8/13/2019 Adjacent Channel Interference in Um Tsf Dd Mode
1/36
Adjacent Channel Interference in UMTS FDD Mode
Document number: UMT/IRC/INF/10Document issue: V01.02 / ENDocument status: ApprovedDate: 30/07/2001
Internal document - Not to be circulated outside Nortel Networks
Copyright2000 Nortel Networks, All Rights Reserved
Printed in France
NORTEL NETWORKS CONFIDENTIAL:
The information contained in this document is the property of Nortel Networks. Except as specifically authorized inwriting by Nortel Networks, the holder of this document shall keep the information contained herein confidentialand shall protect same in whole or in part from disclosure and dissemination to third parties and use same forevaluation, operation and maintenance purposes only.
The content of this document is provided for information purposes only and is subject to modification. It does notconstitute any representation or warranty from Nortel Networks as to the content or accuracy of the information
contained herein, including but not limited to the suitability and performances of the product or its intendedapplication.
The following are trademarks of Nortel Networks: *NORTEL NETWORKS, the NORTEL NETWORKS corporatelogo, the NORTEL Globemark, UNIFIED NETWORKS. The information in this document is subject to changewithout notice. Nortel Networks assumes no responsibility for errors that might appear in this document.
All other brand and product names are trademarks or registered trademarks of their respective holders.
8/13/2019 Adjacent Channel Interference in Um Tsf Dd Mode
2/36
Adjacent Channel Interference in UMTS FDD mode
UMT/IRC/INF/10 V01.02 / EN Approved 30/07/2001 Page 3/3
PUBLICATION HISTORY
23/07/2001 L .BLANCHARD, R DALGLEISH.
Issue 01.01 / EN, Creation
30/07/2001 L .BLANCHARD.
Issue 01.02 / EN, Approved
Update after remarks and comments.
8/13/2019 Adjacent Channel Interference in Um Tsf Dd Mode
3/36
Adjacent Channel Interference in UMTS FDD mode
UMT/IRC/INF/10 V01.02 / EN Approved 30/07/2001 Page 4/4
CONTENTS
1. INTRODUCTION 6
1.1. Object 6
1.2. Limitations 6
1.3. Audience for this document 6
2. REFERENCE 7
3. ABBREVIATIONS 8
4. STATUS OF THE PROBLEM 9
4.1. Emission synopsis 9
4.2. Effect on the networks 10
5. SOME DEFINITIONS 12
5.1. Spectrum allocation 12
5.2. WCDMA Basics 13
5.3. ACLR 15
5.4. ACS 16
5.5. ACP/ACIR 17
5.6. Blocking characteristics 18
5.7. Minimum coupling loss 20
5.8. Summary of ACS/Blocking 20
6. INTERFERENCE CALCULATIONS: THEORETICAL VIEW 22
6.1. Worst case consideration 226.1.1 UE Macro - BS Macro 246.1.2 UE Macro - BS Micro 25
6.2. Ideal intra operator carrier spacing 26
6.3. Conclusion of case study 28
7. SIMULATIONS 29
7.1. Standards simulations 29
8/13/2019 Adjacent Channel Interference in Um Tsf Dd Mode
4/36
8/13/2019 Adjacent Channel Interference in Um Tsf Dd Mode
5/36
Adjacent Channel Interference in UMTS FDD mode
UMT/IRC/INF/10 V01.02 / EN Approved 30/07/2001 Page 6/6
1. INTRODUCTION
1.1. OBJECT
This initial document describes the mechanism of interference induced on a UMTScarrier by its adjacent carrier.
This paper describes the problem of Adjacent Channel Interference in UMTS FDDcase, whatever type of deployment, macro or micro.
The aim of this document is to show the impact of adjacent channel on capacity andcoverage.
After a brief description on Direct Sequence Spread Spectrum technique allowing toseparate carriers within spectrum, the UMTS spectrum characteristics in emission andin reception are described. Notion of UMTS carrier is introduced and developed.
Then case studies are analysed. Potential interference are evaluated in the followingsituation:
- Between two macro layers in adjacent frequencies.
- Between one micro layer below one macro layer in adjacent channel.
Conclusion and recommendations are deduced.
Finally, to complete this theoretical approach, simulations examples are given andeffects quantified.
As conclusion, deployment strategies are given.
1.2. LIMITATIONS
In the present document, worst case of interference is developed i.e.:
- inter-operator interference, where macro base stations are not collocatedor even intra-operator interference between non collocated micro andmacro BS.
- Between mobile and base station.
Collocation of BS has already been studied in [ 4], and for adjacent channel concerns,documents in standards have indeed rejected this scenario as no problematic i f 5 MHz
carrier spacing is respected. A reduction of this spacing is feasible when collocating.
1.3. AUDIENCE FOR THIS DOCUMENT
RF engineering, UMTS working group
8/13/2019 Adjacent Channel Interference in Um Tsf Dd Mode
6/36
Adjacent Channel Interference in UMTS FDD mode
UMT/IRC/INF/10 V01.02 / EN Approved 30/07/2001 Page 7/7
2. REFERENCE
[ 1 ] 25.101 : UE Radio Transmission and Reception (FDD), 3GGP Release 1999.
[ 2 ] 25.104 : UTRA (BS) FDD Radio transmission and Reception, 3GPP Release 1999.
[ 3 ] 25.942: RF System Scenarios, 3GPP Release 1999.
[ 4 ] Co-Location Phenomena with UMTS Introduction ,Moussay L, UMT/IRC/INF/0001.
[ 5 ] Guard Bands Update Report, UKTAG Doc 43/99, 3GPP Submission 1999.
[ 6 ] UMTS & CDMA2000 Voice Capacity Comparison, Dalgleish, R., Nortel 2000.
[ 7 ] 3G Adjacent Carrier Spacing Strategies, Dalgleish, R., Nortel 2001.
8/13/2019 Adjacent Channel Interference in Um Tsf Dd Mode
7/36
Adjacent Channel Interference in UMTS FDD mode
UMT/IRC/INF/10 V01.02 / EN Approved 30/07/2001 Page 8/8
3. ABBREVIATIONS
ACP: Adjacent Channel Protection
ACIR: Adjacent Channel Interference Ratio.
