Small Cell Mobile Backhaul: The LTE Capacity Shortfall

With immense mass-market demand for mobile broadband services, and emergence of new high-capacity mobile devices (e.g., smartphones, tablets) and applications, many of the world’s most advanced mobile networks are struggling to deliver a high-quality consumer experience. Explosion of per-user data consumption, combined with subscriber growth and mobility needs, is putting today’s networks are under tremendous pressure. In addition, as operators continuously evolve networks with the latest technology (e.g., 2G, 3G, 4G) to meet these capacity and coverage demands, network costs are exploding and operators are struggling to keep up profitable businesses.

LTE, representing a 4x capacity improvement over current 3G networks, on its own will be insufficient to address all future capacity demands, as mobile data traffic will double every year equating to a 32x growth by 2014 (Figure 1).

32x backhaul capacity demand jump

Figure 1: The forecast 32x jump in data demand cannot be met alone by LTE, which can only offer a 4x increase over current wireless technologies.

Increasing spectral efficiency with new versions of LTE will help manage the shortfall, but these solutions are not yet available and again will not provide the volume of capacity necessary. Acquiring more spectrum would help but additional spectrum is costly and in most cases not available. Traffic management approaches such as caching and mobile data offloading are emerging to help manage the load but because of limited cache hit rates, these solutions will be insufficient to address the capacity shortfall. Offload techniques, such as in-home femto cells and mobile offload gateways, are emerging to reduce load on mobile infrastructure, but again they will be insufficient. A new approach is required.

Emergence of Small Cells

To meet these capacity challenges, and address ever-prevalent coverage issues, new small cell network architectures are emerging based on a new generation of low power, small cell (i.e., micro, pico, femto) mobile base stations. ABI Research estimates 4 million pico base stations will be shipped per year by 2015. Being deployed into an existing network on lampposts, utility poles and building walls, these base stations offer a way for operators to meet challenges of urban, suburban and in-building locations. Combined with existing base station infrastructure, these small cells are transforming the flat macro mobile network into a multi-level, hierarchical radio access network (Figure 2).

Macro, Pico & Femto base stations

Figure 2: Combined with existing macro base station infrastructure, small cells are transforming the flat mobile network into a multi-level, hierarchical radio access network.

Small Cell Backhaul: Wired or Wireless

When considering IP mobile backhaul options, operators must first ponder the choice between wireline or wireless solutions. There is generally no “one-size-fits-all” solution, and in reality we’re likely to see a mix of mobile backhaul technologies deployed to meet the small cell backhaul challenge. However, because of challenging utility pole and lamppost deployments, operators cannot count on fixed line options (e.g., fiber, cable, copper/DSL) being ubiquitously available. Moreover, more than 40 percent of the world’s macrocell base stations are backhauled wirelessly and because of these challenging locations, we’re likely to see a much higher percentage of wireless-based backhaul in small cell applications.

Wireless Backhaul for Small Cells: Challenges

Small cell deployments present a number of challenges—not the least of which is impact on mobile backhaul. Operators—and equipment vendors—must consider the key factors below when selecting (and designing) wireless backhaul solutions for small cells:

Lower cost solutions needed—Smaller cells mean more cells and thus more mobile backhaul. To meet overall cost objectives, lower cost backhaul solutions will be required to make sure small cells can be deployed cost effectively. Typical macrocell backhaul CapEx is about 50 percent of the total base station CapEx, and similar ratios will be required to ensure a cost-effective solution.

Space-optimized solutions required—To improve street-level coverage and capacity, small cells are being deployed on lampposts and utility poles. These challenging deployment locations place demands on the physical attributes of backhaul solutions. Unlike traditional cellsites, typical dish antennas will not be feasible for such deployments. In addition, because of space constraints and operations costs, backhaul and base station hardware integrated into common enclosures would be ideal.

Line-of-Sight (LOS) not possible—Street level, metro area deployments mean line of sight to backhaul hub locations are not always—in fact—rarely possible. Requiring large antennas, combined with lack of LOS characteristics, makes traditional point-to-point wireless backhaul ineffective for most small cell backhaul applications.

Interference must be carefully managed—When it comes to wireless backhaul solutions, close proximity of cellsites creates possible interference issues for the backhaul system. These interference issues are relatively new for backhaul systems and need to be considered.

High-capacity solutions required—Driven by increasing demand for mobile data, backhaul requirements for small cells are expected to approach macro cell capacity requirements (50-100Mbps per cellsite) in the next three years.

Which challenges matter most will depend heavily on how small cells eventually are deployed. Stay tuned for a followup blog post where I discuss small cell backhaul deployment options and available solutions to address these needs. In the meantime, feel free to leave me your thoughts, or comments.

