PUBLIC DOCUMENT - Europa...more receivers, where each receiver can have a wired or wireless (inductive) connection to a hearing aid or be an integral part of a hearing aid. Hearing

European Commission, DG Communications Networks Content & Technology, 200 Rue de la Loi, B-1049 Bruxelles RSC Secretariat, Avenue de Beaulieu 33, B-1160 Brussels - Belgium - Office BU33 7/09 Telephone: direct line (+32-2)299.66.11 / 295.26.65 switchboard (+32-2)299.11.11. Fax: (+32-2) 296.83.95 E-mail : [email protected]

EUROPEAN COMMISSION Directorate-General for Communications Networks, Content and Technology

Electronic Communications Networks and Services Radio Spectrum Policy

Brussels, 9 March 2017 DG CONNECT/B4

RSCOM17-17

PUBLIC DOCUMENT

RADIO SPECTRUM COMMITTEE

Working Document

Subject: Presentation of the study on Assistive Listening Devices (ALDs) in the

2.3-2.4 GHz band by the JRC

This is a Committee working document which does not necessarily reflect the official

position of the Commission. No inferences should be drawn from this document as to the

precise form or content of future measures to be submitted by the Commission. The

Commission accepts no responsibility or liability whatsoever with regard to any

information or data referred to in this document

mailto:[email protected]

L

JRC StuLTE and

ope

udy on CWirelesrating in

F

09

Europe

Joint R

Coexistss Hearn the 2.

inal Repo

9 March 20

ean Comm

Research C

Ispra (Italy

ence being Aid .4 GHz b

ort

17

mission

Centre

)

etween Systemband

ms

2

Prepared by: Detlef Fuehrer, Jean‐Marc Chareau, Philippe Viaud, Tiziano Pinato, James Bishop

3

ContentsGlossary ................................................................................................................................................... 5

1. Introduction ........................................................................................................................................ 7

2. Definitions ........................................................................................................................................... 8

3. Objectives of this study ....................................................................................................................... 8

4. Summary of findings ........................................................................................................................... 8

5. Hearing Aid Systems ‐ Background ................................................................................................... 10

5.1 Hearing Aids ................................................................................................................................ 10

5.2 Assistive Listening Devices (ALDs) ............................................................................................... 11

6. LTE‐HAS Coexistence ......................................................................................................................... 13

6.1 LTE characteristics ....................................................................................................................... 13

6.1.1 TD‐LTE .................................................................................................................................. 13

6.1.2 FDD‐LTE ................................................................................................................................ 13

6.2 Previous studies on LTE‐HAS coexistence ................................................................................... 14

7. Assessment of the impact of LTE signals on HAS .............................................................................. 15

7.1 Tested devices ............................................................................................................................. 15

7.2 Interference scenarios and use cases ......................................................................................... 16

7.3 Test procedure ............................................................................................................................ 16

7.4 Test signals and metrics .............................................................................................................. 16

7.4.1 Interfering signal characteristics .......................................................................................... 18

7.5 Measurement setup .................................................................................................................... 19

7.6 Measurements and Observations ............................................................................................... 22

7.6.1 Setup 1 ................................................................................................................................. 22

7.6.2 Setup 2 ................................................................................................................................. 25

7.6.3 Setup 3 ................................................................................................................................. 27

7.6.4 Setup 4 ................................................................................................................................. 31

7.6.5 Setup 5 ................................................................................................................................. 34

7.6.6 Setup 6 ................................................................................................................................. 35

7.6.7 Setup 7 ................................................................................................................................. 36

7.6.8 Setup 8 ................................................................................................................................. 37

7.7 Analysis ....................................................................................................................................... 38

7.7.1 Quality assessment methodology ........................................................................................ 38

7.7.2 Results .................................................................................................................................. 39

8. Summary and conclusions ................................................................................................................ 41

Appendix A – List of measurement equipment .................................................................................... 42

Appendix B – Audio recordings ............................................................................................................. 43

List of Tables ......................................................................................................................................... 44

List of Figures ........................................................................................................................................ 45

Bibliography .......................................................................................................................................... 46

4

Note: All trademarks and registered trademarks are the property of their respective owners.

5

Glossary

ALD Assistive Listening Device Bluetooth LE Bluetooth Low Energy BS Base Station BTE Behind‐The‐Ear CCTV Closed Circuit TV CEPT European Conference of Postal and Telecommunications Administrations CF Centre Frequency CIC Completely‐In‐Canal dB Decibel dBm Decibel milliwatt DL Downlink DUT Device Under Test EC European Commission ECC Electronic Communications Committee EHIMA European Hearing Instrument Manufacturers Association ETSI European Telecommunications Standards Institute EU European Union E‐UTRA Evolved UMTS Terrestrial Radio Access FDD‐LTE Frequency Division Duplex LTE FM Frequency Modulation HA Hearing Aid HAS Hearing Aid System IEC International Engineering Consortium IIC Invisible‐In‐Canal ISM Industrial, Scientific, Medical ISTS International Speech Test Signal ITC In‐The‐Canal ITE In‐The‐Ear ITU‐R International Telecommunication Union ‐ Radiocommunication Sector JRC Joint Research Centre LOS Line Of Sight LSA Licensed Shared Access LTE Long Term Evolution MOS Mean Opnion Score MUS Minimum Usable Signal NFMI Near‐Field Magnetic Induction Ofcom UK [UK] Office of Communications OOB Out‐Of‐Band PAR Peak‐to‐Average Ratio PDSCH Physical Downlink Shared Channel PEAQ Perceptual Evaluation of Audio Quality PMSE Program Making and Special Events RF Radio Frequency RIC Receiver‐In‐Canal RITE Receiver‐In‐The‐Ear

6

RSC Radio Spectrum Committee Rx Receive SRD Short Range Device SNR Signal‐to‐Noise Ratio TD‐LTE Time Divivsion Duplex LTE TV Television Tx Transmit UAS Unmanned Aircraft Systems UE User Equipment UL Uplink WBB Wireless Broadband WHO World Health Organisation Wi‐Fi Wireless Fidelity (IEEE 802.11)

7

1.Introduction

The band 2300‐2400 MHz is allocated to the Mobile Service on a co‐primary basis by ITU radio Regulations in all three ITU regions, and footnote 5.384A of the Radio Regulations identifies this frequency band for IMT. Existing use of the 2300‐2400 MHz frequency band in the European Union (EU) includes telemetry (terrestrial and aeronautical); fixed links, other governmental use including unmanned aircraft systems (UAS) and closed‐circuit television (CCTV), program making and special events (PMSE) ancillary video links as well as amateur radio as a secondary service.

Following a positive opinion of the Radio Spectrum Committee (RSC), the European Commission (EC) submitted in April 2014 a Mandate to the European Conference of Postal and Telecommunications Administrations (CEPT) to develop harmonised technical conditions in the 2300‐2400MHz band for wireless broadband (WBB) electronic communication services in the EU.

In response to the Commission Mandate and following a public consultation, the CEPT delivered in November 2014 its Report 55 on the technical conditions for wireless broadband usage of the 2300‐2400 MHz band.

In light of the comments submitted by Member States, the on‐going trials of Licensed Shared Access (LSA) in some Member States, and concerns brought to the attention of the EC regarding possible interference of LTE equipment operating in the 2300‐2400 MHz band with other equipment operating in the 2400 MHz band, the Commission proposed and RSC decided at its July 2015 meeting to postpone the adoption of an Implementing Decision until after WRC‐15.

Despite the variety of studies on coexistence between 2300 MHz TD‐LTE and systems operating in the 2400 MHz unlicensed band (further on referred to as “victims” or “victim systems”) that have been conducted so far no consensus among the stakeholders regarding the severity of interference from TD‐LTE could be reached. One of the perceived shortcomings of these studies was the limited number of potential victim devices that were tested. Therefore, the EC’s Joint Research Centre (JRC) was requested to conduct a comprehensive technical study on the potential impact of TD‐LTE on the population of deployed Wi‐Fi devices. The final results of this study were presented to the RSC at its October 2016 meeting.

At the July 2016 meeting of the RSC it had been agreed that a subsequent study should assess the impact of LTE interference on Assistive Listening Devices (ALDs) operating in the 2.4 GHz band. Within the scope of this study a four‐day measurement event was held in November 2016 at the premises of the JRC in Ispra in which representatives of the European Hearing Instrument Manufacturers Association (EHIMA) and three major manufacturers of hearing aid systems participated. The results of this study are presented in the current document.

8

2.Definitions

A variety of terms and definitions exist for assistive listening devices and hearing aids, with often slightly different connotations. In this document we use the following terms and definitions:

Hearing aid (HA): Medical device comprising an electro‐acoustic amplifier including a microphone and a loudspeaker and having a frequency response and dynamic characteristics specific to each person's individual hearing loss. Some modern hearing aids feature integrated wireless receivers.

Assistive Listening Device (ALD): Radiocommunication device used in addition to hearing aids to make more sounds accessible to people with hearing impairment. It usually comprises a transmitter, which can be handheld, on a table or around the neck of a hearing impaired person, and one or more receivers, where each receiver can have a wired or wireless (inductive) connection to a hearing aid or be an integral part of a hearing aid.

Hearing Aid System (HAS): Comprises the hearing aid(s) plus accessories as well as any type of assistive listening device1.

