Radiocommunication Study Groups
Received: 28 October 2009 Document 6A/278-E
30 October 2009
Source: Documents 6A/215, 6A/211, 6A/176 (Annex 2), English only
6A/141, 6A/140, 1/67, 1A/135 (Annex 1) and
British Broadcasting Corporation (BBC)
COMMENTARY ON WORK IN ITU-R ON POWER LINE HIGH DATA
RATE TELECOMMUNICATION SYSTEMS
(Question ITU-R 221/1)
In PLT systems, RF energy is applied to the live and neutral wires of an electrical mains supply as a
differential signal. However, mains wiring is inherently unbalanced and real world installations
contain many earth loops and stubs. A proportion of the RF energy will therefore be converted into
common-mode signals with respect to ground, which will inevitably radiate from the mains wiring.
This has resulted in ITU-R expending considerable effort in exploring if these leaked emissions
from PLT systems can be modelled at HF in a repeatable and consistent manner, in order to assess
the potential for causing interference to the various radiocommunication services.
After several years work in Working Party 1A, Study Group 1 has approved Report ITU-R
SM.2158* on the “Impact of power line telecommunication systems on radiocommunication
systems operating in the LF, MF, HF and VHF bands below 80 MHz” (Document 1/67).
A companion Report developed in WP 1C, Report ITU-R SM.[PLT-MEASUREMENT],
“Measurement methods for power line high data rate telecommunication systems”, provides the
latest guidance on how to carry out measurements of radiated and conducted emissions from PLT
The main developments at the September 2009 meeting of WP 1A are summarised in Annex 1.
The need now is to determine how to define and apply limits on the leaked interference from PLT
WP 1A has also been considering a working document towards a preliminary draft new
Recommendation. There were no substantive developments on the text developed so far, because of
lack of time, but the material was rearranged in a more logical way. There is no consensus yet on
what an eventual Recommendation should contain. Some views favour giving guidance on the
design and implementation of PLT systems, along with examples of national regulations, while
* The Syrian Administration reserves its position on this Document.
other views favour just setting out the protection requirements for the various radiocommunication
services and leaving national administrations and organisations such as CISPR and ETSI with the
task of applying this advice as appropriate. Moreover, there was an emerging view that a
Recommendation may not be necessary at all. The current draft text is available as Annex 1 to
As regards guidance on what techniques could be applied to the regulation of in-house PLT
systems, three approaches have been considered at various times in ITU and other fora:
1) to define limits on the power supplied by PLT modems to the power lines used for the
distribution of mains electricity;
2) to define limits measured at outlet ports based on the behaviour of mains wiring models
(artificial mains network, an asymmetric artificial network or an impedance stabilization
network) and expectations of how common-mode currents are excited;
3) to define limits on sensitivity degradation caused by leaked emissions from PLT systems
relative to the existing interference environment.
These are not in fact independent approaches because a limit on PLT power supplied to an electrical
distribution system will result in some level of leaked emissions and a direct limit on leaked
emissions will demand that the PLT power supplied is low enough not to exceed the leaked
Conducted versus radiated emissions
PLT systems can disturb the operation of electrical and electronic equipment through:
conducted emissions along the mains wiring, directly into other equipment connected to the
mains electrical power supply;
inductive pick or reception up of radiated emissions from the mains wiring by low level
signal cabling used by other electronic equipment in the vicinity.
In the case of radiocommunication equipment leaked emissions from PLT systems can be received
directly by the RF circuitry and a receiver’s antenna system.
The problem of how to deal with radiated interference in the LF/MF/HF bands is particularly acute
because the mains wiring powering radiocommunication equipment can also behave as part of the
antenna. Thus, radiocommunication equipment is susceptible to interference from PLT systems by
direct injection via common mains wiring, induction into signal/control cabling or directly into a
receiver’s RF input via the antenna. RF energy injected by conduction, induction or direct reception
can degrade the operation of radiocommunication equipment even when the equipment is not
connected to the same mains supply – for example, when using batteries or an isolated clean
The problems of conducted and radiated interference are not confined to PLT systems. Many other
electrical and electronic devices generate RF energy. For example, fluorescent lighting can also
disturb the reception of radio signals. On 8 June 2009, the US Coast Guard issued a Marine Safety
Alert (Alert 02-09) advising the maritime industry that “energy saving Compact Fluorescent Lights
(CFL) or lighting, sometimes known as radio frequency (RF) lighting devices may interfere with
certain communications equipment”. As a result of these and similar concerns, the FCC has
required manufacturers of CFLs to provide an advisory statement, either on the product packaging
or with other user documentation, similar to the following: “This product may cause interference to
radio communications and should not be installed near maritime safety communications equipment
or other critical navigation or communication equipment operating between 0.45-30 MHz.
All the studies on PLT systems so far indicate that the interference potential is considerably greater
than that of other sources of leakage and radiated emissions from electrical and electronic
equipment that generate RF energy. These types of electromagnetic compatibility (EMC) problems
fall within the remit of organizations such as CISPR, for control purposes, and of standardization
organizations, such as ETSI, for producing the related equipment standards.
