CEPT
ECC
Doc. STG(10)10
Electronic Communications Committee
STG #21,
WG SE - SEAMCAT Technical Group
ERO, Copenhagen
02 March 2010
Date issued: 17 February 2010
Source:
ECO
Subject:
Implementation of overloading in SEAMCAT
Implementation of overloading in SEAMCAT
1. Introduction
Overloading threshold (Oth) is the maximum interfering signal level expressed in dBm, where close to that level
the receiver loses its ability to discriminate against interfering signals at frequencies differing from that of the
wanted signal.
The interfering signal can so strongly overload the receiver front-end that the receiver becomes blind and thus
unable to receive anything at all. In most cases, receiver overloading happens beyond the first adjacent channels.
Under overloading conditions (I>Oth), the receiver is interfered with by the interfering signal whatever the
wanted signal level.
In the current version of SEAMCAT it's not possible to take O th into account.
Consequently, 2 modifications are needed in SEAMCAT:
1.
2.
algorithm(s)
user interface
So far STG as request TG4, SE42, SE43 some more information, in particular on
-
a definition of “overloading”,
what are the input parameters for such functionality to be considered.
what output information is needed.
2. Definition
TG 4 indicated that ECC Report 138 contains a definition for the overloading as follows (see section 2.4):
“Overloading threshold (Oth) is the maximum interfering signal level expressed in dBm, where close to that
level the receiver loses its ability to discriminate against interfering signals at frequencies differing from that of
the wanted signal.
In some cases the interfering signal may so strongly overload the receiver front-end that the receiver becomes
blind and thus unable to receive anything at all. In most cases, receiver overloading happens beyond the first
adjacent channels. Under overloading conditions (I>Oth), the receiver is interfered with by the interfering signal
whatever the wanted signal level.”
Where I is defined as the interfering signal level.
3. Algorithm
This is the implementation as proposed by TG4 (STG(09)83) with enhancement during the development. The
text highlighted in yellow presents the modification/clarification made.
1
Current SEAMCAT interference calculation for the blocking and the uwanted can be found in Annex 3.
3.1 Calculation of the iRSSoverloading
SEAMCAT calculates the interfering signal resulting from the possible interfering system(s) as follows:
For each events{
If (dRSS > sensitivity)
{
Case 1: single source of interference
iRSSoverloading = Pinterferer + Gr + Ge – Rx_filter - Attpropagation
where:

Pinterferer: P power emitted at the interferer (dBm).
Note: The Pinterferer is subject to power control.

Gr: Antenna gain at the victim receiver (dBi)

Ge: Antenna gain at the emitting interferer (dBi)

Rx_filter: Filtering of the receiver (if any).
Note: For a start the filtering of the receiver is by default 0 dB (similarly to the default
blocking filtering value).

