Derivation of BEM for TDD and FDD Terminal Station Int He 2500-2690 MHz Band

8/3/2019 Derivation of BEM for TDD and FDD Terminal Station Int He 2500-2690 MHz Band

1/6

Abstract This paper reports on a study undertaken within theCEPT SE42 project team, with the objective of derivingregulatory out-of-block radiation limits for mitigating the impactof adjacent-channel interference between TDD and FDDterminal stations (TSs) in the 2.6 GHz band. We present a novelstochastic approach for the calculation of TS block-edge mask (BEM) out-of-block (or baseline) limits. Appropriate BEMbaseline limits are subsequently derived via Monte-Carlosimulations based on interferer TS spatial densities

commensurate with those observed in busy hot-spots. It is shownthat, for a typical activity factor of up to 12.5%, a TS BEMbaseline limit of 22.5 dBm/MHz is sufficient to ensure that thevictim TS is desensitized by 3 dB with a probability of only 5%.

Index Terms Co-existence analysis, Adjacent-channelinterference, Spectrum management, Block-edge mask.

I. I NTRODUCTION

s a result of its availability for mobile services in theEuropean Union [1][2] and a number of countries

worldwide, the 2.6 GHz band (2500-2690 MHz) provides animportant opportunity for the introduction of next generation

mobile systems. The two main competing technologies for the provision of mobile services at 2.6 GHz include a) WiMAX,for use of unpaired spectrum through time division duplex(TDD); and b) LTE, primarily (though not exclusively) for useof paired spectrum through frequency division duplex (FDD).

Figure 1 illustrates two example frequency arrangements for the 2.6 GHz band as specified by the European Commission 1 (EC) Decision 2008/477/EC [1].

f

2 5 0 0

M H z

2 5 7 0

M H z

2 6 2 0

M H z

2 6 9 0

M H z

FDD-UL TDD FDD-DL

f

FDD-UL TDD FDD-DL TDD

(a)

(b)

R R

R R R

2 5 0 0

M H z

2 5 7 0

M H z

2 6 2 0

M H z

2 6 9 0

M H z

Figure 1. Examples of the European harmonised 2.6 GHz band plans.

The band plan consists of a FDD channeling arrangement of 270 MHz, based on a block size of 5 MHz, and with a duplexgap of 50 MHz. The FDD uplink (UL) starts at 2500 MHz and

1 See the Appendix for a brief overview of the relationship between the ECand various European regulatory organisations.

the FDD downlink (DL) starts at 2620 MHz. As indicated inFigure 1a, the FDD duplex gap (2570-2620 MHz) can itself beused by TDD systems. Outside the duplex gap, TDD usagecan be allowed at a national level, and shall be in equal partsin the upper segments of the FDD UL and DL sub-bands (suchthat the 120 MHz duplex spacing of the band-plan ismaintained). An example of this is shown in Figure 1b. Notethat blocks marked with R are restricted blocks. Restricted

blocks are not subject to the same level of protection asstandard blocks.

Given the availability of both paired and unpaired spectrumin the 2.6 GHz band, one can identify four types of inter -system adjacent-channel interference. These include: a) BS-to-TS, b) TS-to-BS, c) BS-to-BS, and d) TS-to-TS interference,where BS and TS denote base station and terminal stationrespectively.

Categories (a) and (b) above are no different from the typesof interference which occur at the frequency boundaries whichseparate adjacent FDD cellular systems, or indeed, thosewhich separate adjacent TDD cellular systems. Moreover,

similar types of intra -system interference occur at the channel boundaries within any type of cellular system. Consequently,no special regulatory provisions for the mitigation of BS-to-TS or TS-to-BS adjacent-channel interference in the 2.6 GHz

band are deemed to be necessary (other than those that arealready embedded in the relevant mobile technology standardsin order to deal with such interference issues).

Categories (c) and (d) above, however, are specific toscenarios where transmissions in adjacent frequencies aresubject to uplink and downlink phases which are notsynchronised in time. This is characteristic across frequency

boundaries which separate paired (FDD) and unpaired (TDD)spectrum, or across those which separate licensees of unpaired

(TDD) spectrum where the uplink and downlink phases of thelicensees are likely to be unsynchronised.

