Monday, April 4, 2016

It Takes a Village

Delivering a consistent Internet experience to an airline passenger demands adequate provisioning for the peak periods (highest contention ratios).  It may take many satellites with overlapping coverage to aggregate sufficient capacity to meet the demands in any "hotpot".

One satellite lights up a given region with a finite amount of spectrum - simply the capacity available from a single beam from one satellite.
Satellite downlink spectrum powers the forward channel that delivers data to the passenger. At least 500 MHz times two polarities (1000 MHz) is available to any Ku-band or Ka-band satellite operator. Some operators may be able to access as much as 2000 MHz time two polarities (4000 MHz) for downlink.

In the future, the average data rate to a user will eventually approach 1 Mbps or more.  While the timescale is not fixed or certain, the eventuality is.  Combining such a high average throughput with high take rates (50%) and with a large number of airplanes operating in close proximity creates a capacity density that is not serviceable from one satellite beam (a hotspot).  Four airlines in the US could aggregate a need for 14,000 MHz to serve one hotspot.

A need for 14,000 MHz into one region may require many overlapping beams from satellites in multiple slots across the geostationary arc.

It is debatable whether the passenger usage in the terminal area favors interactive applications Versus more demanding applications such as streaming.  This aspect could moderate the peak at an airport, albeit it is not apparent what the usage scenario will be in the future.  Airplanes flying along congested air-corridors also have the potential for creating a traveling hotspot.

Low-Earth-Orbit (LEO) and Mid-Earth-Orbit (MEO) satellite constellations have the potential for overlapping beams from adjacent space vehicles, depending on the inclination of the orbit.  However, it is not apparent any non-geostationary (GEO) satellite can offer more than one satellite to serve a given region - no overlapping services.  Thus LEO and MEO satellite constellations are fundamentally limited in their ability to aggregate capacity to serve a hotspot.  OneWeb will use GEO satellites to fill equatorial gap in coverage, and may offer supplementary service elsewhere, but that is not part of the OneWeb constellation.  My thoughts on MEO are projected around a honeycomb coverage pattern, not a small set of steerable beams.

Spectral assignments for LEO and MEO satellite constellations in Ku band or Ka band are considered non-geostationary (NGSO) and purposely managed to not interfere with GSO spectrum when the NGSO satellites travel across the geostationary arc. While GEO and NGSO satellite networks operate exclusively, they each may utilize some spectrum from the opposing network under favorable conditions.

For example, ViaSat uses a 500 MHz NGSO allocation while no O3B satellite is in view. OneWeb plans to use the GSO Ku-band allocation while avoiding interference to GSO satellites by excluding parts of the coverage (equatorial) and otherwise offering significant diversity in the beam pointing between GSO customers and OneWeb customers.

The collective demand from all the passengers on all the airplanes in a continental region can be met by just one satellite, but that one satellite will not meet the needs of the hotspots in the future.

In summary:

  • Passengers will demand higher and higher average data rates
  • Combining many passengers and many airplanes into a small region creates hotspots
  • A single hotpot may require many overlapping beams to aggregate adequate capacity
  • LEO and MEO satellite constellations hotspot capacity is likely limited to a single beam

Peter Lemme

Copyright 2016
All Rights Reserved

1 comment:

  1. Interesting. Regarding MEO/LEO (amongst the various players) I believe that one can conclude from public sources that some DO plan to have multiple spotbeams (from multiple satellites) over "hot" locations....and some don't...