Attpropagation: Propagation path loss between the emitting interferer and the victim receiver
(dB)
Note: iRSS overloading is calculated at the interfering frequency.
Case 2: multiple source of interference
2.1
If the interferers are operating at the same frequency, iRSSoverloading is sum of each of the
iRSSoverloading calculated for each of the source of interference.
2.2
If the interferers are operating at different frequencies, an iRSSoverloading should be calculated
considering all of the sources of interference for each frequency (i.e. the sum of the iRSS overloading
is performed at the specific frequency). Each of the calculated iRSSoverloading should be compared with
the different overloading threshold extracted from the overloading mask depending on the considered
frequency offset.
}
else
{
skip this event.
}
}
2
3.2 ICE Calculation
In the ICE calculation{
Nall events are passing the dRSS criterion (It would be useful to have some information on the number which did
not pass this step if it is not yet available. To check how those not passing this criterion are treated)
For both Case 1 or Case 2, the probability of interference should be calculated in a two steps process:
 Step 1: calculate the probability of interference resulting from the overloading for a given total number
of events where the dRSS > sensitivity.
Among Nall events for which the dRSS criterion is met, Noverloaded are interfered due to the
overloading, the corresponding probability is:
Poverloading = Noverloaded / Nall or similarly Poverloading = 1 - ( Ngood_overloading /Nall)
 Step 2: for Nall - Noverloaded the events which were not interfered in step 1, calculate the probability of
interference resulting from blocking / unwanted emissions (the reference should be the total
number of events where the dRSS > sensitivity).
i.e. for (Nall - Noverloaded) events which passed Step 1, Nblocked are interfered by unwanted or
blocking or by the consideration of both phenomena (applying the current threshold for
unwanted and blocking)
Pblocking = Nblocked / Nall or similarly Pblocking = 1 - ( Ngood_blocking /Nall)
The total probability of interference is the sum of those two probabilities (step 1 and step 2) of interference.
Ptot
= Poverloading + Pblocking
= Noverloaded / Nall + Nblocked / Nall
= (Noverloaded +Nblocked)/ Nall
= 1- ((Nall - Noverloaded +Nblocked)/ Nall
}
Signal type combination
Description
Calculate the probability of interference resulting from the
overloading only for a given total number of events where the dRSS >
sensitivity.
The resulting probability is
Poverloading = 1 - ( Ngood_overloading /Nall)
and Ngood_overloading = Nall - Noverloaded
Overloading only
where:
 Nall is the number of event where dRSS > sensitivity,
 Ngood_overloading is the number of “good” events from
overloading (i.e. not overloaded).
 Noverloaded is the number of event when the receiver is
overloaded.
e.g. For 100 events simulated, dRSS > sensitivity occurred for Nall
(=70 events). Out of Nall, the overloading occurred for Noverloaded (=30
events).
Therefore the probability of interference of overloading only is
Poverloading =1-(40/70).
3
Calculate the probability of interference resulting from the
overloading and from the blocking for a given total number of events
where the dRSS > sensitivity.
The resulting probability is
Poverloading+blocking = 1 - ( (Ngood_overloading+blocking)/Nall )
and Ngood_overloading+blocking = Nall - Noverloaded - Nblocked
Overloading + blocking
Where:
 Nall is the number of event where dRSS > sensitivity,
 Ngood_overloading+blocking is the number of “good” events from the
combined overloading and blocking mechanism.
 Noverloaded is the number of event when the receiver was
overloaded.
 Nblocked is the number of event which were inpaired by
blocking (i.e. receiver blocked). It is obtained from the
comparison with the iRSSblocking and a specific interference
criterion (C/I, I/N etc..)
e.g. For 100 events simulated, dRSS > sensitivity occurred for Nall
(=70 events). Out of Nall, the overloading occurred for Noverloaded (=30
events). The iRSSblocking is then calculated over the (70-30=) 40
remaining events. Number of good events resulting from blocking,
Nblocked, is based on the comparison between the iRSSblocking and a
specific interference criterion over these 40 events. Let say that for
Nblocked = 25 events out of 40 the receiver is blocked.
Therefore the probability of interference of the combined overloading
and blocking is Poverloading+blocking= 1 - ( (40 - 25)/70 )
Overloading + blocking + unwanted
Same principle as above except that the iRSScomp
(=iRSSblocking+iRSSunwanted) is considered in the comparison to a
specific interference criterion
Overloading + unwanted
Same principle as above except that the iRSSunwatnted is compared to a
specific criterion
Blocking only
Unwanted only
Blocking + unwanted
Unchanged to the official version (3.1.47)
Unchanged to the official version (3.1.47)
Unchanged to the official version (3.1.47)
Table 1: Summary of the calculation of the probability of interference resulting from the overloading
Note: overloading is only implemented in the victim receiving interface of “traditional” SEAMCAT, i.e. the
overloading is not implemented where the victim is either CDMA or OFDMA. However, the impact of the
interference from CDMA or OFDMA system should be taken into account when overloading is activated in the
victim dialog box.
Note: overloading is not implemented for translation. When overloading is selected in the ICE window, then the
translation button should be shaded to show that it is not active.
4. Implementation phases
Phase 1:
Traditional (interferer) vs Traditional (victim) will be considered. So that simple test can be run.
Phase 2:
CDMA/OFDMA (interferer) vs Traditional (victim) will be considered.
5. SEAMCAT User interface:
4
5.1 Input parameter
With regard to the possible input parameters to be considered for such a feature, TG4 proposed to consider the
following list:




overloading mask in dBm (power (dBm) versus frequency offset (MHz)) (new input - see figure) to input
the Oth response of the victim receiver (Oth versus frequency offset)  To add under Victim link /Victim
receiver : Overloading Function )
possibly information on the filtering of the receiver (dB) versus frequency offset (MHz) (new input - see
figure)
antenna patterns/gains for the victim and interfering systems (Gr and Ge)
output power of the interfering systems
Figure 1: input interface
5
5.2 Output results
1) Output 1
Figure 2: output interface
Probability of interference (see table 1)
2) Output 2
iRSS overloading vector = matrix (number of event x number of freq.) should be extractable. This is found
not possible. Instead the following will be implemented:
For each events, at each frequency the sum of the iRSSoverloading (iRSSsum_overloading) will be compared to the
overloading treshold Oth(fi) , at that frequency (fi) and the difference, delta (in dB), will be stored.

When delta > 0, then the receiver is overloaded at that fi (i.e. the rest of the frequency do not matter)

When delta < 0, then the receiver is not overloaded at fi
If only one frequency is present, the iRSSsum_overloading is compared to the Oth(fi) and the difference will be stored
for that event.
If more than one frequency is present, the strongest of the delta value will be present to the vector (i.e.
irrespective of the frequency) , such that
For each event j (where dRSS > sensitivity)
{
for Frequency = i to number of total frequencies
{
delta_max(j) = max(delta(fi));
}
}
Output vector = delta x events
6
Annex 1
Example of overloading mask
(extracted from TG4(09)292rev1_ANNEX05_Draft CEPT report deliverable D)
The average of all measured protection ratios and the average of all measured overloading thresholds for UMTS
interference into DVB-T are listed in Table 4 and Table 5 for the interfering signal defined as the average and as
maximum rms power, respectively. The frequency offset is measured between the central frequencies of wanted
and interfering signals.
Averaged DVB-T PR and Oth in the presence of a UMTS BS interfering signal without TPC in a Gaussian
channel environment
DVB-T PR and Oth for 64-QAM 2/3 DVB-T signal
(UMTS BS TPC off)
fi-fw (MHz)
PR (dB)
0
18
6.5
-31
7.51
-33
8.54
-35
11.5
-41
16.5
-41
21.5
-45
26.5
-51
31.5
-57
36.5
-45
41.5
-66
46.5
-63
51.5
-66
56.5
-67
72
-48
Oth (dBm)
NR
-9
-8
-7
-4
-2
-4
-5
-6
-6
-5
-4
-4
-3
-1
NR: Oth is not reached. That is at this frequency offset PR is the predominant criterion. Consequently,
DVB-T receiver is interfered with by the interfering signal due to insufficient C/I (<PR) before
reaching its Oth
Note 1: PR is applicable unless the sum of all interfering signals is above the corresponding O th. If the
interfering signal level is above the corresponding Oth, the receiver is interfered with by the interfering
signal whatever the PR is.
Note 2: At wanted signal level close to receiver sensitivity, noise should be taken into account, e.g. at
sensitivity + 3db, 3 dB should be added to the PR.
Note 3: PR for different system variants and various reception conditions can be obtained using the
correction factors in Table 3. The overloading threshold is independent of system variant and
reception conditions.
Note 4: Treatment of overloading threshold in calculations when assessing interference from UMTS
into DVB-T is presented in Annex D.
Averaged DVB-T PR and Oth in the presence of a UMTS UE interfering signal with TPC in a Gaussian
channel environment; note that under dynamic conditions (TPC on) the interfering signal level was
measured as the rms average power. However, in the following table DVB-T PR and Oth are also defined
relative to the interfering signal maximum rms power
DVB-T PR and Oth for 64-QAM 2/3 DVB-T signal
(UMTS UE TPC on)
1
These values are derived by linear interpolation.
7
fi-fw (MHz)
0
6.5
11.5
16.5
21.5
26.5
31.5
36.5
41.5
46.5
51.5
56.5
72
PR and Oth relative to the interfering
signal rms average power
PR (dB)
Oth (dBm)
30
NR
-5
NR
-22
-14
-25
-13
-28
-12
-33
-12
-42
-13
-32
-17
-53
-13
-51
-13
-54
-13
-55
-13
-36
-12
PR and Oth relative to the interfering
signal maximum rms power
PR (dB)
Oth (dBm)
18
NR
-17
NR
-34
-2
-37
-1
-40
0
-45
0
-54
-1
-44
-5
-65
-1
-63
-1
-66
-1
-67
-1
-48
0
NR: Oth is not reached. That is at this frequency offset PR is the predominant criterion. Consequently,
DVB-T receiver is interfered with by the interfering signal due to insufficient C/I (<PR) before reaching
its Oth.
Note 1: PR is applicable unless the sum of all interfering signals is above the corresponding O th. If the
interfering signal level is above the corresponding Oth, the receiver is interfered with by the interfering
signal whatever the PR is.
Note 2: At wanted signal level close to receiver sensitivity, noise should be taken into account, e.g. at
sensitivity + 3db, 3 dB should be added to the PR.
Note 3: PR for different system variants and various reception conditions can be obtained using the
correction factors in Table 3. The overloading threshold is independent of system variant and reception
conditions.
Note 4: DVB-T PR and Oth relative to the interfering signal maximum rms power is applicable to MCL
calculation when the interfering UMTS UE signal power is kept fixed to its maximum value.
Note 5: Treatment of overloading threshold in calculations when assessing interference from UMTS into
DVB-T is presented in Annex D.
8
Annex 2
Proposal of overloading implementation received from broadcasters
Implementation provided by TDF (STG)
 to take Oth into account in the evaluation of probability of interference, as follows:
if (dRSS>Rx_sens)

