Transcript Document

WP-4 “Information Technology”
J. Hogenbirk/M. de Jong




Introduction (‘Antares biased’)
Design considerations
Recent developments
Summary
disk
“All-data-to-shore”
photon
f
detection
information
m+1
f
distribution
detector
m
information
1
f
transmission
information
m
f
management
minimum number of
time-position
correlated photons
“All-data-to-shore” *
 Scientifically
– maximise neutrino detection efficiency
– maintain flexibility (also after construction)
– enable different physics (e.g. Magnetic Monopole)
 Technically
– reduce data transmission to a linear problem (scalability)
– locate all complexity on shore (reliability)
– optimisation of data filter (quality)
* Maximum of event related data
 Antares
•
•
•
•
Status
established proof-of-principle of “All-data-to-shore”
excellent time resolution → easy to find tracks
access to L0 data (single photons), see next pages
easy to include external triggers, see next pages
 NEMO
• mini tower operational
• readout also based on “All-data-to-shore”
 Nestor
• readout for 4-floor NESTOR tower ready, NuBE
• evaluation of commercial digitization system
completed
+ L0
● L1
muon
shower
Cherenkov
cone
neutrino
Gamma-Ray Burst
All data
detector
alert
TCP/IP
location of GRB TCP/IP
save
analysis
All data
before, during and after
GRB
number of gamma-ray bursts
GRB time – earliest recorded raw data
Access to data before GRB
delay [s]
delayed messages
Design considerations
Functional geography
 Photon detection
–
–
–
–
High data rate
Uni-directional
Low information density
Timing ~ns
 Instrumentation
–
–
–
–
Low data rate
Bi-directional
High information density
Timing ~ms
separation of functionalities
Separation of functionalities
Detection Units




Instrumentation Units
Optimise implementation
Reduce cost
Parallel design/production *
Reliability versus redundancy
*Critical path relaxation
requires proof of concept for calibration
Photon counting
 Large detection area PMT
– Slow
– Analogue
– Q-integrator
– ADC
 Small detection area PMT
– Fast
– Digital
– single photon counting
– Time-over-threshold
two-photon
purity
photon is digital!
Probability to detect 2 (or more) photons
as a function of
 photo-cathode area
 distance between muon and PMT
Cherenkov light cone
m
1-2 abs
~1 km
( R)  I 0
e
 R / abs sinc
2 R
Probability to detect 2 (or more) photons
P(#g ≥ 2)
QE = 25%
photo-cathode area: – 0.01 m2
– 0.02 m2
– 0.03 m2
– 0.04 m2
– 0.05 m2
R [m]
2 x larger PMT does NOT see twice as far
Time stamping
 Off-shore TDC
– Distributed clock system
•
•
•
•
Master clock
Time calibration
Network
Many slave clocks
– “Store and Forward” readout
 On-shore TDC
– Local clock system
• Master clock
• Time calibration
• ‘smart’ TDCs
– “Real-time” readout
• Software (protocol)
• Hardware (‘analogue’)
minimise off-shore electronics
Time slice
(or “how-to-get-all-data-in-one-place”)
T ~ Ln /c
time muon
takes to traverse detector
time
off-shore
on shore
Ethernet switch
Trigger
Trigger
Trigger
time
off-shore
on shore
Ethernet switch
Trigger
Trigger
Trigger
time
off-shore
on shore
Ethernet switch
Trigger
Trigger
Trigger
Design concepts
 à la Antares
 1-1 mixed: copper riser / fibre backbone
 photonics based
à la Antares






Design of new front-end chip (Guilloux, Delagnes, Druillole)
Design of new FPGA/CPU (Herve, Shebli, Louis)
Design of data transmission (Jelle, Henk, Mar)
New clock (?)
New slow control (Michel)
Network optimisation
– copper/fibre (Louis, Henk, ?)
– Ethernet switch (Louis)
 Both slow control & data acquisition (mjg)
1-1 mixed: copper riser / fibre backbone
 Design of multi-functional FPGA system
– FPGA/CPU integration (Herve, Shebli)
– Slow control (Michel?)
– Front end (Guilloux, Delagnes, Druillole)
 Integration of clock & data transmission system
– Time synchronisation & calibration (Rethore, Herve, Henk)
– Hardware/software (Nemo)
 Network optimisation (Jelle, Nemo?, ?)
Photonics based






Design of front-end electronics-photonics (Sander, Jelle, Mar)
Optical network (Jelle, Mar)
On-shore multi- laser (Mar, Jean Jennen)
Synchronised readout (mjg)
On-shore smart TDC (Saclay, Hervé)
No slow control (WP2)
Per 16-04-2007 new partners show interest to participate
after an upcoming dedicated meeting of WP4
Review presentations WP4 parallel session
1.
2.
3.
4.
5.
NEMO phase 1: Clock distribution
NEMO floor control Module
Progress on DAQ physical layer
Progress on optical components
Commercial Digitizing card
Conclusions Nemo phase 1: clock distribution
• Currently working for 4 floors
• NEMO phase 2: clock distribution to 16 floors
• Setup can be easily adapted to serve a
KM3NeT size apparatus
Conclusions by the WP4 meeting participants
1. Weakness: lack of input from other WP’s
2. Dedicated WP4 meeting is wanted
with inputs from other WP’s
3. Use today’s technology is required
4. Get together to:
define a medium scale demonstrator
roadmap to efficient decision making on technology
General summary
 “All-data-to-shore” is shown to work
– reduce off-shore electronics to minimum
 Communication with other WP’s important
– separation of functionalities
– front-end electronics
 Pursue different concepts
– à la Antares
– 1-1 mixed: copper riser / fibre backbone
– photonics based