Gary Croke
Sr. Product Marketing Manager, Aviat Networks

Comments

    Joseph Mwangi June 24, 2011/ Reply

    Hi Gary,

    The scenario you have described is happening pretty fast, and we are already grappling with possible suitable backhaul solutions. Our thoughts are around:

    1) Use of broadband Point to Multipoint.
    – Are there systems that can support in excess of 500Mbps per hub and support 10 terminal sites, which using statistical MUXing, will provide the much needed backhaul capacity?

    2) WiFi & Low Frequency Systems to beat LOS.
    – Has WiFi technology evolved to a point of accommodating the high BW requirements?

    3) Extending metro fiber.
    – What about extending existing metro fiber along the underground power grid used to light up the street lights and power buildings (that will be used to collocate the micros)?
    Definitely, it will be interesting to follow on this unfolding scene.

    Regards,
    Joseph.

    aviatnetworks June 28, 2011/ Reply

    Thanks Joseph,

    You are right about PMP being a possible key technology for small cell backhaul. However, the current crop of wireless solutions on the market are probably not going to be suitable for a variety of reasons. Cost is the key consideration. With small cell base stations perhaps costing as little as $1000, backhaul solutions will have to be at a fraction of the cost that they are at today.

    While higher capacity PMP systems are available that may do the job from a capacity perspective, they are too high cost to provide an economical solution. And while WiFI or 802.11-based systems are available at much lower cost, they would not support either the capacity or latency requirements for LTE. Fiber is always the easy answer, but new small cells could be in a variety of locations, from lampposts to the sides of buildings, so fiber would be expensive to deploy even relatively short distances to these locations.

    So far we haven’t seen the right solution, but the technology that delivers the right performance at the right price is out there somewhere. We just haven’t seen it yet.

    Regards,
    Aviat Networks

    Laura Byers July 26, 2011/ Reply

    Hi Gary,

    Very interesting conclusions regarding the need for emerging technologies for addressing LTE’s shortfall. How come we’re not hearing more about this?

    Regarding deployment, it seems that Japan’s Softbank is providing each of its smartphone subscribers a free femtocell and a free DSL line. I am told that it is making money off data (and doesn’t charge for voice) by reducing costs dramatically.

    What if caching and processing capability is integrated within the femtocells, picocells and macrocells? Would that further increase the offload?

    Whatever the technology that will alleviate the pressure on the network, when and where does the “aha” moment happen? Is this something that has to be “pushed” by the equipment guys or something that will be “pulled” via requirements and an RFQ by the service providers? Or do we just wait to experience the pain? How do new architectures make their way into the ecosystem?

    Regards,
    Laura

    aviatnetworks July 26, 2011/ Reply

    Hi Laura,

    Thanks for the comment. You raise some very good questions—of which many have no obvious answer. But I’m happy to weigh in with some thoughts.

    I believe offload will certainly play a major role toward reducing the capacity shortfall on mobile networks. There are two types of offload: network-based offload (i.e., offload gateways from Tellabs, Stoke, others) and WiFi/access offload (i.e., femto, home gateways) that you mentioned. Each of these solutions will have different effects because they offload traffic at different points—one before the RAN and one after the backhaul). The sheer volume of mobile traffic growth will mean that any one capacity relief technology (offload—network or access based, small cells, more spectrum, improved spectral efficiency with new versions of LTE with higher order MIMO, etc.) will not be sufficient to solve the problem—a multi-approach strategy will be needed.

    I agree with you that caching is a great solution to offload network traffic and speed up performance (as we’ve seen with the fast growth of CDN companies over the years) especially for video traffic since cache hit rates can be extremely high and high data rates mean amount of offload is very significant. CDNs have experimented with using home devices (computers via peer-to-peer-based CDN solutions), but these generally didn’t get off the ground because they require use of limited upstream bandwidth, and these architectures were difficult for CDNs to control (e.g., people rebooting their PCs, femtos) and difficult for CDNs to create a business model and charge for reliable delivery of content from these devices. For mobile network traffic, it’s a little different as mobile traffic is encrypted in L2 tunnels back to the switching site and not possible to cache—for the moment at least. I think we’ll see some new innovations on the horizon to overcome this obstacle, too.

    In terms of bringing new technology to the market, my experience is that a significant collaboration among vendors and operators is required for viable new technology innovations—ultimately culminating with the RFI/RFP/RFQ process and deployment (through the course of this process there is enough “socialization” of ideas that there are generally multiple vendors with products available when market demand ramps). We’re definitely seeing that with small cells and small cell backhaul, and we’ve also seen it for various offload solutions now on the market.

    Cheers,
    Gary Croke
    Sr. Product Marketing Manager, Aviat Networks


Leave a Comment

Subscribe to our newsletter