Of particular importance to ITU-R is how to protect radiocommunication equipment from
conducted and radiated disturbances. At HF, the impact of leaked emissions is of particular concern
because the length of mains wiring used by typical in-house PLT systems and outside access PLT
systems is such as to form an antenna capable of causing disturbance to HF receivers over a wide
area. Moreover, the leaked emissions at HF can propagate by skywave and thus increase the noise
floor at HF around the world through the accumulation of many individually small noise-like
CISPR is the body charged with protecting the RF spectrum from conducted and radiated
disturbance from electrical and electronic equipment. The IEC CISPR 22 standard, “Information
technology equipment – Radio disturbance characteristics – Limits and methods of measurement”,
is the operative reference in respect of PLT systems.
CISPR 22 applies to information technology equipment (ITE) in general, which includes data
processing equipment, office machines, electronic business equipment and telecommunications
equipment, and sets limits on conducted and radiated emissions for two classes of equipment:
Class A ITE: Not intended for domestic use.
Class B ITE: Intended for domestic use.
CISPR 22 also provides procedures for the measurement of the levels of “spurious signals”
generated by ITE. Limits are specified over the frequency range 9 kHz to 400 GHz for both class A
and class B equipment (NB: no measurements need be performed at frequencies where no limits are
The intention of CISPR 22 is to establish uniform requirements for the radio disturbance level of the
equipment contained within its scope, to fix limits of disturbance, to describe methods of
measurement and to standardize operating conditions and interpretation of results.
CISPR is currently working to determine if the limits of CISPR 22 can be relaxed to accommodate
PLT systems. The current valid edition of CISPR 22 (edition 6) sets a limit of ~56 dBµV1, averaged
over a 9 kHz bandwidth in the range 5-30 MHz, for the maximum differential-mode RF
disturbance voltage level at the telecommunications port (i.e., the main wiring in this case).
This translates into a power spectral density (PSD) level of -93.5 dBm/Hz.
PLT modems typically operate with a PSD output in the range -75 dBm/Hz to -55 dBm/Hz which,
assuming a terminating impedance of 100 Ω, translates into disturbance voltage levels in the range
74.5-94.5 dBµV averaged over 9 kHz (conversion factors: dBm/dBµV = 110 @ 100 Ω and
1 Hz/9 kHz = 39.5 dB). This output level is some 20 to 40 dB above current CISPR 22 limits.
Moreover, the level of leaked emissions will be even greater than assumed in the application of
CISPR 22 limits. These limits were developed under the assumption that ITE operation is over
telecommunication lines with constant, known characteristics – ideally over well-balanced lines.
However, PLT systems operate over the mains wiring, which is certainly not a well-balanced
1 Actually 50 dBµV on each of the live and neutral wires.
telecommunication line and does not even exhibit constant characteristics. The topology of mains
wiring in the home will change according to what other electrical appliances and lights are
connected or switched on.
The problem then is how to apply limits on leaked emissions from PLT systems and then measure
whether PLT systems can comply with such limits. The characteristics and installation of PLT
systems can vary considerably from place to place and country to country. Variations in mains
wiring topology include ring circuits or spur connections, bi-phase or 3-phase, use of switched
socket outlets and grounding of the neutral line to protective earth. Also, in-house PLT system will
normally use 2 ~ 4 modems. All these factors will contribute to how the mains wiring will leak RF
energy and disturb radiocommunication equipment.
Options for assessing PLT impact
Approach 1) This depends on setting standards for the power output level and spectral
characteristics of PLT modems. The Type 1 modem solution being considered in CISPR couples
this approach with various mitigation techniques to reduce modem output power in specific bands
through notching and automatic power reduction to match the attenuation of the mains wiring.
Mitigation measures are necessary in for the Type 1 modem proposal in CISPR because typical
PSD output is -55 dBm/Hz, which is nearly 40 dB higher than the -93.5 dBm/Hz level, which was
noted above as being equivalent to current CISPR 22 limits.
Approach 2) This attempts to overcome the difficulties with Approach 1) by postulating how
expected representative models of electricity distribution networks will perform in practice when a
PLT system is established. This Approach is based on a consideration of the common-mode
currents at the output connection points and is similar to the Type 2 modem solution also being
considered in CISPR. The method attempts to apply the well-established methods of telecom-port
measurement methods and assessments used in CISPR 22 to PLT. Although the typical modem
PSD output of -75 dBm/Hz, for Type 2 modems is 20 dB lower than the Type 1 modems many
question remain on the reasoning underpinning this approach.
Approach 3) In contrast, aims to set direct limits on emissions leaked into the environment in the
same way as standard methods for assessing compatibility between radiocommunication services.
This is a neutral approach that does not give any priority to leaked PLT emissions over comparable
emissions from other sources of RF interference. Moreover, the level of emissions is capable of
being measured or deduced from measurements, which provides the advantages of transparency and
objectivity. Nevertheless, it will be difficult to separate the leaked emissions from individual PLT
systems and the cumulative impact of wide scale deployment also needs to be addressed.
Regulation based on actual emissions will therefore need to make assumptions about the scale of
use of PLT systems.
Analysis of options
In respect of Approach 1), the apparent simplicity is offset by the many uncertainties and resulting
assumptions that need to made in order to ensure that modem PSD output will not exceed
acceptable levels. Although PLT modems feed a differential-mode signal into electricity wiring, the
highly unbalanced nature of electrical supply wiring means that the differential-mode and common-
mode are strongly coupled. The leaked emissions from PLT systems result from common-mode
currents flowing along the length of the wiring in buildings or electricity distribution networks.
