Dark Fiber: The Coming Revolution in Backhaul

January 1, 2014

By Ted Abrams…

This will be the decade of dark fiber; wireless is the reason. Revolutionary developments in connectivity and standards will drive down the costs of operating a wireless network; stimulate innovation, and open markets.

The story of dark fiber importance begins with the widespread adoption of the LTE standard. LTE (Long Term Evolution) describes the set of wireless technical standards developing according to a timetable set by the ITU (International Telecommunications Union). Specification work for the LTE standard is inherently open-source, performed by partners engaged in a global initiative to rationalize the chaos among wireless protocols, achieving highest and best use for society from the magic of wireless.

If you look back through the history of US wireless, there is an alphabet soup of incompatible technology standards including CDMA, TDMA, GSM, iDEN, WiMAX, and so on. From the Telecom Act of 1996 up to 2012, two major camps formed in the US around air interface technology standards: the US subscriber base for the CDMA (Code Division Multiple Access) camp was larger, including those served by Verizon, Sprint, and affiliates. T-Mobile, AT&T, and most non-US operators camped around GSM / UTRA (Universal Terrestrial Radio Access). LTE equips the four US national operators to agree on the same, interoperable wireless protocol. All LTE operators will use the newest schemes, now channelized OFDM, to facilitate the most efficient use of the airwaves.

As a network architecture, LTE is so much better than any other scheme that operators will concentrate on LTE and phase out prior generations of wireless. Public announcements by all US national operators confirm LTE as the unanimous choice for the future. Sprint’s decision to change the ClearWire network from WiMAX to LTE adds weight to the assertion that LTE is a true convergence standard, whether licensed with paired spectrum for duplexing by frequency or un-paired for duplexing by time division. Already developed and published, subsequent releases of LTE, the standards referred to as “LTE-A” (LTE Advanced) for the first time offer a solution to frequency incongruity within carrier license portfolios.

Our hodge podge of licenses blamed on flaws in spectrum policy can now be solved by technology. Previously, to launch meaningful service an operator needed a swath of contiguous spectrum. Smaller operators were often unable to put together a competitive portfolio under the burden of that mandate. Spectrum assets lay fallow; the national brands endured the same inefficiencies that drove smaller operators out of business each year.

LTE-A, beginning with Release 10, provides for carrier aggregation, combining small slices of spectrum (intra-band and inter-band) into a patchwork quilt of air signal, covering the need for very high bit-rate communications. This gives all operators, large and small, the opportunity to harvest spectrum from various bands and carry broadband traffic through the air with an aggregated, composite carrier. Most of the LTE equipment deployed in the USA is Release 9, so it will be a while before all networks incorporate the carrier aggregation features of LTE-A. Samsung Galaxy 4 LTE-A smartphones are online with service through the SK Telecom LTE-A network.

Kaleidoscope mosaics of LTE-harmonized frequencies move through antenna systems, filters and ancillary devices that must be redesigned to accommodate simultaneous operation across more bands than imagined possible a few years ago. Bright engineers will create designs, gadgets will be built, the networks will operate with aggregated frequencies, and consumers will benefit from better service – a sea change in the global ecosystem driven by LTE convergence

So, what does all that have to do with the claim that the next decade of wireless infrastructure development opportunity will center on dark fiber? When chaos prevented compatibility of radios between competing technologies, there was no standardized interface joining the parts of the radio itself. Each manufacturer devised proprietary methods for delivering payload into the mixer where it joins with the radiofrequency carrier for air transport. Early versions of a standardized interface gained consensus among GSM equipment manufacturers. Version 1.0 of that common public radio interface (CPRI) was of no great interest to operators outside the UTRA camp.

Version 6.0 of CPRI, cruising along in the wake of the LTE juggernaut, is embedded by the OEMs. Operators can move traffic between parts of their radios across CPRI for distances of 40 kilometers (24 miles). Toward the horizon, CPRI range will likely increase to 160 kilometers and beyond.

The CPRI link is what allows the split architecture of modern commercial radios to be physically separated by a great distance, which dramatically improves the economy of building a network. You can build your baseband processing unit on a piece of real estate that is economical, provide that location with robust backup electrical power and multiple links of access to public data networks, then run a CPRI link for 20 miles in any direction, presenting to your customers the full complement of services with no compromise.

Capacity of each remote unit can be a third to half as much as a full-blown cell site, using much less electricity and without any local backhaul; no Ethernet connection required. The remotes are distributed away from the central baseband processing unit, so the likelihood of common cause failure is less – a power outage that impacts all the remotes at the same time is less likely. However, even during power failure, customers served by CPRI-linked remotes can enjoy high reliability service. Because the remotes do not use air conditioning and operate with smaller power load than a conventional base station, each remote can be powered for many hours in an emergency by a small portable generator that a technician can carry to the location of the remote and chain to the base of the pole. This is not like DAS where operators are required to give up control over the specific location, focus, and coordination of each access node, compromising their own network performance to comply with the constraints imposed by the system host.

As an operator, you could have hundreds of your own small cells at the end of dark fiber strands activated by CPRI electronics embedded in the baseband-processing unit at the head end and embedded in the radio remote. These remote connections to your own base station would operate without any other active equipment in the middle. Dark fiber can eliminate 90 percent of the biggest piece in the recurring cost of operation. Dark fiber owners would provide a physical layer and nothing else. No bandwidth shaping, no routing, no switching, no electronics to maintain – just the physical layer.

Smart mobile network architects yearn for lower latency, higher bandwidth transport with their cell sites. Lit fiber architectures with ring geometry were best available technology for serving stationary users, but are sub-optimal with respect to modern mobile networks. Fortunes were made selling bandwidth on lit fiber, where the true cost of incremental use is essentially zero. As ARPU falls, operators pressed toward lower total cost of operation are challenging the value proposition of purchased bandwidth. LTE convergence permits fresh consideration of options revealing significant potential for reduced latency, increased capacity, and lower total cost of operation through placement of new, dark fiber infrastructure directly linking cell sites.

A core, not a switch, serves LTE radio access networks. An LTE core is smaller, less costly, and more energy efficient than a 4G/3G/2G switch. Subscriber transition to LTE, after voice over LTE (VoLTE) is embedded in all the smartphones, paves the way for switches to be shut down. The location of LTE core(s) can be optimized with respect to geometry of cell sites, free from restricted geometry of old landline networks. Cell sites connect to the LTE core across an “S1” interface. That interface is analogous to the old idea of connecting cell sites to the switch with T1 links. With LTE, cell sites connect to each other across an “X2” interface so that local traffic can be offloaded site to site. As operators learn to leverage the power of CPRI across dark fiber for improved service, they are recognizing the savings potential of moving other traffic across dark fiber (versus the certainty of increased cost to route traffic through lit fiber).

Dark fiber changes the game. Fast, low latency connections rule. The fastest path between two points is a straight line, not ring around the rosie. Some links are best maintained as purchased bandwidth across carrier grade Ethernet on lit fiber, but that is not the most efficient solution for all links. LTE convergence will achieve higher and better use from the magic of wireless, and create a wealth of opportunity for infrastructure investors who pay the price to understand new technologies.

Ted Abrams, principal of Abrams Wireless Inc., is a professional engineer licensed in many states with special proficiency in electrical engineering and chemical engineering. He holds two advanced technology patents, one for RF radiation shielding and one for power monitoring and control.