ACLR: Adjacent Channel Leakage Ratio
ACS: Adjacent Channel Selectivity
BS: Base station
CAC: Call Admission Control
DL: Downlink
FDD: Frequency Division Duplex
MCL: Minimum Coupling Loss
PA: Power AmplifierRRC: Root raised Cosine
RRM: Radio Resource Management
TDD: Time Division Duplex
UE: User equipment
UL: Uplink
8/13/2019 Adjacent Channel Interference in Um Tsf Dd Mode
8/36
Adjacent Channel Interference in UMTS FDD mode
UMT/IRC/INF/10 V01.02 / EN Approved 30/07/2001 Page 9/9
4. STATUS OF THE PROBLEM
UMTS is an interference limited system. UMTS network is a trade-off between
capacity, coverage and then total interference. Additional interference raised fromadjacent channel can impact the capacity and coverage of BS.
These problems mainly arise when two operators are serving the same area onadjacent channel or on different layers from one operator (microcell and macrocell).
The main problem occurs in uplink when a mobile far from its mother BS but near acompetitor BS transmits at full power on an adjacent channel. If the mobile leakagepower (see Figure 1) in the competitor carrier is too high, noise observed by the BSwill increase and then the cell size will be reduced (cell breathing, see Figure 2)creating outage at the cell edge of this competitor BS.
Outage can also occur in downlink at the bottom of the competitor BS. This BStransmits too high power in the adjacent carrier, the mobile sees high interference and
then drops its link with its mother BS (see Figure 3).
To avoid such scenario, the transmit power and received power in adjacent channelshould be as low as possible, this is called the Adjacent Channel Protection (ACP).
The value of ACP is linked to the hardware of BS and UE: they should be able toreject interference coming from adjacent carrier to avoid blocking and to keep theuseful signal in their allocated band when they transmit.
Most often the problem comes from UE equipment. Due to its small dimension, it'sdifficult to built tight PA or f ilter. BS ACP is generally 10-15 dB higher than UE ACP.
These problems of adjacent channel interferences can also occur for an operatorwishing to reduce the carrier spacing of its own license to protect its carriers fromuncoordinated competitor carriers.
4.1. EMISSION SYNOPSIS
Figure 1 shows a schematic emission and reception mask. In blue the usefultransmitted signal on one carrier and in green the adjacent channel. We can see thatthe green carrier can transmit in the band allocated to blue carrier. This parasiteemission is normally controlled by the natural ability of the transmitter to control itsemission in one carrier (quantified thanks to ACLR parameter).
In the case of non-coordinated operators, where the mobile is quite near thecompetitor BS and where the natural pathloss is low, the level of leakage power in theadjacent channel can remain high.
8/13/2019 Adjacent Channel Interference in Um Tsf Dd Mode
9/36
Adjacent Channel Interference in UMTS FDD mode
UMT/IRC/INF/10 V01.02 / EN Approved 30/07/2001 Page 10/10
Associated in Rx-UMTS Channel, UE side
BlockingEmission
Adjacent Channel
Interference
In Tx-UMTS Channel, BS side
2110 MHZ 2170 MHZ
UMTS FDD DL band
Figure 1 : Schematic view of what can happens in Tx and Rx Side
On another hand, the green signal received by the blue equipment should be enoughrejected by the equipment to avoid blocking.
4.2. EFFECT ON THE NETWORKS
The following figures illustrate the impact of adjacent channel for both UL and DL inthe case of uncoordinated networks (i.e. non collocated sites).
In the Uplink
Figure 2 gives a view of the problem than can occur. This figure is valid for both macroto micro and macro to macro.
The mobile far away its mother BS (macro) and at the step of a competitor BS (macroor micro), will transmit at high power in an adjacent channel. If the mobile transmits inadjacent channel with high power, it can induce an increase of interference on theuplink of the competitor network. This noise rise has a direct impact on coverage (cellbreathing) and then the BS competitor coverage can be reduced. Of course, thisinterference can lead to coverage holes if the design of the network doesnt take intoaccount this interference.
Figure 2 : Illustration of the problem of adjacent channel in the Uplink
Possible "dead zone"
8/13/2019 Adjacent Channel Interference in Um Tsf Dd Mode
10/36
Adjacent Channel Interference in UMTS FDD mode
UMT/IRC/INF/10 V01.02 / EN Approved 30/07/2001 Page 11/11
In the Downlink
Figure 3 illustrates the problem that can occur in downlink. A mobile close to a basestation transmitting on an adjacent channel will see its interference increase and thenwill be unable to recover its useful signal. Then a dead zone around the step of the
site will appear where any mobile will be able to work efficiently if they cant handoveron a further carrier or if the Radio resource management doesnt allow to transmitenough power to these mobile to counteract this noise rise.
Figure 3 : Illustration of the problem of adjacent channel in the Downlink
In both case, this problem leads to additive interference and then plays a role on cellcapacity and cell reliability.
Possible "dead zone"
8/13/2019 Adjacent Channel Interference in Um Tsf Dd Mode
11/36
Adjacent Channel Interference in UMTS FDD mode
UMT/IRC/INF/10 V01.02 / EN Approved 30/07/2001 Page 12/12
5. SOME DEFINITIONS
In this part, key parameters concerning adjacent channel are given.
5.1. SPECTRUM ALLOCATION
Figure 4 gives the allocated spectrum for UMTS in Europe. MSS is the spectrumallocated to satellite system.
FDD ULTDDUL/DL
TDD
UL/DL
1900 1920 1980
MSS
ULMSS
DL
2010 2025 2110 2170 2200
FDD DL
Figure 4 : Spectrum allocation
In the following, we focus on adjacent channel for the FDD mode.
The carrier spacing is usually the frequency difference between the central f requencyof 2 carriers. The guard band is the band that is equal to the carrier spacing minus thespectral occupancy of the signal. In the following we wil l see that it is possible to havea negative guard band i.e. that the spectrum of two carrier can overlap. The aim of
simulation will be to define the point where this overlap is too restricting in term ofcapacity.
Figure 5: Definition of carrier spacing and guard band
In Europe, standards do not specify carrier spacing. Generally, 5 MHz is proposedwith an adjustment based on a multiple of 200 kHz.
By reducing carrier spacing in its own licence, an operator can increase the guardband with other operator and then get extra protection against interference.
Carrier 2Carrier 1
Carrier Spacing
Guard Band
8/13/2019 Adjacent Channel Interference in Um Tsf Dd Mode
12/36
Adjacent Channel Interference in UMTS FDD mode
UMT/IRC/INF/10 V01.02 / EN Approved 30/07/2001 Page 13/13
5.2. WCDMA BASICS
The following figure sums up the Principles of Direct sequence Spread Spectrum for
WCDMA.
The NRZ symbols which represent the useful information for one user has a narrowfrequency spectrum with a power X. The CDMA principle is to modulate this signalwith a high datarate signal: the spreading sequence. The spectrum of this signal isspread widely on the frequency spectrum, and its power density is very low.