3.Objectivesofthisstudy

A number of studies have been published that address coexistence between TD‐LTE and HAS and other types of wireless short‐range devices (SRD) that use Bluetooth, Bluetooth LE and similar technologies [1] [2] [3] [4]2.

The objective of this study was to complement the findings from the aforementioned studies by assessing the additional impact on HAS performance from FDD‐LTE operating in the 2.5 GHz band and by more systematically analysing the effect of concurrent Wi‐Fi operation on different channels. As the group of test devices included a number of prototypes featuring the latest technology the results should also provide an indication of the progress made in terms of HAS performance and robustness against interference.

4.Summaryoffindings

In this study we examine the effects of adjacent‐band LTE signals on the quality of audio signals received by ALDs and hearing aids. For this purpose we conducted measurements with 21 devices from six major manufacturers in 23 different test configurations. Overall, 192 individual measurements were made. We focused on the effect of transmissions from LTE User Equipment (UE) operating in proximity of hearing aid systems.

We observed that when HAS receiver and transmitter were operating at a distance from each other that is representative of typical operating conditions almost all systems proved to be very robust against interference. Even in the presence of multiple high‐power in‐band interferers the HAS which all appeared to employ frequency hopping and detect‐and‐avoid techniques managed to maintain stable connections and provide distortion‐free audio.

When HAS were operating near the receiver sensitivity level, i.e. when their RF signals were highly attenuated, the presence of strong adjacent‐band LTE signals resulted in degradation of the audio signals in a number of cases. Adding in‐band Wi‐Fi signals generally worsened the situation.

In combination, TD‐LTE and FDD‐LTE degraded victim signal quality slightly more than individually.

1 This definition essentially corresponds to that of “aids for hearing impaired” in ETSI TR 102 791 V1.2.1 (2013‐08) [1]. 2 A summary of the results of these studies is provided in section 6.2 of this document

9

For adjacent‐band LTE signals to cause degradation of a HAS audio signal a number of conditions must be fulfilled:

The quality of the RF link between HAS transmitter and receiver is poor, i.e. the signal‐to‐noise ratio (SNR) at the receiver is low.

There is a nearby LTE UE transmitting continuously, e.g. during the upload of a large file to a remote base station.

The LTE UE is located close to the HAS receiver. Depending on the model of LTE UE the distance at which the audio signal is impaired can be between a few centimetres to 1 metre for severe degradation, and up to 11 metres for minor degradation. These values were calculated for free‐space / line‐of‐sight conditions.

LTE is operating at the band edges, i.e. 2390 MHz for TD‐LTE and 2505 MHz for FDD‐LTE3.

We also noted that the RF emissions from certain HAS models can severely degrade Wi‐Fi performance.

Overall, our findings are fully in line with those of the various previous studies. The adaptive frequency‐hopping mechanism that has apparently been implemented in the devices we tested proved to be very effective for interference mitigation. We conclude that while HAS audio signal quality can be impaired by adjacent‐band TD‐LTE signals the combination of prerequisites for this to happen makes the overall risk appear low. Furthermore, we conclude that the additional presence of FDD‐LTE UE signals in the 2.5 GHz band does not significantly increase the degradation of HAS audio quality.

3 Due to time constraints the impact of LTE signals further removed from the 2.4 GHz band edges could not be assessed. While previous studies considered only TD‐LTE and frequencies up to 2390 MHz the conditions created in this study correspond to worst‐case scenarios.

10

5.HearingAidSystems‐Background

According to the World Health Organization (WHO) over 5% of the world’s population – 360 million people of which 32 million are children – suffer from disabling hearing impairment [5]. Although hearing loss mostly affects the elderly the number of children and young adults suffering from hearing loss is growing steadily. It is estimated that in the US around 5% in the age between 5 and 24 are affected [6]. Globally, some 1.1 billion teenagers and young adults are at risk of hearing loss due to the unsafe use of personal audio devices and exposure to damaging levels of sound at noisy entertainment venues, the WHO reports [7].

Hearing aid systems can enable these persons to participate in daily life. Currently, hearing aids are used by about 50 million people4.

5.1HearingAidsWhile their mechanical predecessors have been in use since at least the early 17th century electric hearings aids came into play at the beginning of the 20th century, with the advent of the carbon microphone. The first wearable hearing aid was developed in 1938 [8]. Until the late 1980s hearing aids were based on analogue technology when advances in semiconductor manufacturing and digital signal processing heralded the digital age, initially in hybrid analogue‐digital models in which digital circuits controlled an analogue compression amplifier. Fully digital models debuted in 1996, and programmable models, which allow for greater flexibility and fine‐tuning of the hearing aids according to the patient's needs, became available in 2000 [8]. In 2004, the first wireless hearing aid was introduced [9].

The majority of hearing aids fall under the “air conducted sound” category [10]. The two major groups are “In‐The‐Ear” (ITE), located in the ear canal and “Behind‐The‐Ear” (BTE) located behind the ear, but with parts of the aid located in the ear canal [11]. There are several kinds of canal‐style devices: “Completely‐In‐Canal” (CIC) and “Invisible‐In‐Canal” (IIC) devices fit the deepest within the canal; a tiny extension cord is used to place and remove the instrument. “In‐The‐Canal” (ITC) devices are slightly larger, so they extend farther out but remain hidden [12]. “Receiver‐In‐Canal” (RIC) and “Receiver‐In‐The‐Ear” (RITE) devices are similar in concept to BTE hearing aids, with the exception that the speaker has been detached from the case and fitted in the ear canal or ear and connected to the case of the hearing aid with a thin wire [13].

Various other types of hearing aids exist such as bone‐anchored aids and cochlear implants. These types have not been included in this study and are therefore not covered here. Detailed information on hearing impairments and the different types of hearing aids can be found in [10] and [14].

The global market volume for hearing aids (BTE, RIC, RITE, IIC) is estimated between 4.5B [15] and 6B USD [16]. In 2014, 12 million hearing aids were sold globally [6]. By 2019 this number is forecast to increase to 17 million, [17]. Europe accounts for 41% of units sold [18].

In the first‐half of 2016, about 9 of 10 (87.5%) hearing aids sold contained wireless technology [19]. While still negligible today, shipments of Bluetooth devices are expected to increase to 6 million units by 2019 [20] which corresponds to a market share of 35%.

The hearing aid market is dominated by six major suppliers which in 2014 held a combined market share of approximately 98% [6].

4 Estimate based on figures reported in [18]

5.2AsALDs argatewayvia a necproducereceiverThe audequipmeHearing ALD dire

RF transand heaEnergy. announc

BluetoBluetoIntereswirelesfractiothe he

sistiveLisre, in most y receives thck loop usinges a magnetr coil in the hdio signal canent such as aaids featurinectly (Figure

smitter and ring aids incThe use of ced in 2011.

ooth 4.0 wasooth Low Enst Group. Blss devices won of the powealthcare, fitn

Figure 1: He

steningDcases, compe audio signag near‐field ic field, typihearing aid. n originate fa TV set (Figung integrate3).

receiver typcreasingly usedigital techn

s introducednergy, whichuetooth Sma

within a rangewer of classiness, and ho

aring aid manu

Devices(Aprised of anal from the Rmagnetic indcally in the

rom a smarture 2). d RF receive

ically operate digital wirenologies in t

d as part of h was later art was aimee up to 50 mic Bluetoothme entertain

11

ufacturers‘ glob

ALDs)n audio sourRF transmitteduction (NFM3 to 15‐MH

tphone, a wi

rs do not req

te in the FMeless technothe 2.4 GHz

the main Brenamed Bed at new lom. Devices op and they qnment indus

bal market shar

rce, an RF ter and relaysMI). The tranHz range whi

ireless micro

quire a gatew

, 900 MHz, oologies such aband is rela

Bluetooth stluetooth Smow‐power anperating wituickly becamstries [4].

res (2014) [6]

transmitter as it to the hensmitter coil ich is picked

ophone, or f

way but can

or 2400 MHas Bluetoothatively new;

andard in 2mart by the nd low‐latenh this versiome available

and a gatewearing aid(s), inside the n

d up by an i

rom any oth

be connecte

z band. Receh and Blueto; first produ

2010, incorpoBluetooth

ncy applicatioon consume in products

way. The typically neck loop nduction

her audio

ed to the

ent ALDs oth Low‐cts were

orating Special ons for a small within

A particundergocompressignal, ssignal. Tshould n

Listeneraudio sigparticulathe casequality. the user

ular challengoes on its wassion, codingsuch delay mThe Internatnot exceed −

s’ tolerancegnals are coarly detrimee of well‐venLatencies higr’s ability to l

Figure

Figure 3

ge for this tyay between g and transcmay result inional Teleco40 milliseco

for delay wmbined. Smntal for usented or opegher than 5 locate the or

e 2: Wireless au

3: Direct wirele

ype of hybridsource and

coding from a lack of syommunicationds (audio d

within an auall delays in rs who enjoyn‐canal hearms will affecrigin of the s

12

udio transmissio

ss audio transm

d wireless trthe user’s ethe RF protynchronicity on Union sugelayed) and

udio stream streamed ay listening toring aids, a ct speech intound [11].

on to Gateway

mission to Hea

ransmission eardrum. Thocol to thebetween thggests that +20 millisec

is even smaudio may reo music throdelay of 5 mtelligibility b

y device

ring Aids

lies in the dis delay resuNFMI signale video and audio/video onds (audio

aller when ssult in a perough open‐cmilliseconds ut also othe

elay the audults from au. In the case the streametransmissioadvanced) [

streamed anceived echo canal hearingmay degradr parameters

dio signal udio data e of a TV ed audio on delays 21].

nd direct which is g aids. In de sound s such as

13

6.LTE‐HASCoexistence

In Europe the 2400‐2500 MHz band has been made available to unlicensed wireless systems which comprise a variety of technologies, such as Wi‐Fi and Bluetooth, and services such as wireless broadband, audio transmission, and motion detection. The bands above and below this unlicensed have been allocated to wireless broadband (Figure 4). FDD‐LTE in band 7 has already been deployed in a number of European countries; TD‐LTE in band 40 is so far only being deployed or in preparation to be deployed by operators in Lithuania and Russia [22]. On the band edges the frequency separation between LTE channels and the unlicensed band is very small so that there is a potential risk of interference between the systems deployed in adjacent bands. The hearing aid system community has therefore been concerned that the presence of high‐power LTE systems operating in the adjacent bands may lead to degradation of HAS performance.

