Crossing the deep divide: Improving wireless cell site backhaul in the last mile

Evaluating coverage, price, performance, and time-to-market, it is clear that hybrid fiber/wireless networks meet the immediate backhaul needs of wireless service providers.

By Peter Neill, Level 3 Communications

During the past 15 years, mobile phone usage has skyrocketed. What initially started as a voice services phenomenon is now exploding into a worldwide hunger for both traditional voice and next-generation broadband data services. Where wireless service was once a luxury, it has now become a critical part of everyday life for billions of subscribers. Consumers that were previously happy speaking with their friends and family via telephone are clamoring for "anytime, anywhere" wireless access to movies, music, television, and mapping- and location-based devices. Cell phones—which once included only basic telephony functionality—have morphed into multipurpose mobile devices.

The booming demand for next-generation wireless services, combined with the steadily rising consumption of subscriber minutes and increased data requirements presented by demand for larger file multimedia content, has exponentially increased the backhaul requirements of cell site locations. These factors are driving up expenses for wireless operators, cutting into profitability and putting pressure on margins. Wireless providers, who have been relying on legacy copper infrastructure for more than a decade, have acknowledged that copper is not a scalable option for supporting next-generation wireless requirements or reducing backhaul expenditures. Leased lines not only create network performance issues and lack scalability, but they are also expensive, costing several hundred dollars per line and accounting for nearly 30% of provider operating expenses.

So what can wireless providers do to meet the growing needs of subscribers while simultaneously improving profitability? Efficient, reliable backhaul is the key.

Weighing the options

The wireless industry is working feverishly to implement better backhaul strategies, and there are differing opinions about how to best increase bandwidth and reliability while containing costs. Most operators today focus on deploying infrastructure that is suited for 3G and newer technologies, provides network reliability, and enables them to reduce operational costs while realizing a greater return on their capital investment. There are, however, three separate and distinct approaches, all of which have advantages and disadvantages.

Manipulate existing T1 lines. According to Visant Strategies Inc., leased lines account for approximately 75% of all backhaul connections in the United States. Since copper is currently in place and functional, reason dictates that the easiest and quickest way to increase bandwidth capabilities is to manipulate the existing copper plant through processes such as inverse multiplexing or conversion to DSL. For example, carriers can join six T1 lines together to create a 9-Mbit virtual pipe to accommodate an increased amount of bandwidth in support of protocols such as ATM or Ethernet. In some cases, these same copper lines can be converted to DSL to provide 6 Mbits/sec or more of capacity.

However, while this option does provide a higher overall amount of bandwidth, it still relies on the same outdated and unreliable copper infrastructure. Simply bundling additional lines together or converting them to a different Layer 2 protocol will not remedy the inherent lack of reliability in the copper plant. In addition, the process of implementing inverse multiplexing or DSL conversion does not help to lower the overall cost per megabit for the backhaul and, in fact, adds more operational complexity and cost to the service while lowering overall network efficiency.

Another T1-related option is for wireless providers to purchase additional T1 lines to meet their backhaul needs. Realizing that cell site backhaul represents a $15 billion market from cell site to Mobile Telephone Switching Office (MTSO), ILECs have drastically reduced the cost of T1 line access in order to keep new technologies from encroaching on their business. Wireless providers that are mindful of this market change can currently buy T1 access at lower prices. However, while a temporary reduction in T1 costs may provide a quick fix, it does not address the issues of network scalability and reliability. These two very important criteria can only be addressed through the move to fiber-optic and/or microwave services.

Make a deal for microwave. Microwave has been a workhorse technology in telecommunications for more than fifty years. In the application of cell site backhaul, microwave historically has been used to provide service to those locations unreachable by terrestrial services such as copper or fiber. In the United States, there are two classes of microwave service relevant to cell site backhaul: licensed and unlicensed. Licensed spectrum comes with the full federal guarantee of protection from interference but can also be expensive to purchase upfront in order to guarantee availability. The Federal Communications Commission (FCC) makes this spectrum available to providers through the public auction process or on a first come, first served basis known as "common carrier" assignments. Unlicensed frequencies can be used by anyone for free but are generally within the 2.4- or 5-GHz bands, which are often laden with interference from cordless telephones, Wi-Fi networks, and the like.

Licensed microwave is superior to copper connections in that it can transmit voice and data at significantly higher speeds, up to and including OC-12, while supporting advanced protocols such as SONET, Ethernet, MPLS, and IP. Because microwave services are a wireless technology, the implementation of these services requires specialized knowledge of RF engineering, site acquisition, and service deployment. When architected correctly, microwave can easily achieve 99.99% reliability. In addition, microwave is often, but not always, deployed in a line-of-sight or point-to-point configuration. Microwave is not always a viable option, though, due to topography and weather elements.

Extend fiber. The wireless industry has long-term aspirations to run fiber to every cell site. Fiber's reliability and scalability make it an appealing technology for backhaul. Capacity on fiber is significantly higher than other alternatives, such as copper and microwave. And fiber is an extremely cost-effective way to transport large amounts of voice and data traffic. Why, then, aren't all providers already deploying fiber in the last mile to connect their cell sites to their points of presence (PoPs)?

Cell sites are often located in obscure places, such as corn fields, rural highways, and on the tops of hills. While a significant amount of fiber has already been laid across the United States, it tends to be located in the more densely populated areas where it can serve a greater number of people. Many cable companies have fiber that extends into residential areas and offer unique footprints, enabling them to better serve suburban and sprawling urban locations. However, these networks, which were designed to provide residential television service, were typically constructed in a hub-and-spoke or tree branch topology and not with the redundant ring topology required for true carrier-grade telecommunications. Regardless of the construction and capabilities of existing fiber networks, the fact remains that there are still many cell sites that are unreachable with today's fiber.

Over time, the number of cell sites addressable by fiber will increase as wireless traffic demands justify the high construction costs necessary to reach them and fiber density continues to expand. In the short to medium term, providers must pursue strategies that meet their technical requirements today while providing an evolutionary path toward an all fiber, packet-based network for tomorrow.

A hybrid approach

Across the industry, many wonder which technology will best support the wireless backhaul networks of the future: copper, microwave, or fiber. In the near term, hybrid fiber/wireless networks are the best option.

As point-to-point microwave services have gained a foothold in the United States, the environment has become ripe for networks that leverage the best features of both fiber-optic and wireless networks. In this new world, wireless providers can use fiber services as the core of the network to access as many possible sites with fiber, while microwave delivers service to the outlying locations. The end result is a fiber-optic network with greatly extended reach and access to locations that would not otherwise be addressable.

Hybrid fiber/wireless networks offer reliability that can be engineered to 99.999% and scalability that provides the bandwidth necessary to meet cell site backhaul bandwidth demands for the foreseeable future. Perhaps most importantly, when compared with all-terrestrial alternatives, hybrid fiber/wireless network designs provide the most capacity in the least amount of time at a reasonable cost.

As senior vice president of wholesale offer management for Level 3 Communications (www.level3.com),Peter Neillis responsible for Level 3's wireless, domestic and international carrier, and system integrator businesses.