Considerations for LDPC codes selection

1
SPACE LINK SERVICES CODING AND SYNCHRONIZATION
WORKING GROUP
Montreal Spring 2004
Considerations for LDPC codes selection
M. Chiani, G. Liva, E. Paolini University of
Bologna mchiani_at_deis.unibo.it
2
Outline

1. Consolidation of Requirements for New Channel
   Coding Schemes
2. Comparison of LDPC Coding Schemes a Proposal for
   Benchmarking
3. Overview of solutions under investigation within
   CCSDS
4. CONCLUSIONS

3
I. Consolidation of Requirements for New Channel
Coding Schemes
4
Consolidation of Requirements for New Channel
Coding Schemes
SPACE MISSIONS
Non-Deep Space (Near-Earth)
Deep Space
Earth station-spacecraftdistance gt 2.0 x 106 km
Earth station-spacecraftdistance lt 2.0 x 106 km
CCSDS Mission CategoryB
CCSDS Mission CategoryA
5
Mission Scenarios

1. Near Earth at Lagrange Point L2
2. Mars Missions
3. Low Earth Orbiter Payload Telemetry for Earth
   Observation
4. Low Earth Missions with High Rates

6
Near Earth at Lagrange Point L2

• (e.g. ESA Missions Darwin, Gaia,
  Herschel-Planck)
• X Band at e.g. 10 Mbps
• 50 MHz available in total No more than 10 MHz
  per mission
• GMSK Modulation
• Bandwith efficient codes required
• BWgt10MHz Ka (26 GHz or 37 GHz)

7
Mars Missions

• Similar to 1 but Deep Space
• 50 MHz available in X Band
• Large BW Ka Band (32 GHz)
• Constraints in Bandwidth and Link Budget to be
  evaluated
• GMSK

8
Low Earth Orbiter Payload

• Telemetry for Earth Observation
• Critical problem in X Band 0.35 GHz available
  band ? 8PSK TCM developed to this scope (CCSDS
  ECSS approved)
• Constraints mainly in Band occupation. Bandwidth
  is driver but code efficiency is important too.
• Ka Band (26GHz) for future very high rate needs.
  1.5 GHz wide. BW-eff to be considered from the
  outset.

9
TDRSS Case

• Not addressed by CCSDS so far. Should it be?
• Alternatively,search for commonality?

10
Low Earth Missions with High Rates

• Not for Earth Observation (e.g. successor of
  Hubble Telescope or Lunar Missions)
• Same Band as 1
• Likely to occupy the complete band
• Not yet well defined ? It may be wise to wait for
  Frequency Managers to fix strategy