One of the big debates in mobile backhaul is whether an operator should use wireless or fiber. The choice is not only or primarily about technology or fiber availability, but it involves deep-seated preferences. European operators use fiber where cost-effective, but they are equally comfortable with wireless backhaul, which accounts for 65% of links according to Deutsche Bank. In Japan and South Korea, fiber is widely available and more extensively used, but in overall APAC wireless accounts for 60% of mobile backhaul. In North America, major operators use wireless where they have no other choice, or in 15% of links. New entrants like Clearwire are much more open to wireless backhaul and to experiment with new technologies.
In the long term, virtually all operators are committed to move to fiber where it is cost effective. The open question is how to make fiber cost effective and how long the transition will take.
The advantages of fiber from a performance perspective are clear. Fiber can transport any traffic load, is not affected by spectrum availability, weather, or line-of-sight (LOS) requirements.
It is equally clear that fiber can be expensive, especially if the operator plans to build and operate its own fiber network.
What if instead of framing the wireless versus fiber backhaul as a dichotomy, a mobile operator uses both in consecutive stages within an expansion of fiber backhaul?
Mobile operators already do this, as they deploy wireless where no fiber is available, and fiber elsewhere. But I moved one step further to ask if--and under which circumstances--it makes sense for an operator to use wireless backhaul as an interim solution, where fiber is available but expensive, or when the operator plans to build its own network.
Since wireless backhaul's total cost of ownership (TCO) is lower for wireless than fiber for most operators in many markets, I looked at when it makes sense to deploy wireless backhaul as an operator waits for fiber to become available or affordable, or to build its network at a more leisurely pace in this paper (both papers were sponsored by Cambridge Broadband Networks). I assumed that the operator
- has to deploy a new backhaul solution today, either because the current backhaul does not have sufficient capacity or has new cell sites
- has a long-term commitment to fiber backhaul
- can use wireless backhaul to meet the current and medium-term capacity requirements.
NPV of different backhaul solutions, assuming a gradual replacement period of ten years for the wireless-to-fiber solution. Source: Senza Fili
The results are shown in the graph for both developed and emerging markets, assuming a linear 10 percent per year replacement schedule of wireless backhaul with fiber, with 90 percent of wireless backhaul sites in Year 10 using fiber backhaul. The cost of this solution is compared to wireless, leased fiber, or built fiber backhaul for the entire period.
It may strike you as odd to plan to spend money to install a wireless backhaul solution just to replace after a few years, but it is an approach that can save money for operators. This is because the substantial lower cost of wireless backhaul, results in a lower net present value (NPV) for the wireless-to-fiber scenario.
In the leased fiber case, the cost savings come from the opex. The investment in wireless backhaul in Year 1 is offset by a lower opex in subsequent years. Unless fiber is inexpensive (our sensitivity analysis indicated less than $13,000 per year, but this figure will vary by operator), operators are better off investing in the network deployment (equipment and installation costs), but avoid the recurrent fiber lease for a few years.
If the operator plans to build its own fiber network, the cost savings come mostly from the capex. Capex investment is required during Year 1 to install the wireless backhaul equipment, and during the subsequent 10 years to build the fiber network. The lower NPV for the wireless-to-fiber solution is due to the slower investment pace for the fiber backhaul, which offsets the additional cost of installing the wireless backhaul equipment in Year 1. If building a fiber link costs less $30,000 or less, the operator is better off with fiber from Year 1, but in most markets this figure is higher and the operator can look forward to save money from a backhaul investment that is more gradually spread through time.
As the cost of fiber backhaul is typically higher in emerging countries, the benefits of using wireless backhaul as an interim solution are larger there. In addition, there are substantial differences among countries--and within different markets within each country. As a result, the developed versus emerging country scenarios different offer a starting platform for operators, but the analysis presented here has to be tailored to the prevailing costs in the markets where they operate.
While the data shown here is based primarily on a macro-cell environment, I expect the results to lend a stronger support for small-cell deployments, because the cost of a fiber link is the same, but the cost of the wireless backhaul link is lower (different hardware, throughput requirements, and technologies drive the difference in cost) in a small-cell environment. The greater difference costs between fiber and wireless amplifies the benefits of a wireless interim solution.
Furthermore the dense environments in which small cells are installed are often packed with fiber, but despite the proximity of fiber to the small cell, connecting a small cell on a lamppost to the fiber along the street can be too expensive to justify and the lease costs too high. Instead, connecting a local network of small cells to a macro cell or another aggregation point with a fiber connection is a much more cost-effective solution at least in the short term in most cases. As for the macro network, if the cost of fiber drops to the desired level, mobile operators can switch to fiber, and save money until they do so.
I do not advocate that mobile operators move to interim wireless backhaul wherever they need more capacity or a new link--fiber is bound to be cost effective in varying portion of markets within the operator footprint, especially in developed markets. But where this is not the case, wireless backhaul as interim solution is an option that can save them money over the long term--facilitating, not impeding, their long-term plans to transition to fiber.
Monica Paolini, PhD, is the founder and president of Senza Fili Consulting and can be contacted at [email protected]. Senza Fili Consulting is an analyst and consulting firm that provides advisory services on wireless data technologies and services.