VTb=Eb
VTc=Ec
Tb
Figure 6 : Direct sequence spreading spectrum
The final signal is then spread over the frequency spectrum (in the base band not inUMTS spectrum yet) its initial power density is divided by Tb/Tc(where Tcand Tbarerespectively the symbol time of a chip and a bit, inversely proportional to the datarate)or what is usually called the processing gain.
T
Im{T}
Re{T}
cos(t)
Complex-valuedchip sequencefrom summing
operations
-sin(t)
Splitreal &imag.
parts
Pulse-shaping
Pulse-
shaping
Figure 7: Pulse shaping between spreading process (and scrambling) and QPSKmodulation
8/13/2019 Adjacent Channel Interference in Um Tsf Dd Mode
13/36
Adjacent Channel Interference in UMTS FDD mode
UMT/IRC/INF/10 V01.02 / EN Approved 30/07/2001 Page 14/14
Figure 8: Pulse shaping principle
The spectrum of the NRZ modulation is made of lobes that cut the frequency axis atmultiples of the bit period.
After spreading, the spectrum of the signal is also made of lobes, but lobes now cutthe frequency axis at multiples of the chips period. Thus, the energy of the signal isspread over a larger band. At that point, around 90% of the energy of the signal is inthe first lobe (in the 5 MHz band).
Unfortunately, there is still a small part of the energy that is outside the first lobe andcreates interference to other UMTS carriers. The goal of the transmit pulse shapingfilter (see Figure 7 and Figure 8) is to remove the energy of the signal outside the 5 (orless) MHz nominal UMTS band.
In order to optimise the power budget, a f ilter with a sharp cut-off frequency has to bechosen. In UMTS, a root-raised cosine (RRC) filter wi th a roll-off at 0.22 will be used.
0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
1
0 0.2 0.4 0.6 0.8 1 1.2 1.4 1.6 1.8 2 2.2 2.4 2.6 2.8 3 3.2 3.4 3.6 3.8 4
Frequency shift relative to the carrier central frequency (MHz)
Relativeam
plitudetomax(linear)
roll off 0 (ideal)
roll off 0.22
(UMTS Rec)
Figure 9 : Recommended Root raised Cosine Filter response for pulse shaping
Figure 9 gives a schematic view of what should be the filter before frequencytransposition in UMTS spectrum and transmission in the air interface. Based on a chip
8/13/2019 Adjacent Channel Interference in Um Tsf Dd Mode
14/36
Adjacent Channel Interference in UMTS FDD mode
UMT/IRC/INF/10 V01.02 / EN Approved 30/07/2001 Page 15/15
rate of 3.84 Mcps, the bandwidth is a bit less than 5 MHz around 4.6 MHz. If we allowoverlap between carriers then 4.4 MHz can be an optimal carrier spacing. This will bestudied further in the document.
After pulse shaping, we consider that the occupied bandwidth is the part of thespectrum containing around 99% of the power of the transmitted signal
The manufacturers should then offer equipments that allow diminishing the effect ofadjacent channel and that are at least compliant to the normalisation.
5.3. ACLR
Adjacent Channel Leakage power Ratio (ACLR) is the ratio of the transmitted power tothe power measured in an adjacent channel (see Figure 10). Both the transmittedpower and the adjacent channel power are measured through a matched filter (RootRaised Cosine and roll-off 0.22) with a noise power bandwidth equal to the chip rate.The requirements shall apply for all configurations of BS (single carrier or multiplecarrier), and for all operating modes foreseen by the manufacturers specification.
Figure 10 : Schematic view of ACLR : transmission chain limits its emission inadjacent channel.
Based on [ 1] and [ 2], the minimum required values for ACLR are given in Table 1.
In the first adjacentcarrier
In the 2ndadjacentcarrier
BS 45 dB 55 dB
UE 33 dB 43 dB
Table 1 : Required ACLR
Note:
Nortel Networks BS are compliant with this requirement and go farther
with an ACLR of 59 dB in the first adjacent carrier and 64 dB for the
second.
Frequency
X
Transmit power
in dBm
X -ACLRAdjacent Channel
ACLRUsefulChannel
Adjacent Channel
8/13/2019 Adjacent Channel Interference in Um Tsf Dd Mode
15/36
Adjacent Channel Interference in UMTS FDD mode
UMT/IRC/INF/10 V01.02 / EN Approved 30/07/2001 Page 16/16
UE ACLR is smaller than the BS ACLR. This is of course due to the size ofthe equipment. Note also that 33 dB is a high value for UE manufacturers;this value is a compromise between operators requirements andmanufacturers.
These requirements represent minimum values of ACLR we should obtain inthe case where the PA is used at its maximum. At full power, PA shows oftenless linearity than at medium power (ACLR should then be better at mediumpower).
5.4. ACS
Adjacent channel selectivity (ACS) is a measure of the receiver ability to receive awanted signal at is assigned channel frequency in the presence of an adjacentchannel signal at a given frequency offset from the central frequency of the assignedchannel.
ACS is the ratio of the receiver filter attenuation on the assigned channel frequency to
the receive filter attenuation on the adjacent channel(s).
ACS is related to the adjacent carrier i.e. 5 MHz around the central frequency of thecarrier in interest. For further carrier, we talk about a blocking level defined after.
Figure 11: Schematic view of ACS : in the reception chain the adjacent frequency are
filtered by a factor corresponding to the ACS
The values given in Normalisation are not really clear. ACS value is given for themobile but not for the BS. BS has to fulfil the following test:
Parameter Level Unit
Data rate 12.2 kbps
Wanted signal -115 dBm
Interfering signal -52 dBmFuw (Modulated) 5 MHz
Table 2 : Test parameters for BS ACS
We interpret this requirement as follow: In Speech service, signal level coming from anadjacent channel should not be higher than 52 dBm in the reception chain.
X- ACS
RX filter
Attenuation in dB
X
Adjacent Channel
ACSUseful
Channel
Adjacent Channel
Frequency
8/13/2019 Adjacent Channel Interference in Um Tsf Dd Mode
16/36
Adjacent Channel Interference in UMTS FDD mode
UMT/IRC/INF/10 V01.02 / EN Approved 30/07/2001 Page 17/17
UE has to fulfil l the same type of test :
Parameter Unit LevelDPCH_Ec dBm/3.84 MHz -103or dBm/3.84 MHz -92.7Ioac (modulated) dBm/3.84 MHz -52Fuw (offset) MHz +5 or -5
Table 3 : Test parameters for UE ACS
Then we will consider that the signal in the adjacent channel should never be higherthan 52 dBm.
Note
These tests aim at being reproducible, so conditions are fixed but notnecessary realistic. In particular, wanted signal can vary a lot according to thelevel of interference and the performance of the RRM.
