IEEE Wireless Communications - April 2017 - page 6

IEEE Wireless Communications • April 2017
4
S
pectrum
P
olicy
and
R
egulatory
I
ssues
Most wireless engineers are aware that national regulations
in their country impose requirements on spectrum availability
and equipment parameters, but may be unaware of the tech-
nical problems such regulations were intended to address. The
general goals of such spectrum regulations from the beginning
of most formal regulation in the aftermath of the
Titanic
sinking
in 1912 have been, and remain today, to avoid interference
and maximize the use of available spectrum. In this issue’s col-
umn we will focus on interference issue.
Cochannel interference is the most obvious case, and life
in spectrum policy would be very simple if radio propagation
was monotonic with distance and deterministic and all inter-
ference could be prevented by keeping cochannel stations
far enough away from each other. Unfortunately, in many sit-
uations path loss does not always increase monotonically with
distance. This deviation from monotonicity is due to a variety
of factors, including multipath and other non-free space prop-
agation issues such as ducting, sporadic-E propagation, and in
the case of lower frequencies, skywave propagation. Not only is
the decrease not monotonic, but in many cases path loss varies
stochastically as one travels in a path of constant distance from
a transmitter. Spectrum policy would be much simpler if radio
propagation was dominated by free space loss as light propaga-
tion usually is!
In any case, controlling cochannel interference through care-
ful allocations and frequency assignments is a problem going
back to the construction of Marconi’s third radio transmitter,
and is relatively easy to understand. Cochannel systems are
spaced geographically far enough apart that even with unlike-
ly propagation modes the interference risk is acceptable. In
the case of disparate radio services at microwave frequencies,
where high gain antennas are practical, such as fixed service
microwave and geostationary orbit satellites, frequency assign-
ment rules that keep the transmitter beams of these services out
of the receiver antennas of the other are commonly used.
But cochannel interference is not the only problem. Anoth-
er real world complication of spectrum policy is the fact that
practical
receivers have finite susceptibility to interference from
out-of-band signals, both adjacent channel signals and signals
somewhat further away. These signals can actually interfere in
two ways:
1. Their out-of-band emissions (OOBE) can be in the receiver’s
desired signal bandpass and be impossible to filter out in the
receiver processing after a single antenna input, although
multiple antenna systems such as MIMO technology can
suppress a signal if its arrival direction is different, and spread
spectrum/CDMA technology allows suppression of cochan-
nel signals if they are orthogonal.
2. Strong nearby frequency signals can overload a receiver front
end, driving it into nonlinear modes even if the emissions are
not in the nominal bandpass of the receiver.
OOBE interference is similar to cochannel interference
because it is in the passband of the victim receiver. Thus,
for spread spectrum/CDMA systems it can be suppressed
by the processing gain of the receiver. Multiple antenna
receiver systems can also suppress such signals if they are
not collinear with the desired signals. However, using such
processing gain for OOBE suppression either complicates
receiver design or reduces other design objectives for the
receiver.
The non-OOBE impact of adjacent strong signals is both diffi-
cult to understand and often results in spectrum regulations that
address it in confusing ways. When UHF analog television was
common in the U.S., a complex set of rules to avoid this type
of interference went by the mysterious name of “UHF taboos”
[1]. These rules were based on what receiver performance for
mass produced consumer equipment was projected to be in
1952, and avoided interference to such receivers by spacing
transmitters a minimum distance from reach other based on rel-
ative frequency and extended to offsets of 15x6 MHz channels
or 90 MHz. The indirect impact of these rules was that only one
out of every six TV channels could be used in a given city, a
graphic demonstration of how rules to prevent interference due
to receiver limitations can limit efficiency of spectrum use! With
the advent of new receiver technology and the introduction of
digital television broadcasting at the beginning of the new mil-
lennium, these rules were relaxed so TV stations can be packed
much closer in frequency in a given city.
National and international regulators traditionally seek to
control intersystem interference by setting upper limits on either
transmitter power or effective isotropic radiated power/EIRP, a
combination of transmitter power and maximum antenna gain.
This approach works well for cochannel interference cases
where the possible interference victims are far away from the
transmitter due to licensing regulations discussed above. But
it is more problematic for adjacent channel/band interference
due to receiver overload/nonlinearity. About 20 years ago the
U.S. had a problem with a licensee that received regulatory per-
mission to convert a high antenna wide area mobile dispatch
system to a cellular coverage scheme with many lower anten-
nas covering the area with frequency reuse. The resulting lower
EIRP antennas also had field strength hot spots near their sites
that had not existed in the original high antenna case. These hot
spots created receiver signals of –20 dBm to –25 dBm in adja-
cent channel receivers, which in turn resulted in receiver-gener-
ated intermodulation interference. The high signal strength near
each transmitter power had no direct benefit to the transmitter
operator since his main concern was fringe coverage, but had
a very negative impact on users of nearby spectrum. This prob-
lem might have been avoided or corrected by antenna designs
that sought to limit hot spots near antenna sites, but this is not
an issue commonly addressed in national or international regu-
lation.
More recently a related problem occurred when a cellu-
lar based station designed to improve building penetration in
nearby tall buildings resulted in signal levels at street level in a
major U.S. city of greater than –10 dBm at receiver inputs of
automobile mounted satellite audio broadcasting systems [2].
As in the first case, the direct cause of the interference was
receiver-generated intermodulation resulting from the high
signal strength in a nearby band. This problem was ultimately
addressed through a combination of decreasing the cellular sig-
nal strength at street level through alternative antenna designs
and increasing the broadcast system signal in some locations
with repeaters. This voluntary solution turned out to be faster
in this case than developing a more general policy to prevent
R
egulatory
A
pproaches
for
C
ontrolling
C
ochannel
and
N
earby
F
requency
I
nterference
M
ichael
J. M
arcus
1,2,3,4,5 7,8,9,10,11,12,13,14,15,16,...132
Powered by FlippingBook