Figure 4: LTE and HAS frequency allocations between 2300 and 2690 MHz (Europe)

6.1LTEcharacteristics

6.1.1TD‐LTE

CEPT Report 55 [23] proposes to make the 2300‐2400 MHz band (band no. 40) available for TD‐LTE. The frequency arrangement should be based on 20 blocks of 5 MHz (Figure 5).

TD‐LTE Band 40

2300

MHz

2305

MHz

2310

MHz

2315

MHz

2320

MHz

2325

MHz

2330

MHz

2335

MHz

2340

MHz

2345

MHz

2350

MHz

2355

MHz

2360

MHz

2365

MHz

2370

MHz

2375

MHz

2380

MHz

2385

MHz

2390

MHz

2395

MHz

5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5

Figure 5: Proposed TD‐LTE frequency arrangement in the 2300‐2400 MHz band [23]

LTE user equipment (UE) may transmit with a power of up to 23 ±2 dBm, measured as the sum of the maximum output power at each UE antenna connector [24].

6.1.2FDD‐LTE

The frequency arrangement within the 2500‐2690 MHz band was defined in ECC/DEC/(05)05 [25].

Band 40 ISM and others Band 7 (up) Band 7 (down)

TD‐LTE ALDs FDD‐LTE FDD‐LTE

2300 2400 2483.5 2500 2570 2620 2690

14

In this frequency arrangement which is shown in Figure 6 any FDD uplink block (UL xx) is paired with its corresponding FDD downlink block (DL xx). The minimum block or channel width is 5 MHz5. In Europe, the most common channel widths are 10 MHz and 20 MHz.

Figure 6: Frequency arrangement within the 2500‐2690 MHz band

6.2PreviousstudiesonLTE‐HAScoexistence

Copsey Communications conducted a series of measurements with a number of ALDs and hearing aids using Bluetooth, Bluetooth LE, or proprietary protocols that were exposed to interference from adjacent‐band TD‐LTE signals [1]. Both 10 MHz signals, centred at 2385 MHz and 20 MHz signals, centred at 2380 MHz were used. During all measurements Wi‐Fi was present. Three types of equipment were tested, categorised as streamers, microphones, and smart devices/ experimental systems. The study found that, depending on the type of LTE signal and device under test (DUT), performance degradation/receiver blocking occurred within a distance of 0.15 to 4 metres from the TD‐LTE UE.

In 2013 Mac Ltd. conducted a study for Ofcom UK on the effects TD‐LTE signals in the 2.3 GHz band have on Bluetooth equipment operating in the 2.4 GHz band [2]. Both quantitative and qualitative tests were made. In the first case, a Bluetooth tester was employed to generate a test signal and measure the bit error rate of the device under test. For the qualitative test, an audio signal was streamed from a mobile phone to a Bluetooth DUT in the presence of a TD‐LTE signal. With all three devices that were tested, distortion of the audio signal occurred only when the interfering signal level reached very high levels (15 dBm into the antenna at a few centimetres separation from the headset), and only in one case it was possible to break the Bluetooth link. The authors concluded that a) the most significant interference mechanism is Bluetooth receiver blocking and TD‐LTE OOB emissions would have little impact, and b) Bluetooth devices are robust in the presence of interference and users of Bluetooth devices are unlikely to notice any impact if TD‐LTE services were introduced in Band 40.

Coexistence measurements conducted by Cambridge Silicon Radio (CSR) on Bluetooth devices showed that the tested devices “exceeded the expected performance, continuing to function in the presence of very strong LTE interference with a UE signal at ranges down to 0.05 m” [3]. The interferer was a TD‐LTE signal with a bandwidth of 20 MHz and a centre frequency of 2380 MHz. During all measurements Wi‐Fi was present. With only two devices tested, the authors cautioned that it could not be concluded that all devices in the field would operate as robustly.

In a study published by the EHIMA the possible effects of LTE signals occupying the 2350‐2390 MHz band on Bluetooth LE‐based ALDs were examined, taking into account ALD receiver blocking, LTE out‐of‐band (OOB) emissions, and ALD receiver selectivity noise [4]. Simulations were made for indoor and outdoor LTE base stations, femtocells and user equipment as interference sources. The study found that for an LTE signal occupying the frequency range up to 2390 MHz minimum separation distances between LTE UE and ALD of 0.5 to 50 metres would be required to satisfy the Bluetooth LE bit error rate requirement of 10‐3.

5 As an exception, Annex A of EC Decision 2008/477/EC [26] allows a departure from the arrangement for TDD operation on a national basis. This would result in TDD operation starting in DL and UL blocks 14 and extending downwards the band in contiguous blocks as required.

2500

MHz

2505

MHz

2510

MHz

2515

MHz

2520

MHz

2525

MHz

2530

MHz

2535

MHz

2540

MHz

2545

MHz

2550

MHz

2555

MHz

2560

MHz

2565

MHz

2570

MHz

2575

MHz

2580

MHz

2585

MHz

2590

MHz

2595

MHz

2600

MHz

2605

MHz

2610

MHz

2615

MHz

2620

MHz

2625

MHz

2630

MHz

2635

MHz

2640

MHz

2645

MHz

2650

MHz

2655

MHz

2660

MHz

2665

MHz

2670

MHz

2675

MHz

2680

MHz

2685

MHz

2690

MHz

UL01

UL02

UL03

UL04

UL05

UL06

UL07

UL08

UL09

UL10

UL11

UL12

UL13

UL14

DL01

DL02

DL03

DL04

DL05

DL06

DL07

DL08

DL09

DL10

DL11

DL12

DL13

DL14

FDD Uplink Blocks FDD Downlink Blocks

TDDor

FDD Downlink (External)

15

7.AssessmentoftheimpactofLTEsignalsonHAS

To assess the potential impact of adjacent‐band LTE signals on HAS performance the following approach was taken, in close collaboration with EHIMA and HAS manufacturers:

1. Selection of devices to be tested 2. Definition of usage and interference scenarios 3. Definition of signal parameters and test metrics/performance indicators 4. Development of measurement setups 5. Measurement of the impact of interfering signals on HAS performance 6. Analysis of measurement results 7. Conclusions

7.1TesteddevicesMeasurements were conducted with 22 different devices in 23 configurations6.

DUT ID DUT type Companion device type Victim link

technology 1 Wireless receiver with inductive loop TV audio streamer Bluetooth

2 Remote microphone Wireless receiver with inductive loop Proprietary



5 TV audio streamer Wireless receiver with inductive loop Bluetooth

6 Hearing aid Smartphone Bluetooth LE

7 Wireless receiver with inductive loop Smartphone Bluetooth

8 TV audio receiver & control TV audio streamer Proprietary

9 Hearing aid Smartphone Bluetooth





14 Remote microphone Wireless receiver with inductive loop Bluetooth

15 Hearing aid Smartphone Proprietary

16 Hearing aid Remote microphone Proprietary

17 Hearing aid Remote microphone Proprietary

18 Wireless receiver with inductive loop Remote microphone Proprietary

19 Wireless in‐ear receiver Remote microphone Proprietary



22 Hearing aid TV audio streamer Proprietary

23 Wireless headphone Smartphone Bluetooth

Table 1: List of tested devices

6 DUT21 and DUT22 were identical but tested with different companion devices.

16

7.2InterferencescenariosandusecasesIn this study we focus on the impact of interference from LTE UE signals on HAS operation. The interferer (the UE) operates in proximity of the victim, and both are located indoors, for instance in a room or a vehicle. The LTE base station signal is assumed to be weak in comparison to the UE signal.

Three typical use cases, as proposed by the Hearing Aid industry, have been considered.

Case A: Two persons are sitting in a bus next to each other. One person makes an LTE data link from their phone and uploads a big file or movie. The second person is streaming audio wirelessly (e.g. via Bluetooth LE) to their hearing aid devices.

Case B: Two persons are sitting in a car close to each other. One person, sitting in the passenger front seat makes an LTE data link from their phone and uploads a big file or movie. The second person who sits in the back seat of the car is streaming audio wirelessly from an ALD placed in the front of the car to a gateway device which relays the audio signal to their hearing aid.

