John Keating, P.E., EVP of Market Operations
Every decision has consequences. Like ripples in the surface of a pond that travel outward, each choice influences everything that comes afterward. These ripples intersect with those created by other decisions, forming new patterns in the process. Standing on the shore of a pond and watching the water’s surface probably isn’t the first thing that comes to mind when you think of wireless tower projects, but it’s an apt metaphor for the way antenna-related decisions throughout a tower project impact how the process unfolds and reshapes the next phase of the project.
Antenna Systems & Technology has invited me to write a series of interconnected articles that illustrate this principle in action. In a three-part article series, I will walk through the significance of antenna selection within each stage of the lifecycle for a “typical” wireless new site build project:
- RF design and antenna selection
- Site engineering, leasing and permitting; and
- Tower construction and testing
This series of articles is an expansion on a piece I contributed last year called, “Progress Comes in All Shapes and Sizes: A Practical Guide to Selecting and Deploying Antennas in Wireless Towers.” That article provided a detailed overview of how antennas have evolved dramatically in recent years and how that presents a series of choices and challenges to wireless professionals. The article provides readers with a practical checklist of considerations to keep in mind as they evaluate antenna options at the beginning of a project. For this series of articles, I am hoping to build on that piece by showing how those early decisions reverberate throughout a tower project, shaping the kinds of choices and challenges that arise as a project proceeds to completion.
Before diving into the rest of this article, I should note that no two wireless tower projects are alike, so no single narrative can successfully reflect the full spectrum of challenges that would occur during a project like this. With that said, I have designed this series of articles to focus on a “typical” new site build that is a composite reflecting the vast majority of projects that readers will encounter. My hope is that even if this is not an exact match for the specifics of your project, many of the takeaways will nonetheless prove valid and useful for you.
The First Ripples Come Early in the Design Process
A new site build (NSB) is not cheap. Typically, they come at a cost of tens of thousands to hundreds of thousands of dollars. Because of the steep costs involved, companies do not embark on these projects without good reasons, and the two primary drivers tend to be coverage or capacity.
Coverage – On the surface, coverage, or lack thereof, seems like an easy thing to determine. Either the subscriber can use their phone, or they can’t. In practice, however, it is more nuanced. The days of determining sufficient coverage by simply validating the ability to place and hold a call are long gone. The coverage benchmark today is largely driven by data speeds, and what constitutes “acceptable” data speeds is highly subjective. For instance, one wireless service provider may design their network for a data throughput rate of 20Mb/s over a given service area, while another may design for 2Mb/s.
Location of the users is also a major consideration. Is the majority of usage expected to be served indoor, outdoor or in vehicle? Is the setting rural, suburban or urban? Answering these questions will lead to the appropriate radio link budget and the production of coverage plots which will help to identify coverage gaps, and ultimately suggest possible locations for the new site.
Capacity – In densely populated areas, or other areas of high demand, the need for a new site may be capacity driven. Since the purpose of the capacity site is to offload demand from one or more existing sites, the new sites design and location relative to the demand is critical. Place it too far away, and the new site may not be able to capture enough of the demand to provide sufficient offload. Design it to cover too large an area, and you may simply shift the problem from the old site, to the new one. Service providers can utilize a number of different data sets, as well as local knowledge, to identify the most effective placement for a new capacity site.
Whether designing for coverage or capacity, the antenna selected must support the desired network result, but it must also meet other project requirements unrelated to network coverage and capacity, such as: planning and zoning codes, landlord requirements and structural limitations. As much as the RF Engineer wants to hold firm on having the “ideal” site to meet the design requirements, in the end they will get the site that strikes the best compromise between cost, quality and schedule. The way this compromise is settled has significant impact on later stages of the process and how the antenna selection and implementation will unfold.
What You Want vs. What Is Realistic
There’s an inherent tension to antenna selection during the design phase of a tower build. There’s what you would choose in a vacuum absent of the myriad of constraints and limitations in front of you…and then there’s what’s practical given all of those suboptimal factors.
