Monday 10 November 2014

Small Cells and/or WiFi - The confusion continues...

Its been an interesting last few weeks. Depending on which report you read, you will either come to the conclusion that 'WiFi will be killing off Small Cells' or 'Small Cells will be killing off WiFi'

First there was this report that "You might not need a mobile carrier by 2020". It makes this bold statement: In Europe, many cell phone owners have already ditched their wireless carrier. But Wi-Fi isn't quite widespread or robust enough for most Americans to completely ditch their wireless carrier just yet. In all honesty, I have never come across anyone that has ditched their mobile network operator and now relies entirely on Wi-Fi. I certainly know of people who now don't even bother switching on their WiFi because their cellular coverage is extremely good and have flat pricing.

Joe Madden, a respected analyst of small cells, recently said the following: "Even if we exclude homespot deployments, the number of Wi-Fi access points will reach the level of millions for cable operators and public venues during 2015, outstripping the capacity of new LTE base stations. Several large mobile operators have made a gigantic blunder, by ignoring the opportunity to deploy Wi-Fi or utilize Hotspot 2.0 –so cable operators and other service providers are jumping on the opportunity. Homespots add another dimension, with massive crowdsourcing of capacity. The total Wi-Fi capacity deployed by service providers worldwide could match the 'data tsunami' in terms of raw capacity over the next five years, although of course there are obvious limitations in mobility and QoS."

While you may be thinking Wi-Fi '1' and Small Cells '0' a thing to remember is that WiFi still has some way to go to sort out the security stuff. This article highlights how easy it is spoof a WiFi AP, the one you have trusted in the past and easily access personal Info. I strongly suggest that you read this article. One may argue that some of these issues will be gone with HS2.0 and other new security mechanisms these problems will vanish. One has to remember though that since WiFi uses unlicensed bands, and since the technology has been around for ages, its easy to get cheap equipment and it may not exactly be illegal to have equipment running in this band.

Cellular on the other hand relies on licensed spectrum and has a very strong authentication mechanism which may get around such basic insecurity info (though to some extent this can be hacked, depending on operator policies on the UICC/SIM card).

Dr. Kim Larsen, recently did a presentation where he looked at the economics of Small Cell and WiFi and in what situations both of these make sense. His presentation is embedded below.

Some thoughts from Kim on his presentation on Twitter:

  • Most Smartphone based WiFi traffic happens at Home, believing this traffic is offloaded is rather foolish!
  • WiFi...Why & When to care (at least when you are an mobile operator)
  • Why do we like WiFi so much & why cellular have so many challenges matching consumer expectations!
  • WiFi has the consumer perception of being 1 Fast, 2 Almost Free & 3 Unlimited...Brilliant Branding!
  • Mobile Operator WiFi off-loading strategies should consider mitigating potential & substantial cellular revenue loss!
  • When WiFi makes the most sense for a Mobile Operator; 
    1. Cellular expansion options have been exhausted!
    2. you control fixed & mobile sides of the customer experience & value chain! 
    3. Competitive Pressures .. ultimately is likely to be a loss-loss scenario!


  1. Manoj Das (Via Het Net Linkedin Group)11 November 2014 at 08:29

    Thanks Zahid. This is an excellent slide deck. One thing I am not convinced is: Traffic going over WiFi, over the Internet through the ISP to EPC of service provider to finally reach the Internet (Slide#24). Definitely this is not efficient model for data delivery. LIPA and SIPTO kind of methods have to be incorporated with WiFi AP, once 802.1x auth is done, the traffic can be offloaded to the Internet locally. That would save the cost per bit for the WSP as well as much less latency for the user. Please share your views.

  2. Manoj, agree that its not a very efficient model but that is what is there as part of 3GPP trusted WiFi using TWAG. This way you can access all the services on your mobile as you would when camped on cellular network. The other approach for untrusted uses ePDG. One could argue that certain flows could be offloaded with SIPTO but in some countries there is a very strong requirement for lawful interception which may not always be suitable in all SIPTO cases. This may have changed since I last looked at it, which was a while back.

  3. Michal Jarski (Via Het Net Linkedin Group)11 November 2014 at 13:55

    Powerful solution shall enable the operator to include all those options in one system. Based on the authentication of the user (EAP-SIM/AKA or EAP-TLS/TTLS for non-SIM devices) and their level of subscription their traffic could be either dropped locally (for added capacity and lower cost, e.g. for prepaid customers), dropped at the wireless gateway (to keep all regulations in place) or tunneled all way up to the 3GPP core (for deep packet inspection or easy roaming between 3G/4G and WiFi with keep alive IP transitions or uninterrupted WiFi Calling / VoLTE).

  4. Hi Zahid, thanks for highlighting this issue. I believe that in the long term, small cells and WI-Fi will be combined using License Assisted Access (LAA) and we will have a lot tighter partnerships between fixed and wireless operators. Right now, the deployment of Wi-Fi is ORDERS OF MAGNITUDE larger than small cells, so the mobile operators that are sitting out of the indoor market are missing an opportunity to provide better service. Mobile Experts plans to publish a series of reports on security, cloud/core integration, and semiconductor implementation over the next year so that we can explore some of the details that you refer to.

    A year ago, we thought that small cells including Wi-Fi would cover a high percentage of in-building cases. Now, we've changed our view because of the slow motion deployment in small cells and the rapid deployment in Wi-Fi. The final verdict is still unclear, but we now see competition where before there was convergence.

  5. Claus Hetting (Via Het Net Linkedin Group)12 November 2014 at 14:59

    I think it is a strong POSITIVE that we've now arrived at the point of 'confusion'. I mean - thus far it is has been all 'small cells' from the telecoms industry side. From my point of view (and other analysts) the fog is in fact starting to clear :)
    I still have yet to see what small cells there exists to be killed off by Wi-Fi. In this country (Denmark) there are certainly close to zero. In many other countries, it's the same - there are none or close to none. In a few big markets, yes - AT&T, etc. still have some money to do this. But it's a very far cry from what's been predicted.
    On the other hand we've just seen Ericsson release the details of a 30k AP deal with a WISP in India, for example. I mean - I've not seen any small cell deals announced for India. There's a lot of carrier Wi-Fi deployment going on that is simply not announced because that's not the way the Wi-Fi industry works (although the mobile industry usual does that, yes).