Case C: A hearing aid user is watching TV using a 2.4 GHz ALD to listen to the TV audio. A second person in the same room makes an LTE data link from their phone and uploads a big file or movie.

Furthermore, we took into account the possibility of multiple interferers being active simultaneously. In Case A, for instance, there could be one another passenger sitting in the next row and uploading a file from his FDD‐LTE UE, or in Case C there could be a download ongoing via Wi‐Fi.

In all of the above cases the distance between HAS components is small so that under LOS conditions the RF link between them would be strong. Under certain conditions, however, additional attenuation, for instance from a body blocking the RF signal, can amount up to 50 dB, as outlined in [4].

7.3TestprocedureThe test procedure was relatively simple: An RF link was established between DUT and companion device. An audio signal (also referred to as ‘test signal’) was transmitted to the DUT from a companion device. In some cases the level of the wanted RF signal (also referred to as ‘victim signal’ or ‘victim link’) was set to the lowest level at which, in the absence of interference, no audible degradation of the test signal could be observed7.

At the beginning of each measurement cycle a recording of the audio signal was made without any interfering signal being present. This recording would constitute the reference that the other signals would be compared to. Then, the DUT was exposed to one or more interfering RF signals. The received audio signal was aurally monitored and simultaneously recorded for later analysis.

7.4TestsignalsandmetricsOwing to the volatile nature of the wireless channel audio signals streamed over wireless links may experience degradation of the perceived sound quality, particularly when highly compressed audio codecs are employed to reduce the required transmission rate. The amount of error correction that can be applied is limited due to the low‐latency requirements in case direct sounds are transmitted or synchronisation between audio and video signals needs to be maintained.

7 This level will further on be referred to as “Minimum Usable Signal” (MUS) level.

17

Typical degradation effects include bandwidth reduction, distortion artefacts, and signal drop‐outs. A selection of artefacts, as provided in [26] is shown in Table 2.

Artefact/Attribute Description

Inherent noise Continuous hissing.

Distortion Crackling, short bursts of hissing, buzzing, clipping.

Spatial distortion Wobbling, sound image stability, loss of directionality.

"Birdies" or tweets "Chirps" in mid to high frequencies.

Temporal smearing Pre/post echo, "sound shadow", diffuse onset of the sound.

Tone trembling Sounds like trembling tones, noticeable on longer notes. Sparkling.

Thin sound Timbre artefact related to skewed frequency response, lack of bass.

"Can" sound Timbre artefact related to boost and/or resonances in the mid‐range frequencies.

Table 2: Observed artefacts and attributes of streamed audio signals [26]

Radio interference from other wireless systems can add to the amount and impact of artefacts being generated.

Not all users, however, will be equally sensitive to the aforementioned artefacts which makes objective measures for and statements about sound quality difficult.

Different methods for sound quality measurements exist. Some are based on physical measurements and perceptual models, such as PEAQ (Perceptual Evaluation of Audio Quality, ITU‐R BS.1387), others on structured listening tests. The latter method was proposed for this study by the HAS community.

A reference audio signal commonly used in ALD/hearing aid test and measurement is the International Speech Test Signal (ISTS). ISTS was developed by EHIMA and adopted by the International Engineering Consortium (IEC) in hearing aid standard IEC 60118‐15 [27].

The ISTS consists of fragments of real speech recordings from six different languages, namely Arabic, English, French, German, Mandarin, and Spanish. The resulting signal has all major characteristics of speech, can be recognized by humans as being composed out of real speech, but is not intelligible. The signal bandwidth ranges from 100 Hz to 16 kHz so that hearing instruments with conventional bandwidths up to about 6000 Hz, as well as those with high‐frequency extended bandwidths can be measured.

During the measurement campaign not all manufacturers of ALDs/hearing aids used the ISTS for detecting impairments of the audio signal. Several products were tested with a 1 KHz sine wave signal.

For the aural assessment of the recorded signals the following metrics were chosen: Audible drop‐outs, clicks, and glitches, as well as variations in volume or frequency response, and the occurrence of wobbling or trembling.

Furthermore, the Mean Opinion Score (MOS) for each of the recorded signals should be determined. The values obtained for recordings made in the presence of interference would then be compared to those of the reference signals. The MOS is frequently used to determine the perceived quality of received voice that has undergone processing for transmission over a digital link. On a scale from 1 to 5 the MOS indicates the expected level of user satisfaction in respect to voice quality (Table 3).

18

User Satisfaction Level MOS Very satisfied 4.3 ‐ 5.0 Satisfied 4.0 ‐ 4.3 Some users satisfied 3.6 ‐ 4.0 Many users dissatisfied 3.1 ‐ 3.6 Nearly all users dissatisfied 2.6 ‐ 3.1 Not recommended 1.0 ‐ 2.6

Table 3: Mean Opinion Score (MOS)

7.4.1Interferingsignalcharacteristics

Table 4 lists the main characteristics of the interfering signals that were applied during the measurements.

Due to time constraints only a limited number of waveforms could be evaluated. We therefore selected those TD‐LTE waveforms which in our previous study on TD‐LTE and Wi‐Fi coexistence [28] had been found to cause the highest amount of degradation of victim performance.

Interferer Centre frequency [MHz]

Channel width [MHz]

Waveform

TD‐LTE

2310.0 20 UE UL, Rome B2.3

2350.0 20 UE UL, Rome B2.3

2390.0 20 UE UL, Rome B2.3

2397.5 5 TM 1.1

FDD‐LTE 2505.0 10 Ispra FDD UL

2486.0 10 Ispra FDD UL

Wi‐Fi 2422 (Ch 1+5) 40

2447 (Ch 6+10) 40

2452 (Ch 11+7) 40

Table 4: Interfering signal characteristics

The signal UE UL is a signal recorded close to a TD‐LTE UE during upload a large file to a remote base station. The measured transmit power levels (during transmission) were +19 dBm (maximum) and +16 dBm (mean), resp. The Peak‐to‐Average Ratio (PAR) for this signal is 14.6 dB.

Rome B2.3 is a signal recorded close to two TD‐LTE UEs which simultaneously uploaded large files to a remote base station. The PAR for this signal is 20 dB.

TM1.1 corresponds to test model E‐UTRA 1.1, as defined in [29]. E‐TM1.1 is employed to test various TD‐LTE base station parameters, including output power, unwanted emissions, and transmitter intermodulation. It is based on uplink/downlink configuration no. 3 and Special Subframe (SSF) configuration no. 8 defined in 3GPP TS36.211 [30]. It is a downlink‐heavy configuration with 6 downlink slots and 3 uplink slots per TD‐LTE frame. The Physical Downlink Shared Channel (PDSCH) is fully occupied by a single user, without power variation during transmission. The PAR for this signal is 12.4 dB.

Ispra FDstation conductsignal be

To studsignals wmaximis

7.5MeThe meSpectrumcalibrateLTE and the inteantenna

The audfrom the

Victim aspectrum

The initFigure 9

Figure

DD UL is a siglocated on tted with the eing present

y the impawere selectese spectrum

easuremeeasurementsm Lab in Isped for frequeFDD‐LTE onrfering signaas. Initially, th

dio test signae DUT was a

and interferm analyser.

ial test setu. A list of the

e 7: COLHA hea

gnal recordedthe JRC Isprcentre frequ at the edge

ct of in‐baned which occoccupation t

entsetup were condpra. The chaencies up to HAS operatals. The DUThe distance b

al was streammplified, dig

ring signals

up inside thee measureme

d

d close to anra site. The uency set to of the 2.4 G

nd interferecupied eitherthe channel

ducted in a amber which18 GHz. To tion two broT was placedbetween the

med from thegitised, and r

were monit

e chamber ient equipme

19

n FDD‐LTE UEPAR for this2486 MHz to

GHz band.

nce from Wr the lower, width was se

fully shieldh measuresassess the imadband hornd in line of se horn anten

To moof RF 7) wcharahearintest,attach

The Lwavefrecordon co

FurtheinstaleffectWi‐Fi.traffic

In theplacedexposmeasuthe wwas pthe banten

e companionecorded wit

tored using

s shown inent is provide

E uploading s signal is 6 o study the h

Wi‐Fi three dcentre, or uet to 40 MHz

ded anechoi7 m x 3.5 mmpact of simn antennas wsight of, andnas and the

odel realisticsignal propawas emplocterisation ang aids (COLeither DUThed to the he

LTE signals wform generaded in the cexistence be

ermore, a led in the cts of additio. The iperf3 c between ro

e initial setd behind thesure to the urements wwanted RF sigplaced insidebox was conna via a prog

n device to th a personal

a log‐perio

Figure 8, thed in Append

a large file tdB. A few hypothetical

different IEEupper part ofz.

ic chamber m x 3 m (D multaneous inwere installe at the sameDUT was 3.7

c conditions, agation a phoyed whicand optimisaLHA). DepenT or compead.

were generatators. The wcourse of theetween TD‐LT

Wi‐Fi routehamber to bonal in‐bandtool was us

outer and clie

tup the come horn anteinterfering s

which requiregnal level the a shielded nnected to grammable a

the DUT. The computer.