In 2006, the EC issued a mandate [3] to CEPT 2 to developleast restrictive technical conditions for the use of a number of frequency bands, including the 2.6 GHz band. The resultingCEPT Report 19 [5] addressed the issue of BS-to-BSinterference in the 2.6 GHz band, but did not adequatelyinvestigate the case of TS-to-TS interference. Subsequently,

2 Conference of European Post & Telecommunications Administrations. The proceedings of the various CEPT groups can be accessed at [4].

Derivation of block-edge masks for TDD and FDDterminal stations in the 2500-2690 MHz band

Hamid Reza KarimiOffice of Communications (Ofcom)

London, UK [email protected]

Grard LapierreAgence Nationale des Frquences (ANFR)

Brest, [email protected]

A


2/6

CEPT decided to specifically address this shortcoming. In this paper we describe the Monte Carlo approach that was adopted by CEPT to investigate the impact of TS-to-TS interference.This is also documented in ECC Report 131 [6].

In order to mitigate the impact of TS-to-TS interference, it isnecessary to place regulatory constraints on the out-of-block (OOB) emissions of the TSs. This is achieved through thespecification of block-edge masks (BEMs).

As shown in Figure 2, a BEM is a regulatory emission mask that is defined, as a function of frequency, relative to the edgeof a block of spectrum that is licensed to an operator. It consistsof in-block and out-of-block components which specify the

permitted radiation levels over frequencies inside and outsidethe licensed block of spectrum, respectively. The out-of-block component of the BEM itself consists of a baseline level and,where applicable, intermediate (transition) levels whichdescribe the transition from the in-block level to the baselinelevel as a function of frequency.

Frequency

Power Edge of block

P IB

P BL

In-blockOut-of-block

Baseline level

LicensedspectrumTransition level

In-block limit

Figure 2. Illustration of a block-edge mask.

In the 2.6 GHz band, the TS BEM in-block (total radiated) power limit is equal to 31 dBm/(5 MHz) as specified by theEC Decision 2008/477/EC. The TS BEM baseline limit then

applies at any frequency boundary where a TDD TS mightinterfere with a FDD TS, where a FDD TS might interferewith a TDD TS, or where a TDD TS might interfere withanother TDD TS. Therefore, in the context of Figure 1, theBEM baseline limit for a TDD or FDD TS would apply over those frequencies outside an operators licensed spectrumwhich consist of the unpaired and FDD-DL portions of the 2.6GHz band. The full specification of the TS BEMs is thencompleted by the definition of a number of transitional levelsover the restricted blocks as well as the FDD-UL portions of the 2.6 GHz band [6]. These transitional limits are no morerestrictive than the emission masks specified in the relevantmobile technology standards.

In this paper we present a novel stochastic approach for thecalculation of the TS BEM baseline limit. The approachincorporates a) the statistics of the victim and interferer TSlocations within their respective cells, b) the statistics of thevictims received wanted signal power and the interferersemission levels, and c) the statistics of collisions at the victimreceiver between interfering uplink packets and wanteddownlink packets.

Section II describes the methodology used for derivation of the BEM baseline limit. In Section III we present the parameter values and Monte Carlo simulation results which underpin the

proposed baseline limit. This is followed by conclusions inSection IV.

II. MODELLING OF TS- TO -TS INTERFERENCE

A. Calculation of interferenceConsider an adjacent-channel interferer TS and a victim TS

as shown in Figure 3. For illustrative purposes, and withoutloss of generality, the interferer is assumed to be a TDD TSand the victim is assumed to be a FDD TS. Other combinations are addressed in Section IV-C.

The tolerable, or target , interference power level, P I,FDD,Target in dBm/MHz, at the victim FDD TS receiver caused by a non-co-channel TS interferer may be written (in the logarithmicdomain) as

FDDA,FDDD,FDDI, NTargetFDD,I, GGG P P +++= , (1)

where N P is the thermal noise floor at the receiver 3 in

dBm/MHz, FDDI,G is the noise rise in dB due to the presenceof intra-system interference power in the DL of the FDD

network, FDD,DG is the tolerable increase in dB of theinterference-plus-noise power level (receiver desensitisation)at the cell edge, and FDD,AG represents the increase in dB of the tolerable interference due to proximity of the victim to itsserving base station.

