By 2020, 90% of All Wi-Fi Hotspots Will Support "Carrier Grade" Access

By 2020, it is possible that as many as 90 percent of all Wi-Fi hotspots will feature “carrier grade” quality-of-service, according to Fon. That does not mean those hotspots will “only” support carrier-grade service. Best-effort-only internet access also will operate.

But networks that bond mobile and unlicensed spectrum to support access to carrier or other managed services will operate in quality-assured mode. That might not be a direct rebuke of network neutrality rules that mandate “best effort only” for consumer internet access.

The new qualification is that the bonded carrier access represents use of unlicensed capacity to support a “managed” service that virtually all network neutrality rules exempt from the regime. Also, network neutrality rules virtually always apply only to consumer services, not business services. And managed services are, by definition, business services.

So licensed and unlicensed spectrum will both be necessary, and will coexist in the world of 5G, says AT&T Labs AVP Dave Wolter, while unlicensed spectrum increasingly will achieve “carrier grade” performance.

That especially is illustrated by the development of LTE-U, LTE Licensed-Assisted Access (LAA) and MulteFire, all part of the LTE unlicensed family of protocols. Note that where LTE-U and LTE-LAA are designed to work with LTE, MultiFire actually is designed to be a stand-alone protocol operating strictly over unlicensed spectrum.

source: Fon

LTE WLAN Aggregation (LWA) is another proposed way of combining licensed and unlicensed spectrum while preserving quality of service.

Those developments neatly illustrate why some have objected to the “best effort only” requirements imposed by net neutrality rules. Though such rules inevitably say reasonable network management policies are allowable, nobody ever can figure out what that means, in practice, as “maintaining quality of experience when the network is congested” nearly inevitably involves some rationing of customer use of the network.

And, by definition, that means active blocking or traffic shaping, which is a violation of net neutrality rules.

In arguing that “carrier grade performance” is to be supplied over a mix of licensed and unlicensed connections, one has to admit that “quality” requires management of previously unmanaged resources. In other words, “best effort” does not work for carrier services, that will apply to use of Wi-Fi or other unlicensed spectrum to support carrier services using those local assets.

But if “best effort” does not work, then neither do network neutrality policies that mandate that is the only lawful way consumers get access to internet services.

To be sure, carrier voice and other services are not “internet” apps in a strict sense, even if they use IP networks and protocols. But the differences between managed services and internet “best effort” services are becoming quite muddled, as bonded licensed and unlicensed resources are used to support services that do have a quality-of-experience dimension.

In its original “weak” form, network neutrality was an appropriate framework for ensuring customer access to all lawful apps, without blocking. In its “strong” form--outlawing quality of service mechanisms--the concept inevitably will be eroded.

As a method for dealing with other issues, especially antitrust or predatory behavior, network neutrality is a hammer. But not every internet freedom issue is a nail.