odic monito

he explanatodix A of this

to a commermeasuremencase of a br

EE802.11n‐cof the victim

of the JRCx W x H) h

nterference fed which trae height as t7 metres.

particularly hantom headch is usation of the Rnding on theanion devic

ted by two waveforms he previous JTE and Wi‐Fi

er and cliebe able to sd interferensed to geneent.

mpanion devnnas to minsignal(s). Foed an adjusthe companiobox. The RFan omnidi

attenuator.

e audio outp

ring antenn

ory block diadocument.

rcial base nts were oadband

ompliant band. To

C’s Radio has been from TD‐nsmitted the horn

in terms d (Figure sed for RF link of e type of ce were

arbitrary had been RC study i.

nt were tudy the nce from rate TCP

vice was nimise its r certain tment of on device F port of rectional

put signal

a and a

agram in

For the the DUTup to ‐2LTE sign

RF (LTE)

RF (LTE)

F

initial measuT was located1 dBm couldal levels whi

Companiondevice

Wi‐FiClient

Signal generator 1

Figure 8: Initial

Fig

urements thd in the far fd be reachedch had been

Tx antenna

Tx antenna

n

Signal generator 2

HAS measurem

ure 9: Initial HA

he distance bfield of the Td. The objectn observed in

1

2

Victim link

20

ment setup insi

AS measureme

between TxTx antennastive was to an real‐life situ

k

3.7 m

de the JRC’s an

ent setup (Setu

antennas an and interfeassess the imuations.

Ful

PCUSB

nechoic chamb

p 1)

nd DUT had ring signal lempact of LTE

DUT

ly‐shielded ane

AudioADC

Monitoring antenna

er

been chosenevels at the on HAS ope

Wi‐FiRouter

echoic chamber

Audio Pre‐amp

Spectrum analyser

n so that victim of eration at

r

Audio

21

After no negative impact on HAS performance was observed for any of the evaluated scenarios the setup was modified on request of the HAS industry representatives. The objective was revised to determine the interfering signal levels at which HAS operation would be disrupted. The measurement distance was reduced to 1 m so that the maximum interfering signal levels at the DUT increased by up to 11 dB8.

Depending on the type of DUT and the usage scenario the companion devices were placed in various locations inside the anechoic chamber (Figure 10). For certain measurements which required an adjustment of the wanted RF signal, i.e. the victim link level the companion device was placed inside a shielded box. The RF port of the box was connected to an omnidirectional antenna via a programmable attenuator.

Overall, there were eight variations of this setup which differed in terms of the number of Tx antennas, the Tx antenna polarisation, and the location of the companion device.

Figure 10: Final HAS measurement setup (Setups 2‐8)

8 Taking into account near‐field antenna gain.

Companion device

Wi‐FiRouter

Wi‐FiClient

Fully‐shielded anechoic chamber

Signal generator 1

Signal generator 2 PC

RF (LTE)

RF (LTE)

Audio

USB

1 metre

Companion device

Companion device

DUT

AudioADC

Audio Pre‐amp

Tx antenna 1

Tx antenna 2

7.6Me

7.6.1Se

In this swith six

DUT 1

The phaof 3.7 mantenna

For the f

• • • •

As no auconductlevels ofsignal geand ‐23/

9 “Transmpower m10 Mean p

easureme

etup1

setup which DUTs (DUT 1

antom head metres from as, approxim

first set of m

Interfering sCentre frequTransmit poResulting LT

udible imparted with signf +10 dBm anenerator. Th/‐8 dBm, res

mit power” indmeasured overpower at the

entsandO

corresponds1 to 6) were

with receivethe transmately 4 metr

measurement

signal: TD‐LTuencies: 231wer9 range: TE signal pow

rtments to thnal TD‐LTE Und +15 dBm.he resulting p.

dicates the por the entire duDUT indicates

Observati

s to the initiconducted. A

er (the DUT) it antenna. res from the

Figure

ts the DUT w

E UE UL 0, 2350, and‐30 dBm to +

wer at the DU

he test signaUE UL at a c. For this purLTE signal p

ower setting ouration of the s the mean sig

22

ons

ial setup shoAs audio tes

and HAs waThe companDUT.

e 11: DUT 1 (Se

was oriented

d 2390 MHz+5 dBm UT: ‐67 to ‐32

al could be oentre frequerpose the TDower levels

of the signal gsignal gnal power du

own in Figurt signal the I

as placed on nion device

etup 1)

in the horizo

2 dBm (mean

observed twoency of 2390D‐LTE signal w(mean/peak

enerator whic

uring the perio

re 9 a total oSTS was use

top of the swas placed

ontal positio

n10), and ‐55

o additional 0 MHz and wwas transmitk) at the DU

ch correspond

od of transmis

of 43 measued.

support at a behind the

n.

to ‐19 dBm

measuremewith transmtted from theT were ‐28/

ds to the mea

ssion

urements

distance transmit

(peak)

nts were it power e second /‐13 dBm

n signal

23

To take into account a possible polarisation mismatch between the Tx antennas and the integrated DUT antenna a second set of measurements was conducted with the DUT oriented in the vertical position. Centre frequencies and power levels were set to the maximum values to create worst‐case conditions.

• Interfering signals: TD‐LTE UE UL, Rome B2.3, TM 1.1 • Centre frequencies:

o UE UL: 2390 MHz o Rome B2.3: 2390 MHz o TM1.1: 2397.5 MHz

• Transmit power levels: o UE UL: +15 dBm o Rome B2.3: +10 dBm o TM1.1: +17 dBm

• Resulting LTE signal power at the DUT (mean/peak): o UE UL: ‐23/‐8.4 dBm o Rome B2.3: ‐26/‐8 dBm o TM1.1: ‐21/‐8.6 dBm

On one channel of the reference signal very faint artefacts were observed, originating most probably from a mobile phone coupling into the audio link in the control room. These artefacts were not present in any of the other recordings.

The last set of measurements was then repeated with DUT 1 oriented in the vertical position but rotated by 180 degrees.

In the next step the effect of two adjacent band LTE signals on victim performance was evaluated.

• Interfering signals: o TD‐LTE: Rome B2.3 o FDD‐LTE: Ispra FDD UL

• Centre frequencies: o Rome B2.3: 2390 MHz o Ispra FDD UL: 2505 MHz

• Transmit power levels: o Rome B2.3: +10 dBm o Ispra FDD UL +14 dBm

• Resulting LTE signal power at the DUT: o Rome B2.3:: ‐27 (mean), ‐9 dBm (peak) o Ispra FDD UL ‐28 (mean), ‐23 dBm (peak)

A final set of measurements was made with the companion device placed in a shielded box. The victim RF signal level was reduced to the minimum at which no audible degradation of the test signal could be observed.

• Interfering signal: TD‐LTE Rome B2.3 • Centre frequency: 2390 MHz • Transmit power range: ‐10 dBm to +10 dBm • Resulting LTE signal power at the DUT: ‐51 to ‐31 dBm (mean), and ‐33 to ‐13 dBm (peak)

As no audible effects on the test signal could be observed in any of the above scenarios a few additional experiments were conducted during which the following observations were made:

24

A TD‐LTE signal (Rome B2.3) was transmitted in the 2.4 GHz band (CF= 2442 MHz). The audio link could not be disrupted, even when operating at the MUS level.

Two TD‐LTE signals (Rome B2.3 and UE UL) were transmitted simultaneously in the 2.4 GHz band (CF= 2432 MHz and 2462 MHz, resp.). The audio link could not be disrupted, even when operating at the MUS level.

When the frequency of the TD‐LTE signal UE UL was changed to 2452 MHz there was a reduction of audio quality lasting several seconds but then the HAS recovered.

DUT 2

The DUT, a remote microphone) was placed on top of the support; the phantom head with receiver and HAs was placed on the floor next to the support to simulate a real usage scenario (Case A).

• Interfering signal: TD‐LTE UE UL • Centre frequencies: 2390 MHz • Transmit power level: +15 dBm • Resulting LTE signal power at the DUT: ‐26 dBm (mean), ‐13 dBm (peak)

Observations:

Very faint artefacts were present in all signals, even in the absence of interference. No further degradation was observed when interference was added.

Even with two in‐band TD‐LTE signals present simultaneously (Rome B2.3 and UE UL, at various frequencies and with maximum transmit power) the audio link could not be disrupted.

DUTs 3‐5

Each DUT (remote microphones and audio streamer) was placed on top of the support; the phantom head with receiver and HAs was placed on the floor next to the support.

To save time we maintained the previous setup and conducted measurements with two in‐band interferers.

Observations:

The results were the same as for DUT 2: Even with two in‐band TD‐LTE signals present simultaneously (Rome B2.3 and UE UL, at various frequencies and with maximum transmit power) the audio link could not be disrupted.

DUT 6

DUT 6, a hearing aid, was attached to the phantom head and placed on top of the support. The companion device was placed at the opposite end of the chamber. In this case audio was streamed from companion device directly to the hearing aids.

The link quality between DUT 6 and the companion device (smartphone) was so poor that a piece of absorbing material placed in front of the companion device disrupted the audio stream. Thus, it was inferred that the HAS operated at MUS level. DUT 6 was the only device identified as using Bluetooth LE.