FDD BS TDD BS

Wantedsignal

TS-TSinterference

TDD TSFDD TS

f 1

f 0

Figure 3. Illustration of interference between two terminal stations.

Note that the intra-system interference that is experienced ina FDD network includes multiple-access interference (co-channel), inter-cell interference (co-channel), as well as allforms of intra-system adjacent-channel interference. The term,

FDD,DG , is related to the receiver desensitization in dB at thecell-edge. For a 3 dB desensitization, for example, the targetinterference, P I,FDD,Target , would be equal to the intra-systeminterference-plus-noise power, FDDI, N G P + , in which case

FDD,DG = 0 dB.

The interference allowance, FDD,AG , accounts for the factthat, as a victim TS moves in from the cell-edge andapproaches its serving base station, the wanted DL signalincreases, and so for a fixed signal-to-interference-plus-noiseratio (and hence DL quality), the victim receiver can tolerate a

proportionally greater amount of interference. Specifically,

0FDD,0FDD,1FDD,A = GGG , (2)

3This is equal to 10 log 10(k T B) + NF TS , where k is Boltzmanns constant

(W/K/Hz), T is the ambient temperature (K), B is the receiver noise-equivalent bandwidth (Hz), and NF TS is the receiver noise figure (dB).


3/6

where FDD,1G and FDD,0G are the BS-TS mean path gains indB at the victim TSs location and the cell edge, respectively.

The objective here is to compute the interferer TS out-of- block EIRP for which, given all realisations of TS locationsand TS-TS propagation gains in the envisaged scenario, thelevel of interference experienced at the victim TS does notexceed the target value in (1).

Hence, for a non-co-channel TDD TS interferer radiatingwith an out-of-block EIRP level of TDDOOB, P dBm/MHz inthe vicinity of a victim FDD TS, one may write

TargetFDD,I,CollTDDPC,TSTSPL,TDDOOB, P GGG P +++ , (3)

where TSTSPL, G is the TS-TS propagation gain (includinglog-normal shadowing) in dB, TDDPC,G is a power controlfactor in dB, and CollG accounts for the extent of collision (intime) between a packet transmitted by the interferer and awanted packet received by the victim.

The term TDD,PCG , accounts for the fact that, as theinterferer TS moves in from the cell edge and approaches itsserving base station, the wanted signal level on the UL

increases, and so for a fixed signal-to-interference-plus-noiseratio (and hence UL quality), the interferer can transmit at a

proportionally reduced in-block EIRP, implying acorrespondingly reduced out-of-block EIRP, TDDOOB, P .Specifically,

0TDD,1TDD,0PC,TDD = GGG , (4)

where TDD,1G and TDD,0G are the BS-TS mean path gains indB at the interferer TSs location and at the cell edge,respectively.

Note that the unit of dBm/MHz is used here for simplicity.The BEM baseline limit itself is typically specified as power measured over a bandwidth that is equal to the block size; i.e.,in units of dBm/(5 MHz) for the 2.6 GHz band.

B. The impact of packet collisionsOne important aspect which needs to be considered is the

likelihood of partial overlaps (in time) between the interferingand victim signals. This is relevant in cases where the radiotechnologies used by the victim and the interferer incorporatesome element of time-division multiple-access (TDMA), as isthe case, for example, in packet-based transmission.

Modern radio access technologies increasingly employ packet-based transmissions over the air-interface in order to better deal with the bursty nature of traffic, and to moreefficiently utilise the radio resource by appropriatelyscheduling transmissions to and from those terminal stationsassociated with favourable radio link conditions at any giveninstant in time.

Consequently, the TSs in such systems transmit and receivedata in bursts of finite duration. As a result, the probability of collision at a victim TS receiver between a wanted DL packetand an interfering UL packet (originating from a non-co-channel TS) is inevitably less than unity. Furthermore, theextent of interference experienced by the victim is also afunction of the degree of overlap (in time) between the wantedand interfering packets, as illustrated in Figure 4.

Time0

T 0

victim packetinterval, T P,V

DL

T Sch

UL

interferer packetinterval T P,I

Figure 4. Partial overlap between interferer and victim packets.