Approach 1) therefore demands a rigorous consideration of the architecture and topology of
electricity distribution systems, and a statistical analysis of deployment, in order to be certain that
the level of leaked emissions found in real world situations will not cause harmful interference or
raise the noise floor to detrimental levels. If these problems can be resolved such that the expected
amount of leaked emissions from PLT systems when connected to representative PLT modems can
be estimated with an acceptable degree of certainty and repeatability, then this approach may
provide a useful basis for regulation.
Approach 2) is, however, fundamentally flawed. In “classical” telecom-port applications use cables
of balanced construction to connect one port to another, without stubs or other unbalancing features.
The common-mode current arises as a result of minor imperfections in the cabling; it also represents
the only significant cause of interference emissions. In contrast, mains wiring is not of this tidy
form. As noted for Approach 1) above, it is the common-mode currents flowing along the power-
line network that contribute to the leaked emissions, not just the common-mode currents flowing at
the outlet. The flaw in the reasoning is to assume that if common-mode currents at the injection
point can be avoided then so can interfering leaked emissions. However, even if common-mode
currents at the injection point could be avoided with pure differential-mode injection, common-
mode currents will still be excited elsewhere in the mains wiring, such as stubs and loops, thus
causing corresponding emissions remote from the injection point. Because measurements taken to
support the Approach are taken at outlet points, this method is likely to underestimate the total
amount of leaked emissions along the length of the electricity distribution wiring comprising a PLT
As noted in Annex 1, Document 1A/174 provides new insights on how electricity wiring in the real
world will behave when a PLT systems is connected. The processes involved in both of
Approaches 1) and 2) depend on many assumptions, which will not be the same for different
practices in countries around the world. Important factors are 3-phase versus bi-phase distribution,
overhead line versus underground cable distribution, length of cabling per phase from the final step-
down transformer to or within buildings. There also fundamental assumptions on how much of the
RF energy in a PLT system stays as differential signals on the live and neutral wires used to
distribute the data and how much energy is converted into common-mode signals relative to ground
potential that will leak RF energy from the electricity distribution wiring into free-space by
Document 1A/174 finds that theoretical analyses, such as that of earlier drafts of the PLT Report
(see section 2 of Document 1A/158), are not a reliable guide to amount of leaked emissions found
in practice. Document 1A/174 has now been incorporated in to Section 2 of the published Report
ITU-R SM.2158, but with the acknowledgement that many uncertainties remain on how the mains
wiring supporting PLT systems will leak emissions into the RF environment and interferes with
A further problem with Approaches 1) & 2) is that, in order to overcome losses, the modem output
power necessary to provide service will vary with the length of the distribution network. However,
this factor does not necessarily correlate with the fraction or energy converted into leaked emissions
and there are various opinions on what representative wiring lengths should be assumed. The length
is also likely to vary from place to place and country to country according to the preferred housing
However, there is one element of certainty concerning modem output levels. The liaison statement
from WP 1A to WP 7D of 18 March 2009 (Document 7D/85, copied to WP 6A as
Document 6A/140) noted an opinion within WP1A that certain tests carried out on PLT modems in
2002 might no longer be relevant to the situation today. Nevertheless, there was an acceptance that
the modems in question were “capable of causing severe interference”. The power spectral density
(PSD) delivered to the electrical distribution system by these modems was, in fact, measured at
–50 dBm/Hz (see Document 1A/200 and related comments in Annex 1). This level of PSD may
therefore serve as an absolute ceiling for PLT power output as a PSD figure. In practice, this PSD
level would have to be reduced by a considerable amount, of the order of tens of dB, in order to take
account of the cumulative effect of a dense population of PLT systems and to ensure that the
generality of radiocommunication systems operating at any location should not suffer interference.
However, as noted in Document 1A/157, a representative maximum output power from many
current PLT modems, such as the Type 1 modems being considered in CISPR, is –55 dBm/Hz. This
is only 5 dB less that the acknowledged ceiling of –50 dBm/Hz. Moreover, there is a new ITU-T
Recommendation G.9960, part of which addresses in-house PLT system based on a reference PSD
of -50 dBm/Hz in the 2~30 MHz range and -80 dBm/Hz in the 80~200 MHz range.
The main conclusion on Approaches 1) & 2) are that there are too many uncertainties to accept
these as a reliable and repeatable way of regulating PLT systems. Moreover, the representative
modem PSD output levels being considered in CISPR and ITU-T are so high that they would give
rise to harmful interference to radiocommunication services and raise the noise floor across the
In respect of Approach 3), a commonly used criterion for establishing an acceptable level of
interference from a station in the same radiocommunication service, or in another
radiocommunication service sharing a band with equal status, is a maximal sensitivity degradation
of 0.5 dB. This ½ dB rule is extensively used in terrestrial broadcasting Plans, such as GE75 and
GE84, to assess the impact of a new or modified station on the useable field strength of existing
stations in the Plan.
In mathematical terms the ½ dB rule requires that:
10×log10(I + N) ≤ 10×log10N + 0.5
where I denotes the additional interference and N the previous level of external and receiver noise.
In linear terms, then
(I+N)/N ≤ 10(0.5/10), or I ≤ 0.12 ×N,
i.e., maximum I/N ≈ 10%
Thus the additional interference should not increase the noise level at the receiver by more than
~10%. Below 30 MHz the noise figure of an HF receiver is lower than external noise for HF
systems and external noise is the dominant limiting factor in HF radiocommunications. A simple
I/N ratio of –10 dB could therefore serve as a common preferred protection ratio for many
broadcasting, mobile and fixed HF systems against other equal status radiocommunication systems.