• Interfering signal: TD‐LTE UE UL • Centre frequency: 2390 MHz • Transmit power range: ‐30 dBm to +15 dBm

•

Observa

With thiNo impa

7.6.2Se

In SetupDUT watest sign

The DUTthe Tx acorrespo

The firstpower r

Observa

Resulting LT

ations:

Even withouthe audio sigNo further d

is setup the Dact on the HA

etup2

p 2 Tx antens reduced tonal.

T was placedantenna. Theonds to Case

t set of meaange from +

ations:

From +8 dBmchannel wassecond recowere observdBm (peak).Between +9audio signal Between +1suggesting t

E signal pow

ut interferengnal. degradation w

DUTs were eAS audio sign

nna 2 was reo 1 metre. O

d on top of the companione B defined in

asurements 4 dBm to +1

m on audio qs lost after ording was mved. The cor 9 and +11 dfrom the lef

13 and +15 dhat the HAS

wer at the DU

ce being pre

was observe

exposed to enal quality co

emoved and nly DUT 6 w

he support, wn device wasn Section 7.2

Figure

was made w5 dBm (resu

quality starte45 seconds

made during rresponding

Bm some aft channel wadBm a fewmay have ad

25

UT: ‐73 to ‐28

esent some a

ed when inte

lectrical fieldould be obse

the separatwas measured

with the phas placed on t2.

e 12: DUT 6 (Se

with TD‐LTElting LTE me

ed deterioratand did noat the sameLTE signal p

rtefacts occas lost after artefacts codapted to th

8 dBm (mean

artefacts (glit

erference wa

ds of up to 1erved.

ion distanced with this se

antom head the opposite

etup 2)

E signal UE Uean signal po

ting. At +8 dt recover une power levepower at the

urred (glitch25 seconds b

ould be obsehe interferen

n), and ‐58 to

tches, wobb

s added.

.6 V/m (mea

e between Tetup. ISTS w

reversed, i.ee end of the

UL at 2390 wer at the D

Bm the audintil the end el; this time e DUT was ‐

hes, wobblinbut returnederved (wobbce environm

o ‐13 dBm (p

ling) were p

an) and 7 V/m

x antenna 1was used as t

e. with its bachamber. Th

MHz and a DUT: ‐25 to ‐1

o signal fromof the recoonly minor ‐21 dBm (me

ng). At +12 d 30 secondsbling but no ment.

peak)

resent in

m (peak).

and the he audio

ck facing his setup

transmit 14 dBm)

m the left ording. A artefacts ean) / ‐6

dBm the s later. glitches)

A furtheof +10 d

Observa

For the n

• •

The Wi‐ 1. Wi‐F

Nextto +

11 Averag

er measuremdBm (LTE me

ations:

There was wobbling co(mean) / ‐2 d

next set of m

Distance froDistance fro

Fi signal leve

Fi channel 1+

With only Wno degradat

t, TD‐LTE sig10 dBm). At a Tx poapproximatesignal powe1.2 V/m (me

Figure 13

For Tx pobserve

ge power duri

ment was maan signal pow

a slight degould be obsedBm (peak).

measuremen

om Wi‐Fi AP tom Wi‐Fi clien

el at the DUT

+5

Wi‐Fi presenttion of the au

nal Rome B2

wer level oely 15 seconr at the DUTean) and 12 V

: Distortion of

ower levels d.

ng transmissio

de with TD‐Lwer at the D

gradation oferved. The co

ts additiona

to DUT: 3.8 mnt to DUT: 2

T was approx

a few artefaudio signal c

2.3 was adde

of +5 dBm tnds but recoT was ‐22 dBV/m (peak).

the DUT 6 aud

from +6 dBm

on.

26

LTE signal RoDUT: ‐19 dBm

f the audioorresponding

l interferenc

m .9 m

ximately ‐47

acts were auould be obse

ed (centre fre

the audio sovered 20 sm (mean) / ‐

io signal in the

m to +9 dBm

ome B2.3 at m).

quality. Seg LTE signal

ce from Wi‐F

dBm11.

udible duringerved.

equency: 239

ignal from tseconds late‐4 dBm (pea

presence of W

m no degrada

2390 MHz a

everal glitchepower at th

i was genera

g the first 10

90 MHz, Tx p

the left char (Figure 13k); the elect

Wi‐Fi and TD‐LTE

ation of the a

and a transm

es and instahe DUT was

ated.

0 seconds. Af

power range

annel was lo3). The resuric field stren

E (setup 2)

audio signal

mit power

ances of ‐17 dBm

fter that,

: +5 dBm

ost after lting LTE ngth was

could be

2. Wi‐F

Meafreq

Obs

3. Wi‐F

Meafreq

Obs

7.6.3Se

For Setuand 9), For DUT

DUT 6

The DUTfar end o

At a Tx seconds17 dBm

Fi channel 6+

asurements wuency: 2390

ervations:

With onof wobb

When bobserve

Fi channel 11

asurements wquency: 2390

ervations:

When oduring t

When boccurred

etup3

up 3 Tx anteone Bluetoo

Ts 6‐8 the IST

T was placedof the chamb

power level but recover(mean) / +1

+10

were made w0 MHz, Tx po

ly Wi‐Fi presbling). oth Wi‐Fi and.

1+7

were made w0 MHz, Tx po

nly Wi‐Fi wahe first 12 seboth Wi‐Fi ad.

nna 2 was inoth receiverTS was used a

d on top of thber. The pha

of +10 dBmred 23 secondBm (peak)

with Wi‐Fi ower: +10 dB

sent a few ar

nd TD‐LTE w

with Wi‐Fi ower: +10 dB

s present theconds and tand TD‐LTE

nstalled agair with inducas test signa

he support; tantom head r

27

the audio snds later. The.

only and withm).

rtefacts were

ere present

only and withm).

e audio signthen was lostwere prese

in. Four DUTctive loop (Dl, and for DU

the companioremained in

signal from te resulting LT

h Wi‐Fi plus

e observed (

no degradat

h Wi‐Fi plus

al from the rt permanentent severe d

Ts were meaDUT 7) and UT 9 a 1 kHz s

on device (areverse posi

he left chanTE signal pow

TD‐LTE signa

(two glitches

tion of the a

TD‐LTE signa

right channetly. degradation

asured, two one TV audsine tone.

udio streamition (back fa

nel was lostwer at the D

al Rome B2.3

s and a few i

audio signal

al Rome B2.3

l showed dis

of the aud

hearing aidsdio receiver

er) was placeacing the ant

t after 16 DUT was ‐

3 (centre

nstances

could be

3 (centre

sruptions

io signal

s (DUTs 6 (DUT 8).

ed at the tenna).

28

Figure 14: DUT 6 (Setup 3)

To assess the impact on HAS operation of a broadband signal at the upper edge of the 2.4 GHz band the FDD‐LTE frequency was set to 2486 MHz.

LTE signal characteristics: • Waveforms

o TD‐LTE: Rome B2.3 o FDD‐LTE: Ispra FDD UL

• Centre frequencies o Rome B2.3: 2390 MHz o Ispra FDD UL: 2486 MHz

• Transmit power levels o Rome B2.3: +10 dBm o Ispra FDD UL +14 dBm

• Resulting LTE signal power at the DUT o Rome B2.3:: ‐17 (mean), +1 dBm (peak) o Ispra FDD UL ‐18 (mean), ‐13 dBm (peak)

The first measurement was made with a TD‐LTE signal and an FDD‐LTE signal present at the same time.

During the first 35 seconds there were only minor artefacts audible. After that the audio signal quality was severely degraded.

For the next measurement a Wi‐Fi signal on channels 6+10 was added. LTE frequencies and signal levels were left unchanged.

Audio signal quality was severely degraded from the beginning with multiple glitches occurring on both channels.

The Wi‐Fi channel was then changed to 11+7. The audio from the right channel was lost permanently after 4 seconds. The audio signal

from the left channel remained stable with only minor degradation.

Finally, a measurement was made with only FDD‐LTE present. There was a very noticeable impact on the audio signal in the form of wobbling on both

channels and multiple glitches on the right channel.

DUT 7

The phaplaced o

With DUand FDD

LTE sign•

•

•

•

No d

DUT 8

For DUTchannel

The first

For the s

ntom head won top of the

UT 7 two meaD‐LTE presen

al characteriWaveforms

o TD‐Lo FDD

Centre frequo Romo Ispra

Transmit poo Romo Ispra

Resulting LTo Romo Ispra

egradation o

T 8 the prevs 6+10.

t measureme

No degradat

second meas

was reversede support, an

asurements t at the sam

istics:

LTE: RD‐LTE: Iuencies: me B2.3: 2a FDD UL: 2wer levels:

me B2.3: +a FDD UL +TE signal powme B2.3:: ‐a FDD UL ‐

of the audio

vious setup

ent was mad

tion of the a

surement an

d again so thnd the compa

Figure 15

were made, e time.