Let’s take a hypothetical NSB coverage project and discuss the design specifications and ramifications of the antenna selection. The initial design is often a “desktop design,” meaning the engineer uses modeling software to identify the location of the coverage gap, as well as possible locations for a new site. In figures 1 and 2, coverage is depicted by the presence of color. The relative “heat” of the color, red versus yellow for instance, indicates relatively stronger or weaker signal. The lack of color indicates a lack of suitable coverage. If the terrain in the area is relatively flat, placing a possible new site candidate near the center of the gap is a good place to start. If the terrain is highly variable, the best placement could be far from the center of the coverage gap.
Wherever the optimal site location is, it is generally desirable to design the new site with three antenna sectors (covering 120 degrees per sector), each carrying a similar traffic load. The “best server” plots in Figures 3 and 4 show how demand is distributed amongst the serving antenna sectors before, and after the new site is added. In this view, a distinct color representing the best serving sector will be displayed. In this example, white is a distinct color, not indicative of a coverage gap. Another benefit to using the best server plots is that they can clearly illustrate when a sector is serving outside of its optimal coverage area. In the upper left side of figure 4 there is a small area (bounded in red), where sector 1-3 is overshooting the coverage of the new sector 4-1. This situation is common when new sites are added to the network. Integrating a new site into the network often involves readjusting the antennas of both the new site and the existing sites.
Although each sector is expected to cover 120 degrees in the horizontal plane, the actual antenna commonly selected has a horizontal pattern of 65 degrees between the half power points (+/- 3dB). The purpose of the narrow pattern is to minimize overlap between adjacent sectors. Too much overlap is inefficient at best, consuming more resources than is necessary to serve the user demand. At worse, too much overlap causes harmful intra-cell, and inter-cell interference, reducing signal to interference and noise ratio (SINR) at the user equipment (UE). Too little overlap between antennas can also be a problem, specifically with intra-cell handover (adjacent sectors on same site). Insufficient overlap can be a common problem when antennas are flush mounted to the sides of buildings. In those cases, the mounting surface (building) itself is an obstruction to the antenna. Mounting antennas on platforms at rooftop edges, across building corners etc., can mitigate the obstruction, but those solutions present their own complications e.g, cost, aesthetics and structural considerations.
In figures 5 and 6 you’ll see the horizontal polar plots for a three sector site designed with 65 degree antennas, and 90 degree antennas, respectively. The horizontal pattern of each sector is uniquely identifiable, by color. The 65 degree antennas enjoy >10dB dominance over adjacent sector signal for ~75-80 degrees of their 120 deg. coverage area (+/- 40 from pattern center). For the site configured with 90 degree antennas, >10dB signal dominance for the serving sector occurs in just ~42-45 degrees of the 120 deg. sector.
Another way to illustrate this is to overlay a 90 deg. pattern over a 65 deg. pattern. In Figure 7, you see the relative difference in the amount of energy that is transmitted outside of the intended 120 degree serving sector. For the 65 deg. antenna, approximately 95% of the transmitted energy is contained within the 120 sector. For the 90 deg. antenna, the transmitted energy contained within the 120 sector drops to 80-85%. Please note that the plot has a logarithmic scale, whereby vectors that have a relative magnitude of -10dB are 1/10th the magnitude of vectors that are un-attenuated. Vectors with relative magnitudes of -20dB are 1/100th the magnitude of un-attenuated vectors, and so forth.
As important as it is to manage the trade-offs of coverage and interference in the horizontal pattern of the antenna, it is also important to do it in the vertical plane. One effective way to control coverage and interference is to limit the height of the antenna above ground level. Generally speaking, higher antenna heights lead to larger coverage areas, which leads to the potential for the antenna to be both a producer, and recipient, of harmful interference. In settings where there is a lot of variation in the terrain however, adjusting antenna height alone may not yield good results. In these situations, the ability to fine tune the vertical component of the antenna pattern is crucial. The two common means of accomplishing this is through mechanical and/or electrical tilt. While “up-tilt” is used in some special cases, we will limit this discussion to the more common “down-tilt.”
