Posted on

Are National Broadband Networks Doomed?

Over the years, I’ve worked with national broadband projects in Australia, New Zealand, Qatar and Singapore. More recently in Australia, the National Broadband Network (NBN) has been in the news for all the wrong reasons. Their Retail Service Providers (RSPs) – who sell the NBN services to end customers, have been up in arms about NBN approaching large enterprise customers directly. Today, they announced that they would no longer do that – and that has the RSPs like Vocus, Mactel and Telstra very happy. Arguably, the decicion for NBN to sell direct was in breach of the founding principles that the Australian government put in place when it created NBNCo.

Such controversy is not why I think the NBN and the equivalents in other countries are doomed, although it’s not helping their case in the eyes of the public and end customers.

No, I think the proliferation of 5G networks and more recently global players like SpaceX’s Starlink constellation could be the harbinger of death for NBN.

Slow Rollouts

NBN has been copping a lot of flack lately in the media for taking too long to roll out. I get it, Australia is a HUGE country – even with most Australians living within an hour of the coast, it’s still a lot of physical ground that needs to be covered by the fibre and HFC networks that cover the bulk of the NBN end users. This has lead to a level of dissatisfaction with NBN as a whole.

Slow Network

Those end customers that do have a NBN connection are often complaining to the telecommunications ombudsman about the service they get – and while some of those faults are laid at the doorstep of the RSPs, some of it is due to physical breakages of modems and network termination devices and some are the fault of NBN – in all cases, because in Australia, we include NBN in the product offerings of the RSPs (ie – its customer facing), NBN cops the blame for ALL of the issues. As an example, my RSP (Optus) sold me a 100/40 HFC based NBN connection – which is usually fine. I often get 90-95 Mbps downstream and 30-37 Mbps upstream. However, so many HFC customers were seeing much slower than advertised speeds that Optus removed that speed combination from the market – the fastest they sell now is 50/20. (50 Mbps down, 20 Mbps up).

5G Networks

The 5G rollout in Australia is still pretty limited, but the 4G (LTE) rollout is pretty comprehensive and on 4G, I often see speeds approaching my home NBN based connection. Assuming 5G will bring a significant boost in speed (along with many other advantages including much great density of connections per cell) – which means that a 5G connections promises to deliver faster connections than NBN and without the need to tie the end customer down to their home boundaries.

If you add unlimited plans (in terms of Gb to be transferred up or down) to such as 5G (or even a 4G service) then you have a strong competitor to the NBN.

Some local mobile network providers and even MVNOs are already talking about selling fixed mobile services instead of selling a NBN based home (or office) connection.

Starlink

This morning, SpaceX launched another 60 satellites into orbit, bringing the total to 240 – that’s 120 new satellites within a month – well on the way to 12,000 satellites.

As I’ve mentioned in my previous blog post (see https://telcotalk.online/index.php/2020/01/09/starlink-a-global-csp-disruptor/), SpaceX’s Starlink constellation of communications satellites promise to deliver broadband (up to 10 Gbps) AND low latency (good for gaming) to 100% of Australia (other than the Australian Antarctic Territory). If SpaceX can deliver reasonable plans (in terms of speed, capacity and price) then SpaceX will be a strong competitor for NBN. If the plans are right, it could kill NBN.

Two NBN alternatives – either could kill NBN

Sure, NBN in Australia is facing some significant challenges, but these are exactly the same challenges that all national broadband networks/project face… Customers have zero allegance to NBN – and if 5G or Starlink will provide faster speeds at a competitive price, NBN is doomed.

If you disagree, let me know what you think…

Posted on

Starlink – a global CSP disruptor

As SpaceX reach 180 starlink satellites in low Earth orbit; on the road to 12,000 once the network is complete, it’s becoming increasingly apparent that Starlink is set to become a global disruptor in the telecommunications industry. Once complete, SpaceX’s network will be nearly global (except for polar regions) and will be ubiquitous. The diagram below illustrates the coverage area – basically 100% coverage of most of the populated regions of the Earth :

Starlink network coverage (approximate)

The big benefit that the Starlink satellite constellation over a traditional communications satellite sitting in a geostationary orbit is the time it takes for the signal to get from a point on Earth to another Point on Earth. The Starlink constellation at an altitude of 550km is much closer to us than a geostationary communications satellite at (approx) 35,700km. The Starlink constellation will transmit information between Starlink satellites before downlinking to the destination. In an example of a connection between Australia and the Middle East, the diagram below shows the Starlink connection in White, a Fibre (mainly undersea) connection in Red and the geostationary communications satellite connection in yellow. This is approximately to scale.

