PPRuNe Forums - View Single Post - Malaysian Airlines MH370 contact lost
View Single Post
Old 21st Mar 2014, 20:39
  #7058 (permalink)  
Vinnie Boombatz
 
Join Date: Mar 2014
Location: SW USA
Posts: 67
Likes: 0
Received 0 Likes on 0 Posts
Some (Inconclusive) INMARSAT Info

A few definitions:

In the following the aircraft can be referred to as the UT (user terminal) or the AES (Aircraft Earth Station).

In the FCC documents, "downlink" is from the satellite, "uplink" is to the satellite. Don't guarantee that other documents follow that, have seen the terms reversed elsewhere, so one needs to skim any given document to get the sense of how those terms are used.

A few caveats:

The satellite antennas are not angle sensing, per se (e.g., no 4-horn feed or whatever). Spot beams and regional beams would have nominal angles relative to a satellite-fixed frame that would be known to the satellite operator. The documents linked do not contain a detailed frequency plan, but it seems reasonable that any pair of adjacent beams would employ some frequency separation.


A 4 MB "INMARSAT 101" briefing from 2009:

http://www.satcomdirect.com/connect/...rsat%20101.pdf

Slides 32-33 show coverage for IOR and POR. Slide 17 says that Generation 3 satellites have 7 spot beams, so the numbers in the plots on Slides 32-33 appear to be the spot beams.

Slide 33 is actually moot, since the current POR satellite is Generation 4, located nearly 40 degrees further West in longitude:

LIVE REAL TIME SATELLITE TRACKING AND PREDICTIONS: INMARSAT 4-F1

MH370 should have been closer to the POR satellite at takeoff, and possibly within its field of view throughout much of its flight.

Haven't seen any "credible" statements of which INMARSAT satellite was providing the "ping" data. It could make a big difference, since the Generation 4 satellites (e.g., POR) have many more spot beams, which could help in the aircraft location.

FCC document that describes INMARSAT Block 3 communications (the Indian Ocean Region satellite is a Block 3):

https://licensing.fcc.gov/myibfs/dow...nt_key=-136047

FCC document that describes INMARSAT Block 4 communications (the Pacific Ocean Region satellite is a Block 4):

https://licensing.fcc.gov/myibfs/dow...ent_key=-94644

"The Inmarsat 4F2 satellite, licensed in the United Kingdom, will provide Mobile-Satellite Services to small User Terminals (“UTs”) . . .

The BGAN UT’s start by searching for the global beam signaling carrier. When acquired, the global beam holds information on any underlying regional beam channel the UT can use for registering on the network. No return communication is carried out in the global beam. The selection of regional beam channel is based upon UT GPS position and spot-beam maps or carrier C/No scanning. Once the correct regional beam has been acquired, the UT will attempt to register using either slotted aloha random access or un-slotted (in case the UT does not have its GPS position available) aloha random access on dedicated logical channels.

After registration the UT is handed over to a spot beam whenever a communications session is started. After the communications session has ended the UT is moved back to the regional beam to preserve resources in the spot beams."

"Regional Beam Signaling:
In the regional beam, two 50 kHz signaling carrier types are used for the BGAN. The modulation
is either 16QAM or QPSK.

The Return direction is used for the UT’s to register onto the network. Depending on the UT Class (1, 2 or 3), the UT will register using any combination of burst characteristics that closes the link. The return signaling carriers are either 25 kHz or 50 kHz and the modulation can either be QPSK or 16QAM."

It's not all that relevant, but if anyone gets hung up on the aloha protocol:

ALOHAnet - Wikipedia, the free encyclopedia

A draft RTCA spec with some system description for the latest version of ACARS (SBB, the broadband service with lots of spot beams):

http://www.icao.int/safety/acp/ACPWG...20Material.pdf

On pg. 35, it mentions "performing link tests (keep alives)". Using those keywords in web searches doesn't seem to turn up much more than using "pings".

The document also mentions an "ICAO 24-bit aircraft address". That allows over 4 million unique addresses, though the bits may be allocated less efficiently (e.g., using BCD). Whatever, that identifier is likely part of the aircraft response to a "ping", hence uniquely identifying MH370.
Vinnie Boombatz is offline