The Curious Case of Small Cells/HetNets

Posted · Add Comment

First a brief history of time… (props to Stephen Hawking)

I remember the curiosity and excitement for the future I had when I first went hands on with the mobile wireless data freedom that IBM field technicians had back (first recognizable outside IBM as the Advanced Radio Data Information System [ARDIS] 800MHz 2 way network) in the 1990’s. It was cool to me that the precursor device to the RIM/Blackberry, a small black keyboard with a simple display, allowed such real time information retrieval from the field. ARDIS was an interesting thrifty technology that relied upon retransmissions to ensure data throughput since the overall 800MHz network was limited in footprint. Special features like macro diversity and so on gave it amazing in building performance all things being equal. In the cellular world, the bag phone was about to give way to devices like the Motorola DynPJ-BG188_WORKFA_G_20120327182717aTAC 8000X (brick phone) just so you know where the rest of us stood.

At another interesting point, I was demonstrating revolutionary Short Message Service SMS(text messages) as this new amazing thing that came with the Global System of Mobile (GSM) communications networks. Carriers didn’t come running. They liked the exploding voice revenue and SMS was seen to them as lost in the noise of the balance sheet. A couple of guys like Eric Ensor, former President of BellSouth PCS shared my excitement but it was a long time until, well, the rise then the fall of SMS to the WhatsApp level of appreciation by the market.


So for the 1990’s – early 2000’s (I’ll skip all the false starts…)I basically held my breath just waiting for wireless data to be big and available.

Carriers claimed wireless data was not as important as mobile voice and just didn’t have any enthusiasm for mobile data until EVDO. They reluctantly started to deploy WCDMA-HSPxA/EVDO (known as 3G mobile data.) Viva EVDV… but that story is for another day. If memory serves, we finally started rolling out EVDO commercially in 2003.  Mobile data everywhere finally became reality but it was a long journey.

The one thing in Radio Frequency (RF) engineering that hasn’t changed much since the 1990’s is the perception/reality of the high cost to put signals indoors.

Indoor network performance was a major complaint by end users trying to access mobile broadband data indoors.  For this and the costs mentioned by the carriers previously, (at Samsung) we developed femtocells (now referred to as HetNets) and started to deploy them commercially in 2007. Carriers resisted using 1 of these 4 phrases.

(A) There’s no budget to buy a femtocell/small cell and give to users since it’s not marketing and not network’s fault (department that gets the blame pays)

(B) There’s no need with our [sub 1000MHz] spectrum/we can’t with our limited spectrum position

(C) Why deploy these when there’s WiFi?

(D) We are very interested but need you to price them to be cheaper than any known piece of electronics ever and they need mobile broadband data capability before we attempt this.

Note, there are some other issues like cost of infrastructure/effort of integration, internal support/logistics etc… but these are not the first lines thrown out.

Next important milestone: We were visited by a space alien that left behind something called an iPhone.

Ok, generalizing…but that’s how the carriers must have felt after putting up so much resistance and exerting so much control over devices only to be leapfrogged in so many dimensions by a Silicon Valley company, not a telecom company. Suddenly everyone was clamoring for a device that drove mobile wireless demand but the bad news was it required ubiquitous, high performance mobile broadband.

Now the carriers were fielding tons of questions and complaints about their indoor coverage and they quickly went through phases of alternatives such as justifying offloading their traffic onto WiFi, increasing their 3G data resources, or massive operational expense related optimization efforts but mostly Wifi was the tool that helped the most…which was a bit like falling on their own sword.

Darn, the carriers had some really groovy, big plans and investments into that real expensive and slow 3G data and voicesanta-workforce-tablet-analytics


So at this point…I mean c’mon, I waited for Santa for > 15 years to bring mobile wireless data….and got nothing to show off to my friends yet.

However, one of the gifts that came before LTE’s near ubiquitous high speed mobile broadband that we didn’t expect to be so good all around was Wifi. It was a game changer for networking at home and office. Just not very mobile.

Also important, in subsequent iterations, iPhones became better with and actively engaged WiFi as did Android. So one by one, globally, carriers began to realize that when data leaves their customer’s device onto a foreign/WiFi network, 3 things can happen and 2 are bad and zero involve additional revenue.

So remember, since 2007 carriers have this wonderful tool that is not WiFi that allows them all the benefits of protecting their network investments…the small cell, that facilities the creation of a HetNet and they have pretty much all tested or deployed them in very limited numbers.

Ok, that’s the history… here’s the curious part:

Where are the HetNets, the saviors of the carriers and customer expectations?

You may say they’re out there, but the reality is I had to become a Sprint, Verizon and ATT customer, then in some cases had to pay ~$300 each after begging and digging to get some of these devices into my home, and here’s the part that truly shows the carriers don’t get it…I churned off of some of the carriers and they turned their devices off!


Free ride for their traffic that makes their end users happy and they totally don’t get it. So I’m curious, what are we exactly waiting for?

In the meantime, Qualcomm has been also taken note of this as they could certainly use some large revenue stream, say of HetNet devices out there… so yes, they have been taking actions.

One of them is to promote a technology called LTE-U for LTE-unlicensed. If you look at the diagram from their website, you can catch the drift that it’s a HetNet but could potentially remove the roadblock (called the carrier) in the deployment path to the end user. It does this by allowing unlicensed spectrum (as used by WiFi or home cordless phones) to be used by your device. It’s 50/50 strictly speaking, if you need the carrier for this, but the tech could help if you add their spectrum and your locally unlicensed spectrum into a bigger heap for faster speeds. The problem is it’s a bit of a replacement for WiFi, not a seamless partner. You would struggle to operate these devices side by side as is defined today.

Not to say problems are insurmountable or this technology is a bad idea, no, it’s a good idea, but is it worth waiting for this? Investing all the time and money in development, standardization, testing and deployment to wait for this? What in the heck is the deal with carriers and HetNets?

If I wanted to disrupt the carriers, I would simply bundle at HetNet device with new devices for free or subsidized based on usage based incentives (upfront free.) I could offload a crazy amount of my start up network’s limitations, problems and expenses.

Qualcomm could take those recent investments into the consumer silicon companies such as Atheros and make some crazy new revenue streams by combining these things together, a cheap WiFi/HetNet chipset so every device is always maximally connected.

Network operators: OK so you missed supplemental services like SMS, Over the Top services like YouTube and Skype, WiFi, iPhone and the small cell…What happens to every single organism on Earth that fails to adapt?

10 more years?