Will Wi-Fi rescue 3G?

A post by Stephen Rayment, CTO of BelAir Networks, in the Financial Times urges mobile operators to use Wi-Fi to offload traffic from their overcrowded 3G networks. The CTO of a company that makes Wi-Fi base stations would obviously say this, but it is something that many others have been wondering about for a long time. If the iPad becomes as successful as the iPhone, expect your mobile phone service to get even worse.

But I wonder if the operators will go for the Wi-Fi solution any time soon, as proposed by Mr. Rayment. There are reasons why they won’t do this. They distrust Wi-Fi (which they consider to be more susceptible to interference and security problems) but the most obvious reason to me is that they are looking for ways to squeeze more out of their 3G/4G networks by, among other things, metered billing, putting lower data caps, etc.

Moreover, a cellular operator such as AT&T who deploys a Wi-Fi network in a city like San Francisco will have to limit access to AT&T customers (this is what Cablevision is doing in New Jersey and Cablevision uses BelAir equipment), otherwise, there’s no reason for customers to stick to AT&T.

Now imagine if T-Mobile, AT&T, Verizon and Sprint all hang their own Wi-Fi access points on city lamp posts, building rooftops and other locations creating a rather nightmarish landscape of structures that seem to have hanging Wi-Fi fruit. The only answer is to have one large citywide Wi-Fi network with one set of base stations, used by all the operators, managed by a company like Boingo, which is already a Wi-Fi hotspot roaming partner of one or more of the operators I have just mentioned. The problem with this proposal is there’s no exclusivity anymore over the network and if there’s one thing I learned from mobile operators, they love to “own the customer” (the words of a Verizon executive at a SXSW Interactive panel in 2005).

What do you think?

Read the article: Is Wi-Fi the answer to network overload?


  1. Esme;

    Having previously worn a BelAir badge, you might accuse me of bias, but as I have spoken and blogged about extensively (http://www.xchangemag.com/blogs/suter/blogdefault.aspx), Wi-Fi is an excellent offload option for wireless operators. The rationale you cite as negatives are easily rebutted.

    The interference argument is a red herring. Interference exists in licensed bands as well, which operate in much narrower bands than does Wi-Fi. And while the 2.4GHz band has been the unlicensed band of choice, modern 802.11n radios will also operate in the 5GHz unlicensed bands, where there is less interference and much more “free” spectrum. Other arguments can be made about data traffic/applications being very different from a low latency, real-time app like voice, Wi-Fi protocols, etc., but I won’t make them here.

    The second point regarding metering and billing is easily solved by vendors like IntelliNet (http://intellinet-tech.com/) with its I-WLAN products, which allow wireless customers to roam seamlessly back and forth across 3G and Wi-Fi, but which also allows the mobile operator to maintain control of the customer, track data usage, etc. Of course, AT&T offers unlimited data plans, so for them, freeing up wireless capacity is the primary objective, but other operators with data caps will want to look at something like the IntelliNet solutions.

    The solution, as you described, remains rooted in a “muni” view of the world, with ubiquitous, city-wide Wi-Fi. The industry (and customers) are smarter now, and understand that attempting to cover the wide-area with Wi-Fi is unrealistic. It’s about adding capacity, surgically, in areas where it makes sense to do so. High foot traffic areas (like the Riverwalk in San Antonio), sports stadiums, events, college/university campuses, etc…It’s still a long list, but it falls well short of the old muni fantasy.

    That’s my .02!

    Martin Suter

  2. I am glad you pointed out the interference red herring. And I do agree that anyone who creates a network should prioritize the locations – deploy first in places that are suffering from overload and where there is high demand, e.g. in densely populated areas. However, I wonder if the telcos think this way. Why haven’t the operators worked faster to reduce overcrowding in downtown SF, in central London, Paris, NYC, and areas like that?

  3. Esme;

    The questions as to whether telcos are/aren’t already doing so is an interesting one. Looked at offload via an outdoor lens, I would agree that uptake is underwhelming, but I think you also need to look at their indoor deployments. Every iPhone user sitting in Starbucks and using Wi-Fi is a user that’s not on the 3G network. In this respect, Wi-Fi Hot Spots are also a key part of the offload story, and one that is often overlooked. By the same token, enterprise and home APs accomplish the same thing, at least from an AT&T perspective, and cost them nothing per bit.

    The challenges to outdoor Wi-Fi deployment haven’t really changed. Site acquisition, power and backhaul are all required, making urban deployments a challenge. Privately controlled venues, like sports stadiums, arenas, campuses, etc., are far better candidate locations as a result (or at least locations with less friction to deploy). This is also one of the reasons that Cablevision is able to deploy so readily – they control the infrastructure. Overhead cable strand is powered, wired backhaul, at the same height as light poles and runs in front of commercial, high traffic locations throughout their footprint.

