Juggling spectrum…continued

Posted in Financials, Inmarsat, LightSquared, Operators, Regulatory, Spectrum at 9:01 am by timfarrar

Yesterday’s announcement that LightSquared plans to use the lower 2x10MHz block of its L-band spectrum (1526-1536MHz downlink, paired with 1627.5-1637.5MHz uplink) for initial operations, leaves a number of questions unanswered. In particular, the most critical issue is when Inmarsat will make this block of spectrum available. Half of the spectrum (1526-1531MHz downlink) was part of the Phase 1A plan, which was part of the rebanding plan that LightSquared was paying Inmarsat $250M to undertake, and completion was scheduled for February 2012, though Inmarsat also potentially could have extended this date by 9 months in the event of an “excusable delay”. The other half of the block (1531-36MHz downlink) was part of the Phase 2 spectrum leased by Inmarsat to LightSquared for $115M per year, which was scheduled to be available in July 2013.

Inmarsat has indicated that under the amendment to the Cooperation Agreement signed on April 25, for which LightSquared paid an additional $40M, it will work to make this lower block of Phase 2 spectrum available somewhat earlier than July 2013. It is also possible (although unconfirmed) that Inmarsat may reduce the amount of potential excusable delay in making the Phase 1A spectrum available.

LightSquared has indicated that it plans to “be able to test its service in early 2012 and launch commercial services around mid 2012″. This would presumably imply testing in the Phase 1A 2x5MHz block and launching commercial service in the 2x10MHz block, i.e. Inmarsat would have no excusable delay in making the Phase 1A spectrum available and would advance the Phase 2 lower block spectrum availability by roughly one year. However, the lack of clarity around definitive dates tends to suggest that Inmarsat will be making “best efforts” to move up the Phase 2 spectrum availability but has not made any definitive commitment that it will be able to do so.

This would not be surprising, because Inmarsat has to accommodate its maritime and aeronautical customers who rely on its satellite network for distress and safety services (GMDSS) by fitting filters to these devices. I understand that Inmarsat-C GMDSS terminals use a block of spectrum around 1537-38MHz for their safety-critical communications (basically broadcasting distress messages and other alerts from the Inmarsat satellite to ships in the surrounding area), and I was told that some of Inmarsat’s terminals using this spectrum may not be frequency agile. Even if they are (per the comment below), then after the Phase 2 migration Inmarsat will concentrate its satellite operations within the 1536-45MHz range and so the 1531-36MHz downlink block has the highest potential for interference with these safety services.

Inmarsat also plans to close its Aero H and I services, which are apparently not compatible with LightSquared’s operations, and transition these customers to SwiftBroadband safety services (which will be better protected from interference). However, the SBB safety services are only expected to start flight trials in early 2013, which would then lead to safety certification during 2014. Whether this timetable is compatible with making the Phase 2 spectrum available at an earlier date is also unclear. Notably I understand that there may be as many as 1000 US government aircraft that rely on Aero H in North America, and so there may be considerable pushback if these users are forced to transition without either compensation or an approved alternative aeronautical safety service.

With respect to Inmarsat’s land services, there is no intention to protect these services (which include the John Deere Starfire GPS augmentation broadcasts that use the Inmarsat satellites). In fact I’m told that Inmarsat is specifically prohibited by the terms of its agreement with LightSquared from fitting filters to its land terminals to allow them to operate in the presence of LightSquared interference. The reason for this is that the uplinks from Inmarsat’s relatively high powered satellite terminals (such as BGAN or ISatPhone Pro) could produce significant interference to the sensitive LightSquared receivers on its terrestrial towers if they were to operate within range of a LightSquared base station. However, without an input filter these terminals will be overloaded by the terrestrial LightSquared signal and will shut down, thus preventing them from transmitting and causing interference to LightSquared.


  1. Clive Packer said,

    June 21, 2011 at 9:44 am

    Inmarsat-C terminals are frequency agile over 1530 to 1545 MHz / 1626.5 to 1645.5 MHz so there may be some scope for moving C service closer to 1545 MHz. When I was at Inmarsat in 1994 we moved Inmarsat-C frequencies in order to accommodate the then TMI Communications launch of MSAT services.

  2. krbarker said,

    June 21, 2011 at 12:20 pm

    We’ve done a quick examination of the interference created by the GPS receivers due to LightSquared carriers (indirectly caused by distrition of the LightSquared carriers within the GPS receivers, NOT directly by the presence of LightSquared), and esimate that the new frequency plan will significantly reduce the impact of LightSquared on GPS receivers (see http://www.questinygroup.com/qgi-blog). That does not mitigate the GMDSS and Aero issues mentioned above, but basically what LS now claims is true. This is, as many of us predicted, an admission that their Phase 2 frequency plan was untennable for the GPS community.
    God love LightSquared! They have provided some of the most interesting discussions for wireless engineers in years!

Leave a Comment

You must be logged in to post a comment.