Sprint has started provisioning its QChat Push-To-Talk (PTT) service over its CDMA network, and there is a bridge between it and the Nextel iDEN network so PTT can be used between them-a CDMA QChat customer can talk to an iDEN PTT customer or group and vice versa (they are branding the service on their CDMA network as Nextel Direct Connect also)

QChat only runs over CDMA2000 1xEV-DO Rev A, not Rev 0 or CDMA2000 1X, so Sprint had to build out EV-DO Rev A everywhere it wanted to deploy PTT, including to the very edges of its network. This expanded EV-DO footprint might provide some interesting additional revenue opportunities.

So far, the three U.S. networks that offer high-speed data services are AT&T Wireless, Sprint Nextel and Verizon Wireless with T-Mobile bringing up its network later this year. None of them have wanted to offer much in the way of fixed wireless broadband services, which are services that could be used at home or in an office in a fixed installation. Each network has its own reasons for not offering fixed wireless-some don't want to compete with their own wire line and fiber business for broadband services, while others are concerned about bandwidth consumption.

Fully mobile networks are designed with most of the cell sites divided into three 120-degree sectors. We have all seen the cell towers with flat-panel antennas pointed in different directions. By sectoring cell sites, operators have more capacity at each site than if the sites were built with omni-directional antennas (360-degree coverage). A number of companies are building smart antennas that can further segment this 120-degree cell sector many times, but today's norm is three cell sectors.

Traffic loads are predicated on mobility, that is, devices that move from sector to sector and from cell site to cell site. Introducing a lot of fixed services into this type of network can result in diminished capacity for a sector and perhaps slower data speeds for mobile customers. For example, if there were twenty, thirty or more people using wireless broadband in a fixed location, and all of them received their coverage from the same cell sector, that sector's capacity could be impacted, perhaps affecting the equations used to calculate capacity and bandwidth allocations.

There is no doubt that fixed wireless devices are being used. If someone has a wirelessly-enabled notebook instead of a desktop, and the notebook doesn't move from the home or office, it is, in essence, a fixed-location wireless device. But network operators are not selling devices that are designed to enable fixed broadband data services. In fact, the all-you-can eat data pricing models have been replaced with metered services (50 MB or 5 GB per month, for example). And where there is a growing demand for wireless broadband mobility services, it probably makes little sense for a network operator to offer fixed services, especially since in high demand areas there are other choices available for fixed data services.

However, building EV-DO Rev A or UMTS HSPA to the very edge of a network offers some new opportunities for network operators. The cell sites located at the edge of the network will, most likely, be lightly used for data and PTT services and will therefore have a lot of capacity that could be used for fixed wireless services. So it would make sense for a wireless operator such as Sprint to believe it could begin offering fixed wireless broadband services in these more rural areas. In this case, Sprint does not have to worry about competing against its own wired DSL or cable services (although it does have some cable company relationships) and perhaps in some areas where it has excess wireless data capacity, there is no cable or DSL service.

One of the key issues would be how to offer fixed wireless broadband services in portions of the network where the operator knows it will have plenty of capacity but not in areas where its network is already busy and adding fixed broadband services would not be appropriate. If the equipment were listed on its website or available via the company and associated stores, the network operator would have no control over who, when and where issues. The answer, at least for Sprint, seems to be to let someone else sell the service. The word on the street is that Sprint is looking for companies that might want to resell wireless broadband on its EV-DO networks in areas where it has the capacity. This would keep the Sprint customer population happy and would enable small ISPs or others to offer fixed wireless broadband to homes, apartments and small and medium businesses at the edges of the Sprint network.

Perhaps it would work something like this: Each franchise would be given a certain section of the country and would be limited to selling fixed broadband equipment within that area using only local marketing and sales activities. The devices would have to be provisioned so Sprint could keep track of where they are, or at least where they started out being used, and perhaps the name of the network. "Sprint," in this case, would not even be mentioned to the customer. You may remember that Amazon's Kindle uses Sprint's EV-DO network to deliver books, magazines and other content, but that Amazon pays the bill and calls the network Whispernet. It isn't easy for customers to find out who the network provider really is, not that they care.

