SE24 Meeting M80
Mainz, 8 – 10 December 2014
M80_33R0_SE24
Date issued: 07 12 2014
Source:
WI42-2 DG chairman
Status:
For consideration
Subject:
WI42-2: working responses to doc. M80_24R0
Password protected:
yes
No
x
Summary: The in line provided marked texts below are responses to comments made
by doc M80_24R0 (ANFR) on WI42-2 working draft ECC Report; these are evidenced
by track changes tool.
While welcoming ANFR contribution, as same as any other ones, the WI42-2 DG
chairman, considering in one hand the only-one SE24#80 meeting session for WI42-2
and in the other hand further complex inputs to be considered. In the spirit of good will
he therefore provides a facilitator response document herewith. This also seeking a
quick processing.
A number of ANFR comments/proposals are due to misunderstanding and response
explanations are provided accordingly. Other text proposals showed fine. They are
therefore marked OK and would be included in the next version of the draft ECC Report.
Proposal: For SE24#80 consideration
Background: M80_24R0_SE24
SE24 Meeting M80
Mainz, 8 – 10 December 2014
M80_24R0_SE24
Date issued: 05 12 2014
Source:
France
Status:
For consideration
Subject:
WI42: comment to the working document
Password protected:
yes
No
x
Summary: Comments are proposed about the current version of the draft report on WI
42.
Proposal:
For consideration
Background: WI 42.
Introduction
In this documents comments are proposed about the current version of the draft report
on WI 42.
Comment 1 (section 4.2 of the Report)
Section 4.2 deals with the possibility of introducing 25 mW SRD applications in the
band 862-863 MHz. The results of the preliminary simulations indicate that:
 The simulation of interference from LTE UE into SRD indicates non negligible
probability of interference. This is not surprising, in the light of the results of
ECC Report 207.
 One the other side, it shown by preliminary simulations that the impact of SRD
on LTE is tolerable.
 The interference into audio applications in the band 863-865 is also tolerable.
The draft report concludes (page 37):
By looking at all the above results it may be concluded that adjacent band
interference from LTE Ues to SRDs deployed in 862-863 MHz would result
in highly challenging co-existence in Scenario 1...
The comment is that, while it is true that an emitting UE will likely interfere an SRD in
the same room, one can imagine that specific SRD applications that do not need a low
latency and a quasi-real time communications , can wait for a time window when the
interference from the LTE UE is not present. This capacity can be implemented by a
LBT mechanism, possibly complemented by a MAC protocol that involves handshaking
functionalities.
It is therefore proposed to modify the draft report in the following way:
The sentence:
By looking at all the above results it may be concluded that adjacent
band interference from LTE UEs to SRDs deployed in 862-863 MHz
would result in highly challenging co-existence in Scenario 1...
Becomes:
Considering all the above results, it follows that the band 862-823 MHz
will constitute a challenging environment for SRD, because of potential
interference from LTE UEs in the adjacent band. Simulations indicate
that existing SRDs would suffer from a significant level of interference. It
is expected that the design of new SRDs applications for the band 862863 MHz will have to take into account the relatively high level of
interference from LTE UEs.
Response: ok
Comment 2. Section 5.2, of the Report
The paragraph considers the impact of the introduction of 500 mW SRDs in the band
862-863 MHz considering as the potential victim the broadband SRD terminal (Tx
power 20 dBm, bandwidth 1000 kHz).
The main comment is that the simulation done in SEAMCAT is mainly an assessment
of the probability of ‘collision’ of the two systems in accessing the spectrum. It does not
give results about the probability of successfully completing the communication within
given QoS metrics. Clearly, from the point of view of spectrum management, this
second case should be considered.
The ideal way would be to simulate spectrum access protocols (otherwise referred to, in
the context of SE-24) as time domain analysis, for instance via MATLAB simulations.
