
-1 -0.8 -0.6 -0.4 -0.2 -0 0.2 0.4 0.6 0.8 1
χ 2 / ndf
3.294 / 25
Constant
9.297
-0.002059
1.2
Mean
Sigma
0.01314
σ = 0.013
8
0
1
0
-1 -0.8 -0.6 -0.4 -0.2 -0 0.2 0.4 0.6 0.8 1
χ 2 / ndf
0.3244 / 27
Constant
1.071
-0.001114
Mean
2.5
Sigma
0.01493
σ = 0.015
0.408 / 27
2.306
-0.007125
0.0182
σ = 0.018
2
0.8
Level-3:
Order of
magnitude
improvement
1.5
6
0.6
1
4
0.4
2
0
-0.1
0.5
0.2
-0.05
0
0.05
CPT Annual Review, 18 Sept 2002
0.1
0
-0.1
-0.05
0
0.05
13
0.1
0
-0.1
-0.05
0
0.05
0.1
PRS/Muon Status, D.Acosta
Efficiency vs. PT Threshold
Single muons with |η|<2.1
1
Efficiency
Efficiency
Level-1, Level-2, Level-3
0.9
0.9
0.8
0.8
0.7
0.7
0.6
0.6
0.5
0.5
0.4
0.4
10 GeV (a)
threshold
0.3
0.2
0
0
0.2
0.1
10
20
30
40
50
60
0
0
70
80
µ
pT [GeV/c]
Efficiency
1
0.9
0.7
0.7
0.6
0.6
0.5
0.5
0.4
0.4
30 GeV(c)
threshold
0.2
20
30
40
50
60
70
80
µ
pT [GeV/c]
0.9
0.8
0.3
10
1
0.8
0.3
(d)
40
0.2
GeV threshold
0.1
0.1
0
0
20 GeV(b)threshold
0.3
0.1
Efficiency
1
10
20
30
40
50
60
70
80
pµT [GeV/c]
0
0
10
20
30
40
50
60
70
80
pµT [GeV/c]
Thresholds defined at 90% efficiency
CPT Annual Review, 18 Sept 2002
14
PRS/Muon Status, D.Acosta
Cumulative Efficiency vs. Eta
Efficienc
Level-1, Level-2, Level-3
Single muons, no threshold, no pile-up
1
0.98
0.96
0.94
0.92
Eta coverage
limited to 2.1
(Limit of RPC,
and ME1/1a of
CSC does not
participate @
L1)
0.9
0.88
0.86
0.84
0.82
0
0.5
1
1.5
Suppressed zero
CPT Annual Review, 18 Sept 2002
15
2
2.5
|η µ|
PRS/Muon Status, D.Acosta
Muon Isolation
General framework for isolation released since ORCA 6
Î
Î
Based on ΣET or ΣPT measurements in cones around the muon
Cone sizes and thresholds are optimized
‡ To get maximal rejection on “reference background”
(Minimum Bias muons with PT above the trigger threshold)
for a given nominal efficiency on reference signal (W→µν)
‡ Optimization provides flat ε(η) on signal by construction
Calorimeter Isolation
Î
ΣET from calorimeter towers in a cone around muon
(sensitive to pile-up)
Pixel Isolation
Î
Î
Î
ΣPT of 3-hit tracks in the pixel detector in cone around muon
Studies done for full pixel detector (no staging)
Requires that contributing tracks come from the same primary
vertex as the Level-3 muon (to reduce pile-up contamination)
Tracker Isolation
Î
ΣPT of tracks in the tracker (regional reconstruction around
Level-3 muon)
CPT Annual Review, 18 Sept 2002
16
PRS/Muon Status, D.Acosta
Isolation Performance
Efficiency on minbias events as a