Sometimes it is assumed that UE and BS have the same ACS, i.e. UE one butsometimes we can see 45 dB for BS in some discussion papers. As BS ACLR isaround 45 dB we can expected that ACS is of the same order.
UE BS
ACS 33 dB 45 dB
Table 4 : ACS Value
5.5. ACP/ACIR
ACP (Adjacent Channel Protection) or ACIR (Adjacent Channel Interference Ratio) forUplink or downlink is a combination of ACS and ACLR.
Standard works agreed on the following formula:
ACSACLR
ACIR
11
1
+
=
Of course, the lower of the two parameters (ACLR and ACS) will give the ACIR toconsider.
Then the 45 dB of BS ACS/ACLR will be hidden by the UE performance. In thefollowing, we will consider each time the figures for ACS and ACLR for bothequipments. Then it can avoid misunderstood.
8/13/2019 Adjacent Channel Interference in Um Tsf Dd Mode
17/36
Adjacent Channel Interference in UMTS FDD mode
UMT/IRC/INF/10 V01.02 / EN Approved 30/07/2001 Page 18/18
In Tx-UMTS-Band
In Rx -UMTS-Band 2170MHZ
1980 MHZ1920 MHZ
2110MHZ
In Tx UMTS Channel
In Rx UMTS
ACLR
ACS
Figure 12 : Synoptic view of ACS and ACLR
With UE and BS characteristics and the formula above, we can get a composite figure for the ACP inthe system as a function of carrier spacing (Figure 13).
-40.00
-35.00
-30.00
2
-25.00
-20.00
-15.00
-10.00
-5.00
0.000 1 3 4 5 6 7 8 9
Carrier Spacing (MHz)
d
B
relativetoCarrier
Figure 13: ACP function of the carrier spacing [ 5]
5.6. BLOCKING CHARACTERISTICS
The blocking characteristics is a measure of the receiver ability to receive a wantedsignal at its assigned channel frequency in the presence of an unwanted interferer onfrequencies other than those of the first adjacent channels. Characteristic on firstadjacent channel is specified through ACS requirements.
Based on 3GPP recommendations [ 1] [ 2], the blocking characteristics are given inthe following table.
What is called In-band is the UL band (1920-1980 MHz) for the BS Rx side and DLband (2110-2170 MHz) for the UE Rx side.
8/13/2019 Adjacent Channel Interference in Um Tsf Dd Mode
18/36
Adjacent Channel Interference in UMTS FDD mode
UMT/IRC/INF/10 V01.02 / EN Approved 30/07/2001 Page 19/19
Center Frequency ofInterfering Signal
InterferingSignalLevel
WantedSignal Level
Minimum Offsetof Interfering
Signal
Type of Interfering Signal
1920 1980 MHz -40 dBm -115 dBm 10 MHz WCDMA signal with onecode
1900 1920 MHz1980 2000 MHz
-40 dBm -115 dBm 10 MHz WCDMA signal with onecode
1 MHz -1900 MHz, and2000 MHz 12750MHz
-15 dBm -115 dBm CW carrier
Table 5 : In-band blocking for BS Rx (at antenna connectors)
In this case, for the second adjacent (10 MHz offset) carrier the signal level should belower than 40 dBm. This is coherent with Table 2.
Parameter Unit Offset OffsetDPCH_Ec dBm/3.84 MHz -114 -114
or dBm/3.84 MHz -103.7 -103.7
Iblocking (modulated) dBm/3.84 MHz -56 (-45) -44(-33)Fuw (offset) MHz +10 or 10 +15 or 15
Table 6: In-band blocking for UE Rx
Note: Compared to Table 3 specifying the first adjacent channel maximum levelthrough ACS, we can notice that required values in Table 6 are more restricting in thesecond adjacent channel than in the first. This is due to the differences in conditions oftests that cannot strictly be compared. In particular the level of wanted signal andinterference signal are 11 dB higher in Table 3. So logically, values given in Table 6could be increased by 11 dB. Then for calculations we will consider values in bracketsin Table 6.
Parameter Unit Band 1 Band 2 Band 3
DPCH_Ec dBm/3.84 MHz -114 -114 -114
or dBm/3.84 MHz -103.7 -103.7 -103.7
Iblocking (CW) dBm -44 (-33) -30 (-19) -15 (-4)Fuw MHz 2050
8/13/2019 Adjacent Channel Interference in Um Tsf Dd Mode
19/36
Adjacent Channel Interference in UMTS FDD mode
UMT/IRC/INF/10 V01.02 / EN Approved 30/07/2001 Page 20/20
5.7. MINIMUM COUPLING LOSS
An important parameter to be defined is minimum coupling loss (MCL), i.e. the
minimum loss due to physical separation between the base station and the UE. Wecan find the same notion in cositing rules document [ 4] where the loss in concern is
the isolation between two antennas.
Minimum Coupling Loss (MCL) is defined as the minimum distance loss includingantenna gain measured between antenna connectors; the following values areassumed for MCL [ 3]:
70 dB for the Macrocellular environment
53 dB for the Microcell environment
These values are the values recommended in the Standard 25.942 [ 3] .
Figure 14 : Illustration of coupling loss (assuming O dB UE antenna gain)
If we assume that the path loss is in f ree space. We have the well-known formula:
Pathloss = 32.5+ 20*log(d) +20 log(f) with d in km and f in MHz.For example in the macro case where the minimum coupling loss is 70 dB, if we havean antenna gain of 9 dB (under the main antenna lobe - with 4 of beam width - wehave not the maximum gain) and 4 dB cable loss, we need a separation distance of 65m which is quite higher than the antenna height difference.
Note that free space loss is very optimistic, there is always building; roof mask effectsand immediate environment shadowing that allow having the required value even ifseparation distance is small.
5.8. SUMMARY OF ACS/BLOCKINGThe two following schema sum up the values given in Normalisation. ACS values arenot given clearly, just a test case. 33 dB of ACS for the UE and 45 dB for the BS willbe assumed.