Rome B2.3 Ispra FDD UL

2390 MHz 2505 MHz

+10 dBm +14 dBm wer at the DU‐17 (mean), +‐18 (mean), ‐

signal qualit

was mainta

de with only W

udio signal q

n FDD‐LTE sig

29

hat it faced tanion device

5: DUTs 7 and 8

with Wi‐Fi o

L

UT: +1 dBm (pea‐13 dBm (pe

y was observ

ained. Durin

Wi‐Fi presen

quality was o

gnal was add

he LTE Tx ane was placed

8 (Setup 3)

only (Channe

ak) ak)

ved.

ng all measu

nt.

observed.

ded.

ntennas (Figuat the far en

el 6+10), and

urements W

ure 15). The nd of the cha

with Wi‐Fi, T

Wi‐Fi was pre

DUT was amber.

TD‐LTE,

esent on

• • •

For the t• • • •

Further and sign

DUT 9

The phaon top oin line o

For DUTperiodic

Centre frequTransmit poResulting LT

A number of

third measurSignal: RomeCentre frequTransmit poResulting LT

When the Ttransmit podropouts.

measuremenal levels.

When a stroa relatively sufficient to An increase in the imme

ntom head wof the supporf sight (LOS)

T 9 a 1 KHz sic pitch shift w

uency: 2505 wer level: +1E power at D

f short drop‐

rement the Fe B2.3 uency: 2390 wer level: +1E power at t

TD‐LTE signalwer was red

ents were co

ong FDD‐LTE weak TD‐LTcause dropoof the TD‐LTdiate and pe

was reversedrt; the comp.

ine wave wawhich occurr

MHz 14 dBm DUT: ‐18 dBm

‐outs in the a

FDD‐LTE sign

MHz 10 dBm the DUT: ‐17

l was switchduced to +5

onducted wi

signal was pTE signal (Txouts in the aTE transmit ermanent los

d again so thpanion device

as used as aured approxim

Figure

30

m (mean) /

audio signal

nal was repla

dBm (mean

hed on the a5 dBm the a

th various c

present (Tx px power: ‐2udio signal. power to +5ss of the aud

at the back we was placed

udio test signmately every

e 16: DUT 9 (Se

‐13 dBm (pe

occurred.

aced with a T

n) / +1 dBm

udio signal wudio signal

combinations

power: +14 d26 dBm, pow

5 dBm (powedio signal.

was facing thd at a distanc

nal. It was no2 seconds.

etup 3)

eak)

TD‐LTE signal

(peak)

was lost immwas receive

s of TD‐LTE

Bm, power awer at the

er at the DU

he antenna. ce of approxi

oticed that th

l.

mediately. Wed but suffer

and FDD‐LT

at the DUT: ‐DUT: ‐53 dB

T: ‐22 dBm)

The DUT waimately 2 me

his signal dis

When the red from

TE signals

‐18 dBm) Bm) was

resulted

s placed etres and

splayed a

Three mpresent

The firstbursts o

For the s• • •

Again, th

For the t

• • • •

As in the

7.6.4Se

A total o1 KHz sin

In each cbox at asignal lecould be

The LTE would b

For all Dand FDD

measuremenon channels

t measuremef low‐level n

second measCentre frequTransmit poResulting sig

here were se

third measur

Signal: RomeCentre frequTransmit poResulting si

e previous tw

etup4

of five DUTs,ne wave was

case the DUTa distance oevel was adjue observed (M

Tx antennase more susc

DUTs measurD‐LTE.

nts were mas 6+10.

ent was madnoise, approx

surement anuency: 2505 wer level: +1gnal power a

everal occurr

rement a TD

e B2.3 uency: 2390 wer level: +1gnal power a

wo measurem

, all hearing s used as aud

T was placedof approximausted to theMUS).

s were rotateptible to ho

rements wer

ade with th

de with onlyximately 0.3

n FDD‐LTE sigMHz 14 dBm at the DUT: ‐1

rences of sho

D‐LTE signal w

MHz 10 dBm at the DUT: ‐

ments severa

aids, were mdio test signa

d on top of thately 1 metre minimum le

ed by 90 degorizontally po

Figure 1

re made for v

31

his configura

y Wi‐Fi preses in length.

gnal was add

18 dBm (me

ort bursts of

was added so

‐17 dBm (me

al short burs

measured wital.

he support; tre from the evel at whic

grees into tholarised inte

17: DUTs 9‐13 (

various com

ation. Durin

ent. There w

ded.

ean) / ‐13 dB

low‐level no

o that all thre

ean) / +1 dB

sts of low‐lev

th this setup

the companiDUT (Figureh no degrad

he horizontaerfering signa

Setup 4)

binations of

g all measu

were several

Bm (peak)

oise.

ee interferer

Bm (peak)

vel noise wer

p (DUTs 9 – 1

ion device we 17). For eadation of the

al polarisatioals.

interference

urements W

occurrences

rs were activ

re observed.

13). For thes

was placed in ach DUT thee audio signa

n plane, as t

e from Wi‐Fi

Wi‐Fi was

s of short

ve.

e DUTs a

shielded e wanted al quality

the DUTs

, TD‐LTE,

32

• Interfering signals: o Wi‐Fi: Channel 6+10 o TD‐LTE: Rome B2.3 o FDD‐LTE: Ispra FDD UL

• Centre frequencies: o Rome B2.3 2390 MHz o Ispra FDD UL: 2505 MHz

Transmit power levels were varied between ‐10 dBm and +10 dBm for the TD‐LTE signal, and between ‐10 dBm and +14 dBm for the FDD‐LTE signal.

Several additional measurements were made with an FDD‐LTE signal centred at 2486 MHz.

DUT 9

Even in the absence of interference there was noticeable background noise indicating that the system was working at MUS level.

Four measurements were conducted with different configurations in which all active interferers transmitted at maximum power level.

When only Wi‐Fi was present the audio signal was lost after 33 seconds but recovered 3 seconds later.

With only TD‐LTE present the background noise level increased noticeably and short dropouts occurred frequently.

When TD‐LTE and Wi‐Fi were active the audio signal was lost temporarily. When TD‐LTE, FDD‐LTE, and Wi‐Fi were active the audio signal was lost permanently.

Further measurements were conducted with different combinations of interferers and transmit power levels.

At a TD‐LTE transmit power level of ‐10dBm there was no impairment of the audio signal, even with FD‐LTE operating at maximum power and Wi‐Fi active at the same time.

The audio signal started being disrupted when the TD‐LTE transmit power reached ‐8 dBm which corresponds to a signal level at the DUT of ‐35 dBm (mean) / ‐17 dBm (peak).

DUT 10


When only Wi‐Fi was present there was a strong increase in background noise, accompanied by frequent pitch shifts and dropouts.

When only FDD‐LTE was present audio quality started deteriorating at a transmit power level of ‐3 dBm. At maximum transmit power (signal level at the DUT: ‐18 dBm) the audio signal was lost immediately.

When both TD‐LTE and FDD‐LTE were present the audio link could only be maintained when the FDD‐LTE signal level at the DUT was lower than ‐40 dBm.

When TD‐LTE and Wi‐Fi were present the audio link broke down after a few seconds. It could only be maintained when the TD‐LTE power at the DUT was lower than ‐32 dBm.

Additional measurements were conducted with an FDD‐LTE signal at 2486 MHz and a maximum power level at the DUT of ‐34 dBm. No degradation of the audio signal quality was observed.

33

DUT 11


When only Wi‐Fi was present the audio signal was lost twice, for durations of 5 and 6 seconds.

When only TD‐LTE was present (Tx power = +10 dBm) the audio link was lost almost immediately.

When both TD‐LTE and FDD‐LTE were present the audio link could only be maintained when the FDD‐LTE signal level at the DUT was lower than ‐41 dBm and the TD‐LTE signal level was lower than ‐42 dBm.

When TD‐LTE, FDD‐LTE, and Wi‐Fi were present the audio link could only be maintained when the FDD‐LTE level was lower than ‐43 dBm and the TD‐LTE level was lower than ‐42 dBm.

Additional measurements were conducted with an FDD‐LTE signal at 2486 MHz.

The audio signal was lost immediately when the FDD‐LTE signal was switched on (signal level at the DUT: ‐18 dBm).

Audio quality started deteriorating at an FDD‐LTE signal level at the DUT of ‐38 dBm.

Although Wi‐Fi signals alone caused disruptions to the audio signal the addition of FDD‐LTE at 2486 MHz (signal level at the DUT below ‐38 dBm) appeared to improve the situation. No degradation of the audio signal could be observed, probably because the Wi‐Fi system adapted or the HAS completely avoided the upper part of the 2400 MHz band.

DUT 12


When only Wi‐Fi was present short disruptions and glitches were observed. When only FDD‐LTE was present audio quality started deteriorating at a transmit power

level of ‐3 dBm. At maximum transmit power (+14 dBm) the audio signal was lost immediately.

When both TD‐LTE (signal level at the DUT: ‐17 dBm) and FDD‐LTE were present the audio link could only be maintained when the FDD‐LTE level at the DUT was lower than ‐41 dBm.

When both TD‐LTE (signal level at the DUT: ‐17 dBm) and Wi‐Fi were present the audio link broke down immediately. The audio link could only be maintained when the TD‐LTE level was lower than ‐36 dBm.

When both FDD‐LTE and Wi‐Fi were present the audio link could only be maintained when the FDD‐LTE level was lower than ‐36 dBm.

DUT 13

Due to time constraints no audio recordings were made for DUT 13.

When only Wi‐Fi or FDD‐LTE (signal level at the DUT: ‐18 dBm) were present no degradation of the audio signal was observed.