Meanwhile, it is necessary to warn the reader that SAMCAT simulations do not take
into account spectrum access techniques, like CSMA that will be used by broadband
SRDs for the IoT.
It is proposed to add the following sentence to the conclusions of section 5.2 of the draft
report:
Response: OK, but I guess it is meant here to say “do not take” in the first sentence
The simulations performed in the report do take into account the
impact on the possibility to share spectrum (i.e. operate with a
given QoS) of spectrum access techniques and communication
protocols.
Comment 3. Section 5.2, of the Report
The paragraph considers the impact of the introduction of 500 mW SRDs in the band
862-863 MHz considering as the potential victim the broadband SRD terminal (Tx
power 20 dBm, bandwidth 1000 kHz).
The simulation is certainly pertinent and interesting and basically it serves the purpose
of answering the question “can high power (500 mW) SRD and broadband SRDs share
the “862-863 MHz” frequency slot?
It has to be noted, however, that the initial question was different. The rationale was to
check whether the 500 mW SRD could be introduced in the band 862-863 MHz at all,
i.e. in the following conditions:
Response: the point of whether band 862-863 MHz could be shared at all was
considered in section 4.2 of the draft Report. So the context of this Comment 3 is
misplaced and is apparently a result of misunderstanding of the objectives of Section 5.2
(and these different sections are structured according to progression of questions in the
TO DO list).



under the assumption that the band 862-863 MHz is empty from other SRD
applications (or, in the worst case, used by other low power, low DC
applications)
checking, and this is the most relevant point, whether the high power of the SRD
(up to 500 mW) can help withstanding the interference from LTE UE OOB.
Checking the impact of OOB emissions of high power (500 mW) SRD into
adjacent bands, namely LTE up-link below 862 MHz and audio applications in
the band 863-865 MHz
In order to investigate the issues described above, the following simulations need yet to
be performed:
Response: not withstanding the wrong context of this comment as mentioned above, the
below proposals may be considered as follows…
 Impact of LTE EU OOB emissions into 500 mW SRDs. This simulation is
totally missing in the report, but it does not seem that it can be skipped – This
case is similar to the scenario that was studied in Section 4.2, just with different
representative victim SRD and on the assumption that the conclusions would be
generally applicable to other SRDs as well. In that respect, the France’s very
own proposal in Comment 1 at the beginning of this document is addressing this
very point and the text proposed by France could be equally applicable to the
case of 500 mW SRD. Nevertheless, if there is a wish to check additionally this
specific case, surely it could be done. So far nobody focused/requested on this
specific case, hence it was not included on presumption that what is in 4.2 now
is enough.
 Impact of 500 mW SRD into LTE UL. A preliminary version of this simulation
has been provided by ERICSSON. The SE-24 still need to review the

assumption and results, but definitely the Report needs to cover this case – sure,
this is already being addressed
Impact of high power SRDs (500 mW) into SRDs operating in the band 863-865
MHz. – this is studied in Section 5.3
Comment 4. Section 5.2, of the Report
The following files were provided for the simulation of the high power (500 mW)
applications in the band 862-863 MHz:
NewSRD in 862-863
NewSRD in 862-863
MediumDC.sws
NewSRD in 862-863
HighDC.sws
NewSRD in 862-863
MediumDC.sws
MHz - ILK 500mW SRD VLK BB SRD - HighDens-HighDC.sws
MHz - ILK 500mW SRD VLK BB SRD - HighDensMHz - ILK 500mW SRD VLK BB SRD - MediumDensMHz - ILK 500mW SRD VLK BB SRD - MediumDens-
By inspecting them, one finds that the power control is not set in the simulations (figure
below gives a screen shot from one of the files):
Figure 1. Screenshot from SEAMCAT files for the simulation of HP (500 mW) SRDs in the band
862-863 MHz)
It is necessary to repeat the simulations by inserting the power control as well and
compare the results. Should APC prove a necessary condition to assure compatibility,
this needs to be highlighted in the report and possibly inserted as a prerequisite in the
regulation. The argument that APC is a costly feature that does not necessarily appeal
manufactures is not a sufficient argument not to do the simulation.