BS
Coupling Loss= Path LossAntenna gain+ Cable loss
Path Loss
Cable Loss
Antenna Gain
8/13/2019 Adjacent Channel Interference in Um Tsf Dd Mode
20/36
Adjacent Channel Interference in UMTS FDD mode
UMT/IRC/INF/10 V01.02 / EN Approved 30/07/2001 Page 21/21
Figure 15 : Downlink RX maximum level at UE Antenna port to avoid blocking (based on [ 1])
Figure 16 : Uplink RX maximum level at BS Antenna port to avoid blocking (based on [ 2])
1900 1920 1980
-15 dBm
-40 dBm
RXChannel
-52 dBm
ACS (45 dB)
f0+5 2000
Frequency (MHz)
f0f0-5
In RX Band Blocking
Out RX
Band
Blocking
Out RX
Band
Blocking
-19 dBm
-33 dBm
-45 dBm
RXChannel
ACS (33 dB)
-4 dBm
2185f0+10f0+5 f0+15 2230 2255
Frequency (MHz)
f020952050 f0-152025 f0-10 f0-5
-52dBm
In RX Band Blocking
Out RX
Band
Blocking
Out RX
Band
Blocking
8/13/2019 Adjacent Channel Interference in Um Tsf Dd Mode
21/36
Adjacent Channel Interference in UMTS FDD mode
UMT/IRC/INF/10 V01.02 / EN Approved 30/07/2001 Page 22/22
6. INTERFERENCE CALCULATIONS:THEORETICAL VIEW
In this chapter, we give a roughestimation of the interference level due to ACP. In thefirst part, the worst caseis assumed: i.e. a mobile is transmitting on a adjacent carrierat full power (assumed 21 dBm) at the step of a competitor BS (each BS is at the celledge of the competitors one). The acceptable level that avoids blocking in theadjacent channel is evaluated.
In the second part, for a collocated case we will show that problems are lessimportant.
6.1. WORST CASE CONSIDERATIONWe give here the main value to consider for worst case calculation.
Note that this worst case is unlikely to happen (mobile at the step of BS, full powertransmission), and maybe the mobile will hand off on another carrier or will be droppeddue to downlink interference from the competitor BS.
The notation used in the current paragraph will be :
- UE macro for a mobile belonging to a macro site
- BS Micro for a micro BS.
Macro BS and micro BS are not supposed to be collocated as far as we supposed tobe in a worst case.
We will consider 4 cases that go from the additive noise due to adjacent channel, theACS case, the in-band blocking with 10 MHz and 15 MHz offset.
For additive noise we will consider as a threshold the thermal noise plus noise figureto have an idea of maximum noise. We can then assume a noise floor at 103 dBmfor the mobile and 105 dBm for BS (i.e. we consider that the sensitivity for speechservice for example is then around 117 dBm). If we accept an interference level thatdegrades the sensitivity of 0.5 dB then we have an acceptable level of interference at112 dBm for the UE and -114 dBm for BS.
To recap the parameters involved in calculations, see Figure 17 and Table 8:
a: an adjacent channel Tx leakage falls in the Rx band: ACLR of thetransmitter is concerned.
b: the first adjacent channel pollutes the out of Rx band of theadjacent channel: ACS of the Rx is important
c: The second adjacent channel out of band Tx pollutes an out ofRx band: In band blocking at 10 MHz offset is important.
8/13/2019 Adjacent Channel Interference in Um Tsf Dd Mode
22/36
Adjacent Channel Interference in UMTS FDD mode
UMT/IRC/INF/10 V01.02 / EN Approved 30/07/2001 Page 23/23
d: a far carrier pollutes the out of band of the receiver : out of bandblocking is considered.
As minimum coupling loss and requirements (Table 8) are given at the antennaconnectors, on the receiver side we must not take into account the ACS. This oneshould be included in the level to respect.
For cases b, c and d, as we have to respect test conditions, any value of ACP areused. We just supposed that the material is compliant with the tests.
Figure 17: Synoptic view of interference induced by adjacent carrier
Values considered for calculations are sum up in Table 8.
Case Type ofinterference BS UE
a Cochannelinterference
-114 dBm -112dBm
b Inband blockingwith 5 MHz offset
(ACS)
-52 dBm -52 dBm
c Inband blockingwith 10 MHz offset
-40 dBm -45 dBm
d Out of band
blocking
-15 dBm -33 dBm
Table 8 : Blocking characteristics at the air interface
c
b
InterferingCarrier 1
(5 MHz
Victim Carrier
a
Interfering
Carrier 2(10 MHz
Interfering
System
(out of band
d
8/13/2019 Adjacent Channel Interference in Um Tsf Dd Mode
23/36
Adjacent Channel Interference in UMTS FDD mode
UMT/IRC/INF/10 V01.02 / EN Approved 30/07/2001 Page 24/24
6.1.1 UE Macro - BS Macro
In macro-to-macro worst case, as the antenna is placed above the mean rooftop, we
will assume:
- A minimum coupling loss of 70 dB.
- Mobile EIRP of 21 dBm.
- BS output of 43 dBm (20 W).
UE macro to BSMacro
BS Macro to MSmacro
Minimum couplingloss between mobile
and BS
70dB
Equipment fullpower
21 dBm 43 dBm
Case a
ACLR 33 dB 59 dB (NortelProducts)
Interference level -82 dBm -86 dBm
Required value -114 dBm -112 dBm
Extra Isolationneeded
32 dB 26 dB
Case b= -49 dBm -27 dBm
Required value -52 dBm -52 dBm
Extra Isolationneeded
3 dB 25 dB
Case c
Interference level = -49 dBm -27 dBm
Required value -40 dBm -45 dBm
Extra Isolation
needed
no 18 dB
Case d
Interference level = -49 dBm -27 dBm
Required value -15 dBm -33 dBm
Extra Isolationneeded
no 6 dB
Table 9 : Worst case calculations for macro to macro
8/13/2019 Adjacent Channel Interference in Um Tsf Dd Mode
24/36
Adjacent Channel Interference in UMTS FDD mode
UMT/IRC/INF/10 V01.02 / EN Approved 30/07/2001 Page 25/25
In this case, the most severe problems occur for the mobile. Too close to a competitorBS, it will suffer of very high interference and will be blocked. This is mainly due to aweak UE ACS.
BS can suffer also (case a and b) but less severely and if we move away from theworst case the problem will quickly disappears.
From case a, we can see that the level of interference can be very high, giving a veryhigh noise rise in the BS and in the mobile. Further simulations will give more realisticfigures.
All these values are given with particular conditions, in a real network features likepower control will allow avoiding this blocking in the limits of power dedicated to onelink.
6.1.2 UE MACRO - BS MICRO
In this case where usually, micro BS antennas are mounted below the mean roof top,the distance between a macro mobile and the micro BS is very small.
For calculations purpose, we will assume:
- 21 dBm transmit power for macro UE.
- 37 dBm (ie 5 W) EIRP for micro BS.
- 53 dB Minimum coupling loss.
- Same noise floor and same ACP than in the macro case.
