1、m 3404583 0080277 410 m Released: July 1 1993 GSM 05.90 Version: 4.1 .O Date: July 1993 Work item No: Key words: European digital cellular telecommunication system (phase 2); GSM EMC considerations ETSI European Telecommunications Standards Institute ETSI Secretariat: F - 06921 Sophia Antipoiis Cede
2、x. France TP. + 33 92 94 42 00 TF. + 33 93 65 47 16 Tx. 47 00 40 F This is an unpublished work the copyright in which vests in the European Telecommunications Standards institute. Ail rights reserved. The information contained herein is the propetry of ETSI and no part may be reproduced or usai exce
3、p as authorised by contract or other written permission. The copyright and the foregoing restriction on reproduction and use extend to ail media in which the information may be embodied. I 3404583 O080278 357 Page 3 GSM 05.90, version 4.1 .O. July 1993 Contents 1 . scope 5 2 . Informationavailable 5
4、 3 . Cauce o potential EMC interference 5 4 . LaboratoryresuRs 5 4.1 Hearingaids . 5 4.2 Cardiac pacemakers . 6 4.3 Domestic Equipment . 7 5 . Modellingresuits 7 Soli onc . 7 6 . 7 . Non-ionizing radiation 8 8 . Conclusion 9 9 . OtherEMCreports . 10 AnnexA 11 AnnexB . 12 AnnexC . 13 AnnexD . 14 Anne
5、xE . 15 AnnexF . 16 AnnexG 17 AnnexH . 18 Previous page is blank 3404563 0060279 293 Page 5 GSM 05.90, version 4.1.0: July 1993 1. Scope A considerable amount of work has been conducted, mainly in the UK, to investigate the effect of wanted radio frequency transmissions from GSM MS and BTS on other
6、equipment. This report aims to summarise this work and to look at the implications for GSM. Since GSM EMC considerations extend outside the GSM arena, it is thought essential that GSM considers the implications of EMC and produces this report. 2. Information available A number of European organisati
7、ons have conducted extensive investigations into GSM EMC. These investigations looked at the potential of a GSM transmission to interfere with a wide range of electrical apparatus. Having conducted both objective and subjective investigations, it was discovered that personal audio equipment (e.g. Wa
8、lkmans) and hearing aids were most susceptible and most likely to be in close proximity to GSM apparatus. of these two types of apparatus, hearing aids were considered the greatest potential problem and thus a considerable amount of modelling work was conducted in order to assess the likely incidenc
9、e of interference in various scenarios. Interference with pacemakers was considered of utmost seriousness and consequently tests were made to investigate the possibility of interfering with certain types. 3. Cause of potential EMC interference The source of GSM interference is the loOO! amplitude mo
10、dulated RF envelope introduced by burst transmission necessary for TDMA. Audio apparatus having some non-linear component able to demodulate this AM envelope will be subject to interference in the audio pass-band since the frame and burst rates for GSM are 220 Hz and 1.7 kHz. Another source of inter
11、ference is the DlX (Discontinuous Transmission) mode of operation in GSM. In the DTX mode there are two signal components with much lower frequencies than the normal GSM transmission: a component with a frequency of 2.1 Hz corresponding to the transmission of the 8 timeslots of the SID (Signai Descr
12、iptor) message block, and another with a frequency of 8.3 Hz corresponding to the repetition rate of SACCH. 4. Laboratory results 4.1 Hearing aids Objective laboratory results from the United Kingdom, Department of Trade and Industry, Radiocommunications Agency (DTI/RA) Annex A showed that a typical
13、 behind the ear hearing aid in normal (amplifying) mode was susceptible to peak GSM field intensities of; - - between 1 O V/m and 17 V/m in order to produce the same audio power as speech, 0.5 m in front of the hearing aid, and between 5 V/m and 8.5 V/m to produce audible, slightly annoying“ interfe
14、rence. It was noted that the group of hearing aids tested showed a 4 dB spread in susceptibiltty in the normal mode and a 13 dB spread in susceptibility in the inductive loop mode. Subjective investigation conducted at BTRL with the hearing aid worn by the user showed that audible, slightly annoying
15、 interference was perceived when subject to a peak field intensity varying between 10 V/m and 4 V/m depending upon the orientation of the head. This was modelled by a peak field intensity of 10 V/m for a 270 arc and 4 V/m for the Wo arc not shielded by the head inferring an 8 dB attenuation provided
16、 by the head. This directional susceptibility corresponds to an average of 6.6 V/m and thus agrees with the DTI/RA objective results. Previous page is blank = 3404583 O080280 TO5 Hearing aid type Page 6 GSM 05.90, version 4.1 .O: July 1993 Field strenght for noticable interference Distance for notic
17、able interference 2WMS 8WMS t These results were subsequently used for modelling activities to assess the consequences of this susceptibility in various scenarios. It should be noted that the susceptibility without head attenuation used in the model (4 Vim) is somewhat worse than the DTI measurement
