Sunday, July 6, 2014

Blueprint: Street-Level Small Cell Wireless Backhaul For Outdoor Small Cells

by Erik Boch, MSEE, Peng
CTO, VP of Engineering and Co-founder
DragonWave Inc

The global small cell market can be represented as being partially “indoor” small cells (pico, femto cells), or “outdoor” small cells (“micro” cells). Backhaul availability is of critical/primary interest to the success of this mobile network segment (see Figure 1). In the indoor environment, small cells can often capitalize on existing backhaul infrastructure, however in the outdoor small cell case the picture is quite different.

In the outdoor environment, fiber “close” to a micro-cell site doesn’t generally mean that there is a point-of-presence which allows cost effective or timely deployment of a fiber spurline to the micro-cell site (located on a store front, or lamp-pole for example). As a result, wireless backhaul technology is a very important contributor to the expected build-out strategies in small cell networking. Fiber penetration statistics sit about ~16% , which includes FFTH and FTTB/E, so it’s logical to assume that Fiber to the Street-Lamp-Pole (SLP)” or “Traffic-Light-Pole (TLP)” is a very small value.

A growing market consensus is that micro-cellular network segments will tie into local macro-cellular points-of-presence (PoPs) . These PoPs tend to be on high point locations in the dense urban environment. The problem with servicing these from the macro PoP is that the “street furniture” onto which the micro-cell equipment is mounted generally doesn’t have a clear LoS connection path. Assuming installations at/near roadway intersections, only 5% - 15%  of these locations have clear LoS to the elevated macro PoP locations.

As a result of this reality, conventional Line-of-Sight (LoS) radio link technology has been seen as somewhat limiting, despite its other positive attributes, namely;
  • Many spectrum choices. Lots of available spectrum in site-licensed, area-licensed, or unlicensed segments.
    o At high frequencies, antenna sizes can be made very small whilst beam shaping for optimized spatial filtering of static and dynamic multipath contributors in the path.
  • Very high degree of deployment success certainty.
  • Predictable over-life availability performance.
  • Use of FDD technology which allows for low delay and low delay-variability.
  • The use of very high frequency radio systems in street-level backhaul links allows the use of the low-multipath channel just above the vehicular traffic in the roadways, assuming they have properly designed antenna beam shapes. This avoids, or largely mitigates flat and selective fading impacts to the radio link (which results in superior link performance stability).
Despite some of the attractive strengths of LoS radio systems, the path blockage reality has incented various non-Line-of-Site (so called nLoS or NLoS) radio system products to have been brought to market. Generally, these systems rely on low RF frequencies (i.e. < 6GHz) and the use of modem/waveform techniques that allow varying degrees of tolerance of the harsh propagation environments involved (i.e. OFDM, MIMO). There are several residual artifacts that the operator has to accept when adopting this technology, namely;
  • There is very limited spectrum available.
    o Difficult to achieve the needed high capacities.
    o If operating unlicensed, there is a significant risk of interference. This can negatively impact capacity, but also can severely impact delay and delay variability.
    o Spectrum allocations often drive the use of TDD technology, which negatively impacts delay and delay variability.
  • It is not possible to predict with certainty the success/outcome of the installation of a non-LoS radio link.
  • It is not possible to know the over-life radio link availability.
  • Small physical form factors [usually] dictated by microcell installations cause larger beamwidth operation. This in turn leads to an increased vulnerability to flat and selective fading, particularly acute when considering street-level links in which the dynamic impacts of vehicular traffic can have severe impacts to the radio channel performance.
Higher frequency waveforms have also been used to build nLoS/NLoS links using reflected/bounced path geometries. 28 GHz was used in the 1990’s by CellularVision to deliver consumer TV services. Recently similar types of links have been demonstrated successfully at frequencies above 6 GHz and as high as 60 GHz . However, these suffer from difficulties in predicting performance in advance of installation, similarly to sub-6GHz.

The resulting reality is that the various wireless solutions have different attributes that need to be optimally combined such that the resulting “networking” solution provides a predictable, reliable and available backhaul; function.

So how does this get done? One backhaul networking solution (see Figure 2) can be composed of the following:
  • LoS high capacity radio shot from a macro [high] site to a street level PoP.
  • Once at the street level, build rings, zig-zagging LoS radio shots up the center of the roadways. The use of high frequency delivers high capacity, low delay & delay variability, and stable operation in the presence of vehicular (and other) multipath.
  • The use of self-healing rings is desirable because in order to keep availability high … as there are unforeseen propagation impairments, and network node outages that are a reality. Downed poles, interruptions to power, elevated maintenance equipment (etc.) need to be realistically considered.
  • Spur shots used to pull in local base station sites which can’t be directly designed onto the ring path.
  • Use of nLoS radio technology to address [the odd] blocked link which can’t be deigned out of the solution.



In summary, understanding the beneficial attributes and limitations of various wireless technology solutions can allow optimized combinations of LoS and n/NLoS technologies into reliable and deployable backhaul networking solutions – a key enabler for outdoor small cell deployments.

About the Author
Erik Boch holds a Masters degree in Electrical Engineering from Carleton University in Ottawa and is a registered professional engineer. Erik has held senior engineering or technical management positions at a number of communications and aerospace companies namely Litton Systems, ComDev, Lockheed Martin and Alcatel Networks (formerly Newbridge). While at Alcatel, Erik was AVP of the Wireless Systems Group and was involved in various aspects of microwave & millimeter wave subsystem and system design for more than 22 years. Erik led the R&D team at Alcatel (formerly Newbridge) that introduced the first ATM-based Fixed Wireless Access System in the industry.

Erik has been published extensively in major networking publications, including Telephony, Microwave Journal, Wireless Review, Internet Telephony and America's Network. Erik has spoken at numerous industry events including IEEE sessions, WCA, Broadband Wireless World and IWPC. Erik has served on the Technical Advisory Board of the NCIT (National Capital Institute of Technology). Erik holds several approved RF design patents, with numerous patents pending.

About DragonWave
DragonWave is a leading provider of high-capacity packet microwave solutions that drive next-generation IP networks. DragonWave's carrier-grade point-to-point packet microwave systems transmit broadband voice, video and data, enabling service providers, government agencies, enterprises and other organizations to meet their increasing bandwidth requirements rapidly and affordably. The principal application of DragonWave's portfolio is wireless network backhaul, including a range of products ideally suited to support the emergence of underlying small cell networks. Additional solutions include leased line replacement, last mile fiber extension and enterprise networks. DragonWave's corporate headquarters is located in Ottawa, Ontario, with sales locations in Europe, Asia, the Middle East and North America. For more information, visit http://www.dragonwaveinc.com.

1  Data from www.OECD.org , June 2013
2  Small Cell Forum
3  DragonWave field data
4  Ericcson at 28 GHz, DragonWave at 60 GHz