In this study, the above effects are captured by the factor,

CollG , where

)/(log10 VP,010Coll T T G = , (5)

where T 0 is the overlap interval between a wanted DL packet(of duration VP,T ) and an interfering UL packet (of duration

IP,T ) at the victim receiver. Naturally, in the case of acomplete overlap, VP,0 T T = (i.e., dB0Coll =G ), and thevictim experiences the full effect of interference. Conversely,in the case of no overlap, 00 =T (i.e., =CollG ), and thevictim experiences no interference.

The extent of packet overlap, and hence the value of CollG ,is re-calculated at each Monte Carlo trial. The FDD DL packetdestined for the victim TS is assumed to be received (at auniformly distributed time of arrival) within a schedulinginterval SchT . The TDD UL packet transmitted by an adjacent-channel interfering TS is then assumed to be received by thevictim TS (at a uniformly distributed time of arrival) withinthe UL phase of the TDD network.

In order for the modelling of interference to equally apply toa TDD victim as well as a FDD victim, we assume equal

packet durations, IP,VP, T T = , and a TDD UL/DL ratio of 1.The latter assumption implies that the TDD TS is a potentialinterferer during half the scheduling interval, and is a potentialvictim during the remaining half of the scheduling interval.

III. SIMULATION METHODOLOGY

A. Examined geometriesAs depicted in Figure 5, a scenario is considered where a

TDD cellular network is deployed in the same geographicalarea as a FDD cellular network. It is further assumed that theTDD network operates within 5 MHz frequency blocks thatare neighbouring those used by the FDD network in thedownlink direction, thereby giving rise to the possibility of TS-to-TS interference from the TDD network to the FDDnetwork.


4/6

FDD cells

TDD cells

FDDbase station

r 0

TDDterminals

FDDterminal

TDD cellradius(1 km)

FDD cellradius(1 km)

TDDbase station

Figure 5. Urban macro-cellular deployment.

A macro-cellular deployment is considered in order tocapture situations where both the interferer and victimterminals are far from their serving base stations. In suchcircumstances, the interferer causes most interference (byradiating at high powers) and the victim is most susceptible tointerference (by receiving wanted signals at low powers).

B. Simulation process

The following steps are performed at each Monte Carlo trial:1) Drop the victim FDD TS at a random (uniformly distributed) location

within the FDD macro-cell.

2) Drop M TDD TS interferers at random (uniformly distributed) locationswithin a hot-spot surrounding the FDD TS.

3) Drop the TDD hot-spot at a (uniformly distributed) location within theTDD macro-cell.

4) Calculate the TDD power control factor, -40 dB GPC,TDD 0 dB, based on the location of the TDD TS within the TDD macro-cell,according to (4).

5) Calculate the interference allowance factor, GA,FDD , based on the locationof the FDD TS within the FDD macro-cell, according to (2).

6) Calculate the tolerable interference, P I,FDD,Target , at the victim FDD TS, based on (1).

7) Calculate the path gain (including log-normal shadowing), GPL,TS-TS , between the victim TS and each of the M TS interferers.

8) Calculate a collision factor, GColl , for each victim-interferer pair based on (5).

9) Select the dominant TS interferer (i.e., that which would cause thegreatest interference).

10) Compute the out-of-block EIRP, P OOB,TDD , for the dominant interferer, asindicated in (3).

Note that this formulation corresponds to the case where theFDD TS is always in the proximity of a high density of TDDTSs (i.e., a TDD hot-spot). A 5% probability that the victimTS is desensitized by 3 dB is considered to be an appropriate

protection criterion in this context.

Following a sufficiently large number of Monte Carlo trials,the statistical distribution of the TS out-of-block EIRP,

P OOB,TDD , can be derived. This process fully identifies thecharacteristics of a TS interferers out-of-block EIRP subjectto the requirement that, across the ensemble of all realisationsconsidered, the victim TS does not experience interferencethat is greater than a defined tolerable (target) level.

C. Parameter valuesA list of all parameter values used in the derivation of the

results in this paper is presented in Tables I to III below.

TABLE I. LIST OF FDD RECEIVER PARAMETER VALUES .