    Keep up the good work…

    That’s my .02!

    Martin Suter

  4. AT&T’s network problems also affect people who have only a voice subscription. While I mention AT&T here, it’s not just AT&T having these problems. Mobile operators around the world are grappling with the same issues. What baffles me is how long it takes for AT&T to improve the service. Example: at the Ferry Building in SF, which is very crowded on weekends and even weekdays with tourists and locals, AT&T’s service is terrible. It’s very slow. This isn’t a location in the Santa Cruz mountains, it’s right by downtown and the Financial District. This problem has been going on for at least a couple of years. I suspect AT&T has known about this for a long time. If they can’t even improve service in downtown SF and around the Financial District, it’s hard for me to see how they can improve it throughout SF and in other large US cities any time soon (i.e. in the next 3 years). Maybe they’re waiting to deploy LTE . . .

  5. more on offloading data from 3G networks:

  6. A quick word to share the perspectives we’re gaining on this here at Ruckus — where we’ve been working with mobile operators on 3G offload via hotspots for the last couple years, and where we’re seeing *very* substantial acceleration of operator activity in this category over the past few months. Key findings to date: (1) many operators are fast-tracking plans to use more Wi-Fi in both indoor and outdoor modes, in the high-traffic urban areas you mention, so stay tuned, (2) the interference issue is actually more than a red herring in those urban environments already busy with Wi-Fi in many forms — even with the additional 5 GHz channels — and mobile operators are taking this very seriously. [Not to turn this into too much of an advertisement, but we have active interference rejection capabilities the other guys don’t, so we address this known problem head-on rather than just asserting it’s not one], and (3) regarding your question of build v. partner, there are currently disadvantages to shared networks in terms of support for a seamless subscriber and operator experience (which requires good client provisioning, authentication, security, and control) that are motivating build in many cases — don’t forget that the mobile operators already have lots of points of presence in these urban environments (their existing sites) that offer a good head start on the mounting assets dimension.

    Generally, we’re finding mobile operators moving to use every tool they have at their disposal — more spectrum, LTE sooner, traffic management, tiered plans, femtocells (usually in a controlled way, rather than a broad and random consumer play), and Wi-Fi — because the smart-phone data traffic issue is growing so rapidly worse, and no single tool (including LTE) will solve the problem all by itself.


    Steven Glapa

  7. Esme,

    This is a very interesting topic. For years, the conventional wisdom has been that Wi-Fi is a disruptive technology and presents a competitive challenge to conventional mobile networks. The truth is that Wi-Fi (as a proxy for unlicensed operation) may offer some opportunities for competition, but it is more of a complementary technology than a disruptive one. Using licensed spectrum, carriers can always offer more robust, secure and reliable networks than Wi-Fi. I have recently read that some of the new Wi-Fi networks can offer five nines reliability (99.999%), but even so they will always have to content with a higher probability of interference. Neither licensed nor unlicensed operation holds an absolute advantage over the other. However, there is the chance for carriers to offer an entry level service and a premium service, or a portable and a truly mobile service.

    I, however, disagree that a network of unlicensed radio equipment is some sort of natural monopoly, esthetics be damned. Granted, due to lower power and less in building penetration smaller sizes are cell and more network equipment is needed to provide unlicensed networks. But, this does not necessarily mean that there will be an antenna attached to every lamppost or traffic light. Even so, overbuilding networks is inherently a good thing. We want to build competition into the access infrastructure. To the extent that overbuilding is not economically viable, roaming arrangement between networks will eventually arise. In such cases and in smaller markets, a “muni” network can make a lot of sense.

    That said, an integrated Wi-Fi and LTE network is an interesting prospect.

    Here is my take on the subject: http://kennethrcarter.com/CoolStuff/2010/02/wi-fi-wi-not/

    Ken Carter

  8. more on offloading data from 3G networks at http://www.theruckusroom.net

  9. I am wondering why this attention- any attention- is being given to private companies making vast profits? AT&T et al are savvy businesses. Let them figure it out on their own and leave, please leave, the locally owned WiFi networks alone. That is the answer. I believe that smaller is better. When people sign up at my humble company for instance they actually get what they pay for and not “up to” speeds. People forget that AT&T etc. are not into business to make their clients happy- they are in business to make their investors happy period and their goal is to grow once more into the monopoly they once were.

  10. peter,

    who are you? write to me at thejoff@gmail.com?

  11. WHITE SPACE WiFi –


    FCC Approves White Space “Wi-Fi on Steroids”

    By Tony Bradley, PCWorld

    With a unanimous, bipartisan vote of the FCC commissioners, the unused television broadcast spectrum has been unleashed for use in wireless networking. Using the broadcast white space for wireless networks will usher in billions in investment and innovation, and could fundamentally change the nature of wireless networking.