18、s (5 Vim - 8.5 V/m) and thus the modelling results will be very much worst case. It was found that metallising the hearing aid case reduced the susceptibility with no head attenuation from 4V/mto12V/m (10dB). Laboratory measurements have been carried out also in Australia by Telecom Research Laborat
19、ories and National Acoustic Laboratories Annex HI. In these measurements the field strenght level causing useful “annoyance“ threshold level of 10 dB above the noise floor of the hearing aids was measured and then compared to measured field strengh of 2 W and 8 W GSM MS to determine the distances wh
20、ere the threshold levels can be expected. Both behind-theear and imtheear type hearing aids were measured, the former ones both with microphone input and telecoil input. The results are shown below. Behind the ear, microphone input Behind the ear, telecoil input In the ear 0.7 - 3.1 Vim 0.4 - 4.9 Vi
21、m 4.9 - 32.3 Vim 2.0-10m 1.5 - 20 m 0.2 - 0.6 m 3.5 - 20 m 2.5 - 40 m 0.4 - 1.5 m Table 1. Field strenght and safety distances for noticeable interference. Note 1 : The distances in Table 1 can not be compared directly with those in Table 2 because Table 1 distances are approximate real-life distanc
22、es whereas Table 2 is based on theory. 4.2 Cardiac pacemakers Work was carried out by CSELT Italy to investigate the effects of GSM type burst structure on cardiac pacemakers Annex E. Unipdar and bipolar types from one manufacturer were tested. The results show that, although it was possible to inte
23、rfere with pacemaker operation in free space, it was not possible, with the equipment power used, to interfere with operation when the pacemaker, leads and electrodes were placed in a phantom simulating realistic use in the human body. The equivalent maximum field strength used for this test would n
24、ot normally be exceeded at further than 0.5 m away from any allowed GSM transmitter except the maximum power base station. For information the field strength required to defeat the pacemaker in free space was in excess of 40 V/m for the most sensitive class of pacemaker. As there does not appear to
25、be a problem with defeating of pacemaker operation by a normal GSM signal, the remainder of the work done by GSM, and thus the remainder of this report, is restricted to scenarios for audible interference with hearing aids. Page 7 GSM 05.90, version 4.1.0: Juiy 1993 3404583 O080281 741 4.3 Domestic
26、Equipment Tests carried out by various laboratories and collected together by the Radio Technology Laboratory (RTL) of the Radiocommunications Agency Annex F ch that for a limited number of devices under test the cassette decks, television receivers and portable radiosIcasette players etc. are the m
27、ost susceptible domestic equipment with the mean field intensities causing “visiM4audible, but not annoying interference being 2.9, 4.0, and 5.6 V/m, respectively. For exampie for 8 W MS the fieid strenght of 4 V/m will be found at distances less than 5 m (worst case assuming 100 % efficiency and fr
28、ee space path loss) as can be seen in Tabie 1. This means that in practice, due to building attenuation etc., interference will not occur unless the transmitter and the victim equipment are in the same room. This is likely to occur if the GSM terminal is transportable (8 W output power for instance)
29、. 5. Modelling results A wide range of scenarios were modelled Annexes B and Cl to include the possible interference to hearing aid users from base stations, mobiles and handportables. Not surprisingly, by far the highest incidence of interference was caused in crowded urban environments where heari
30、ng aids and handportable transceivers are likely to be in closest proximity. It was found that a hearing aid user would experience 3 seconds of interference every 8 minutes whilst walking on a London street and would be subject to a 2.4% probability of interference whilst travelling on a commuter tr
31、ain for a GSM system occupying 2 x 25 MHz. Further results shed that with 1% of the train passengers using GSM transmitters (0.1% previously) and an average susceptibility uf 4 V/m, the probability of interference was 5%. Reducing the susceptibility by 10 dB, 8s found posible by simpie hearing aid m
32、odification, reduced the probability of interference in all scenarios by a factor of ten. It should be noted that the modelling work is based on free space path loses. The effect of, for exampie, people in a crowded train has not been measured, but in general it is expected that the presence of peop
33、le or objects between the MS and the hearing aid will be to reduce the interference in most cases. It should be noted thai all the scenarios examined assumed the hearing aid was active all the time. Clearly, there will be instances where the user will switch off the aid when not required to communic