However, while this formulation is widely accepted where the potentially interfering stations are in
the same or an equal status radiocommunication service it would not be appropriate where the
potential interferers are in a lower status service or do not have a corresponding allocation in any
radiocommunication service. In this case the number, location and characteristics of the potential
interferers are not known in advance (as would be the case with a new or modified broadcasting
station in a Plan) and their distribution and characteristics may change over time. A lower limit on
sensitivity degradation should therefore be applied, which takes account of the aggregation of the
effects of a multiplicity of interfering sources, as is the case with PLT systems.
For the protection of broadcasting services in general, Recommendations ITU-R BS.1786 and
ITU-R BT.1786 were established in order to cover the circumstance of maintaining the integrity of
terrestrial sound and television broadcasting services when a multitude of low-level interfering
signals from non-broadcasting sources are present. Because these Recommendations were intended
to be as general as possible, no assumptions were made as to the number, type and location of the
sources of potential sources. On the contrary, they are considered to be unidentified, indeed
unidentifiable, with no set locations or distribution.
Recommendations ITU-R BS.1786 and ITU-R BT.1786 therefore set a conservative limit for the
total excess noise from non-broadcasting sources of 1% relative to the total receiving system noise
of systems operating in the broadcasting service. This contrasts with the figure of ~10% commonly
used to assess the interfering potential of an additional, or modified, broadcasting station in
a terrestrial broadcasting Plan. However, in the case of a broadcasting Plan modification, the
location, characteristics and identity of the potential interfering source are all known precisely. In
effect, Recommendations ITU-R BS.1786 and ITU-R BT.1786 incorporate an aggregation factor of
around 10 dB in order to take account of a multiplicity of unknown interfering sources outside the
broadcasting service. This is consistent with the extensive studies carried out in Study Group 1,
covered by Recommendations ITU-R SM.1757 and Report ITU-R SM.2057, on the impact of ultra
wideband devices (UWB). These studies include protection criteria and aggregation factors
applicable to the broadcasting and other radiocommunication services. In many instances, a factor
of around 10 dB is included in order to cover the impact of interference from multiple UWB
sources. The impact of interference from multiple PLT sources is surely no less harmful than that
from multiple UWB sources. Further support for a figure of around 10 dB comes from
Document 1A/157, which addresses the impact of aggregation interference on the aeronautical
mobile service at HF (see also the further discussion in Annex 1). This notes that a correction factor
of 7 dB is appropriate when going from a density of 50 to 250 PLT interfering sources per km2.
One consequence of Approach 3) is that the leaked emissions of any particular PLT installation
should be well below the ambient noise level. This requirement appears to be supported by the
subjective listening tests contained in Document 1A/117, which has now been incorporated in
Annex 2 to the PLT Report ITU-R SM.2158, as Section A2.5. The same information was also
submitted directly to WP 6A by Japan as Annex 2 to Document 6A/175.
A practical limit of 0.05 dB sensitivity degradation or ~1% increase in I/N from any particular PLT
system could therefore form the basis of a recommendation on leaked emissions from PLT systems.
This would provide a sufficient margin to take account of the cumulative effect of a high density of
Unfortunately, the situation may not be that simple .The above analysis assumes that sources of
noise and interference can be combined in a simple power summation. This would be valid if all the
sources are Gaussian in nature, but there are indications that the noise from PLT systems has a
strong impulsive component. As reported by WP 6A to WP 1A in Document 6A/141, attempts were
made to reformat the data presented in Document 1A/117 using power summation/decomposition in
order to uncover if scaling effects were present and to assess both the individual and combined
effects of PLT noise with ambient noise. However, these efforts failed to uncover any obvious trend
and it was concluded that power summation involving PLT noise does not deliver consistent results.
A possible explanation is that, within the receiver bandwidth, PLT noise appears to be impulsive in
nature rather than Gaussian. As explained in Report ITU-R P.2089, “The analysis of radio noise
data”, simple power summations are not valid for a mixture of Gaussian and impulsive noise
sources. Moreover, the effect of impulsive noise is not scalable with bandwidth. The impulsive
nature of PLT noise may mean that an aggregation factor of 10 dB may not provide sufficient
protection, because aggregation has been assumed to follow simple power summation over
contributing Gaussian noise sources.
Although there are some uncertainties on whether an aggregation factor of 10 dB is sufficient to
take account of the cumulative impact of a dense population of PLT systems, this sensitivity
degradation approach gives more certainty to users of radiocommunication services. Admittedly,
this leaves the users and national administrations responsible for regulating PLT systems with the
problem of how to know in advance, and without recourse to on-site measurements, whether a
particular PLT system or installation will raise the level of leaked emissions above the
recommended limit. That immediately leads back to the same problems, as noted above for
Approaches 1) & 2), in trying to predict what level of leaked emissions will result from a particular
type of PLT system or arrangement of PLT devices.
This assessment indicates that there may be no simple means of ensuring that the wide scale use of
in-house PLT systems will not cause harmful interference to radiocommunication services. An
alternative would be to treat leaked emissions for PLT systems as a source of environmental
pollution. There are many examples of human activities and technology that initially seem harmless
or even beneficial that have been prohibited or rigorously controlled because the cumulative impact
of a myriad of small effects has been found to damage the environment to an unacceptable degree.