if (Ilevel > Oth) 
Rx interfered;

Else 
if (dRSS/iRSS<Protection criterion) 
Rx interfered;



Note that in this proposal the Overloading threshold is assumed (with a negligible error) to be
independent from the useful signal level.
Information from the Media Broadcast as a Description of a potential procedure (TG4(09)306)
The following steps would be needed:
a) define the receiver overloading response (Oth (dBm) in function of frequency)
b) calculate the “headroom” HR which is available for interfering signals, due to overloading threshold Oth and
the sum of all useful signals (Ci). Note: In case of DVB-T, usually there will be more than one DVB-T signal
on-air.
HR = Oth – i (Ci)
c) This result needs to be checked:
 If HR is less or equal to 0, than overloading is reached already by wanted signals, no additional signal
can be tolerated without attenuation.
 If HR is above 0  continue with d)
d) calculate the sum of all interfering signals, within the victim receiver bandwidth, and com-pare this with HR
from b), OL = HR –j (Ij)
 If OL is below zero – interference / overloading occurs, since the sum of all incoming interferences
together with the wanted signals will exceed the overloading threshold.
 If the result is above zero – continue with standard interference treatment
9
Annex 3
Current SEAMCAT interference calculation
In SEAMCAT, the interference is calculated as the ratio between the desired/wanted signal (dRss) and the
interfering signal which is compared to a criteria (C/I, I/N).
iRSS unwanted: The interfering power received by the Victim receiver within its bandwidth is called the iRSS
unwanted (Bandwidth correction factor need to be applied to the calculation of the iRSS unwanted when the Vr
and It have different bandwidth).
iRSSunwanted = Pe+Ge+Gr-L
iRSS blocking: When the receiver captures some unwanted signal because its filtering is not ideal, it is called
blocking. In SEAMCAT the Blocking is provided either in dB (user defined and protection ratio mode) or in
dBm (sensitivity mode). It represents the attenuation of the receiver at a given frequency offset. Then, the iRSS
blocking at the interferer operating frequency may be calculated as follows (Note: The It bandwidth is not
considered in the iRSS blocking calculation):
iRSS blocking (fit) = Pe+Ge+Gr-L-Attenuation (fit)
user defined mode: Attenuation(f) = blockUser defined(f) [dB]
sensitivity mode: Attenuation(f) = blockMax Interf Signal (f) [dBm] – sensvr [dBm] + C/(N+I) [dB]
protection ratio mode: Attenuation (f) = blockProtection Ratio(f) [dB] + C/(N+I) [dB] + 3 dB
SEAMCAT allows to calculate the interference considering the two above phenomena as independent from each
other but also as a result of a combined process.
10
Annex 4
Answer/Questions during the development of the algorithm
Question 1: How to simulate the i (Ci) in SEAMCAT as proposed by MB. Currently only one victim (wanted)
link is simulated and multiple interferer are generated.
ANS1: To implement the wanted signal in SEAMCAT it could be interpreted as "interfering" (for overloading,
only). Therefore “extra” intereferers (with the victim transmitter characteristics) could be added to the “real”
intereferers. Note that there is an option to define the distance between some of the Stations to be considered in
