TMCnet - World's Largest Communications and Technology Community





[January 25, 2005]

WiFi vs. Cell or WiFi + Cell?


Two technologies have grown rapidly in the last several years, and their combination was inevitable: WiFi and VoIP. The ability of VoIP to make telephony lose network boundaries is being combined with WiFi's ability to make devices free of their physical boundaries - it's a match made in heaven... as long as it works.


The first forays of WiFi devices into the market were fairly lackluster(*): short battery life, poor user interfaces, impossible roaming characteristics, and few advanced features have made them gimmicks at worst and very fragile office or home tools at best. The next generation of devices is now upon us: hybrid 802.11x VoIP delivery systems embedded with traditional mobile phone handsets. As a device moves from place to place, it uses the most convenient network. Roaming between alternate WiFi hotspots is a bigger trick than it may appear, and is one of the first obstacles that must be overcome. There are several solutions to roaming, both from the commercial standpoint (Logisense, Boingo, iPass, goRemote, etc.) and technical standpoint (802.1x, Mobile IP, SIM access, etc.) but it does appear that 802.1x is the emerging standard. Hardware vendors for both end devices and access points will have to embrace a method and stick with it.


(*Caveat: some WiFi based VoIP networks have worked very well for quite a while, but are for very specific purposes, and even some of the hardware was nice. Hospitals typically come to mind, where budgets are high and the environment is tightly controlled. Moving WiFi VoIP out into the 'wilds' of small offices and even residential users in uncontrolled radio spectrum situations is the trick, and where the large bulk of the market resides. My comments herein focus specifically on that type of deployment.)


In short, I think that 802.11x VoIP delivery is an excellent technology. However, its usefulness is more a wedge than it is a tool to be used by itself. 802.11x deployment will only become seriously useful when combined with some other delivery mechanism that fills in the gaps of coverage. Nobody wants to have a mobile phone that stops working when they move a few feet outside of their office environment.


The coverage area of mobile providers is tens of thousands of times as large as that of 802.11x hotspots, and will be that way for the foreseeable future. Getting the mobile carriers to turn more of their spectrum into IP network traffic will allow devices to use that data network to communicate to various IP endpoints in order to create and terminate calls. This is the end goal - the transition of telephony calls into simply streams of data which the user can have serviced by any endpoint in the network. Just as we are seeing the separation between the physical wires into a facility now separated from the telephony services (see "Tax the Pizza Man, not the Pizza",) this same model needs to be applied to the cell phone networks.


Devices that can operate on several different (and possibly revenue-competitive) spectrum segments will quickly force this change. (When I say "revenue-competitive", I mean license-free bandwidth such as that which 802.11x uses, versus licensed and expensive spectrum like that which traditional mobile carriers use.) Carriers will simply have to expect this shift, and create survival models that match those requirements. WiFi VoIP devices, or hybrid cell/WiFi devices are the embodiment of this transition to creation of yet another "dumb" network, which will ultimately benefit the customer. 802.11x devices will thrive in smaller environments, but to deliver the promise of wider mobility they will need to be more than simply 802.11x systems: they will need to be selective on a number of different spectrums, each designed for long or short haul IP data delivery depending on circumstances. "802.11x + cell" is the ultimate destination on this path.


How will cell phone providers react to this transfer?


It's unclear how traditional wireless service providers will make money in these situations where calls may become simply a stream of data which is indistinguishable from other data. I would doubt that mobile providers will give up the lucrative per-minute revenue that they obtain, but they may have no choice if their customers start circumventing those per-minute costs by moving calls to alternate providers via IP channels. Perhaps devices with dual-mode capabilities will have a higher cost per minute when roaming outside of an 802.11x area. Perhaps cell phone carriers should start to supply SIP termination as quickly as they possibly can roll it out, but this doesn't seem to be the case if looking at the volume of product release announcements is any metric. The mobile providers are very much lagging in announcements for SIP deployments, and even the slow-moving RBOCs seem to have made more headway in this area than their mobile phone counterparts. This is all speculation; it's unclear where the mobile providers will find themselves in five years from now, but I'm fairly certain that the years will find them looking more and more like ISPs and less like telcos.


I anticipate that the dual-mode phone issue has yet to truly sink in for service providers, or if it has, there is a misguided understanding that the 802.11x features will be "locked" to their service somehow. This will not be the case; customers will want to have multiple service providers, and multiple numbers, ringing on the same device, and this control will need to be put in the hands of the end user and not the network.


This separation between device and service will become more apparent in the coming years. Most mobile providers assume that the device is locked to their service because of a contract or because of a technology (CDMA, iDEN, GSM, etc.) that makes it difficult for the customer to transfer portions of their minutes (aka: recurring revenue) to an alternate provider.


Of course, the ideal end goal would be to move away from the "traditional" delivery methods of mobile calls and present the entire session as an IP transaction, so that the already-existing methods of call transfers and re-invites which exist in protocols like SIP could be used to move calls between different IP networks when they became available or more preferred. Using TDM-like protocols to transmit voice signals seems to be unnecessarily complex if there are IP networks available, and certainly roaming between IP networks is far easier than trying to switch between two entirely different delivery methods. So, as soon as the pure data capability of 802.11x is available in phone-like devices, it will become much more obvious to everyone that the mobile provider is now the slowpoke, and conversion of the mobile portion of the device to a pure IP data transmission scheme is inevitable.


So, to look into the future, let me end this discussion by making some guesses:


1) Hybrid devices will quickly drive mobile carriers to an all-data network, and specifically, an all-IP network.

2) A second generation of hybrid devices will arise, which have no "TDM" circuitry, and simply include only what is required to transmit data across mobile carrier's networks. All calls from these devices will be via VoIP.

3) In order to make money, mobile carriers will need to start to charge more for customers who use less.

4) Regardless of hype, 802.11x hotspots will still have miserably small geographic coverage, and will only be useful in the office or at home. However, since this is where >~90% of the minutes are generated it will be OK. The remaining 10% of the minutes will have to bear the burden of the cost of the non-802.11x mobile data or TDM-like network. Customers will not see a significant shift in the "baseline" costs of their mobile phone bills; there will be a short drop to a "minimum" cost floor, but that's as far as will be possible without bankruptcy.

5) 802.11x hotspots will embrace a standard for sharing roaming credentials to allow for scaleable cost and revenue distribution. An industry consortium will form to handle this.

6) There will be a period of time when mobile carriers will refuse to allow users to configure their own SIP settings for call delivery on the 802.11x portion of the device, in the hopes of keeping the call revenue. This will be short-lived due to consumer demand.

7) Third-party firms doing nothing but call presence management and call filtering will develop as the technology becomes more universally accepted and a wider variety of IP-only gateway firms rush to provide lower-cost and higher-feature services to mobile users.

8) iPBX systems will need to become more open at the edges to allow for roaming extensions.

9) A swarm of WiFi/VoIP devices will hit the market, which will drive a massive backfill of hotspot dead areas in office-wide or campus-wide deployments, but this will take time and will not quite provide the wide-scale coverage required outside of small areas without use of traditional wireless carriers.

Purchase reprints of this article by calling (800) 290-5460 or buy them directly online at

Respond to this article in our forums!

Technology Marketing Corporation

35 Nutmeg Drive Suite 340, Trumbull, Connecticut 06611 USA
Ph: 800-243-6002, 203-852-6800
Fx: 203-866-3326

General comments:
Comments about this site:


© 2018 Technology Marketing Corporation. All rights reserved | Privacy Policy