UE macro to BSMicro
BS Micro to MSmacro
Minimum couplingloss between mobile
and BS
53 dB
Equipment fullpower
21 dBm 37 dBm
Case a
ACLR 33 dB 59 dB (NortelProducts)
Interference level -65 dBm -75 dBm
Required value -114 dBm -112 dBm
Extra Isolationneeded
49 dB 37 dB
Case b
= -32 dBm -16 dBm
Required value -52 dBm -52 dBm
8/13/2019 Adjacent Channel Interference in Um Tsf Dd Mode
25/36
Adjacent Channel Interference in UMTS FDD mode
UMT/IRC/INF/10 V01.02 / EN Approved 30/07/2001 Page 26/26
Extra Isolationneeded
20 dB 36 dB
Case c
Interference level = -32 dBm -16 dBm
Required value -40 dBm -45 dBm
Extra Isolationneeded
8 dB 29 dB
Case d
Interference level = -32 dBm -16 dBm
Required value -15 dBm -33 dBm
Extra Isolationneeded
no 17 dB
Table 10 : UE Macro BS Micro worst case calculation.
Macro UE risks to be blocked in the vicinity of the micro BS and inversely. Theproblem is more severe for the UE than for the BS. Even if we move away the worstcase, the problem for UE will remain at the bottom of BS. This is mainly due to a weakACS.
These calculations demonstrate a high level of interference so that cell coverage willdecrease and capacity of the network will be affected. We insist that this worst case isunlikely to happen but even if interferences are lower, problem will occur.
Static simulations will show what is the correct ACP to avoid too many problems. As
micro BS is not necessarily coverage limited, maybe the impact is less. Thats why adesensitisation of BS is of ten proposed.
Note:
All The previous calculation were focused on macro UE to BS micro and tomacro to macro, for the micro UE to macro BS, we are in the same case thanmacro to macro.
6.2. IDEAL INTRA OPERATOR CARRIER SPACING
Based on the previous considerations and definitions (5.2), we can see that theoccupied bandwidth by a UMTS emission could be lower than 5 MHz around 4.6 MHz.
So why not reducing carrier spacing to maximise spectrum usage. Basically, carriersfor a same operator are always co-located then there wont be any problem ofadjacent carrier and no need to have too much protection between carriers: thanks topower control,
in uplink the signal from adjacent carrier is received with a power l inked to theservice (i.e. Eb/No and Processing gain). With 2 mobiles and a differentservice on each one, the power difference can be around 11 dB, then an
8/13/2019 Adjacent Channel Interference in Um Tsf Dd Mode
26/36
Adjacent Channel Interference in UMTS FDD mode
UMT/IRC/INF/10 V01.02 / EN Approved 30/07/2001 Page 27/27
ACLR rejection from the mobile side of 25 dB is enough i.e. the interferencesignal received at BS in the worst case will be 14 dB lower than the usefulsignal.
In downlink, lets imagine two carriers equally loaded i.e. same powertransmitted from the same BS. A mobile can receive the useful compositesignal sent by its assigned carrier and the composite signal on another carrier.As the mobile has an ACS higher than 30 dB it will see this second signal asan interference 30 dB lower than its intended composite signal. Then the ratioof the useful signal over this interfering one is at least 30 dB, which is highenough to demodulate the useful signal. The ACLR of the BS transmitter ishigh enough (more than 45 dB) to avoid one carrier interfering on the other.
In this case, even if ACP is quite less than the required value, we can hope that therewould be no problem. To complete this approach, if we reduced further the carrierspacing from 5MHz to 4.4 MHz (Figure 13), ACP is reduced by a factor of 10 dB.
Such a reduction in ACP can induce a potential degradation in term of capacity and/orcoverage. Actually, there will be a compromise between the reduction of intra-carrierperformances and the rejection of the effects of competitors carriers.
For the downlink case, there is incomprehension with the normalisation concerning theACS test. In that case, a mobile close to the BS will see the signal on the other carrierhigher than the required value (see above) but its useful signal will be strong too. Itsimportant to note that for this test, the useful signal is low near the sensitivity level. If abase station transmits 43 dBm on an adjacent channel, the signal received by amobile should never be higher than 52 dBm (ACS Test see [ 1]) which mean that 109dB isolation is required. Such an isolation cannot be achieved at a distance lower than100 m which mean that mobile could not communicate when they are too close to the
BS. But in the collocate case, the signal sent by 2 carriers are not very different andanyway ACP will be large enough to reject the interferer and to avoid problem in thatcase.
In 6.1, we have seen that in worst case (i.e. uncoordinated operators), problems canoccur. The theoretical approach of adjacent carrier interferences leads to therecommendation that the better way to avoid interference from adjacent channel is tocollocate carriers. Of course this is obvious for one operator but for many competitorsoperating in the same area and using adjacent channels, the solution is to collocate asoften as possible their stations.
As concerning, intra-operator carrier spacing we have seen that it is possible toreduce from 5 MHz to 4.6 MHz and maybe less (if we accept little capacity loss).
Simulations must confirm this approach.
Note that this study can apply to carriers belonging to an operator or to two collocatedoperators using adjacent channel.
8/13/2019 Adjacent Channel Interference in Um Tsf Dd Mode
27/36
Adjacent Channel Interference in UMTS FDD mode
UMT/IRC/INF/10 V01.02 / EN Approved 30/07/2001 Page 28/28
6.3. CONCLUSION OF CASE STUDY
We have seen that we could have a blocking problem: a micro BS is blocking a macroUE and inversely if both are working on adjacent channel. Now depending on RRMalgorithm, we can assume that only the mobile will be dropped: experiencing too manyinterference, it will be blocked and then will not transmit. As a consequence, the microBS will not be affected in the uplink.
We also have seen that in the case of 2 competitors with macro layers, blocking of UEis possible at the step of each other site.
The previous calculation for a worst-case scenario show the noise rise due to adjacentchannel interference can be very high. The impact is a coverage reduction (cellbreathing) that can lead to outage if this interference is not taken into account in theinitial network dimensioning. More accurate simulations given in the following part willallow to estimate the impact of interference from adjacent channels on coverage andcapacity.
These worst cases are unlikely to happen and even if some mobile are close enoughto the BS antenna to create problem, with the level of interference in the DL, a handoffon a farther carrier can solve this problem (of course if the other carrier is not tooloaded).
What is the solution?
Collocation and site sharing between operators become a key advantage. Theidea for one operator having its own sites proposed to competition is perhapsshocking, but beside the economic aspects, it can prevent the users to bedropped when approaching a competitor BS
If collocation is impossible, downlink capacity will be degraded (for alloperators). This degradation would be directly dependent on UE
performances.