function of efficiency on W signal
1
efficiency MB
efficiency
Efficiency on minbias events
as a function of PTgen for 97%
efficiency on signal
L2 Calorimeter Isolation
0.9
L3 Pixel Isolation
0.8
L3 Tracker Isolation
0.7
High Lumi
0.6
0.4
High Lumi
L2 Calorimeter Isolation
L3 Pixel Isolation
L3 Tracker Isolation
e(W) ≥ 97%
0.4
p T ( µ gen ) > 22 GeV, | η | < 2.4
0.3
| η | < 2.4
0.5
0.5
0.2
0.3
0.1
0.2
0.1
0
10
10
pT gen
0
2
0.8
0.85
0.9
0.95
1
efficiency W
“reference background”
CPT Annual Review, 18 Sept 2002
0.75
“reference signal”
17
PRS/Muon Status, D.Acosta
10
10
10
10
10
6
5
Sensitive to entire inelastic
cross section at LHC, since
every π/K/b/c can decay into a
muon and multiple-scatter to
appear as high PT
Î Procedure applied to force
muon decays and assign
event weight based on the
decay and the generated PT
Î
MB : p T > 0 GeV , pµT > 1 GeV
MB : p T > 0 GeV , pµT > 4 GeV
4
MB : p T > 10 GeV , pµT > 10 GeV
W→µ+X : pµT > 3 GeV
3
Z/γ*→µ+X : pµT > 3 GeV
2
tt→µ+X : pµT > 3 GeV
10
1
10
10
10
10
-1
-2
Rate [Hz]
d σ/dp Tµ [nb/GeV]
Muon Samples for HLT Rate Estimates
-3
-4
0
10
20
30
40
50
60
Differential cross-section
Integrated muon rates
vs. PT threshold,
L=1034, |η|<2.1
Neutron background not
included in simulations
CPT Annual Review, 18 Sept 2002
70
80
p µT [GeV/c]
10
6
±
10
10
10
10
K /π±
0
KL
c
b
τ
5
4
0
Z /γ
W±
all
3
*
2
10
1
10
10
18
-1
-2
0
10
20
30
40
50
60
70
80
p µT threshold [GeV/c]
PRS/Muon Status, D.Acosta
Possible Level-1 Working Points
|η| < 2.1
GeV/c
50 kHz DAQ
41.0kHz for µ, µµ
100 kHz DAQ
8 kHz for µ, µµ
14
2
cut
1.5
12
2.0
10
2.5
8
3.0
4
4.0
5.0
2
6.0
3
12
12, 8;8
εW=91.4 %
εZ =99.7 %
εBs→µµ=14.5 %
3.5
6
Symmetric di-muon pT
Symmetric di-muon pT
cut
14
L1 single and di-muon combined trigger rates, high luminosity
GeV/c
trigger rates in kHz
L1 single and di-muon combined trigger rates, low luminosity
14, -;εW=89.6 %
εZ =99.8 %
εBs→µµ=27.1 %
4
6
8
10
12
8
15
6
18
25, 5;5
εW=74.1 %
εZ =99.5 %
εBs→µµ=14.3 %
4
2
7.0
0
0
10 12 14 16 18 20 22 24 26 28
Single muon pcut
T / GeV/c
L = 2x1033cm-2s-1
20, 6;6
εW=82.3 %
εZ =99.6 %
εBs→µµ= 9.9 %
10
15
20
25
30 35 40 45 50
Single muon pcut
T / GeV/c
L = 1034cm-2s-1
Thresholds similar to those listed in Level-1 Trigger TDR, but acceptance limited
CPT Annual Review, 18 Sept 2002
19
PRS/Muon Status, D.Acosta
HLT Analysis
Level-2
PT (@vertex) > 0 (valid extrapolation to vertex)
Î Number of used RecHits > 3