Examples include large screen plasma TV displays, which are set to be banned in Europe and
California because of excessive power consumption, the refrigerant gas Freon 12, which was
prohibited by international treaty in 1996, because of the impact on the ozone layer, and the
prohibition on ships from dumping oily waste at sea after cleaning cargo and ballast tanks. A more
directly relevant example is that of light pollution. Astronomers have long complained about
exterior artificial lighting blurring their observations, but now the adverse environmental effects of
badly designed artificial lighting have come to be more generally recognised. Several countries
have introduced legislation to reduce and control light pollution.
However, it is the purpose of ITU-R to ensure that radiocommunication services are accorded the
proper degree of protection, not to design non-radiocommunication systems such as PLT. It would
be more appropriate for national administrations to provide guidance on the provision of PLT
systems to suppliers of PLT devices relevant to national practice in electricity distribution and the
resulting probability of the RF output from PLT modems being converted into leaked emissions.
It is clear that any ITU-R Recommendation should be "in the field of radio communications". Since
the PLT devices and power lines are not the radiocommunication devices, it would not be possible
that an ITU-R Recommendation is applied to non-radiocommunication services and devices.
Because of the intractable nature of trying to assess and limit the amount of RF energy leaked into
adjacent electrical and electronic systems and directly radiated as noise like emissions into the
environment, ITU-R and CISPR have considered mitigation measures to protect the frequency
bands used by some radiocommunication services. A mitigation method that has been developed in
conjunction with the Type 1 modems being assessed in CISPR is to notch out blocks of PLT
carriers in order to keep the RF emissions down to a level consistent with CISPR 22 over certain
The location, depth and width of the notches all have to be sufficient to protect the reception of the
radiocommunication services involved. The depth of the notch should at least match that of the
current CISPR 22 limits, i.e., a modem PSD output level of no more than -94 dBm/Hz. So far
notching has been proposed for the bands used by the amateur and broadcasting services only. In
addition, the latest studies in WP 1A concerning the aeronautical mobile and radio astronomy
services, indicate that the maximum PSD output of Type 1 PLT modems needs to be reduced by
around 50 dB to -105 dBm/Hz by a combination of notching or power management (see
Documents 1A/157 and 1A/171 and the further discussion in Annex 1). The width of the notches is
also a factor because receiving equipment does not have infinite selectivity. Therefore, there has to
be some degree of control of interference in nearby frequencies as well as reducing the interference
within the selected frequency bands.
A further consideration is that the notching process may not be sustainable over mains wiring in the
real world. Because of all the reactive and non-linear elements contained in typical household
electrical and electronic equipment connected to the mains wiring, intermodulation products will be
generated from all the PLT carriers. This will cause the notches to lose definition and fill, thus
comprising their effectiveness.
The application of notching to the amateur, broadcasting, aeronautical mobile and radio astronomy
services would require at least 25% of the HF band to be notched out. This would add to the
complexity of modems and reduce their effective bandwidth and hence their throughput.
A simpler solution for protecting the RF spectrum as a whole would be a general limit on modem
power output of around -105 dBm/Hz.
Developments in ITU-T
The ITU-T also has a role in setting limits on PLT systems that will ensure that multiple systems
can co-exist on common or adjacent electricity distribution wiring. An example is ITU-T
Recommendation K.60 – Emission levels and test methods for wireline telecommunication
networks to minimize electromagnetic disturbance of radio services (2008-02). The scope of this
Recommendation has now been clarified (see Document 6/164) to confirm that its purpose is to
guard against conflict between telecommunication systems, not to imply that the Recommendation
also serves to protect radiocommunication services, with the following text:
“The purpose of this Recommendation is to guide administrations when considering
complaints of interference between telecommunication systems and is not intended to set
compliance requirements or recommendations for protecting the radio spectrum.”
However, ITU-T Recommendation G.9660, “Unified high-speed wire-line based home networking
transceivers – Foundation”, was approved by ITU-T Study Group 15 in October 2009 has gone
considerably further in recommending standards for PLT. The limits proposed do take account of
the possibility of radiated interference, but only in respect of the amateur bands, for which notching
is recommended. This Recommendation has set maximum PSD output levels for PLT modems at
level of -50 dBm/Hz, which is the same level as the modems acknowledged to be disruptive by
WP1A in connection with Document 1A/200 from WP7D, and are indeed higher than the maximum
PSD levels being considered in CISPR (-55 dBm/Hz for Type 1 and -75 dBm/Hz for Type 2)
This level is also higher than the -55 dBm/Hz PSD level assumed in Document 1A/157, which
assessed the interference potential of PLT systems to the aeronautical mobile service and concluded
that PLT modem output would need to be reduced by 49 to 65 dB, depending on frequency, in order
to maintain the desired receiver sensitivity or by 45 to 55 dB in order to avoid increasing the noise
floor by more than 0.5 dB.
In addition, this Recommendation allows for a high band version of PLT, which will operate from
80~200 MHz. The maximum modem PSD output in the high band version is -80 dBm/Hz at around
100 MHz. This configuration has the potential to interfere with several important
radiocommunication services operating in the VHF bands, including Band II broadcasting, ILS,
GBAS and VOR systems in the aeronautical radionavigation services and amplitude modulated
VHF communications in the aeronautical mobile service.
