Your browser is not supported. Please update it.

3 July 2014

5G for cellular just a fat-pipedream

We often say this, but if you have a long memory then you can see the future. Ericsson, Nokia, Qualcomm, Huawei and Alcatel, to name just a few, have a mission to retain the cellular status quo for as long as is humanly possible.

But consumers, once they taste what is possible on a network, want at least that all the time and they want it cheap. Consumer pricing, like water, has a way of flowing down the easiest path.

The cellular community wants to keep subscription monies rolling in at the current $45 to $70 a month levels which cellular subscriptions run to. But much of that payment is spent keeping the CPE up to consumer performance expectations and keeping it free, and very little of it is used on improving the network or even on actual delivery costs. Smartphone handsets are now replaced routinely in under 2 years, as we move to an era which expects the network to cope easily with video.

Of late we have seen statements made about 5G cellular from both Ericsson and ZTE, although it is interesting that Broadcom already calls 802.11ac 5G WiFi, as if the same technologies are already here.

There are many parallels to be drawn here between the two 5Gs, including MIMO and SON for instance, but Ericsson, taking up a large unspecified chunk of the 15GHz band to find 5Gbps, in a laboratory environment is nothing like the daily performance of WiFi in the real world. If it rains, then a 15GHz signal gets blocked or slowed to a crawl. Ericsson this week has cutely intimated that 5G cellular networks would be with us by 2020. Oh yeah? How on earth is that going to happen?

Ericsson also talked about M2M being enabled by 5G. Again we wonder how. M2M is mostly about one device sending a few bits of information to another, over a low power network, just a usage number or an instruction to close an actuator, or a sensor with a tiny element of data to volunteer, because it is over a threshold. The whole myth that this will include free flowing 4K video, is a massive piece of over-hype. Let’s cope with Internet 1.0 uses of video first, when a person wants to interact with video, before we take on the potential for two machines to send it to one another, for absolutely no benefit whatsoever.

One of the things that is already laid out in the 3GPP view of the world, is that there will be more and more tiny cells used for increasing network capacity, these come under a number of technologies which make up LTE Advanced. Essentially if there are lots and lots of smaller cells, they will interfere with one another and any given handset will be in reach of many of these, and not know which is best to use. A bit like WiFi. So cellular has to solve the problems that WiFi has already solved.

Cellular adopts a strategy of complete control and certitude, centralizing this type of decision. It uses dynamic interference management which is espoused in the 3GPP idea of eICIC (Enhanced InterCell Interference Coordination) and the idea of coordinating which device talks to which base station through CoMP (coordinated multi-point networks). If you add virtualization to this and Active Antenna Systems, where the heavy processing of a base station is separated from the antennas and their orientation and their orthogonality being actively managed, then you have most of the benefits for LTE Advanced. Whew. You also have a nightmare of centralization and complexity.

Occam’s Razor tells us how we should treat complexity, even if it has not been mathematically proven. The more complex a system is, the more likely it is to go wrong. The more assumptions it makes, the less perfect it will be. And it seems to us that each of these new layers of complexity will take time to bed in. A lot of time. And not be complete by 2020, not close.

Another thing that will take a long time are the creation of a huge number of backhaul points. It is not only that there will be more small cells than there were macro cells, by an order of magnitude, but the connections between elements of RAN have to be made, between antennas and the demodulating logic. This would take a long time to build, even if cellcos had an unending supply of cash, but given that this cash has to come out of revenues, it will either take an inordinate amount of time to fill out, or it will mean a rise in the cost of cellular services and we all know that is not going to happen.

All of this compares with humble WiFi. It has interference orchestration built in, it has graceful degradation and MIMO already in rollout, and it has multiple sources of wealth to pay for its existence. It is always backwards compatible. And it has already gone past 5 Gbps, with 7Gbps demonstrations. Largely it is the power limitations of WiFi that prevent interference, and some advanced implementations have dynamic channel hopping to evade interference further. It has a simplicity, in that each Access Point it is a network on its own, and the rise in complexity for WiFi will be within multiple forms of authentication, from multiple points in the cloud, as aggregators, MNOs and site owners swap rights to let their customers be authenticated for seamless WiFi use. That’s an acceptable layer of complexity.

Cellular wants to take these advantages away from WiFi. It wants to borrow the technology of MIMO and its ability to self-organize, while retaining its ability to centrally control, manage and bill all of its clients, as well as authenticate roaming, as well as billing while roaming. And all along it is planning to install all of these new technologies that take it towards LTE Advanced, and yet still contemplate what’s next in 5G, which we presume will have more continuous upgrades required, as networks become hyper dense and there are multiple base stations all around us.

The thing that doesn’t sit well with us is why consumers will go along with this and pay for it. Smartphones have had to have WiFi chips put in them because of the weakness of in-building penetration of cellular, and the high price of cellular data. WiFi has now taken about 70% to 80% of all data usage off the cellular network. That percentage can only rise, because at home and at work are the two places where you are confident that you will not have to pay for cellular, and where you know that the WiFi is free. As video viewing shifts to smaller devices (we get more and more evidence of that each day) this percentage in terms of raw data will get higher and higher. It will breach 90% in no time at all.

So when you are paying $70 a month for a cellphone subscription, which includes all you can eat data up to an LTE cap of say 5 GB per month or higher, you will find you are using 50 GB or 100 GB regularly, but most of it over WiFi.
At this point you will be ‘encouraged’ to use cellular data when you are out of home, in order to keep your bill that high. WiFi First services will make it fairly easy to go along with this, for instance using pre-pay for cellular, so that you cannot get bill shock.

This will only accentuate how little data you are getting for your cellular money, and how much you are getting from fixed broadband attached to WiFi, and this will breed resentment, which is what continuously drives down cellular pricing. It will lead to people re-assessing their cellular needs and limiting it to voice and some social media data spending. That $70 a month will drop, to perhaps $50 or maybe $40. Which means that over time consumers can cut cellular spending by as much as 40%, and if that happens to all the data ‘high rollers’ as well as average cellular contracts, what need will cellular have to introduce 5G by 2020. And anyway they won’t have the money to spend that amount of capex. They will need a smart development plan than the ones the engineers are planning for them.

The single most important thing that cellular operators can do right now is to treasure their relationship with customers. When a customer is roaming to another operator, fight on his behalf to keep the data roaming bill down, instead of gleefully jumping up and down at the $200 bill shock for using data in another territory, and sharing the spoils with the local incumbent. When customers get close to their data cap, cut them off or inform them, or put them on another cheap rate, instead of a punitive one, or better still, offload them to Wifi. When data is seen as a universal right that it should be possible to use everywhere, cellular will once again be a threat to WiFi. Until then its dreams of there even being a 5G cellular network are nothing but pipedreams.