Cell radius 1000 metresBS antenna height 30 metres

Minimum BS-TS separation 50 metresBS-TS mean path loss model Extended (urban) Hata [7]

TS antenna gain 0 dBi Noise-equivalent bandwidth, B 5 MHz

TS noise figure, NF TS 9 dB

Desensitization 3 dB ( GD,FDD = 0 dB)Intra-system noise rise, GI,FDD 6 dB

Downlink packet duration, T P,V 2.5 ms

TABLE II. LIST OF TDD TRANSMITTER PARAMETER VALUES . IS PEOPLE DENSITY .

Cell radius 1000 metresHot-spot radius 25 or 50 metres

BS antenna height 30 metresMinimum BS-TS separation 50 metres

BS-TS mean path loss model Extended (urban) Hata [7]TS spatial density (per 5 MHz) /(10 2210) metre 2

Number of interferers in hot-spot, M 2 or 1 (per 5 MHz)Uplink packet duration, T P,I 2.5 ms

Uplink/downlink ratio, uUL/DL 1:1

TABLE III. LIST OF GENERIC PARAMETER VALUES .

Operating frequency 2.6 GHz Number of Monte Carlo trials 5000

TS-TS separation 25 or 50 metres (max),1 metre (min)

TS-TS propagation loss model IEEE 802.11 Model C [8]Scheduling interval, T Sch 20 ms

TS antenna height 1.5 mBoltzmanns constant, k 1.3804 1023 (W/K/Hz)Ambient temperature, T 290 K

The hot-spot interferer TS spatial density is calculated byconsidering a people spatial density of m2, and assumingthat only 10% of the people use their wireless device at anygiven time. It is then assumed that 50% of the TSs operate inthe 2.6 GHz band (the rest operating in other bands). Of thoseTSs operating in the 2.6GHz band, it is assumed that half operate in FDD mode and half in TDD mode. The TDD TSsare then assumed to be uniformly distributed across a total of 10 available unpaired (TDD) 5 MHz blocks. The interferer TSspatial density per 5 MHz block is then /(10 2210) m 2.

Note that, in practice, the duration of radio packetstransmitted and received by a TS is a function of the multiple-access technique, the modulation and coding scheme used atany given instant, and the details of the multi-user schedulingalgorithm implemented by the base station across the cell. Inline with the principle of technology neutrality, a packetinterval of 2.5 ms is adopted for the purposes of this study.This value can be justified by noting that many of the packet-

based radio access technologies today use transmission timeintervals of the order of 1 to 2 ms. Furthermore, the consideredscheduling interval of 20 ms is consistent with the encodinginterval of many audio/video compression algorithms. The

packet duration of 2.5ms, in conjunction with a schedulinginterval of 20 ms, implies a TS activity factor of 12.5%.


5/6

IV. SIMULATION RESULTS

A. Impact of packet collisionsFigure 6 shows the CDFs of the TDD TS out-of-block EIRP,

P OOB,TDD , where the probability of packet collisions is nottaken into account (i.e., GColl = 0 dB), and where the

probability of packet collisions is incorporated throughrecalculation of GColl at every Monte Carlo trial, as describedin (5). The CDFs here are derived for an intra-system noiserise of G I,FDD = 6 dB, a 3 dB desensitization ( GD,FDD = 0 dB) of the victim FDD TS, and a 100 m separation 4 between the FDDand TDD base stations. The results are also for a hot-spotradius of 25 m and a people density of = 1/3 m 2, implyingM = 2 interferer TSs in the adjacent 5 MHz channel (SeeSection III-C).

Note that the right-most CDF in Figure 6 does not approacha value of unity, since there exist Monte Carlo trials where no

packet collisions occur (i.e., GColl = dB). More specifically,the M = 2 TDD packets each of 2.5 ms duration can together occupy up to half of the available time in the 10 ms UL phaseof the TDD network. For this reason, the probability that nocollisions occur is mainly determined by the number of trialswhere the received FDD packet falls within the 10 ms DL

phase of the TDD network.

Figure 6. CDF of interferer TS out-of-block EIRP, for a 3 dB desensitizationof the victim TS ( GD,FDD = 0 dB), intra-system noise rise of GI,FDD = 6 dB, and

accounting for the likelihood of packet collisions ( GColl 0 dB).