Î Convert PT (@vertex) to 90% efficiency scale
Î Apply calorimeter isolation at 97% efficiency point
Î Di-muons:
‡ Require isolation of either of the two candidates
Î
Level-3
Number of silicon hits > 5 (pixel + strips)
Î Convert PT (@IP) to 90% efficiency scale
Î Apply pixel isolation at 97% efficiency point
Î Apply tracker isolation at 97% efficiency point
Î Di-muons:
‡ Reject ghosts (∆ PT < 0.1 and ∆η < 0.01 and ∆ϕ < 0.05)
‡ (Zµ1 – Zµ2) < 2mm @ IP
‡ Require isolation of either of the two candidates
Î
CPT Annual Review, 18 Sept 2002
20
PRS/Muon Status, D.Acosta
Rate [Hz]
L1, L2, L3 Trigger Rates @ Low Lumi
10
generator
4.1 ± 0.1 kHz
4
L1
L2
L2 calo isol
L3
L3 calo + pixel isol
L3 calo + pixel + tracker isol
10
10
4 kHz
3
2
30 Hz
L = 2×1033 cm-2s-1
10
10
CPT Annual Review, 18 Sept 2002
15
20
21
25
30
µ
p T threshold [GeV/c]
PRS/Muon Status, D.Acosta
Rate [Hz]
L1, L2, L3 Trigger Rates @ High Lumi
10
generator
5
L1
6.1 ± 0.3 kHz
L2
L2 calo isol
10
4
L3
8 kHz
L3 calo + pixel isol
L3 calo + pixel + tracker isol
10
10
3
2
30 Hz
10
L = 1034 cm-2s-1
10
20
CPT Annual Review, 18 Sept 2002
30
40
22
50
60
µ
p T threshold
70
[GeV/c]
PRS/Muon Status, D.Acosta
10
Before isolation,
all
4
±
K / π± → µ ν
c/b → µ + X
10
and after
τ→µ+X
3
±
W → µν
Z0 / γ* → µ +µ -
10
2
10
1
10
-1
10
20
30
40
50
60
70
µ
p T threshold [GeV/c]
Rate [Hz]
Rate [Hz]
Physics Content after Level-3
10
4
all
±
K / π± → µ ν
c/b → µ + X
10
3
τ→µ+X
±
W → µν
0
Z / γ* → µ+µ-
L=
1034 cm-2s-1
10
2
30 Hz
10
1
π/K/b/c strongly suppressed
10
CPT Annual Review, 18 Sept 2002
23
-1
10
20
30
40
50
60
70
µ
p T threshold [GeV/c]
PRS/Muon Status, D.Acosta
High luminosity
Low lumi
L3 Single µ Rates with Isolation (Hz)
12 GeV
14 GeV
16 GeV
18 GeV
20 GeV
22 GeV
total
166.1
79.8
45.0
29.9
23.0
18.4
K/π
38.1
15.2
6.8
3.5
2.1
1.0
b/c
104.9
43.5
18.5
8.0
3.7
1.8
tau
2.5
1.7
1.3
1.0
0.7
0.5
W → µν
16.9
16.4
15.7
14.9
14.1
13.0
Z → µµ
3.5
3.0
2.7
2.5
2.3
2.1
20 GeV
22 GeV
24 GeV
26 GeV
28 GeV
30 GeV
32 GeV
34 GeV
36 GeV
38 GeV
total
121.4
93.3
81.2
69.3
62.9
56.9
48.5
43.1
35.1
29.9
K/π
11.2
5.3
2.8
1.7
1.3
1.1
0.5
0.4
0.3
0.3
b/c
26.4
12.2
7.2
3.7
2.2
1.4
1.0
0.8
0.4
0.3
tau
3.4
2.6
2.0
1.5
1.2
1.0
0.8
0.6
0.4
0.4
W → µν
69.1
62.9
59.5
53.5
49.8
45.5
38.9
34.3
27.8
23.2
Z → µµ
11.3
10.3
9.7
8.9
8.4
8.0
7.4
7.0
6.2
5.7
CPT Annual Review, 18 Sept 2002
24
PRS/Muon Status, D.Acosta
Efficiency for W, top
Effic @ high lumi
tt → 1µ + X
1
0.9
L1 @ 90%