It is a rather serious matter that ITU-T SG 15 has produced Recommendation G.9660 apparently
without liaison with ITU-R or CISPR. It might also be questioned whether setting standards for
in-house PLT systems in this way fall within the remit of ITU-T. SG 15 is responsible in ITU-T for
the development of standards on optical transport networks and access network infrastructures,
systems, equipment, optical fibres and cables, and their related installation, maintenance, test,
instrumentation and measurement techniques, and control plane technologies to enable the
- 10 -
evolution toward intelligent transport networks. This encompasses the development of related
standards for the customer premises, access, metropolitan and long haul sections of communication
ITU-T Recommendations are intended as a means of setting standards that define how
telecommunication networks operate and interwork. This does not obviously cover matters of
consumer choice for distributing data from international or national electronic communications
services around homes or businesses. Setting standards for networks delivering electronic
communications services to the public and business is certainly a necessity, especially where
international connectivity is concerned. However, the rationale for ITU-T SG15 for developing a
Recommendation covering home networking therefore seems tenuous at best.
Role of National Regulation
Several examples of national practice for the regulation of PLT systems have been reported to
WP 1A. The intention in WP 1A is to publish this material, either as a new ITU-R report or as an
Appendix to a new Recommendation on PLT systems. This discussion is summarised in the WP 1A
Chairman’s Report (see Annex 1 to Document 1A/207) and the material itself is available in
Annex 1 to Document 1A/135.
It is therefore worth examining the role of national authorities and the conflicting pressures on them
to facilitate broadband access using PLT while observing their obligations to ensure the continued
availability of the RF spectrum and guard against harmful interference, as set out in Nos. 78,197,
198 & 199 of the ITU Constitution.
NB: CS 199 reads:
3 Further, the Member States recognize the necessity of taking all practicable steps to
prevent the operation of electrical apparatus and installations of all kinds from causing
harmful interference to the radio services or communications mentioned in No. 197 above.
This aspect of the debate on PLT in both WP 6A and WP 1A has shown that national regulatory
authorities often find difficulty in achieving a balance between the demands of ensuring protection
of the RF spectrum and satisfying the demand for providing broadband internet connectivity. In
many countries the regulatory authority that manages the radio spectrum is not the same as the
authority responsible for electronic communications services such as broadband internet
Often the national regulatory authority for electronic communications services has statutory duties
in relation to the provision of network access and service interoperability. In particular, such duties
as securing efficiency and sustainable competition in the markets for electronic communications
networks, electronic communications services and associated facilities, with the aim of providing
the maximum benefit for the persons who are customers of communications providers and of
persons who make such facilities available. In this framework, the national regulatory authority is
normally obliged to act in a technology and application neutral way as regards the means used to
deliver electronic communications services.
In the case of access PLT, such duties require the national authority/ies involved to balance the
provision of broadband with the need to protect radiocommunication services. However, the scope
for interference to radiocommunication services, notably the broadcasting service is greater in the
long term with in-house PLT systems. Given the bandwidth advantages and cheapness of fibre-optic
systems, access PLT is unlikely to be more than a stop-gap solution for broadband delivery.
Speaking to the February 1982 edition of International Fiber Optics and Communications,
- 11 -
Dr. Charles Kao2, the moving force in fibre-optics for over 40 years said that “The economic
advantages of fiber systems is so overwhelming that the telephone industry will move into the
future using fiber systems in almost everything”.
In the case of in-house PLT, it is less likely that national regulatory authorities have a statutory duty
to facilitate deployment. However, this may have been overlooked as a result of earlier involvement
with access PLT, where there could be an interest in securing competition in the delivery of
electronic communications services to consumers. In-house PLT systems only distribute data from
an outside network around the home or a business, and may be viewed instead as an issue of
consumer choice. Their use on private premises is a discretionary action by the consumer that does
not impact on the delivery of the service to the home, or the choice in service providers, or their
technology. The use of in-house PLT systems does not therefore obviously qualify as part of a
network or service that would require national regulatory authorities to facilitate the use of
discretionary consumer equipment at the expense of compromising the availability of the RF
This document has reviewed recent development in ITU directed towards assessing the impact of
PLT systems and considered how to ensure that radiocommunication services do not suffer harmful
interference. The main conclusion has to be that there is no satisfactory way of assessing the impact
of in-house PLT systems on the RF environment or ensuring the protection of individual
radiocommunication services. This is because the basic design and use of in-house PLT systems
leads to a harmful level of unavoidable leakage of RF energy into adjacent electrical and electronic
systems and direct radiation of noise like emissions into the environment. This situation is
fundamentally incompatible with maintaining the availability of the RF spectrum on the frequencies
used by in-house PLT systems.
The immediate task for WP 6A is to determine how to proceed and, in particular, on whether to
develop its own Recommendation on the impact of PLT emissions. The question is all the more
pressing now that there is some doubt as to whether WP 1A will produce an overarching
Recommendation on PLT.
If WP 6A will continue to develop a Recommendation then the next step is to decide what it should
contain. The analysis of options for controlling or measuring the disturbance caused by PLT
systems suggests that the best course of action is to set out a simple list of protection requirements
for the broadcasting service with references to existing relevant Recommendations. This can then be
used by the relevant national authorities, and organisations such as CISPR and ETSI, in order to
determine how to guard against the spread of interference from PLT systems and to avoid raising
the RF noise floor around the world.
