The Challenge of Wireless Infrastructure:
Time to “Think Outside the Box”

Wireless systems, almost by definition, need antennas. In recent years local zoning and permitting issues have been a major barrier to the rollout of new wireless services and limiting factor in many cases for quality of service and size of service area. Thus while FCC keeps creating new bands for both mobile service and fixed wireless access (FWA) that each require new antennas spaced throughout their service areas, little has been done to facilitate installation of new antennas in the real world.

What are long term needs for wireless antennas? We know there are the following CMRS spectrum blocks: the original 2 cellular blocks at 800 MHz, 6 blocks for broadband PCS at around 1900 MHz, and 6 for AWS-1 at 1700/2100 MHz. That makes 14 although in practice there are fewer independent systems in a given area due to licensees holding adjacent blocks. To this one must add other CMRS spectrum such as the BRS/EBS systems that are replacing MDS/ITFS, the new 700 MHz licenses, etc. Now larger Part 90 systems are also going to cellular architecture to get the spectral efficiency associated with it so it is likely that there will be a need for some Part 90 antennas on a cellular like grid throughout suburbia also. I am guessing that the long term need in suburbia will be in the range of 20-30 independent mobile and fixed antennas located on a grid of about a mile.

The antenna problem is basically a suburban one. In urban areas there are usually buildings around to mount antennas that can blend in reasonably well. In rural areas there are few antennas needed and they generally can be away from any neighbors. But in suburbia we get the need for a lot of antennas with neighbors. Let’s face it, the majority of suburban antenna installations look like “they were designed by engineers”. Now a few antennas are camouflaged to look like trees, but this is a very expensive process and it is not clear if it can be expanded to meet long term antennas needs.

The main action of the wireless industry and FCC has been to try to fine tune local preemption issues. But it is clear that there is little political appetite in this country for federal preemption of local zoning – not matter how appealing it is to the wireless industry. After all, FCC Chairman Ferris’ mentor was “Tip” O’Neill, famous for the phrase “all politics are local”.

I don’t know how far preemption will go, but I doubt it will go far enough to get enough antennas built quickly enough. The vast majority of suburban antennas built today are of the “endoskeleton” variety with a central tower, multiple platforms leased to different operators, and antennas one each platform mounted to optimize the cell plan of that operator. The visual impact is terrible and it appears that no one is in charge – no wonder the neighbors wage zoning battles.

The B-2 stealth bomber and F-117 stealth fighters were both designed by teams of aeronautical engineers and electromagnetic engineers who through out conventional design practices to design a plane that flew well and has virtual no radar reflections. I believe there is an urgent need to try a similar approach to suburban antenna design to help break the log jam that threatens the growth of our industry. The tower and antenna elements in current manufacturers catalogs are not the only designs that can be built! Indeed, most antenna elements sold are much more massive than they need to be because each is built for an “endoskeleton” system where it is assumed that it must be able to cope with stresses from wind and precipitation. We need to find ways to get architects/urban planners to collaborate with electromagnetic experts to explore new design approaches that will allow large numbers of systems to cover suburbia efficiently with less visual impact on their neighbors.

In Europe, Nokia sponsored a university design contest in this area a few years ago. In 2007 Ericsson unveiled its innovative “Tower Tube” design concept based on a sculptural “exoskeleton” made with an exotic dielectric concrete that would have antenna elements mounted on its inside surface. But in the US, carriers, trade associations and FCC seem to focus on almost solely on preemption as the solution to getting new towers built quickly, probably not realizing that the problem is accelerating due to new services that will add to the required number of antennas in suburbia.

While exoskeleton antenna designs are promising, they are not the only alternative to the current endoskeleton concept. Going back to technical basics, the radiation pattern of an antenna system is the 3-D Fourier transform of the currents in all the antenna elements. There are a lot of alternative current distributions that could yield the desired coverage. However, a natural limit that comes from the Fourier relationship is that antenna designs that are small in the horizontal plane, e.g. come from a modest diameter antenna system, will have less effective beam patterns and hence will require more sites to cover an area. Similarly, antennas that are limited in vertical length will have less control over their elevation pattern and are likely to have nearby “hot spots” – something that Motorola showed was an issue in the 800 MHz Nextel/public safety problem.

All the key players need to take a bigger look at the problem and start to understand design alternatives and the tradeoffs. I urge FCC to start making estimates of long term antenna requirements as it makes new allocations and changes service rules. I urge industry to start sponsoring innovative design contests for architect/electromagnetic engineer teams to start fleshing out more innovative design concepts for multiple systems. Identifying new antenna design as a key research area could also lead to federal R&D funding in this area. Finally industry might want to fund an academic study of fundamentally new alternatives to antenna system design.

[This post was originally written as for a new guest column in RCR Wireless' website and was scheduled for publication this week. Unfortunately with the demise of RCR it was never published there.]

Ericsson Unveils a Bold New Approach to the Cell Tower Issue

A major problem impeding the rollout of new wireless services is the compatibility of the antenna towers/infrastructure required and the local environment. This affects mobile operators, fixed wireless access providers, and even private land mobile (Part 90) operators who want to improve their efficiency by switching to a cellular structure. All of these systems need antenna sites in suburbia* roughly every mile or so. The basic endoskeleton system used today, a tower with multiple antennas bolted to multiple platforms at different heights is visually objectionable to a large sector of the public.

The wireless industry has been amazingly insensitive to this issue: building more and more visually discordant structures while they press - with limited success - legal approaches like local preemption. Chairman Ferris' mentor "Tip" O'Neill famously said "all politics is local" - a truism that in general dooms the preemption approach to tower siting.

Traditional endoskeleton antenna systems: No wonder the neighbors don't like them!




How many antennas are needed on each tower? FCC CMRS competition report says that 51% of US population in 2006 lived in counties with 5 or more CMRS carriers.

Wireless Competition Study Data

These need 5 or more antenna systems. And this is before the 700 MHz auctions and AWS auctions are finished! The BRS build out and other point-to-multipoint systems at higher frequencies will also need multiple antenna systems in suburbia. Finally, public safety agencies and other large Part 90 operators may well go to cellular architectures creating more antenna systems needed in a given area. I estimate that in a given area there will be a long term need for more than 12 mobile systems and 6 point-to-multipoint systems. That's a lot of antennas on a grid of 1 mile or less in suburbia if the traditional endoskeleton technology is used.

Ericsson's Tower Tube takes basically an exoskeleton approach, using a smooth exterior structure, in their case an advanced cement with fiber reinforcement, and antenna placement inside the structure so individual antenna elements are not visible. This type of approach could in general accommodate a large number of antennas and still keep a smooth, hopefully attractive design. As an added bonus, the Ericsson design keeps the equipment inside the tower, eliminating the usual security fence. However, a downside of the narrow diameter of this particular structure implementation is the limited horizontal width possible for each antenna with then limits sectorization of each tower. A larger diameter top, as on some water towers, would lessen this problem.

All sorts of variations on exoskeletons are possible besides the Ericsson one, the key issue is a nonconducting strong exterior. Other approaches with closely spaced flag pole-like structures are possible.

So my congratulation to Ericsson for "thinking outside the bun" and possibly convincing industry that new technology and improved design is a better approach than confrontation with local officials on preemption issues.




* While these wireless systems also need sites in urbanized areas and rural areas, the suburban problem is the most complex one. Urban areas generally have buildings that can be used to mount antennas with little visual impact. Rural areas generally have sparse population so antennas can be sited away from residences. While camouflaged antennas like fake trees have been used in limited instances, they are very expensive, not that convincing, and not readily adaptable to mounting multiple antennas