L2 @ 90%
0.8
L3 @ 90%
L3 + isolation
0.7
Efficiency
Efficiency
W → µν
1
0.9
0.8
0.5
0.5
0.4
0.4
0.3
0.3
0.2
0.2
0.1
0.1
20
30
40
CPT Annual Review, 18 Sept 2002
50
L3 @ 90%
L3 + isolation
0.7
0.6
10
L1 @ 90%
L2 @ 90%
0.6
0
0
(W → µν)
0
0
60
70
80
µ
p T threshold [GeV/c]
25
10
20
30
40
50
µ
pT
60
70
80
threshold [GeV/c]
PRS/Muon Status, D.Acosta
10
4
generator
L1
10
L2
3
L2 calo isol
L3
10
L3 calo + pixel + tracker isol
2
L = 1034 cm-2s-1
10
Rate [Hz]
Rate [Hz]
L1, L2, L3 Di-muon Trigger Rates
1
10
10
0
5
10
15
20
L = 2×1033 cm-2s-1
25
30
µ
p T threshold [GeV/c]
5
generator
L1
L2
4
L2 calo isol
L3
10
10
L3 calo + pixel + tracker isol
3
2
10
1
Inclusive rates
10
CPT Annual Review, 18 Sept 2002
-1
26
0
5
10
15
20
25
30
35
40
45
50
µ
p T threshold [GeV/c]
PRS/Muon Status, D.Acosta
L3 Single and Di-µ Rate with Isolation
60 Hz
18
30 Hz
16
70 Hz
14
20 Hz
40 Hz
12
50 Hz
10
28.9 ± 0.5 Hz
80 Hz
L = 1034 cm-2s-1
8
2
12
14
L=
16
18
20
22
24
26
Single muon p T threshold [GeV/c]
2×1033 cm-2s-1
T
4
threshold [GeV/c]
30
6
Di-muon p
Di-muon p
T
threshold [GeV/c]
20
20 Hz
25
20
40 Hz
5
30
27
32.7 ± 0.4 Hz
15
10
CPT Annual Review, 18 Sept 2002
30 Hz
50 Hz
60 Hz
32
34
36
38
40
42
44
46
48
50
Single muon p T threshold [GeV/c]
PRS/Muon Status, D.Acosta
W and Z Efficiency
Di-muon
1
0.9
W ± → µν
0.8
0
Z → µ+µ-
Efficiency
Efficiency
Single muon
1
0.9
0.8
0.7
0.7
0.6
0.6
0.5
0.5
0.4
0.4
0.3
0.3
0.2
0.2
0.1
0.1
0
0
10
20
30
40
50
Effic @ low lumi
0
0
60
70
80
µ
p T threshold [GeV/c]
0
Z → µ+µ-
10
20
30
40
50
µ
di-muon p T threshold [GeV/c]
Combined Z efficiency is 98% (88%) for low (high) lumi
CPT Annual Review, 18 Sept 2002
28
PRS/Muon Status, D.Acosta
Efficiency of Higgs →WW → 2µ 2ν
L3 Singleµ after Iso
L3SingleMuIsoEff_histo
M Higgs
= 160
Entries
356993
Mean= 120
21.27
M Higgs
RMS
13.92
M Higgs = 200
0.9
0.8
0.7
0.6
0.5
0.4
0.3
0.2
0.1
0
0
10
20
30
40
50
60
PµT Threshold
Efficiency at low lumi:
MH = 120: Single µ = 73% Di-µ exclusive: 14% Combined: 87%
Î MH = 160: Single µ = 87% Di-µ exclusive: 5% Combined: 92%
Î
Efficiency at high lumi:
MH = 120: Single µ = 30% Di-µ exclusive: 34% Combined: 64%
Î MH = 160: Single µ = 45% Di-µ exclusive: 32% Combined: 77%
Î
CPT Annual Review, 18 Sept 2002
29
PRS/Muon Status, D.Acosta
CPU Usage
W sample
Single muon trigger
2x1033 cm-2s-1
Treshold = 10 GeV
1034 cm-2s-1
Treshold =18 GeV
ms/(event passing previous steps)
ms/(event passing previous steps)
Total
GEANE excl.