Response: the APC was enabled in the simulations. From the provided screenshot in
Figure 1 it appears that the author was looking at settings for Interferer: RFID (seen in
the left-most column). If choosing there the “500 mW SRD”, then the APC setting
should be visible.
Important note: it needs to be checked with the SEAMCAT Task Group (possibly via a
Liaison Statement) whether by checking the Power Control feature in SEAMACAT this
disenables or overwrite the probability distribution of power specified by the user, so far
used as a surrogate way to simulate the DC and activity factor of the SRDs.
Response: this does not seem necessary because the distribution of power is used by the
program when first initiating a given interferer instance and assigning it transmit power.
Then the EGE engine further checks the power control condition and if it is set, then the
internal path loss on interfering link is calculated and the resulting APC effect is
expressed as additional negative gain applied to the interferer’s transmitter initial power
in further iRSS link budget calculations. This algorithm could be easily observed by
choosing Debug/Info function and inspecting the resulting print out of snapshots.
Comment 5. Section 5.2
The nominal values for the AP TX are (see Table 27 in the report):
 Tw power: 27 dBm (500 mW)
 Antenna gain: 0 dBi
with a resulting overall eirp of 27 dBm.
However, if we open the SEAMCAT files (Figure 2) the following values are used as
input:
 Tw power: 20 dBm (500 mW)
 Antenna gain: 8 dBi
with an overall e.i.r.p. of 28 dBm.
Figure 2. HP SRD in the band 862-863 MHz. Power and antenna settings in the SEAMCAT file.
There is, therefore, a 1 dB discrepancy between the table and the SEAMCAT
simulations. It is suggested that this discrepancy is corrected.
Response: again it seems that the author was still inspecting the settings of RFID
interferer which indeed has transmit power of 20 dBm and antenna gain of 8 dBi. So
everything seems to be correct.
In addition, France wants to raise the point that below 1GHz the radiated power levels
are expressed in dBm Effective Radiated Power, not in e.i.r.p. All simulations have
therefore to be corrected with an antenna gain of -2.15dBi instead of 0 dBi.
Response: I would humbly disagree with France on this point. Traditionally, I would
say in the old days, the 2.15 dB correction factor was more frequently encountered as it
would apply to stations using conventional half-wavelength dipole antennas which have
2.15 dB gain and thus 2.15 was the difference between antenna gains expressed in dBd
(dB related to dipole) and dBi (dB related to isotropical). And especially this would be
typical for VHF bands. However these days when antennas are far from halfwavelength, it would seem the manufacturers specify antenna gains explicitly in dBi
values, so any corrections do not seem necessary. So when applying antenna gain
explicitly expressed as dBi to the output power, then the result would seem to be the
e.i.r.p.
Comment 6. Section 5.3
This section deals with the introduction high power SRDs in the band 863-865 MHz.
Again, here the simulation is done without considering APC (power control) in the
SEAMCAT simulations. It is proposed to repeat the simulations with APC and compare
the results, along the same line of reasoning as in Comment 3.
Response: same response – was author sure looking at the right interferer?
Comment 7. Section 5.3
In Table 28 of the draft Report it is indicated that, for a wireless audio link, the
maximum distance between TX and RX is 105 m. This value seems reasonable (it
roughly corresponds to possible size of an arena). It is highlighted that is is possible to
have audio application with distances up to about, for instance 150 m. However if we
look into the following SEAMCAT files:
NewSRD
NewSRD
NewSRD
NewSRD
in
in
in
in
863-865
863-865
863-865
863-865
MHz
MHz
MHz
MHz
-
ILK
ILK
ILK
ILK
500
500
500
500
mW
mW
mW
mW
SRD
SRD
SRD
SRD
VLK
VLK
VLK
VLK
WirelessAudio
WirelessAudio
WirelessAudio
WirelessAudio
-
HighDens-HighDC
HighDens-MediumDC
MediumDens-HighDC
MediumDens-MediumDC
it is found that the coverage distance set in the simulations for wireless audio
applications is 40 m (Figure below).