8/13/2019 Adjacent Channel Interference in Um Tsf Dd Mode
28/36
Adjacent Channel Interference in UMTS FDD mode
UMT/IRC/INF/10 V01.02 / EN Approved 30/07/2001 Page 29/29
7. SIMULATIONS
The worst-cases studied before give very pessimistic figures. Then simulations have
been done to see the impact of adjacent channel on a network where mobiles arerandomly generated and where sites are non uniformly positioned, collocated or at thecell edge of each other.
The results given here below are for the macro-macro case. Furthers simulation willgive results for the macro-micro case.
7.1. STANDARDS SIMULATIONS
The results given here after are those found in [ 3]. They were obtained with differentcases:
- Worst case where BS are the cell edge of each other.
- Intermediate case where BS are located at a half-cell radius shift.
Simulation conditions are sum up in 25.942 [ 3]. The aim of this recommendation wasto be sure that all simulations given by standard participants would be done on thesame basis and it would be easier to make comparison.
Here we give the results for UL and DL for speech and as a function of ACIR. ACIR ismainly influenced by UE performance. Figure 13 gives the relation between carrierspacing and ACIR.
Simulations are performed relatively to the case where there is no problem of adjacentchannel. The network is loaded until 5% of users are dropped.
ACIR (dB)DL UL
25 89,12 % 91,15 %30 95,30 % 97,09 %35 98,21 % 98,98 %40 99,41 % 99,65 %
Table 11 : Simulation results for the intermediate case
ACIR (dB) DL UL
25 86,72 % 87,75 %30 93,84 % 95,81 %35 97,68 % 98,66 %40 99,01 % 99,57 %
Table 12 : Simulation results for the worst case
As expected in the case study, adjacent channel has more impact on the downlinkthan in the Uplink.
We can note than an ACIR of 35 dB, reasonable figure, gives few capacity loss.
And of course the worst case gives less capacity as it has been demonstrated in thecase study.
8/13/2019 Adjacent Channel Interference in Um Tsf Dd Mode
29/36
Adjacent Channel Interference in UMTS FDD mode
UMT/IRC/INF/10 V01.02 / EN Approved 30/07/2001 Page 30/30
These two tables show that the ideal solution to avoid too much capacity loss is toplan the networks in coordination with competitors. Of course this notion of distance islinked to the environment and the service considered. Then we can see how crucial isthe accuracy of a link budget and the parameters chosen.
The capacity losses given above are function of the ACIR. Reducing the ACIR isequivalent to reduce the carrier spacing (see Figure 13). Then we can have an idea ofthe capacity loss if we reduce the carrier spacing.
Reducing the carrier spacing is acceptable in the intermediate case but not for theworst case (more than 13 % capacity loss for an ACIR of 25 dB or 4.7 MHz of carrierspacing).
The next part will gives some figures for the carrier spacing in the case of collocatedoperators or for intraoperator carrier.
7.2. NORTEL SIMULATIONS
Simulations led by Nortel are mainly focused on Speech Service. One difference withregards to previous simulations is the bandwidth used. Standards simulations aredone with the old values of 4.096 Mcps. But the main difference is in site positioningwhere Nortel simulations try to be close to a real network where site spacing will notbe regular. Then the network is more self-interfered and then the effect of additiveadjacent channel interference as less impact than in a regular grid. Moreover, in a
regular grid, capacity tends to be overestimated in comparison with IS-95 network(speech mainly) and then interferences created by an adjacent channel have a bigimpact on capacity.
7.2.1 SIMULATION SCENARIO AND PROCESS
In these simulations, two cases have been studied:
- Collocated case for simulating carrier spacing between 2 operators orfor 2 carriers of an operator.
- Realistic positioning of BS.
The baseline voice simulation process and method of determining capacity and areaavailability are described thoroughly in [ 6] and a sample output is shown in Figure 18.Traffic load is incrementally increased until total availability just falls below 95%.
8/13/2019 Adjacent Channel Interference in Um Tsf Dd Mode
30/36
Adjacent Channel Interference in UMTS FDD mode
UMT/IRC/INF/10 V01.02 / EN Approved 30/07/2001 Page 31/31
Figure 18: Single Network FER Plot Figure 19: Dual Network Checkerboard Plot
.
Figure 19 shows a checkerboard plot of Forward FER with system A users in thecolored bins representing FER operating from the Black sites and system B users inwhite bins operating from the grey sites. (Blue = 3%) Thegrey sites are collocated with the black but use a different antenna grid orientation.The propagation conditions are identical for both networks and inter system
interference can be modelled as an adjustment to the pathloss between system Asites and B bins and between B sites and A bins. This adjustment can be determined
dynamically from predefined TX emission and RX filter characteristic curves.Traffic load is applied to both networks and performance analysed independently foreach network to model the performance/capacity degradation as the frequencyspacing between the networks is decreased. Traffic load on the interfering network isheld constant at the nominal maximum while traffic load on the interfered network isprogressively increased until the 95% availability objective is no longer met.
Sites for independent networks A and B can also be separately located as shown inFigure 20 and Figure 21. System A sites are black in Figure 20 and colored binsrepresent System A Forward FER performance. (System A sites are plotted in grey atthe same locations and system A bins appear white in Figure 21). System B sites areblack and colored bins represent system B FER performance in Figure 21. Both plots
are the result of one simulation run modelling the simultaneous performance of bothnetworks.
The average site spacing is the same for both networks but the grid orientation isdifferent to achieve random inter system spacing between particular site pairs. Thiscauses worse interference than collocated sites but not as bad as having all system Asites on system B cell boundaries.
8/13/2019 Adjacent Channel Interference in Um Tsf Dd Mode
31/36
Adjacent Channel Interference in UMTS FDD mode
UMT/IRC/INF/10 V01.02 / EN Approved 30/07/2001 Page 32/32
Figure 20: Non Colocated System A Figure 21: Non Colocated System B
Figure 21 illustrates areas of maximum degradation (Pink & Grey circled near top left)occur on system B for users at system B cell boundaries lacking a dominant server butvery close to a System A site. This near-far interference effect is quantified in theoutput by a reduced traffic throughput at the 95% area availability threshold.
Traffic load is approximately 1.5 erlang per sector per 200KHz channel resulting inapproximately 20% loading per channel.