Total
GEANE excl.
430
100
460
105
95
25
110
40
330
160
560
336
Pixel Isolation
75
75
340
340
Tracker Isolation
70
70
220
220
Level-2
Calorim. Isolation
Level-3
L2 time/L1 event
Total L2 time
20
CPU time on PIII 1GHz
W→µ+X events
“Signal”
Prop. Excl.:
0.105 s
Prop. Incl.:
0.459 s
18
16
14
W→µ+X
L=1034
12
10
8
6
4
2
0
CPT Annual Review, 18 Sept 2002
30
0
0.5
1
1.5
2
T (s)
PRS/Muon Status, D.Acosta
CPU Usage: Realistic Muon Spectrum
Realistic Spectrum
Single muon trigger
2x1033 cm-2s-1
Treshold 10 GeV
ms/(event passing previous steps)
Total
GEANE excl.
1034 cm-2s-1
Treshold 18 GeV
ms/(event passing previous steps)
Total
GEANE excl.
Level-2
840
100
Calorim. Isolation
120
40
Level-3
735
425
Pixel Isolation
340
340
Tracker Isolation
370
370
1020
200
Total/L1 muon
CPU time on PIII 1GHz
CPT Annual Review, 18 Sept 2002
31
PRS/Muon Status, D.Acosta
Conclusions & Beyond the TDR
Muon data volume ≈2GB/s
Possible HLT working point at high luminosity
Single PT threshold
Î Di-muon symmetric PT threshold
Î Combined rate
Î Nearly all W/Z after isolation
Î
= 38 GeV : 29.9 ± 0.4 Hz
= 12 GeV : 2.7 ± 0.2 Hz
: 32.7 ± 0.4 Hz
Further studies/solutions of problematic regions needed
Inefficiency in DT/CSC overlap region and showering “TeV” muons
Î Fast track propagator and replacement for GEANE
Î
Further optimize L1→HLT→offline for physics
Disparity between L1 and L3 inclusive muon thresholds leaves room
for more exclusive HLT triggers (correlations) at lower thresholds
Î Some cuts should be relaxed at high PT (e.g. isolation)
Î
Refine/distinguish HLT tools and offline reconstruction
HLT should be fast and robust. Just need to get event to tape, which
means reconstructing just one or two muons
Î Offline you want all muons reconstructed (e.g. H→4µ)
Î
CPT Annual Review, 18 Sept 2002
32
PRS/Muon Status, D.Acosta
Backup Slides
Low Lumi Working Points
Low Lumi:
P1:
ÎSingle
PT threshold = 19 GeV : 25.7 ± 0.4 Hz
ÎDi-muon symmetric PT threshold = 7 GeV : 3.2 ± 0.2 Hz
ÎCombined rate = 28.9 ± 0.5 Hz
P2:
ÎSingle
PT threshold = 19 GeV : 25.7 ± 0.4 Hz
ÎDi-muon symmetric pT threshold = 6 GeV : 8.5 ± 0.7 Hz
ÎCombined rate = 34.2 ± 1.0 Hz
P3:
ÎSingle
PT threshold = 20 GeV : 23.0 ± 0.3 Hz
ÎDi-muon symmetric PT threshold = 6 GeV : 8.6 ± 0.7 Hz
ÎCombined rate = 31.6 ± 1.0 Hz
CPT Annual Review, 18 Sept 2002
34
PRS/Muon Status, D.Acosta
Low Luminosity Rates Comparison
97% isolation cut (upper) and 90% isolation cut (lower)
total
12 GeV
14 GeV
16 GeV
18 GeV
20 GeV
22 GeV
166.1
92.2
79.8
45.0
45.0
28.0
29.9
21.5
23.0
17.8
18.4
14.8
K/π
b/c
38.1
23.1
15.2
8.9
6.8
4.1
3.5
2.3
2.1
1.3
1.0