Based on the above results, one can conclude that, if the permitted TDD TS out-of-block EIRP is limited to amaximum of 27 dBm/MHz, this would exclude 95% of therealisations in which the victim FDD TS is desensitized by 3dB (referenced at cell-edge). Stated differently, a regulatory

TDD TS BEM baseline limit of 27 dBm/MHz would imply a5% probability of the victim FDD TS being desensitized by 3dB (referenced at cell edge).

B. Sensitivity to hot-spot user spatial densityThe results presented in Figure 6 are based on a hot-spot TS

interferer spatial density that is derived with reference to a

4 This separation is consistent with the base-to-base coordination distance of 100 m considered in CEPT Report 19 [5]. A randomized base-to-baseseparation results in higher BEM baseline limits.

people density of = 1/3 m2, and with 1 person in 10 usingtheir wireless communication device at any instant. Thisimplies a substantial number of 65 TSs simultaneouslyoperating within a radius of 25 m from a victim TS. It isevident that, while such TS densities might be plausible invery high-density hot-spots, these occur rarely, and where theymight occur regularly, it is likely that the terminals would beserviced by pico-cells rather than macro-cells. For the abovereasons, it is interesting to evaluate the TS BEM baselinelevels required in what might be considered to be more typicalgeometries experienced across a macro-cell, as outlined inTable IV.

TABLE IV. HOT-SPOT CATEGORIES WITHIN A MACRO -CELL .

PeopleDensity (m2)

Hot-spotradius

(m)

Number of TSinterferers, M , per 5 MHz in hot-spot

Very high-density hot-spot 1/3 25 2High-density hot-spot 1/5 25 1

Hot-spot 1/10 50 2

Figure 7 shows the variation in the CDFs of the TS out-of- block EIRP, P OOB,TDD , for each of the three hot-spot categories

described above. Also shown is the CDF of the TS out-of- block EIRP when aggregated (with equal weighting) over thethree hot-spot categories. Table V shows the corresponding TSBEM baseline limits for a 5% probability of the victim TS

being desensitized by 3 dB (referenced at cell-edge).

Figure 7. CDFs of interferer TS out-of-block EIRP for three different hot-spot categories in a macro-cell. Results are for a 3 dB desensitization of the

victim TS ( GD,FDD = 0 dB), and intra-system noise rise of GI,FDD = 6 dB.

TABLE V. TS BEM BASELINE LIMITS FOR DIFFERENT HOTSPOT CATEGORIES .

Hot-spot geometryBEM baseline limit

(dBm/MHz)Case (a) Case (b)

r 0 = 25 m, = 1/3 m ( M = 2) -38 -27r 0 = 25 m, = 1/5 m ( M = 1) -33.5 -20.7r 0 = 50 m, = 1/10 m ( M = 2) -29.4 -15.9Average over three hotspot categories -34 -22.5

C. Interference from FDD TSs to TDD TSs and betweenunsynchronised TDD TSs

The TS BEM baseline levels derived above were calculatedin the context of adjacent-channel interference from TDD TSs

(a)Not accounting

for likelihoodof packetcollisions

(b)Accounting


(a)

(b)

(a)Not accounting for

likelihood of packet collisions Probability

of nocollisions

(b)Accounting for likelihoodof packet collisions

T P = 2.5 ms, T Sch = 20 ms

-27

5%


6/6

to FDD TSs. Figure 8 shows the corresponding CDFs of theTS out-of-block EIRP for cases where interference is i) fromFDD TSs to TDD TSs, and ii) from TDD TSs to TDD TSs (inunsynchronised TDD networks). Simulation parameters areunchanged.

Simulation results for case (a) are unchanged since the probability of collision is not taken into account. Simulationresults for case (b) indicate that, while the precise shape of the

CDFs is different for the latter two interference scenarios (e.g.,due to different probabilities of no collision), this has littleimpact on the lower tail of the CDFs. Hence, the BEM

baseline limits derived in Section IV(B) for interference fromTDD TSs to FDD TSs broadly apply to the case of interference from FDD TSs to TDD TSs, and TDD TSs toTDD TSs.

Figure 8. CDFs of interferer TS out-of-block EIRP averaged over threedifferent hot-spot categories. Results are for a 3 dB desensitization of thevictim TS ( GD,FDD = 0 dB), and intra-system noise rise of GI,FDD = 6 dB.