UMTS
Hardware Configuration
System Noise B/W & Chip Rate(mhz) 3.84
Nominal Cosited Channel Spacing (MHz) 3.84
RX Noise Figure(dB) 3.3
PA Blocking Power Per Carrier (W) 20
Pilot + Paging Channel Overhead Pwr (W) 2.8
Time Average Bit Rate/User (KB/s) 4.32
Maximum Mobile TX Power (dBm) 21
Channel Model Definitions
Environment Mix Veh/Ped
Forward 1Way Eb/No Target (dB) 10
Forward Code Inorthogonality (ratio) 0.1
Reverse 1Way Eb/No Target (dB) 5.5
Mobile TX Headroom Margin (dB) 1
Power Control Error 0.5dB
RF Performance Characteristics at Load
Reverse Link Area Availability % 98.2
Nominal Carrier/Pa Load (%) 70
Loaded Sectors Per User (average) 1.52
Nominal Traff/Carrier at 95% Area Avail (Erl) 60
Table 13: Technology Specific Configuration Parameters
8/13/2019 Adjacent Channel Interference in Um Tsf Dd Mode
32/36
Adjacent Channel Interference in UMTS FDD mode
UMT/IRC/INF/10 V01.02 / EN Approved 30/07/2001 Page 33/33
7.2.2 TX EMISSION AND RX FILTER MODEL DEFINITION
The TX emission spectra and RX filter responses are modelled as polynomial
approximations to N pole filters according to the following rules. Published TXemission masks are not used as they represent upper bounds for narrow spurious
energy and would over represent the power falling into adjacent channels.
The UMTS TX Spectrum is based on measurements from Prototype BTS hardwareshowing 55dB falloff of TX spectral power density between 1.92 and 2.5 MHz carrieroffset measured in 30KHz resolution bandwidth. The RX filter response was adjustedto achieve 33dB ACLR according to [ 1] UE specification section 6.6.2.1.1. The BTSACLR specification is tighter at 45dB in [ 2] section 6.6.2.2.1 however the moreconservative 33dB is used for both directions. 25.101 also specify that the prototypereceiver passband for ACR testing should be modelled using a raised cosine filter withrolloff coefficient of 0.22.
The UMTS RX filter responses used in this study are more conservative (less sharp)and are similar to a 7-pole filter to just achieve 33dB Adj Chan Rejection at 5MHz (see
Figure 22).
-40-38-36-34-32-30-28-26-24-22-20-18-16-14-12-10
-8-6-4-20
0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 4.5 5.0
Freq (MHz)
dB
U_U ACI
Figure 22: ACP Composite Response used for simulation
7.2.3 ADJACENT INTERFERENCE RESULTS
The simulation process and filter masks described in the last section are used toestablish the relationship between loaded capacity at 95% availability and frequencyoffset between interfering groups of carriers. The dotted curves in Figure 23 representnetworks of collocated sites which avoid the near far interference effect and cantolerate 400 KHz closer carrier spacing for the same capacity loss as the noncollocated cases indicated by solid curves. The initial falloff from 100% is affected bythe shape of the floor of the emission spectra while the region of fast falloff isdominated by the RX filter falloff.
8/13/2019 Adjacent Channel Interference in Um Tsf Dd Mode
33/36
Adjacent Channel Interference in UMTS FDD mode
UMT/IRC/INF/10 V01.02 / EN Approved 30/07/2001 Page 34/34
70
75
80
85
90
95
100
3.4 3.6 3.8 4.0 4.2 4.4 4.6 4.8 5.0
Carrier Spacing(MHz)
%
Um->Um Cl
Um->Um
Figure 23: Relative UMTS Capacity vs Adjacent Interferer Spacing (MHz) for collocated case (doted line)
and uncoordinated case
Then, it is possible to reduce the carrier spacing in the case of collocated network(intra or interoperator) without losing too much capacity.
7.2.4 DEPLOYMENT STRATEGY GUARD BAND CALCULATIONS
Guard band and carrier spacing are directly dependant.
The curves from Figure 23 can be used to define guard bands to manage interferencebetween interfering networks for an optimal balance of spectrum usage andperformance.
UMTS Carrier
Spacing
Guard band
Figure 24:Guard Band Relationships
The guard band width is calculated by taking the inter system carrier spacing fromFigure 24 and Figure 23 and subtracting of the nominalcarrier spacing (specifiedTable 13 i.e. 3.84 MHz for UMTS). This remaining Guard band gap representsunoccupied (wasted) spectrum that would have been used if a continuous block of
8/13/2019 Adjacent Channel Interference in Um Tsf Dd Mode
34/36
Adjacent Channel Interference in UMTS FDD mode
UMT/IRC/INF/10 V01.02 / EN Approved 30/07/2001 Page 35/35
carriers of 1 technology were deployed. Negative values indicate that carriers cantolerate partial overlap with spacing of less than half the sum of the 3dB bandwidth.
With such consideration, this is the result of capacity loss as a function of guard band.
% Capacity Loss 1% 5% 10%
UMTS -> UMTS 960 260 -40
UMTS -> UMTS Col 360 -140 -340
Table 14: Calculated Carrier Spacing (in KHz) vs % Capacity Loss
Table 14 can be used to determine a required guard band size depending on theamount of capacity loss that can be tolerated and whether the networks have co-located sites. Values outside the range of this table can be determined directly fromthe curves in Figure 23.
In uncoordinated case, simulation gives then a capacity loss of 5% if the guard band is260 kHz. This corresponds to a carrier spacing of around 4.1 MHz (Figure 23).
8/13/2019 Adjacent Channel Interference in Um Tsf Dd Mode
35/36
Adjacent Channel Interference in UMTS FDD mode
UMT/IRC/INF/10 V01.02 / EN Approved 30/07/2001 Page 36/36
8. CONCLUSION
This document has given an overview of problems that adjacent channels can create.
Adjacent interference has to be well defined before initial deployment to avoid badsurprise as dead zone or low QoC.
Calculations interpreting Normalisation works give very pessimistic figures where thenatural minimum coupling loss is not high enough. This calculation have been done forworst-case scenario, the aim of simulation is to give a more realistic view of theproblem (for macro-macro case).
The main conclusion of such calculation is that collocation of competitor BS is stronglyrecommended to avoid too much severe interference. This will avoid the near-fareffect in uplink and will avoid a big unbalanced scenario in the downlink.
Simulations have shown in the same way that capacity loss is lower when sites are
collocated.Theoretical calculations have also shown that adjacent channel interference problemsare mainly linked to UE performances. Simulations have demonstrated that winningsome dB in ACP allows to reduce the capacity loss in all cases (collocated,intermediate and worst).
Simulation has shown also that it is possible to reduce the carrier spacing of oneoperator without losing too much capacity and then to keep away from competitorscarriers.
Future work will consist to examine the impact of the deployment of a TDD layer overa FDD layer and to evaluate by simulation the capacity of a micro/macro network withdifferent strategies of carrier allocations.
8/13/2019 Adjacent Channel Interference in Um Tsf Dd Mode
36/36
Adjacent Channel Interference in UMTS FDD mode
END OF DOCUMENT