Spark spectrum buy tilts broadband power balance

On Thursday the Commerce Commission gave Spark permission to buy 70 MHz of the 2300 MHz band from Craig Wireless and Whoosh.

That’s a lot of bandwidth. It changes the balance of power in mobile spectrum holdings.

It could change the broader economics of broadband in New Zealand.

More fixed wireless broadband

Spark says it intends to use the spectrum to extend its fixed wireless service.

At the time of writing Spark offers rural users a fixed wireless service under its own brand. The company’s subsidiary, Skinny, offers a separate, low-cost fixed wireless broadband service.

Vodafone also has a rural fixed wireless broadband service.

Spark say it plans to expand wireless broadband services across New Zealand. A Commerce Commission press release announcing the decision confirmed this:

“…Access to the spectrum will enable Spark to provide a wireless alternative for rural customers and those urban customers currently unable to access fibre internet”.

— Dr Mark Berry, Commerce Commision chair

Those customers “Unable to access fibre internet” is part of the likely market for a new Spark wireless broadband service. So are customers unable or “unwilling to pay the high cost of a fixed line connection”.

Wireless broadband alternative

Thanks to a Commerce Commission ruling the wholesale price of fixed-line access in New Zealand is high by international standards. Wireless is a cost effective way for an ISP like Spark to bypass copper or fibre local loops.

When the government first decided to subsidise a nationwide fibre roll-out, fixed wireless broadband offered feeble speeds closer to dial-up than landline broadband.

Everyone knew wireless broadband speeds would improve, but the rate of improvement has been faster than most expected.

Fibre offers greater speeds and will continue to do so. It is more reliable, although that advantage is diminished with newer cellular technologies. Wireless, on the other hand, is more flexible. It’s easier for network operators to connect customers, whole suburbs can be connected overnight.

Fixed wireless broadband is subject to an entirely different set of regulations. This makes it attractive to the more wealthy service providers, especially those already offering mobile voice and data.

Need for speed

Spark has to move fast. Under the existing contract, if the spectrum rights are not used by the end of 2016, they expire. No doubt the government would love auction them again for a financial windfall.

Before it can use the spectrum for fixed wireless broadband Spark must enter a fresh Management Rights Agreement with the government. That’s a mere legal formality.

Spark must then commit to serving at least 30 percent of New Zealand’s population and provide services in at least 15 local authority areas. It already has the towers in place to meet that condition.

What 2300 MHz brings

While quickly building a new network shouldn’t be hard for a large telco, the cost of deploying in the 2300 MHz band is higher than, say, in the 700 MHz band.

That’s because 2300 MHz band signals cover less distance and are less able to penetrate barriers than signals at lower frequencies. 2300 MHz is microwave spectrum, that means it is line-of-sight. This could be an issue in hilly places like Wellington.

Long story short: Spark will need to use more cell sites to cover the same number of people with 2300 MHz services than it does for, say, 700 MHz services. And, there may be coverage black spots.

In most urban areas this won’t be a problem. You can take it as read, Spark will focus on the low-hanging fruit and service city dwellers first.

Shaking the broadband market

Despite these minor limitations, data services in the 2300 MHz band could have a profound effect on New Zealand’s broadband market. That’s because Spark now holds such a large chunk of 2300 MHz spectrum. A 70 MHz holding in this band is enough to deliver high-speed data.

Higher speeds than you might imagine. In Australia Optus offers what it describes as 4G plus services on the 2300 MHz band. To get an idea of its capability, last year Optus paired three 20 MHz blocks from this band with one from the 2600 Mhz band to deliver download speeds of more than 400 Mbps.

Spark’s New Zealand fixed wireless customers are unlikely to get that speed. For a start, Spark has less spectrum than Optus used: 70 MHz compared to 80 MHz.

Moreover, wireless spectrum is shared between users, so individual customers don’t often see peak speeds.

TD-LTE is the key

Nevertheless, In performance terms a Spark fixed wireless service will be competitive with most everyday UFB fibre plans. At least in terms of download speeds. Spark will probably use a version of TD-LTE. This is a standard which can be configured for asymmetric services. In other words, Spark may offer faster downloads than uploads.

Before long network operates should be able to offer gigabit speeds for fixed wireless broadband. The yet to finalised 5G mobile technology promises 10 Gbps. That’s not scheduled to drop until after 2020 at the earliest. In the meantime the 4.5G technology promises 1 Gbps.

Until now wireless network operators have been careful not to market their services as a direct competitor for land-based networks. That’s changing.

Fixed wireless economics

First, the economic balance is swinging in favour of wireless broadband. The cost of consumer equipment is low, a fixed wireless router is only NZ$100 or close.

Resource consents are needed for towers, but plenty are already in place. Despite simplified paperwork requirements, getting fibre to every home and apartment is a bureaucratic nightmare.

Likewise the cost of network equipment is falling. Competition to sell wireless network hardware is brutal. The biggest spenders at last month’s Mobile World Congress in Barcelona were the network hardware companies aiming to woo carriers.

The main barrier is not speed, but the volume of data. Skinny Broadband, along with the Vodafone and Spark RBI services have data caps while fixed line broadband contracts are moving to unlimited plans.

Wireless data caps

Skinny points out the 60 GB data cap on its Skinny Broadband plan is far more data than 40 GB or thereabouts consumed by the average household.

Data consumption is climbing fast, reports say it doubled in the last year. We haven’t heard the last word on fixed wireless data caps yet. Expect it to follow the same pattern as landline broadband data caps. These increased by leaps and bounds before unlimited plans became standard.

To date fixed wireless broadband hasn’t put price pressure on landline, copper or fibre-delivered broadband. That could be set to change. It’s worth remembering the regulated copper and fibre wholesale prices fixed by the Commerce Commission are maximums. If Spark fixed wireless takes off, expect to see landline wholesalers sharpen their pencils.