Until now, the prime consideration in WP 6A has been the protection of the broadcasting service in
the HF bands. In view of the uncertainty surrounding the mechanisms of how and in what
proportion RF energy from PLT systems is converted into leaked emissions at HF, the view of the
BBC is that any eventual Recommendations on protecting radiocommunication services from
leaked PLT emissions should be based on the impact of radiated emissions rather than conducted
emissions and use hard limits on sensitivity degradation, or equivalent formulations in terms of I/N
ratio. The intent would be similar to anti-pollution measures adopted elsewhere such as the oil
pollution at sea regulations made by IMO. Methods of blocking conducted interference would,
however, be useful include as reference material, if feasible and effective.
2 Dr Kao has been awarded the 2009 Nobel Prize in Physics for this work on fibre-optic communications.
- 12 -
The introduction of ITU-T Recommendation G.9660 will, however, extend the potential spread of
PLT interference to higher frequency bands. WP 6A should therefore be prepared to extend its
consideration of the impact of PLT interference to Band II and III, where analogue and digital
sound and television broadcasting will also be put at risk.
- 13 -
Development of Report ITU-R SM.2158 on PLT in WP 1A
In respect of PLT matters the priority for WP 1A during its meeting from 16 to 22 September 2009
was to make further progress on its preliminary draft new Report, “Impact of power line
telecommunication systems on radiocommunication systems operating in the LF, MF, HF and VHF
bands below 80 MHz”, with a view to presenting it to Study Group 1 for approval in 2010 at the
latest. In the event, there was sufficient progress to send it as a draft Report to the September 2009
meeting of SG 1 (see Document 1/67). The PLT Report was approved as Report ITU-R SM.2158,
albeit with an objection from one administration, Japan, which considered the approval to be
premature. All acknowledge that some parts of the PLT Report are in need of improvement at the
forthcoming meetings of WP 1A. However, the prevailing view was that the importance of the PLT
issue requires that the studies conducted to date should be made available to wider audience as soon
a possible, even though further work remains.
The major developments of the text of the PLT Report were in response to the following
Document 1A/200 (copied to WP 6A as Document 6A/211), WP 7D provided WP 1A with a
revised version of a report on experiments and measurements carried out at Mt Akagi in 2002
on leaked emissions from PLT systems. This was in response to a liaison statement from
WP 1A (Document 7D/85, copied to WP 6A as Document 6A/140), which explained why the
previous version of the study had been removed from the draft PLT Report.
The main reason given by WP1A for removing the earlier version of this study from the PLT
Report was that the tests dated from 2002 and the modems tested were no longer permitted in
Japan because they had been recognised as being “capable of causing severe interference”.
In view of this, WP 6A provided a liaison statement to WP 7D (Document 7D/97), with a copy
to WP 1A (Document 1A/144), which noted the study was still important because it could serve
to define an upper limit for the power output of PLT modems.
WP 7D therefore rewrote the technical study derived from experiments and measurements
undertaken at Mt Akagi in 2002. This more comprehensive technical study took account of
additional material from the original experiments and was included in the PLT report. Part of
the additional information provided is that output level of the modems used was -50 dBm/Hz, in
terms of Power Spectral Density (PSD). The recognition that this power output level is “capable
causing severe interference” must be set against the limits now being considered in CISPR,
which include a “Type 1” PLT modem typically producing a PSD output of -55 dBm/Hz.
The 5 dB reduction in more recent PLT modems appears to be a quite inadequate margin to
guard against interference, especially when the effects of multiple modems are considered. On a
basic power sum basis, assuming Gaussian noise characteristics, two equidistant PLT systems
using modems of -55 dBm/Hz PSD would be equivalent to one of -52 dBm/Hz and 3 would be
equivalent to the acknowledged disruptive level of -50 dBm/Hz. Typical household PLT
systems consist of 2 to 4 modems.
The study also looked at interference caused to radio astronomy observations in the band
322-328.6 MHz. Although, this is outside the study range of 2-80 MHz it should be noted that
the cause of the interference in this higher frequency range was not obvious and that the
interference actually extends over all of the upper adjacent band 328.6-335.4 MHz allocated to
the aeronautical radionavigation service. This band is used by the glide slope component of the
- 14 -
ILS. Broadcasting organisations have devoted considerable effort over the last 25 years on
ensuring compatibility between broadcasting in VHF Band II and ILS/VOR systems in the
adjacent band, 108-117.975 MHz. It is therefore a cause for concern that a possible interference
mechanism to ILS systems exists that has received no attention.
This study now forms Section A3.4 of Report ITU-R SM.2158.
Document 1A/182 from IUCAF examined the distance dependence of the electromagnetic field
radiated from house wiring connected to PLT modems and was added to the PLT Report.
Previous studies included in the Report have used an attenuation rates in the range
20~40 dB/decade of distance for free space propagation in order to assess the effects of PLT
This study used test situation of an isolated house in an isolated low RF noise environment used
for radio astronomy in order to minimise the impact of other extraneous sources of RF noise and
interference. The study concluded that distance dependence cannot rely on single representative
figure, e.g., 20 dB/decade or 40 dB/decade. The free space attenuation of leaked emissions from
in-house PLT systems does not exhibit a simple behaviour and needs to be determined carefully
according to the ratio of the wavelength to the distance
This study also included direct measurements of the common currents on the in-house wiring.
The measurements revealed the presence of HF broadcasting signals which means that power
lines act as a good receiving antenna and, by reciprocity, must also serve as an effective antenna
for radiating the RF signals from PLT modems. This study now forms Section A3.5.3 of
Report ITU-R SM.2158.