the simulations. Therefore the option WTx – ITx (Wanted Transmiter – Interfering Transmitter distance can be
put equal to “0”) would have to be considered.
Finally TDF SAS and MB agreed with the initial implementation. However; it is considered as a temporary
workaround – partly caused by missing measurement results - as it does not really describe the physical
background. Therefore, it is proposed to add a warning for users within this initial implementation as follows:
“ Warning! If the interferers are operating at the same frequency, iRSSoverloading is sum of each of the
iRSSoverloading calculated for each of the source of interference (single frequency power sum).
Interferers operating at different frequencies are not summed “
For the sake of clarity, the summing method used in the initial implementation proposed by TG4 is called
“single frequency power sum”.
There is a possibility to revisit this issue and propose a “multiple frequency power sum” method to be able to
sum interferers operating at different frequencies.
Question 2: What is the unit to consider for the overloading and is it a scalar or a function?
From MB contribution and TG4 CEPT deliverable D (see Annex 1), the overloading Oth is expressed in dBm
and is a function of frequency separation (fi-fw).
ANS2: function dBm vs MHz
Question 3: Should the overloading be calculated as option 1 or option 2?
Option 1: at the interfering frequency, as it is done with blocking currently (i.e. for one point)
Option 2: should it be the integral over the all frequency response of the overloading mask (i.e. integration over
the ref. bandwidth of the overloading mask)?
11
ANS3: option 1 is prefered by TG4 and detailed algorithm for various interference frequency is provided
Question 4: Should a iRSSoverloading “unwanted” (view from the interferer transmitter) and a iRSSoverloading “blocking”
(view from the victim receiver) be considered in the calculation? Or should we have only one iRSS overloading ?
iRSSoverload = Pe + Ge +Gr –L-Att (filter)
where the Att(filter) is a user defined function (different from the blocking response) (see figures above).
ANS4: iRSSoverloading = Pinterferer + Gr + Ge – filtering of the receiver (if any) - Attpropagation
i.e. iRSS overloading is calculated at the interference frequency
Note: For a start the filtering of the reciver is by default 0 dB
Question 5: From both proposals (TDF, MB), the overloading is not based on the interference criteria (C/I or
I/N etc..) but a direct comparison between Interference level and a Overloading treshold. Remember that in
SEAMCAT, the interference analysis of the blocking +unwanted is possible. How can we extract the
interference probability from?
ANS 5: the following combination that SEAMCAT should produce are as follows
- Overloading only
- Overloading + blocking
- Overloading + blocking + unwanted
- Overloading + unwanted
- Blocking only (as in ver 3.1.45)
- Unwanted only (as in ver 3.1.45)
- Blocking + unwanted (as in ver 3.1.45)
Question 6: After the overloading calculation, is the “Protection criterion” (TDF) or “continue with standard
interference treatment” (MB) only protection ratio mode or can it be any of the three SEAMCAT mode?
ANS 6: there is no restriction on which blocking mode should be used
12