3 Challenges Backhaul Networks Must Address
You do not have a “cell phone”.
Reach into your pocket, purse, backpack, or briefcase. There is probably a smart device of some kind in there. You may even call it a smartphone. But, it is not just a phone. That device is much more than that, having become the lifeblood of managing our lives on a daily basis. To most, it is a necessity.
“To deliver this low latency network, backhaul design teams need to plan for the most direct route possible and avoid simply putting backhaul traffic on a large network ring that can take hundreds or possibly thousands of miles to reach the mobile carrier’s hub. Every 180 miles of fiber optic cable distance will add 1ms to the latency, no matter how efficient the equipment tied to the fiber is.”
As you likely know, the network requirements to support this reliance have also had to change. Mobile network operators have all spent a tremendous amount of time and capital dollars deploying more macro cell towers or small cells. On top of that, the cellular companies have been upgrading those from 2G, 3G, to 4G (LTE), and soon to 5G. But, that part of the mobile network is only one half of the network equation. The other half, backhaul, will also need to continue in this evolution.
Early on, mobile backhaul once was the domain of the incumbent local exchange carriers (ILECs) and they provided a T-1 to a cellular site over the existing copper telephone cables. If it was a large cell site with multiple carriers, maybe a few T-1s were needed. And while each T-1 could support 24 simultaneous voice calls, most towers only needed a few T-1’s at the most.
When digital communications came to the cellular industry in the form of 2G, it brought the promise of data capabilities like CDMA2000 1x and GPRS. While they only offered around 144 kbps of speed, this was the beginning of mobile data connectivity that started a mobile data race for the fastest, newest technology.
However, the small data potential of 2G networks still meant the traditional T-1’s provided by the ILECs were more than sufficient to meet demand, and changes in backhaul infrastructure were not yet needed.
As the wireless companies continued their data speed “arms race” with the launch of 3G services like EVDO, UMTS, and EDGE, wireless networks were now able to provide near 1 Mbps speeds. With a T-1 offering only 1.4 Mbps of throughput, backhaul now had to change. However, this change was not very radical, and, in most cases, the solution was to add more T-1’s.
In the largest capacity sites, fiber optic cables were being placed to support DS3 levels of capacity (50 Mbps). Even with the addition of fiber directly to the cell sites, the overall network had still not changed very much, consisting primarily of synchronous optical network (SONET) rings either feeding the site directly with OCx level services or feeding multiplexers (MUXs) that served individual T-1’s. All of these were still provided by the same telephone company networks.
The real changes were yet to come.
Finally, 4G and its associated services like WiMAX, or the more common long-term evolution (LTE) technologies were deployed. Mobile networks were now capable of offering speeds to a device of 50 Mbps or more. No longer was a SONET service like DS3 or multiple T-1’s sufficient. It was now time for the traditional telephone company network to step aside and make room for Ethernet backhaul.
For the last 4 to 5 years, fiber has been extended to cell sites. Ethernet services of 100 Mbps up to 1 Gbps are being provided to cell sites in support of all the Facebooking, emailing, gaming, and other non-phone things that smart devices can do. In fact, according to the Pew Research Center, making actual phone calls is not even one of the top two most-used features on smart devices. (Source: http://www.pewinternet.org/2015/04/01/us-smartphone-use-in-2015/)
Even with all this progress and the upgrades being done on both the wireless and backhaul side of the cellular network, there is still a substantial amount of change required to support the future. Although backhaul networks are now supporting 1 Gbps of throughput to a tower over fiber optics, backhaul operators can’t kick their feet up on the desk just yet.
Today’s backhaul network is not a future-proof architecture for 3 reasons: latency, resiliency, and dark fiber availability.
Challenge #1: Latency
The backhaul network of the future will need to be able to support very low latencies. No longer is it acceptable to deploy a large Ethernet pipe to the tower that has to physically route packets hundreds of miles out of the way on its return path to the mobile switching center. Latencies of 10ms or less are now table stakes and will continue to be driven down to the time required for light to travel through the fiber optic cable glass.
The growth of real-time data communication is driving this and even those seldom-used voice calls switching to data-stream voice over LTE (VoLTE) will continue to require low latencies. To deliver this low latency network, backhaul design teams need to plan for the most direct route possible and avoid simply putting backhaul traffic on a large network ring that can take hundreds or possibly thousands of miles to reach the mobile carrier’s hub. Every 180 miles of fiber-optic cable distance will add 1ms to the latency, no matter how efficient the equipment tied to the fiber is.
Challenge #2: Resiliency
While the performance of the backhaul service is important, the resiliency of the network is even more so. To truly enable all the things these magical devices can do, the most critical thing is a working connection. Society is becoming totally reliant on mobility, so failures in the backhaul network can have serious consequences — from inconveniences to potential medical emergencies. Deploying the backhaul network of the future must have network resiliency at the core of the design so as to eliminate any single point of failure from cell site to the switching center. Varying levels of risk in different segments of backhaul networks have been accepted up to this point, but that will not suffice in the mobile-reliant, data driven society that is being cultivated for the future.
Challenge #3: Dark Fiber
The final piece to enable the backhaul needs of the future may be a bit uncomfortable for some within the industry: dark fiber. Many providers have avoided providing dark fiber, and have instead focused on offering lit service options because offering lit service offerings allows for future revenue growth as bandwidths or service needs increase. This is very understandable, given the expense to construct fiber throughout difficult urban or rural terrain. But, some cellular networks are starting to slowly migrate to cloud-controlled platforms (Centralized Radio Access Network, or C-RAN, for example), and there simply is no ability to use an Ethernet service with those architectures.
The process of deploying a cloud-controlled platform is akin to moving the brains of the cell site into a central location. It extends the connection that would typically run between the base of the tower and the radios at the top, to running between the same radios at the top of the tower and all the way back to the cloud location (usually either a central cell site or mobile switching center). To do that, dark fiber is required with some incredibly low latencies in the nature of 2 milliseconds or less. Building backhaul networks that are fiber-dense, with options to provide dark fiber for these future transitions, will enable backhaul providers to be a part of this technology shift by the wireless operators.
To meet those challenges, Horizon Network Partners, has been deploying cellular backhaul networks that meet these requirements in place for a few years now. Delivering solutions to address the above challenges enables the long-term growth of the wireless carriers. The best way to do that is through a highly capable, resilient, and fiber-dense network architecture.