This would give Sprint additional data revenue for sites that are lightly used, and it would provide local companies with an income from selling the fixed wireless devices and receiving a percentage of the monthly revenue. These small companies would be able to compete with existing DSL and cable operators if such services were available in their territories. This business model would also keep Sprint at an arm's length from the fixed wireless customers so it would not appear to be competing with Clearwire's WiMAX service (Sprint currently owns 51% of Clearwire). Sprint certainly would not want to be seen as competing with Clearwire, especially using EV-DO. If you remember, when it first announced it would build WiMAX on 2.5 GHz, Sprint made it very clear that EV-DO and WiMAX were complementary, not competitive and that the two networks would serve different customer populations with different customer needs.

Now it appears as though Sprint may, in fact, be setting up others to compete with Clearwire, at least in the smaller cities where Clearwire has been building. Although the new Clearwire will be concentrating more on urban areas with larger populations now and not as many of the smaller markets as it was planning just a year or so ago, it would still be interesting to hear what Clearwire has to say about Sprint's possible new direction.

I also have to wonder about Verizon Wireless, which is reportedly about to announce its own PTT service (also based on EV-DO Rev A from what I have heard). That would mean that EV-DO Rev A will also be (or has already been) built out to the edge of Verizon's network. I am sure it doesn't want to complete with its own DSL business, but since its wireless network is nationwide and its wired footprint is not, Verizon, too, could invoke the same business model, using third parties to sell fixed wireless broadband in areas without wired infrastructure and with lightly loaded EV-DO Rev A cell sites.

This is only a guess on my part, I have not ventured into the AT&T Wireless world of UMTS/HSPA, but AT&T still has an advantage when it comes to PTT. It employs Kodiak Networks' technology that runs on the voice channel, not as VoIP packets on a data network, so AT&T's PTT service works everywhere it has coverage, regardless of whether the coverage is GSM or UMTS/HSPA.

I am sure Spint's EV-DO Rev A build-out went faster and covered deeper into its network because it really wanted to launch QChat PTT, which is part of a broader master plan. At some point, Sprint wants to migrate all iDEN (Nextel) customers to its CDMA network and finish the rebanding on the 800-MHz band that is necessary because Nextel was causing interference to first responders using channels in the same band. Once Sprint has successfully migrated all Nextel customers to CDMA and has completed the rebanding, it will have contiguous spectrum in the 800-MHz band and could use this band for CDMA and/or CDMA EV-DO Rev A and improve its CDMA coverage. Right now, Sprint's entire CDMA network is built out on 1900 MHz while AT&T and Verizon's networks are a mixture of both 850 and 1900 MHz, giving them the ability to cover more territory with fewer cell sites.

The name of the game is to get more customers and therefore more revenue from the data portion of all of the networks. Perhaps adding fixed wireless broadband is another way to help drive data revenue and see a faster payback for the EV-DO Rev A upgrades at the edges of the networks. I know there are a number of companies selling fixed equipment into the wireless voice marketplace-I have several devices set up and operational-and there are a few companies that make fixed environment, wireless data devices. However, they are not widespread and for all of the reasons mentioned earlier and more, I don't think we will see any of the wireless networks pushing fixed data services themselves.

Femto cell and Wi-Fi solutions require a DSL or cable connection for backhaul, so their rapid adoption will help unload the wide-area networks. Since they are already connected via a wired broadband connection, they won' t have any impact on fixed broadband, but I believe we will be seeing fixed wireless broadband soon at the edges of the networks, and this can become a valuable tool in solving Internet access where dial-up is all that is available today.

Let's see what Sprint and the other wireless broadband providers have up their sleeves and how long it will take them to have fixed wireless broadband services available at their edges.

Andrew M. Seybold