34、ate. A further modelling exercise indicated that it was unlikely thai a hearing aid user will be able to use GSM handportable terminals due to the interference effects. 6. Solutions The generic immunity standard, EN 50082-1, produced by CENELEC, calls foc immunity to RF electromagnetic fields of 3 V
35、/m. This work has shown that current hearing aids have immunities close to this proposed levei and that a handportable GSM transmitter is likeiy to present a field strength greater than this at regular intervals in a crowded environment and thus Muse interference to the hearing aid user Annex DI. Th
36、e actual fieid strength from a dipde, as calculated from IEC 801 -3:1984, is shown in Table 2 (the values are independent of frequency), M 3404583 O080282 888 power CNattS) 0.8 2 Page 8 GSM 05.90, version 4.1 .O: July 1993 class lm 2m 5m 5 6.3 3.1 1.3 4 9.9 5.0 2.0 I I I 1 5 8 20 Peak field strength
37、 (v/m) 1 Peaktransmit I GSMMSpower I 3 15.7 7.8 3.1 2 19.8 9.9 4.0 1 31.3 15.7 6.3 0.25 1 2 3.5 1.8 0.7 1 7. O 3.5 1.4 Tabie 2: Close proximity field strengths A solution to this potential problem could be achieved by a combination of increased hearing aid immunity and constraints placed on the GSM
38、system in urban environments. Due to the likely peak field strengths that will be experienced from GSM transmitters in crowded urban areas, it is proposed that the immunity of Mure body worn apparatus, such as hearing aids, should be increased to 1 O V/m since this has been found to significantly re
39、duce the probability of GCM interference (this 10 V/m figure is derived from considerations of frequencies around 900 MHz and may not be applicable to frequencies significantiy higher or lower than 900 MHz). Further to this, a number of simple constraints for urban GSM system design should be adhere
40、d to:- - dynamic power contrd to be implemented at the MS such that only the minimum required transmit power is used at all times (BS interference was shown not to be a problem) - urban cell sizes limited to reduce required transmit powers - discontinuous transmission (DTX) to be implemented where p
41、ossible - GSM base site and mobile pay phone (e.g. on train) transmit antennas should not be located in close proximity to electrical apparatus likely to be susceptible to this type of interference. It is assumed that DTX will provide a reduced interference potential aithough this has not been verif
42、ied. 7. Non-ionizing radiation The major effect from exposure to RF radiation is due to the transfer of energy from the electromagnetic field to bidogical tissues, resulting in a temperature rise. The heating is caused by the fact that centres of negative and positive charges do not coincide in many
43、 biological tissues including water. The charge separation causes molecules to oscillate in a microwave field, generating heat. Guideline levels for exposure to non-ionizing RF radiation have been published by many organisations including Non-Ionizing Radio Committee (INIRC), the UK National Radidog
44、ical Protection Board (NRPB), the Institute of Elecrtical and Electronics Engineers (IEEE) and the German Electratechnical Comission of DIN and VDE (DKE) Annex GI. Table 3 shows the levels of power density given by each organisation which they believe will ensure that no heaith hazard exists. Some o
45、f the figures given in Table 3 are m 3404583 0080283 714 m Organisation Page 9 safety levei at 900 MHZ 01V/m2) 1 safety levei a i .a GH (w/4 I GSM 05.90, verdon 4.1 .O: July lk3 under review by the above organisations. NRPB INIRC IEEE DKE 22.5 4.5 6.0 4.5 45 9 12 9 Table 3. Power density safety valu
46、es me differences in the figures depend on the time how long the victim is assumed to be exposed by the radiation, and is for the most stringent figures up to 24 hours a day. When the safety distances, which can be derived from the above safety levels with some ssswnpions of the antenna configuratio
47、ns used, have been agreed by the appropriate organisations they will be included in this report. 8. Conclusion Extensive research has highlighted a potential compatibility problem between GSM transmitters and body worn audio apparatus; in particular hearing aids. An increased immunity for future bod
48、y worn apparatus, enforced through the Communitys EMC Directive (89/336/EEC), combined with some urban cellular design constraints aimed at ensuring the minimum transmit power is employed should ensure incidences of interference from GSM apparatus is kept to a minimum. The studies made have shown th
49、at the immunity levei of 3 V/m which has been propcsed for EMC directive does not protect hearing aids very well from the interference of GSM phones. Also, it has been shown that increasing the immunity to 10 V/m, as found possible by simple hearing aid modification, will reduce the probability of interference considerably. Concerning the domestic equipments it can be concluded that GSM transportable 8 W mobile stations are likely to cause problems to domestic equipment being used in a domestic environment. Further, it is recommended that the users data (l