A scale comparison of Starlink vs Geostationary Satellite and Fibre connections (Melbourne to the Middle East)

The white Starlink connection can pretty closely approximate a great circle path and thus is a shorter path than the red fibre connection (which must travel where the undersea cables have been laid). Yes, there are potentially more hops in a Starlink connection than a traditional satellite connection, but the distances involved are MUCH shorter. The other issue with traditional geostationary satellite connections is because of the distances involved, the signal strength is relatively very weak and because of that, these signals suffer from higher data loss, which means that these comms don’t just use standard TCP/IP, rather they use protocols that have much greater error correction and parity capabilities – this has the cumulative effect of slowing down the connection. Combine this slow connection speed and high latency (because of the distances involved), traditional satellite carriers have a big challenge ahead of them when compared to Starlink which promises to deliver much greater speeds at much lower latency – competitive with fibre for all but the largest bandwidth consumers.

To give you an idea of the comparison between Starlink and Geostationary satellite communications, I built a quick animation – remember this is only half of the connection (one way only) ; a real connection would be double the times as the response needs to back to the initiator.

A visualisation of the speed difference between Starlink and traditional geostationary satellite communications

What about for connections between two systems in the same country, as opposed to international connections? Even local in-country CSPs face significant competition from Starlink.

If we look at the Telstra coverage map for Australia+, easily the largest Telco in Australia with the best coverage, and yet we have lots of space that has no coverage at all. Contrast that with a 100% coverage that Starlink would provide with lower latency and much higher throughput… what do you think ? Will Telstra or any other CSP in the world face a challenge from Starlink? I think so…

January 2020 Telstra Mobile Coverage Map

+ Yes, I know Australia has a very large area and has a relatively small population, so the problem is not as big in other countries, but just imagine 100% coverage 100% of the time and in any country you ever visit…

Posted on

SpaceX launch new batch of Starlink satellites

I’ve just watched the SpaceX launch of the latest batch of 60 starlink satellites into low earth orbit – aimed at providing low latency internet services all over the world. Initially, SpaceX are targeting the North American market – I mean, why wouldn’t they? The US has such a disjointed connectivity marketplace with a mixture of Metro Area Networks (WiFi and Wimax based) in small towns, LTE/5G in larger population centres, HFC cable and Fibre connectivity options for fixed services and probably still a bit of xDSL running around… Not to mention the oft complained about mobile network coverage. Starlink (despite being Internet rather than voice focused) has the potential to steal a lot of the subscribers that live in or travel to marginal coverage areas. Think of it – 100% coverage of North America at up to 10Gbps – if the price is competitive, why wouldn’t you as a subscriber go with that option!

There were a few things that peaked my interest with this launch in particular:

  • The launch of these Starlink satellites in close succession from the December’19 launch of the Kacific comms satellite (ironically on a SpaceX Falcon 9), a more conventional geostationary communications satellite, targeting at providing services to the South Pacific, SE Asia and Himalayan nations (not Australia) via Ka band radio (thus the name). They plan to provide services to over 600 million subscribers – from the following countries (from https://www.kacific.com):
    • American Samoa
    • Bangladesh
    • Bhutan
    • Brunei
    • Cook Islands
    • East Timor
    • Federated States of Micronesia
    • Fiji
    • French Polynesia
    • Guam
    • Indonesia
    • Kiribati
    • Malaysia
    • Myanmar
    • Nepal
    • New Zealand
    • Niue
    • Northern Mariana Islands
    • Papua New Guinea
    • Philippines
    • Samoa
    • Solomon Islands
    • Tonga
    • Tuvalu
    • Vanuatu

Obviously, the bulk of those subscribers are going to be coming from Indonesia, being the highest population country in their target list. It makes me wonder about the competition between Kacific and Starlink for those same subscribers once SpaceX establish their services in the north American market and spread their wings to the rest of the world…

  • The Starlink swarm of satellites have had astronomers up in arms because of the additional light and radio pollution these satellites have been adding to the night sky making it difficult for both visual and radio astronomers to get good observations. With more than 12,000 (!!!) Starlink satellites planned to go into orbit, we’re just seeing the beginning of this problem.
Telescopes at Lowell Observatory in Arizona captured this image of galaxies on May 25, their images marred by the reflected light from more than 25 Starlink satellites as they passed overhead.
Victoria Girgis/Lowell Observatory – image linked from astro.princeton.edu

I noted during the latest launch coverage, the SpaceX presenter said that one of the satellites launched today had been ‘darkened’ to reduce reflections in the hope that it would lessen the affect on visual (at least) astronomy. Let’s hope it works.

If you want to read up on the Starink’s effect on Astronomy – I’d suggest you read this article on Nat Geo – https://www.nationalgeographic.com/science/2019/05/elon-musk-starlink-internet-satellites-trouble-for-astronomy-light-pollution/

For sure, these launches are great to watch and remind me of when I watched Apollo 17 launch as a boy (that’s the only one I remember from way back then) and the excitement I felt when I watched that launch…