Document 1A/174 from IUCAF provided a critical review of the previous material on
modelling RF leakage from in-house wiring. This analysis was accepted as an improvement on
the previous material, which could be viewed as too simplistic. Discussion revealed that the
enhanced analysis was still not sufficiently comprehensive for the purpose, not complicate
enough in fact.
In particular this study addressed the assumptions surrounding the longitudinal conversion loss
(LCL), which is used as determining factor for estimating how much of the differential-mode
current applied to the live and neutral wires is converted to common-mode current with respect
to ground, and in turn how much RF energy is radiated .The study finds that the LCL measured
at the outlet overestimates the effective conversion loss by the amount of the common-mode
loss between the outlet and the remote unbalanced element.
The outlet LCL cannot therefore be used as an effective measure of the imbalance of the power
line and, in turn, as the conversion loss from the differential-mode current to the common-mode
current generated in the power line network. These findings cast doubt on the methodology
being considered in CISPR for “Type 2” PLT systems, which is based on limits to be applied at
the PLT system outlet. In brief, the measured or estimated conditions at the PLT system outlets
cannot be treated as giving a valid estimate of the leaked RF emissions from the house wiring as
This study also investigated coupling from in-house PLT systems to the external service lines
outside the house. Depending on the design of the in-house distribution panels and meter and
the configuration of exterior service line, in particular the grounding arrangements the common-
mode currents converted and transferred to the exterior service lines from the differential-mode
PLT signals may be larger or smaller than the in-house common-mode currents by a
considerable amount. Depending on circumstances the PLT emissions radiated from outside
service lines could be some 30 dB higher or lower than those inside the house.
- 15 -
This study was incorporated into Section 2 of Report ITU-R SM.2158.
Documents 1A/196 from Japan and 1A/171 from IUCAF both used the same methodology to
address the cumulative effect of widescale deployment of in-house PLT systems as a result of
the aggregation of a myriad of contributions propagated by skywave via the ionosphere. These
contributions were combined into a new composite section of the PLT Report. The contribution
from IUCAF examined the particular impact on radio astronomy observations. The conclusion
was that radio astronomy observations could only be safeguarded by:
1) implementation of fixed notch filters to the radio astronomy bands, i.e., a PLT system do
not use the frequency bands allocated to the radio astronomy service in the HF band; or
2) reduction of the radiation power from a single PLT system by more than 50 dB.
The study from Japan was also reported to WP 6A as Document 6A/176. Several delegations
pointed out that the skywave contributions between areas in Japan were at high elevation angles
and that the cumulative field would be lower than in other cases, particularly where
contributions at greater distance and at oblique angle are concerned. The conclusions in respect
of the protection requirements for radio astronomical observations are, if anything, not stringent
These two studies now form Sections A2.4.3 and 2.4.4 of Report ITU-R SM.2158.
Document 1A/186 from the USA (NTIA) addressed another example of assessing the
cumulative impact of PLT interference by skywave propagation, though for access PLT
systems. The study showed that a large number of access PLT systems using overhead power
lines will raise the noise floor. However, the number of PLT access systems is likely to be
limited as this method of broadband internet distribution is not competitive with fibre-optic
distribution in terms of bandwidth. This study now forms Section A2.4.2 of Report ITU-R
Document 1A/117 submitted to the previous WP 1A meeting contained results of subjective
listening tests carried out by Japan and were included in an earlier draft of the PLT Report.
These studies had then been the subject of a liaison statement from WP 1A to WP 6A (see
Document 6A/141). The same study was also submitted directly to WP 6A by Japan as Annex 2
to Document 6A/175. The reply liaison statement from WP 6A to WP 1A (Document 1A/142)
made a number of comments on these listening tests and, in response the administration of
Japan has promised an improved presentation of the study (see the reply liaison statement given
as Document 6A/215). The current version forms Section A2.5 of Report ITU-R SM.2158.
Document 1A/157 from Germany contained a study on the cumulative impact of PLT systems
on HF communications in the aeronautical mobile service and a proposal for some definitive
advice to CISPR Subcommittee I (CISPR/I) in respect of how to define limits on leaked
emissions from PLT systems. The proposal to liaise advice to CISPR/I on limits and mitigation
measures such as notching out the HF aeronautical mobile bands was considered premature and
a suitably edited version of the study now forms as Section A2.6 of Report ITU-R SM.2158.
This study demonstrates yet again that leaked emissions from PLT systems can exceed the
maximum permissible interfering field strength for the airborne receiver, as measured in the
laboratory; and raise the noise floor to an unacceptable level above that found during measuring
flights. The adverse impact also scales up as the density of PLT deployment increases. For
250 PLT devices per km2 (equivalent to between 125 and 62 households) the PLT output would
need to ne reduced by 49 to 65 dB, depending on frequency, in order to maintain the desired
receiver sensitivity or 45 to 55 dB to avoid increasing the noise floor by more than 0.5 dB.
- 16 -
The corresponding figures for 50 PLT devices per km2 (equivalent to between 25 and 12
households) are 7 dB less, i.e., 42 to 58 dB in respect of receiver sensitivity or 38 to 48 dB in
respect of noise floor increase. The reduction in modem PSD would have to be achieved by a
combination of power management and notching.
WP 1A has provided a reply liaison statement to WP 5B on progress on PLT work
(Document 5B/322), which will provide an opportunity for WP 5B to consider the sections of
the PLT Report relating to the aeronautical mobile service.