As operators are confronted with insane demand for mobile data, there are at least four distinct areas where congestion may appear:
• the access radio network
• the signaling and control portions of the network
• the network packet core and
• the backhaul network
Each of these choke points pose a unique challenge to the operator and generally can be addressed in one of three ways:
1. increasing capacity of the affected network resource
2. offloading the network resource to relieve congestion or
3. doing both
The move to smaller cells to augment existing macro networks is widely viewed as a potential panacea to the access radio network congestion these problems but also creates a new one: backhaul. This has become one of the telecom industry's hottest debates.
Mobile operators are planning their LTE networks as a combination of macro cells and an 'underlay' network of smaller micro and picocells. To achieve the capacity density required by rapidly rising mobile Internet bandwidth demand, these small cells will need to be much larger in number in a given area than is the case in current cellular networks. This presents a new and very significant backhaul challenge — because the mounting locations of these small cell nodes (such as utility poles or other street-level assets) will very rarely be a natural fit for fiber or microwave solutions.
The concept of LTE self-backhaul or meshing is one possible solution, but as with early mesh Wi-Fi networks that attempted to provide access and meshing all within the same spectrum band, this approach rapidly consumes scarce (and expensive) LTE access-capable spectrum with backhaul traffic. The option of using a 5 GHz 802.11n point-to-point solution is a very attractive alternative — easily delivering the more than 100 Mbps of backhaul capacity an LTE cell will need.
Small cells are low-powered, multi-radio access points (cellular/Wi-Fi/backhaul) that improve indoor and outdoor coverage to increase capacity and offload traffic - as much as 80 percent during peak times. While small cells benefit 3G service deployments today, their importance will only grow as the industry moves towards higher capacity 4G/LTE, especially in urban environments. According to In-Stat's latest report, Femtocells and Small Cells: Making the Most of Megahertz, small cell shipments will reach $14 billion in 2015.
The problem is, as network operators continue to increase coverage and capacity and look to offload data to relieve traffic pressures, they also increase the stress on their cell site backhaul connectivity. In this small cell world, conventional point-to-point microwave, bonded copper and fiber-based backhaul solutions can quickly become impractical or uneconomical.
While microwave point-to-point equipment costs have come down in recent years, it generally requires a line-of-sight (LOS) link with the connecting backhaul hub, a condition many small-cell locations will be unable to meet. Sub-6 GHz NLOS solutions using a point-to-multipoint architecture are better suited for dense underlays, but when using licensed spectrum, narrow bandwidth channels put strict limits on backhaul capacity, and most sub-6 GHz spectrum bands are expensive and frequently not available for licensing.
Another choice, fiber, is clearly the preferred backhaul option for mobile operators (if you can get it). But pulling fiber to every small cell location is, well, just not going to happen. It's simply too expensive, disruptive and time consuming. Consequently, traditional cellular backhaul solutions must now be rethought in the context of moving to smaller cells.
Wanted: New Backhaul Options
New backhaul options, well suited for dense urban environments and for close-to-the ground equipment (both line of sight and non-line of sight), are required to make small cells viable.
Counterintuitive to most, unlicensed smart Wi-Fi has become a viable and affordable option to solve this problem and looks to play a crucial role in backhauling licensed small cell traffic. Yes, cellular traffic. Here's why:
Assume a mobile network operator (MNO) deploys an infill underlay radio network of small cells to add access capacity to areas where there is a high density of mobile data users, perhaps in an urban city center such as in London, New York or Hong Kong.
Today this small cell network would likely be comprised of lower-powered 3G and/or Wi-Fi nodes, or possibly in the future LTE radio nodes. No matter what the access radio technology is used, how does the operator get the data from the access radio node back to the network?
One obvious high performance solution is fiber, assuming that it's available. The operator may have to lease this fiber from a fixed line carrier which drives up operational costs, but perhaps more significantly there is the very real possibility that the fiber POPs will not exist in specific locations where the MNO needs to place the small cell.
The reality is that small cells only increase network capacity if you place them in close proximity to subscribers trying to access the network. Therefore site acquisition becomes a major determinant in the relative effectiveness of the small cell deployment.
But this then poses a very real problem – given the constraints of where operators must place small cells. It is highly unlikely that a fiber POP will exist in all of those locations. And given the cost and time delays of provisioning new fiber runs to each small cell location, an alternative solution is clearly needed.
Microwave radio links are of course a well-understood alternative technology that can be used to at least partially address the problem. But while microwave point-to-point (PtP) links are high performance, reliable workhorses for backhauling data and voice traffic, they have issues.
First and foremost, PtP microwave solutions generally rely on licensed radio bands for transmission. This improves reliability, however, acquiring new licensed spectrum takes deep pockets filled with lots of cash. Also radio capacity is directly related to how much spectrum is used for the radio transmissions. This means deploying more capacity on the access radio side exacerbates both the cost and the shortage of spectrum for the backhaul radio network. Add to this the problem that PtP radio links require highly skilled installation to aim or align the radio nodes. In a crowded urban area or near street level, this quickly becomes an onerous task.
Using the Unlicensed Band for Transporting Licensed Band Traffic?
Exactly. Wi-Fi has evolved to become an ideal solution for this small cell backhaul problem – if done properly.
New Wi-Fi technology has been developed that combines integrated adaptive directional antennas with smart meshing technology and predictive channel management – all used within the channel-rich 5 GHz 802.11n spectrum. The combination of these technologies makes the use of Wi-Fi for both line-of-sight and non-line of sight backhaul applications advantageous.
Adaptive antenna arrays deliver more reliable connectivity at longer ranges by focusing and steering RF energy only where it helps deliver the best throughput across a specified link. As the environment changes, these smart antennas mitigate Wi-Fi and non-Wi-Fi interference, constantly selecting better signal paths that yield the highest data rates and lowest latency at any given time. When used within the 5 GHz band, these antenna arrays become ideal for constructing highly resilient, long range, adaptive backhaul connections between Wi-Fi nodes.
Predictive channel management is then used to optimize RF channel selection by maximizing network capacity specifically in high-density, noisy public Wi-Fi environments. It does this by measuring actual channel throughput and building a statistical model that allows access points to learn over time what channel will yield the highest capacity. By relying on real-time, observed capacity on all 2.4 and 5 GHz frequencies, backhaul links can be automatically moved to a better channel with less interference thereby realizing higher data rates.
Utilizing smart mesh techniques with adaptive antenna arrays as an alternative to fixed PtP links eliminates much of the complexity associated with aiming and alignment during the installation process. This also results in a much more affordable solution with greater resiliency in crowded urban environments given its intrinsic capabilities to dynamically adjust to changing conditions by choosing alternate paths to the network.
In live field trials with multiple network operators today, this small cell Wi-Fi backhaul approach has proven to deliver reliable, carrier grade transport of 3G mobile data and circuit switched voice traffic along with the prioritized transport of timing signals (eg. IEEE 1588v2/PTP or NTP) necessary for small cell network synchronization.
Wi-Fi backhaul technology is currently being built into small cell nodes housing cellular and Wi-Fi access – within a fairly small footprint. This allows operators to deploy a single box to provide Wi-Fi access, cellular access and backhaul together.
Ultimately with small cells and better backhaul, mobile subscribers should enjoy higher speeds with more coverage in more places. In turn, mobile operators can reduce subscriber churn and increase revenue by having visibility into both cellular and Wi-Fi traffic – giving the customers more options to connect in more places.
Steven Glapa is director of field marketing at Ruckus Wireless.