Bandwidth: just how much is enough?

Ian Miller, spectrum and technical coordinator for ARCIA, has responded to the ACMA’s recent article on its approach to the emergency services spectrum. The authority has said it will allocate 10 MHz in the 800 MHz band for public safety mobile broadband.

For several years now there have been discussions on the requirements for emergency service organisations (ESOs) to have ‘mobile broadband’ facilities to continue to provide for our safety using ‘world best practice’ methods, something we all have every right to expect.

These discussions have come to the fore over the past couple of years as the government, through the Productivity Commission, has directed the Australian Communications & Media Authority (ACMA) to generate more revenue from the electromagnetic spectrum.

As a result the ACMA has been reviewing selected frequency bands within the spectrum to increase the efficiency of their use and so allow for more productivity. In general this is a good thing and is welcomed by many sectors of industry.

Some two years ago, the ACMA released a discussion paper outlining options for the review of the ‘900 MHz band’, a sector that contains various user sectors and gave the appearance of perhaps being not as effectively segmented as it could be.

Within the initial discussion paper released in July 2011 there were plans for the revision of the band; however, at that stage there was no mention of any provision being made for ‘public safety mobile broadband’ (PSMB) and it was only after representations from the ESOs and industry that the ACMA had to revisit the proposals.

As outlined in several of the comments in recent days, there is very little spectrum allocated to PSMB use around the world at this stage. Even in the US they are still working their way through the issues and, although they have 20 MHz of spectrum available as part of their ‘digital dividend’, many of the systems being installed are still not much beyond the evaluation stage and actual products for use by the ESOs are still in the developmental stage. Because of the product still being in the early stages, it is hard to find any real ‘experts’ to speak with authority and I certainly don’t claim to be one.

The ACMA has made several statements regarding the allocation of 10 MHz of spectrum and supposedly its ‘experts’ have decreed that it is more than sufficient for the ESO requirements. At a very basic level I am inclined to agree with them; for normal usage, and with the proper infrastructure in place, it would seem that there will be some latency in the system over and above normal usage requirements.

Perhaps I am showing my age a little, but I have good memories of a popular song from the brothers Finn back in the early 1980s. It was called Six Months in a Leaky Boat and contained one lyric that is quite pertinent - “the tyranny of distance”.

This is quite important, as the laws of physics outline that as you move away from the fixed infrastructure point the data rate to a mobile unit falls away. Try it with your own Wi-Fi connection - move away from the service point and watch the data rate fall down fairly quickly. The same will happen with any PSMB system, so as the vehicles move away from the main transmitter location the data rate will fall. Effectively the 10 MHz will quickly be down to 5 MHz, with message retries and error correction techniques.

This can be quite easily overcome with more transmission sites but that is an expensive option (as outlined below). So now the 10 MHz that seems to have latency in-built has suddenly become only just sufficient at best.

Maybe we should look at what the situation is in the US. They have been allocated 20 MHz of spectrum; however, our regulator contends that with a smaller population and perhaps less requirements we don’t need as much spectrum as the US has allocated. In very simple terms that sounds feasible, but when you consider that our major cities of Sydney, Melbourne and Brisbane/Gold Coast would rank within the top 20 if located in the US, and then calculate that our population density in our cities is less per square kilometre than most of the similar sized cities in the US, the comparison changes dramatically. Maybe we have a greater spectrum need than many locations in the US?

The ACMA (in information released this week) outlines that it believes 10 MHz is plenty. As indicated above, there should be some concerns at that blanket statement. The ACMA has also stated that it has allocated 50 MHz of spectrum at 4.9 GHz for PSMB usage. This is to be welcomed but it is also the proverbial ‘red herring’ when you go back to those laws of physics again.

The next concern is that at similar power levels, as the frequency of operation increases the distance the radio waves will travel decreases. This means that although there is a large amount of spectrum allocated, the actual costs to use it over a wide area like one of our major metropolitan areas will be enormous - far more than for the 800 MHz sector.

In addition to the above segments, the ACMA also outlines in its release information that the ESOs have 25 MHz of spectrum in the 400 MHz band they can use, but again this is not fully correct; the frequencies at 400 MHz are for all government users and carry the voice communications which, at this stage, are absolutely essential for any ‘mission critical’ organisation and cannot be compromised. This section of the spectrum might be able to be used for some data communications in some areas, but in real terms it would only be a stop-gap arrangement.

Perhaps it is time to pose some questions that may or may not be pertinent:

• Given that long-term evolution (LTE), which is the data transmission format that will be used for PSMB, is still very much in its infancy, should we just assume that ‘today’s experts’ are going to be correct, or will time and product development change the vista completely - maybe for the better but also maybe for the worse? I am sure we all remember government statements like “there will be no GST” and “there will be no carbon tax”. Hopefully statements about the absolute suitability of 10 MHz of spectrum for PSMB won’t also be proven wrong in the long term.
• Why are the public communications carriers (Telstra, Optus and Vodafone) so very keen to have access to the 800 MHz spectrum? The answer is, of course, that it provides them with the business case for providing mobile broadband to their users with best possible coverage at lowest infrastructure cost. This raises the question: should we be allowing this ‘sweet spectrum’ to be auctioned off to commercial interests who generate revenue to increase their infrastructure requirements, or should we reserve it for public safety where the governments have to fund the infrastructure costs?
• On the basis that the geographical footprints of our major cities are equal to or greater than any in the US, why would we believe that less spectrum will suffice to give greater coverage? If it is not correct and there is a need to increase coverage with more infrastructure, to get from a 5 MHz availability back up to the 10 MHz for efficiency I estimate the cost will increase five-fold. A huge investment will be required.
• Should our government and the regulator (ACMA) grab the few million dollars from a short-term auction of the ‘sweet spectrum’ and ignore the potential long-term investment of many billions of dollars if the ‘expert opinion’ is not correct?

Like I said before, I don’t claim to be an expert; just a grumpy old man with 50 years of radio experience and a healthy cynicism of ‘experts’ deciding on such new technology requirements and not having anything in reserve, just in case. Let me hark back to that song again - it was about spending a long time in a leaky boat. We don’t want to condemn our PSMB project and ESO users to that situation.

Ian Miller is an independent radio communications consultant at Orange Horizons (www.orangehorizons.com.au) with many years of practical experience in radio communications. He is also the spectrum and technical coordinator for the Australian Radio Communications industry Association (www.arcia.org.au).