V. CONCLUSIONS

A novel stochastic approach was presented for the derivationof regulatory BEM out-of-block limits with the purpose of mitigating adjacent-channel interference between TDD andFDD terminal stations (TSs) in the 2.6 GHz band. Macro-cellular geometries involving hot-spots of interfering TSslocated in the vicinity of a victim TS were investigated. Thestatistics of the TS out-of-block EIRP were calculated for arange of typical hot-spot sizes and interferer terminaldensities. These were then used to define the terminal BEM

baseline limit, subject to the requirement that the victimterminal is desensitized by 3 dB with a probability of only 5%.The results indicate that, given a terminal activity factor below

12.5%, a TS BEM baseline limit of 22.5 dBm/MHz isappropriate. The purpose of BEMs is to manage the risk of undue interference between operators in a manner which isindependent of the technology used by each operator. In thisway, licensees can more easily deal with future evolutions intechnology.

APPENDIX A BRIEF OVERVIEW OF THE RELATIONSHIPBETWEEN THE EC, CEPT, AND ETSI

The main institutions which coordinate spectrummanagement throughout Europe are described below.

European Commission (EC) : Spectrum policy in theEuropean community is driven in particular by the ECs RadioSpectrum Decision 676/2002/EC of March 2002. Based onthis Decision, two complementary bodies were set up tosupport radio spectrum policy, namely the Radio SpectrumPolicy Group (RSPG) and the Radio Spectrum Committee(RSC). Conference of Postal and Telecommunications

Administrations (CEPT): Its main task is to develop policieson electronic communications activities in a European context,forward plan and harmonise the efficient use of the radiospectrum, satellite orbits, and numbering resources withinEurope. Another of its important tasks is the development of common European positions and proposals for use in theframework of international and regional bodies. EuropeanTelecommunications Standards Institute (ETSI): It producesstandards for equipment including fixed, mobile, radio,converged, broadcast and internet technologies. It isrecognized by the EC as a European Standards Organization.The EC issues mandates to CEPT for the development of technical implementation measures. In response, CEPTdelivers CEPT reports which are used by RSC to develop, if necessary, EC legislation (e.g., an EC Decision ), which then

becomes binding for the 27 Member States of the EU.

ACKNOWLEDGEMENTS

The authors wish to thank their colleagues in the CEPTSE42 project team, in particular Martin Fenton (Ofcom),Simon Pike (Vodafone), Guillaume Lebrun (Qualcomm),Jamshid Khun-Jush (Qualcomm), Johnny Dixon (BT), MikeChartier (Intel), and Karl Loew (T-Mobile) for their manyhelpful comments and suggestions.

R EFERENCES [1] EC Decision 2008/477/EC on the harmonisation of the 2500-2690 MHz

frequency band for terrestrial systems capable of providing electroniccommunications services in the Community, June 2009,

www.erodocdb.dk .[2] ECC Decision (05)05 on harmonised utilisation of spectrum for IMT-

2000/UMTS systems operating within the band 2500-2690 MHz, March2005, www.erodocdb.dk .

[3] EC mandate to CEPT to develop least restrictive technical conditions infrequency bands addressed in the context of WAPECS, July 2006,http://ec.europa.eu/information_society/policy/ecomm/radio_spectrum/manage/eu/rsc/rsc_subsite/mandates .

[4] European Communications Office. http://www.ero.dk/ .[5] CEPT Report 19, Report from CEPT to the European Commission in

response to the Mandate to develop least restrictive technical conditionsfor frequency bands addressed in the context of WAPECS, December 2007, www.erodocdb.dk .

[6] ECC report 131, Derivation of a block edge mask (BEM) for terminalstations in the 2.6 GHz frequency band (2500-2690 MHz), February2009, www.erodocdb.dk .

[7] European Radiocommunications Office, SEAMCAT user manual(Software version 2.1), February 2004.

[8] TGn Channel Models (IEEE 802.11-03/940r2), High Throughput Task Group, IEEE P802.11, 15 March 2004.

(a)Not accounting


(b)Accounting


(a)

(b)

Documents

Derivation of BEM for TDD and FDD Terminal Station Int He 2500-2690 MHz Band