Long anticipated, it now appears that the boundaries between fixed and mobile network infrastructure are fading. Demand for steadily increasing bandwidth coupled with service providers' urgent need to control costs is forcing convergence of network infrastructure, while widespread and enthusiastic adoption of wireless devices in all market segments (wireline, mobile, enterprise and consumer) is further blurring the service boundaries.
Smartphones are clearly not substitutes for home broadband (fixed) connections. A recent Pew Research Center survey found that 56 percent of Americans have a smartphone, while 46 percent have both a home broadband connection and a smartphone. Viewed from the home broadband perspective most devices including laptops, smartphones, tablets, printers, and game players ship with a Wi-Fi interface.
About 60 percent of all smartphone traffic is off-loaded to Wi-Fi home, enterprise or public networks. The motivation for this is rather obvious, given roughly a 2 GB/month data limit for many mobile services versus 100s of GB limits, if any, for fixed broadband services. An even stronger preference for Wi-Fi over mobile broadband is shown in the tablet market where global shipments of Wi-Fi-only tablets have enjoyed a 4:1 advantage over those with embedded mobile broadband cards in recent quarters. Tablet users' strong preference for Wi-FI/fixed is driven by the tablets bigger screen which requires more bandwidth and data capacity to deliver high quality, data-intensive video apps.
It is only the radio access network (RAN) that distinguishes a mobile broadband network from a fixed network. Mobile and fixed broadband operators use the same optical-fiber based network infrastructure in the aggregation and core portions of their networks. Mobile network operators impose tight monthly data limits on their subscribers compared to those of fixed broadband networks because RAN bandwidth capacity scales poorly relative to fixed access networks. The aggregation and core network infrastructure scales excellently with bandwidth because it exploits the extreme bandwidth capacity of optical fiber. Fiber optical modulation schemes have a long way to go before they exhaust the theoretical bandwidth capacity of optical fiber. This is not the case for the RAN. The strategy employed by fixed and mobile operators to affordably increase bandwidth capacity, therefore, is to get the fiber closer to the user. This need to affordably scale bandwidth capacity is what is driving fixed/mobile convergence.
The best strategy to affordably increase the bandwidth capacity of the RAN is to move from mobile macro-cells to small cells. Macro-cells have a range of up to several tens of kilometers while small cells have a range from 10 to several hundred meters. The Small Cell Forum describes four strategies for increasing the bandwidth capacity of the RAN—frequency division, modulation techniques, access to a wider range of frequency spectrum, and frequency reuse through more cell sites. It shows that the capacity gain for the four strategies to be 5, 5, 25, and 1600 times respectively. While small cells reduce cost by using radio spectrum more efficiently, they also are much less expensive to build than macro cells. For example, an Alcatel-Lucent (NYSE: ALU) discounted cash flow study of the revenue and cost benefits of using residential DSL and femtocells to expand capacity rather than macro-cells finds a 144 percent IRR and 32 month discounted payback period.
The most important technical difference between Wi-Fi solutions—whether residential access points or enterprise WLAN systems—and small cells is that Wi-Fi employs unlicensed spectrum while small cells use licensed spectrum. The technical barriers to developing small-cell solutions, therefore, are low due to the maturity of technologically similar Wi-Fi systems. This technological maturity also lowers cost.
Small-cell deployments increase the number of cell sites and reduce the number of concurrent subscribers per cell site. Backhaul of moderate bandwidth capacity is required over widely dispersed residential and commercial areas. Incumbent wireline operators' residential broadband and business Ethernet offerings most cost effectively meet this requirement. This leverages the economies of scope of existing fiber-based access networks by concentrating residential, enterprise, and mobile services on the same optical network. In the extreme example of residential femtocell deployments the network infrastructure investment is nearly eliminated because the cell is owned by the subscriber. Sprint (NYSE: S) was the first to offer a residential femtocell, called Airave, which is sold through retail stores and an online portal. It has about one million units in service today.
This trend toward network infrastructure convergence greatly benefits mobile operators by making them less capital intensive and positioning them to affordably provide higher speed and higher quality mobile broadband services. It provides mobile broadband subscribers a more affordable and higher capacity mobile broadband service. It increases wireline operators' addressable market for broadband and business Ethernet services. This is essential to achieving the economies of scale and scope they need to profitably operate fiber-optic-based infrastructure. It also offers something for the 20 percent of Americans that the Pew survey finds do not use the Internet. Small-cell voice services provide a cost effective way of migrating legacy voice traffic away from copper and TDM based infrastructure.
Michael Kennedy is a regular FierceTelecom columnist and is Principal Analyst at ACG Research, www.acgresearch.net. He can be reached at [email protected]esearch.net.