0.6
104.9
49.3
43.5
17.8
18.5
6.8
8.0
3.2
3.7
1.5
1.8
0.6
CPT Annual Review, 18 Sept 2002
tau
2.5
1.8
1.7
1.3
1.3
1.0
1.0
0.8
0.7
0.6
0.5
0.5
35
W → µν
16.9
14.8
16.4
14.3
15.7
13.7
14.9
13.1
14.1
12.4
13.0
11.3
Z → µµ
3.5
3.1
3.0
2.6
2.7
2.4
2.5
2.2
2.3
2.0
2.1
1.8
PRS/Muon Status, D.Acosta
Isolation: performance
0.5
pT (µ
0.4
gen
efficiency MB
efficiency MB
Efficiency on Minimum Bias events as a function of
nominal efficiency on signal
) > 16 GeV, | η | < 2.4
Low Lumi
L2 Calorimeter Isolation
L3 Pixel Isolation
0.3
0.5
p T ( µ gen ) > 22 GeV, | η | < 2.4
0.4
High Lumi
L2 Calorimeter Isolation
L3 Pixel Isolation
0.3
L3 Tracker Isolation
L3 Tracker Isolation
0.2
0.2
0.1
0.1
0
0.75
0.8
0.85
0.9
0.95
0
1
efficiency W
CPT Annual Review, 18 Sept 2002
36
0.75
0.8
0.85
0.9
0.95
1
efficiency W
PRS/Muon Status, D.Acosta
Isolation
1
efficiency
efficiency
Efficiency on Minimum Bias events as a function of PTgen and η
for 97% nominal efficiency on signal
L2 Calorimeter Isolation
0.9
L3 Pixel Isolation
0.8
L3 Tracker Isolation
0.7
0.3
High Lumi
0.6
0.4
| η | < 2.4
0.5
e(W) ≥ 97%
0.4
0.2
p T ( µ gen ) > 22 GeV
0.3
0.1
0.2
0.1
0
10
CPT Annual Review, 18 Sept 2002
10
pT gen
0
0
2
37
L2 Calorimeter Isolation
L3 Pixel Isolation
L3 Tracker Isolation
0.5
1
e(W) ≥ 97%
High Lumi
1.5
2
pseudorapidity
PRS/Muon Status, D.Acosta
Alignment using Muons
Use prompt collision muons with PT>50 GeV/c from GMT
Î
0.2–1 Hz muon/sector
Require ~4 days @ L= 2×1033 to reach 200 µm precision
~ 1 TB dedicated data stored offline for alignment analysis
Î Must have sufficient precision on ∆B/B
‡ See F.Matorras et al., CMS-TN 96-005
Î
For the endcaps, might make use of a special
accelerator muon trigger implemented in the
CSC Track-Finder and GMT
Triggers on halo muons traveling parallel to beam axis
Î Expect a halo muon to trigger just one endcap or the other
Î Needs investigation
Î
CPT Annual Review, 18 Sept 2002
38
PRS/Muon Status, D.Acosta
Misalignment Effects on HLT Trigger
Study degradation of the muon
identification and Pt resolution after
applying different misalignments
Use modified ORCA package to perform
misalignments
Î Great advantage: Already existing
MC is used, without being
necessary to simulate distorted
geometries.
Based on existing Tracker ‘misalignment
tools’:
Basic concept: displace chambers
hosting the reconstructed segments
while leaving its local position
unchanged
ORCA study: CMS IN 2002/049
Calderon et al.
L2
barrel
L3
Example: Vertical displacement due to
gravity. Units in cm.
∆y = 0
∆y = 1 cm
CPT Annual Review, 18 Sept 2002
39
PRS/Muon Status, D.Acosta