Much talk in the mobile industry today focuses on how demand for services is concentrated inside buildings, such as homes and offices, with those two locations making up easily 70% and often as much as 90% of all mobile traffic.
While that’s true, there are other locations where people spend extended periods of time sitting still with their mobile devices and would like to get (better) access to services:namely trains and planes. We looked at trains extensively on behalf of the GB rail industry in a previous project, but have more recently been turning our thoughts to providing enhanced Internet and mobile services to planes.
Challenges in getting this working abound:
- Regulatory: Getting permission to deploy picocells on planes for 2G services took many years and involved careful technology development to ensure spectrum wasn’t used inappropriately close to the ground. There are deployments, but it seems that using mobile data is what people want on planes rather than talking on phones, so 2G seems a rather limited medium. Repeating this for 3G, let alone 4G/LTE doesn’t yet seem to have happened.
- Backhaul: There are quite a few services providing Wi-Fi on planes, but prices are fairly high and data rates not always impressive. One reason for this may be that existing services are mainly satellite backhauled using some fairly capacity-constrained satellites. There is a whole new generation of satellites which currently have significant spare capacity and which could help with this (e.g. Hylas 2), but keeping up with the rate of mobile growth is arguably not easy to do with satellites alone. On the other hand, satellites offer a great way of ensuring the entirety of the flight path is well covered even in some very remote areas.
- Demand: Infrastructure supporting aviation necessarily has to be planned and maintained many years in advance: for example, getting approval for a new type of equipment to go on board could easily take 5 years. This timescale is very much at odds with the rapid growth – and more importantly the rapid fluctuation – of demand forecasts for mobile broadband. Arguably it would be better for the capacity to serve such demand to be ground-located, where it can easily be increased and upgraded.
- Spectrum: Aviation frequencies are harmonised internationally and with very good reason: no-one wants the pilot of a plane coming into an international airport to be unable to communicate with air traffic control. But do the frequencies used for passenger data services need the same degree of harmonisation? And if so, which frequencies should they be? Perhaps they could be the same frequencies harmonised for LTE, allowing reuse of existing LTE equipment. Trials (such astheose by Deutsche Telekom and Airbus) have already shown that a fast-moving aircraft can be served via LTE from the ground. LTE frequencies vary hugely around the world however, leading some to suggest that it would be best to adopt dedicated frequencies for this application. Qualcomm is currently promoting the use of 14-14.5 GHz in the US and the FCC is examining this possibility.
Clearly there is a lot to consider, but much scope for a substantial opportunity for passengers, airlines and operators. It seems to us as wireless network folks that this is a problem involving an unusual profile of very wide-area coverage and very intense capacity and systems need to be efficient across all of these needs.
The video below shows an animation we put together of a day of aviation in Europe. Notice the significant routes across locations where terrestrial services would struggle to reach, yet very intense demand near hub airports, which might exhaust satellite capacity density. This suggests that a hybrid solution involving both satellites and a terrestrial network might well be necessary.
A Day in the Life: European aviation in the course of a single day