Figure 3
It is proposed to correct this inconsistency re-running the SEAMCAT simulations with a
distance of 105 m and to calculate also the impact on applications operating with a
distance of 150 m.
Response: the author here is mixing up the operational range of given SRD link and the
impact range between TRX-RX belonging to different SRD families. The snapshot
above shows the setting for operational link distance and the 40 m was for quite some
time now (at least since Report 200 if not before) considered a reasonable assumption
for this parameter. Whereas 105 m impact range is used for setting of maximum
placement radius between VLK-ILK, in the “Interfering Links” tab, sub-tab
“Transmitter to Victim Link Receiver Path”.
Comment 8. Section 5.3
In the simulation, it is supposed that wireless audio devices have a bandwidth of 200
kHz. However, other wireless applications exists with a bandwidth of 300 kHz, 600
MHz and 1.2 GHz. A larger bandwidth means a higher probability of collision with the
interfering service (and, if the power is e.i.r.p. is kept equal regardless of the bandwidth,
a lower SNR ratio). It is therefore sensible to perform simulations covering also these
bandwidths.
Response: it is in general true that the wide variety of SRDs produced by a large number
of small manufacturers have all range of differing parameters. Therefore it was the
purpose of Chapter 2 of the draft report to establish some most typical uses and their
representative parameter sets in order to limit the number of simulations. Moreover, I
may remind that the effect of wider bandwidth is actually analysed by looking at the
wideband SRDs as one of the subjects of this study.
Comment 9. Section 5.4
Section 5.4 deals with the assessment of the interference from 500 mW SRDs into
RFIDs.
The SEAMCAT files are set with APC. The draft report thus concludes that APC is
necessary to ensure coexistence:
Quote from page 56:
“These results demonstrate that 500 mW SRD deployed in the RFID
interrogator channels (plus one unused channel centred at 865.1 MHz)
may be able to co-exist with RFID applications w/o any additional
mitigation measures except the APC and CD limit.”
However, since simulations where not done yet with SRD not employing APC, it is not
possible to tell whether the APC condition is necessary. One may as well find that 500
mW SRD can operate in the band without APC.
Response: in this case the APC is considered as default feature of 500 mW applications
according their initial specifications. So there does not seem to be the point for doing
simulations without APC. But if it would be requested to consider operation of 500 mW
SRDs in given band without APC, then surely it would be easy to provide this
additional simulation case.
Also, since in the high DC scenario the interference probability into RFID is about 1% it
is not appropriate to conclude that low DC is necessary. It actually looks like a higher
DC is quite possible.
Response: this seems like a misunderstanding and probably the text in the draft report is
indeed not explicit enough. What was attempted to be said is that if taking the initial
assumptions of operational parameters of 500 mW: APC plus DC limited up to 10% for
NAPs (and lower for other device types), then the simulations do not show problems in
this scenario. So no additional limits were requested, just the confirmation that
originally suggested limits are OK.
It is also necessary to consider the opposite scenario: RFID versus high power (500
mW) SRDs.
Response: this could be done if felt necessary, so far the studies were focused on impact
on incumbent applications as the ones having “right of way”. So if that condition is
considered cleared and the candidacy of 500 mW SRDs is confirmed by SE24 as viable,
then the other direction of interference could be studied as second phase.