Mechanical down-tilt is accomplished by simply tilting the physical antenna in the vertical plane, from perpendicular (90 degrees with respect to foundation), to something less than 90 degrees. Most antenna manufacturers sell mechanical down-tilt brackets that have a fixed pivot point at the bottom of the bracket, and a series of mounting holes at the top that correspond to a fixed tilt amount. The holes are usually set at 1 degree intervals and typically range from 0-10 degrees. While mechanical tilt can be effective, it comes with potential drawbacks, which makes electrical tilt a potentially better option.
Electrical tilt is accomplished by changing the phasing of individual antenna elements within the antenna array, relative to each other. In doing so, the antenna pattern can be tilted along a full 360 degrees of azimuth. Antennas can be manufactured with a fixed amount of electrical tilt, or made to support variable tilt through the use of a tilt actuator. The antennas most commonly deployed today use a remotely controlled, electrical tilt actuator. Assuming that the antenna was installed perfectly plumb, and everything is operating normally, an engineer or operations technician can access the tilt controller while at the site, or from the office. In doing so, the antenna tilt can be adjusted, or simply verified. The h-plane antenna plots in figure 8 show the dramatically different effects of mechnical and electrical tilt.
The H-Plane and E-Plane polar plots for a 0°, 4° and 8° electrical tilt are shown in figures 9 thru 11.
It may seem as though remote tilt antennas are a panacea for RF engineers. But, as useful as remote tilt antennas can be, they can’t solve all of the coverage and interference problems that engineers face, and also come with some potential negative side effects of their own. Antennas are subject to tremendous environmental stresses; wind, water, sun and ice, to name a few. As such, antenna, actuator and/or cabling failures can, and do, happen. It’s not unheard-of for a tilt actuator to become uncommunicative, or for the actuator to fail or jam. When those situations occur, the network operator may not know for certain how their antenna’s electronic tilt is configured. Assuming that the problem persists, the only solution is to send climbers up the tower to replace the offending part.
With the proper knowledge of the network and sophisticated modeling software [See footnotesand.], an RF Engineer can come up with a very thorough preliminary RF design. At a minimum, a preliminary RF design should include information such as preferred site location(s), radio equipment configuration, antenna performance parameters (gain, vertical beam width, horizontal beam width) and antenna installation parameters (height, azimuths, tilts). Armed with the preliminary design, the development team will then do a field visit to begin the final RF design stage and begin the Architectural & Engineering and the site acquisition phases (leasing, zoning and permitting). These next two phases of the site development project will determine how much of the preliminary design can be carried forward, if any! The next installment of this article series will discuss the site development phases in detail, emphasizing how the preliminary design choices may change in light of site engineering, leasing and permitting factors.
About the Author
John C. Keating is an Executive Vice President at Centerline Solutions, the trusted turnkey provider of network services to the wireless telecommunications industry. In this role, John has responsibility for all professional services nationally, as well as construction services in the Southeast, Desert Southwest and New York state regions. His portfolio of services encompasses: RF engineering; A&E; site acquisition; field technical services; Construction and network maintenance. John holds a B.S. in Electrical Engineering from the University of Colorado. With more than 20 years of experience in the wireless industry, he knows the ins and outs of managing all aspects of building and operating wireless networks.
Note: The graphic in Figure 8 is from an article found at www.rfwireless-world.com entitled, “Antenna Mechanical Tilt Versus Electrical Tilt.”
 Forsk’s ATOLL modeling software was used for the coverage and best server plots shown in this article.
 Commscope Planet viewer was used to create the antenna polar plots, unless otherwise noted.