H. Bischof, A.N. Belbachir (TUVIE)

1
SPU High Level Software

• H. Bischof, A.N. Belbachir (TUVIE)
• F. Kerschbaum, R. Ottensamer, P. Reegen, C.
  Reimers (UVIE)

2
Data Compression/Reduction Scheme
Figure 1. Data Compression/Reduction Scheme
3
ASW Requirements - SPU HLSW Data Flow

• Spectroscopy (400 detectors, 50 test channels, 18
  empty channels per SPU Module)
• 2000 kbits/s (4000 kbits/s for both SPUs)
• Photometry (512 detectors per sub-image)
• 5 sub-images (1700 kbits/s for both SPUs)
• 340 kbits for the LWL SPU
• 1360 kbits/s for the SWL SPU
• Telemetry rate
• 120 kbits/s are available for science data
• Transparent mode max. of 28 selected detectors
  in Spectroscopy or
  max. of 185 selected detectors in Photometry
• Default mode Spectroscopy 97,14 kbits/s
  Photometry 105,3 kbits/s

4
ASW Requirements - Telemetry rates
Figure 2. Telemetry rates for the SWL and LWL SPU
5
ASW Requirements - Default Mode in Spectroscopy
Figure 3. Default Compression Mode in Spectroscopy
6
Data Compression/Reduction - Spectroscopy

• Ramp Fitting Methods (1)

7
Data Compression/Reduction - Spectroscopy

• Ramp Fitting Methods (2)

8
Data Compression/Reduction - Spectroscopy

• Evaluation of ramp fitting algorithms and test
  results

9
Data Compression/Reduction - Spectroscopy

• Ramp 1

Ramp 2
Ramp 3
10
Data Compression/Reduction - Spectroscopy

• Analysis with simulated and real test data
  
• Status
• Preprocessing, glitch detection
• Ramp Fitting cumulative errors ? difference
  scheme RANSAC (2 point with least square errors)
  implemented
• Integration mean algorithm implemented
• TRR SRR Reference value Difference values

11
ASW Requirements - Default Mode in Photometry
Figure 4. Default Compression Mode in Photometry
12
Data Compression/Reduction - Photometry

• Evaluation of averaging algorithms and test
  results

13
Data Compression/Reduction - Photometry
Voltage
Voltage
Plateau 1
Plateau 3
Time
Time
Voltage
Mean Median Sample Difference Real readout
Error Plateau 1 Plateau 2 Plateau 3
Mean 0,2 2,5 16,7
Median 0,0 0,3 0,0
Sample Difference 0,1 0,1 0,1
Plateau 2
Time
14
Data Compression/Reduction - Photometry

• Analysis with simulated data from our data
  generator
• Status
• Preprocessing, glitch detection
• Robust Averaging mean algorithm implemented
  (calibration on ground)
• Integration mean algorithm implemented
• TRR SRR Reference value Difference values

15
SPU HLSW Context Diagram
Figure 5. SPU HLSW Context Diagram
16
SPU HLSW Concept

• HLSW consists of three main parts
• Communication Interfaces to DPU and to DEC/MEC
• Watch Process (Command Acknowledgement)
• Application Software (Reduction/Compression)

Figure 6. SPU HLSW Concept
17
SPU SW Interfaces

• DPU TO SPU SW Interface
• Communication is bi-directional (commands,
  response, HK and compressed data)
• All SPU SW activities are commanded by DPU(e.g.
  start, stop, )
• SPU SW acknowledges the reception of all DPU
  commands according to the communication protocol
• SPU SW sends telemetry packets to DPU
• DPU checks the life of the SPU SW via the HK
• DEC/MEC to SPU SW Interface
• Communication is unidirectional (DEC/MEC to SPU)
• Packet from DEC/MEC to SPU consists of science
  data and a header
• Science data are detector readouts and test
  channels
• Header contains the instrument configuration and
  the compression parameters

18
Memory Description

• 1.5MB EEPROM
• 32KB DPRAM
• 7MB RAM
• 1 Mbytes for program storage
• 0.25 Mbytes for SW tables storage
• 1 Mbytes for input science data buffering
• 0.25 Mbytes for the DEC/MEC header buffering
• 0.5 Mbytes for output data buffering
• 4 Mbytes are for processing, etc.

Figure 7. Memory Distribution for the SPU HLSW
19
SPU HLSW Status

• SPU HLSW design frozen
• SPU HLSW Interfaces with DPU and DEC/MEC
• Interface Control Document are under
  configuration control
• Software interfaces have been verified under test
  environment (PCemulator)
• Detailed description of LLSW drivers are
  available since 2001, 13 Dec. for the integration
  on the real HW
• Application Software
• Mechanism has been verified under test
  environment (PCemulator)
• Performance not tested (individual compression
  modules tested Prelimi.)
• No real data
• No representative development HW

20
PA/QA Activities and Schedule

• PACS PA Plan is adopted (from IFSI)
• SPU Test Plan is under configuration control
• SPU HLSW Interface tested under test environment
  (PC Emulator Spacewire Board)
• SPU HLSW individual module complexity tested in
  Sigma board
• Schedule
• SPU SW Interfaces and application SW mechanism
  have been tested
• SPU HLSW functionality will be tested at IAC
  (Spain)
• Test at IAC is planned in March/April 2002
• Delivery of SPU HLSW to project by begin May

21
Summary and Perspective

• SPU HLSW Interfaces tested with OBS Simulator
• Preliminary Application SW modules are ready for
  integration and performance tests in real HW
• Several ramp fitting and averaging algorithms are
  tested
• It is still possible to add new algorithms to
  this library
• New algorithms will be tested (functionality and
  performance)
• Verification of SPU HLSW functionality will be
  done at IAC (Spain) in March/April 2002