Comment 10. Section 5.4
France wants also to remind SE24 that the initial request in not only to use RFID high
power channel as the bandwidth is too tight. Simulations have also to run on the entire
865-868MHz. It’s obvious that a detection of a reader operating means a band that is
occupied over 200+200+200= 600kHz when including both tag answer bands and the
reader band
Response: this wish contradicts a very strong initial opinion from RFID industry. An
appropriate LPRA statement to that effect is provided in the draft report. Furthermore,
this proposal from France seems to contradict the general logic because outside high
power channels the transmitters are the battery-less RFID tags with very weak
rudimentary transmitters. So it is counter-intuitive to mix those in co-channel scenarios
with 500 mW transmitters unless some mitigation measures are envisaged from the start.
What France says about needing only 600 kHz to operate both for RFID interrogators
and tags I would re-direct to RFID industry colleagues to answer.
Comment 11. Section 5.5
The first sentence is too categorical, considering that the simulations do not take into
account, at least so far, the impact of the spectrum access protocol used for the IoT. The
sentence should be modified in order to be more factual.
A proposal is given here:
Original text:
Considering the results of simulations reported in Section 5.2-5.4, it
appears that introduction of new 500 mW SRDs, as proposed to be used
for M3M/SN type of professional utilities network infrastructure, would in
general have moderate impact on incumbent narrow band applications
and very large negative consequences for future development if
broadband IoT type of applications in the band 862-868 MHz.
Becomes:
Considering the results of simulations reported in Section 5.2-5.4, it
appears that the introduction of new 500 mW SRDs, as proposed to be
used for M3M/SN type of professional utilities network infrastructure,
would in general have moderate impact on incumbent narrow band
applications. It may however have a significant impact on the deployment
of broadband SRD applications for the IoT. Actual coexistence also
depends on the medium access protocols and communications protocols
adopted by the IoT that were not considered in this report.
Response: sure, that would be fine amendment
Comment 12 on section 6.2
Section 6.2 of the report considers the introduction of new broadband applications in the
862-863, with power up to 100 mW.
The main focus is the impact of those applications on high power (500 mW) SRDs (see
Table 31 in the draft report).
This simulation is certainly interesting, but other simulations are even more relevant and
are not considered yet, namely:

The impact of LTE UE OOB into the broadband SRD applications. Considering
the results of ECC Report 207, this simulation cannot be skipped. – well,
frankly, it is hard to add anything in addition to the tons of information already
provided in Report 207. Anyway, the impact from LTE UE OOB into SRDs is
subject of analysis in section 4.



The impact of broadband SRD into LTE UL. A preliminary study in this
direction has been presented by ERICCSON and need to be inserted in the
report, after refinement, if needed. – this is being addressed
Impact of broadband SRDs into the audio services in the band 863-865 MHz. –
this is subject of section 6.3
Comment 13 on section 6.2
In the simulation of impact of SRD applications into 500 mW SRD applications, the
interfering link (nor the victim) are set to use APC (see Figure 4 below).
Figure 4. SEAMCAT for broadband SRDs interfering 500 mW SRDs
The simulation should also be performed applying APC to the 500 mW, to check
sensitivity of results on this parameter.
Response: here again it seems that author is looking at the wrong interferer – RFID. As
regards the mention of APC for victim, actually there are no provisions for simulating
APC on Victim. If that is felt necessary, typically for highly sensitive victims, this could
be implemented by either setting appropriate distribution for dRSS or choosing
“Receive power dynamic range” option.
Comment 14 on section 7
The fourth bullet is not yet supported by the analysis (relevant simulations need yet to
be performed).
All Section 7 is highly tentative…
The fifth bullet reads:
as regards the 500 mW SRD (e.g. M3M or SM networking
infrastructure), its prospects would be very difficult, however it would
seem possible to designate for them 5x200 kHz channels within 865868 MHz....
Should be modified to read instead:
It seems possible to designate for 500 mW SRD applications (e.g.
M3M or SM networking infrastructure) 5x200 kHz channels in the
range 865-868 MHz
Response: personally I would still think the original version gives more nuanced picture
of results and goes some way to infer why it is only 5x200 kHz that are identified as
feasible. But if there is support for the proposed simpler text I would not object.