When only TD ‐LTE (signal level at the DUT: ‐17 dBm) was present minor glitches were observed.

When any two of the interferers were present and transmitting at maximum power the audio signal was lost within 4 seconds.

7.6.5Se

This setusupport,receiversignal.

Two meactive.

•

•

•

•

No a

When both Tthe FDD‐LTEdBm. When TD‐LTwhen the FD

etup5

up was used, audio was r, was placed

easurements

Interfering so Wi‐Fo TD‐Lo FDD



Resulting sigo Romo Ispra

audible degr

TD‐LTE and E signal leve

TE, FDD‐LTE,DD‐LTE level

d to measuresupplied fro

d at the oppo

were condu

signals: Fi: CLTE: RD‐LTE: Iuencies: me B2.3: 2a FDD UL: 2wer levels

me B2.3: +a FDD UL: +gnal power ame B2.3: ‐a FDD UL: ‐

adation of th

FDD‐LTE werl at the DUT

, and Wi‐Fi was lower th

e DUT 14, a om a smartposite end of

Figure

ucted, one w

Channel 6+1Rome B2.3 Ispra FDD UL

2390 MHz 2505 MHz

+10 dBm +14 dBm at the DUT: ‐17 dBm (me‐18 dBm (me

he audio sign

34

re present thT was lower

were presenhan ‐36 dBm

remote micphone via cathe chambe

e 18: DUT 14 (Se

with only Wi‐

0

L

ean) / +1 dBean) / ‐13 d

nal could be

he audio linkr than ‐33 dB

nt the audiom and the TD‐

rophone. Thable, and thr (Figure 18)

etup 5)

Fi present, a

Bm (peak) Bm (peak)

observed.

k could only Bm and the

o link could ‐LTE level wa

he DUT was he companio). ISTS was u

and one with

be maintainTD‐LTE less

only be maas less than ‐

placed on toon device, a sed as the a

h all three in

ed when s than‐35

aintained ‐35 dBm.

op of the wireless udio test

terferers

7.6.6Se

This setuthe comDUT (Fig

Four me

1. 2. 3. 4.

•

•

•

•

No d

etup6

up was usedmpanion devigure 19). ISTS

easurements

Wi‐Fi‐only Wi‐Fi + TD‐LWi‐Fi + FDD‐Wi‐Fi + TD‐L





degradation

d to measureice, a smartpS was used a

s were condu

LTE ‐LTE LTE + FDD‐LT


me B2.3: +a FDD UL: +gnal power ame B2.3: ‐a FDD UL: ‐

of the audio

e DUT 15, a phone, was as the audio t

Figure

ucted with th

E


2390 MHz 2505 MHz

+10 dBm +14 dBm at the DUT: ‐17 dBm (me‐18 dBm (me

o signal could

35

hearing aid.placed at a test signal.

e 19: DUT 15 (Se

his setup:

0

L

ean) / +1 dBean) / ‐13 d

d be observe

The DUT wdistance of a

etup 6)

Bm (peak) Bm (peak)

ed.

as placed onapproximate

n top of the ely 1 metre f

support, from the

7.6.7Se

Five DUreceiverplaced a

For each

1. 2. 3. 4.

No d

An additLTE signpresent.

etup7

UTs were mers). In eachat the opposi

h DUT four m






Field strengt

degradation

tional experials were mo.

easured with case the Dite end of th

measurement



me B2.3: +a FDD UL: +gnal power ame B2.3: ‐a FDD UL: ‐th at the DUT

of the audio

iment was cooved inband,

h this setupDUT was place chamber (

Figure 20:

ts were cond

E


2390 MHz 2505 MHz

+10 dBm +14 dBm at the DUT: ‐17 dBm (me‐18 dBm (meT: 2.1 V/M (m

o signal could

onducted wiTD‐LTE to 24

36

p, DUTs 16‐1ced on top Figure 20). IS

DUTs 16‐19, 2

ducted:

0

L

ean) / +1 dBean) / ‐13 dmean) / 21 V

d be observe

ith DUT 20. T422 MHz and

18 (hearing of the suppSTS was used

2 (Setup 7)

Bm (peak) Bm (peak) V/m (peak)

ed during any

To test the rd FDD‐LTE to

aids) and Dport; the comd as the audi

y of these me

robustness oo 2475 MHz,

DUTs 19 andmpanion deio test signal

easurements

of the RF linkwith Wi‐Fi s

d 22 (RF vice was l.

s.

k the two still being

7.6.8Se

Three D(headph(smartp

For each

1. 2. 3. 4.

Audio qualitWi‐Fi througWhen DUT 2the monitorhopping signmoved by a10 Mbits/s. When TD‐LT

etup8

DUTs were hone). In eahone) was p

h DUT four m





ty deteriorateghput was re22 was testering spectrunal did not about 50 cm,

TE was added

measured wch case thelaced at a di

measurement



me B2.3: +a FDD UL: +

ed slightly. educed from ed with only m analyser avoid the ba Wi‐Fi starte

d, Wi‐Fi thro

with this see DUT was stance of ap

Figure 21: DU

ts were cond

E


2390 MHz 2505 MHz

+10 dBm +14 dBm

37

55 Mbits/s tWi‐Fi preseit could be nd occupieded working a

ughput dete

etup, DUTs placed on t

pproximately

UTs 20, 21, and

ducted:

0

L

to to 2 Mbitsnt the Wi‐Fi observed th

d by Wi‐Fi. Wagain but at

riorated furt

20 and 21top of the s 1 metre fro

d 23 (Setup 8)

s/s. link broke dhat the ALDWhen the coa reduced ra

ther, to less t

1 (hearings support; them the DUT (

down compleD system’s frompanion deate of appro

than 1 Mbit/

aids), and e companioFigure 21).

etely. On requency evice was oximately

/s.

DUT 23 n device

38

Resulting signal power at the DUT: o Rome B2.3: ‐17 dBm (mean) / +1 dBm (peak) o Ispra FDD UL: ‐18 dBm (mean) / ‐13 dBm (peak)

Field strength at the DUT: 2.1 V/M (mean) / 21 V/m (peak)

No degradation of the audio signal could be observed during any of these measurements.

7.7Analysis

The objective of this study is to assess the impact of adjacent band TD‐LTE and FDD‐LTE signals on the performance of HAS equipment operating in the 2.4 GHz band. For this purpose any degradation of the quality of the received audio signal should be identified that might have been caused by LTE and/or Wi‐Fi signals.

During the measurements the audio signals were recorded and monitored, in real time, for flaws such as glitches and dropouts. Subsequently, the recordings where aurally examined to identify more subtle degradations such as wobbling and trembling. It should be mentioned that audio quality varied considerably between DUTs, particularly in terms of background noise and frequency response. Further information on this subject is provided in Appendix B of this document.

We also tried to make a quantitative assessment of the degradation of signal quality caused by different levels of interference. For this purpose we analysed the ISTS recordings by means of an audio quality analyser software tool, namely AQuA [31].

During the process, however, we encountered a number of problems which have prevented us from obtaining conclusive results as of now.

Audio files may contain drop‐outs or show other signs of degradation at random locations. If the total duration of these impairments is short compared to the overall duration of the audio file they may not be considered critical for overall file quality by the audio analyser. As a consequence, the calculated MOS score for the degraded file may be not much lower than that of the original file although audio quality was severely degraded. In order to reflect the severity of such impairment more accurately it was proposed to split reference and test audio into short segments (e.g. of 10 seconds duration) and compare those separately.

AQuA features a large number of variable parameters to adapt its signal processing algorithms to the audio material and the test environment. In the time available for conducting this study it has not been not possible to fully understand and verify how variations of these parameters affect the results of the signal quality calculations. We therefore restricted our analysis to the qualitative method described below.

7.7.1QualityassessmentmethodologyDepending on the level of audio quality degradation we defined three impact categories: “Zero”, “Minor” and “Severe”.

Zero: No audible degradation

Minor: Up to two glitches or short dropouts, light wobbling or trembling, minor increase of background noise.

AQuA is a tool for end‐to‐end voice and audio quality testing.

It determines the Mean Opinion Score (MOS) and relative degradation of an audio signal by comparing a reference or source file and a (degraded) received file.

39

Severe: Temporary or permanent loss of signal, more than two glitches or short dropouts, strong wobbling or other distortions reducing speech intelligibility, strong increase in background noise.

For each category we determined the interferer signal levels at the location of the DUT. We then calculated the corresponding separation distances between DUT and interferer by applying the free‐space propagation model according to Friis.

For calculating the separation distances we had to take certain assumptions on the power transmitted by LTE User Equipment. In [28] we had characterised four different types of TD‐LTE UE and measured transmit power levels12 between ‐9 dBm and +20 dBm (median: +17 dBm). These measurements had been conducted during a file upload from the UE to the network when the UE was located deep indoors so that the base station signal was highly attenuated and the UE had to increase its transmit power to maintain the connection. In the calculations both cases were taken into accounts.

We further categorised the victim links as “Strong” or “Weak”. As the actual level of the wanted signal at the DUT was not measured the categories were defined as follows:

─ A link was considered “strong” when the signal was not artificially attenuated by placing the companion device in the shielded box.

─ Conversely, a link was considered “weak” when the companion device was placed in the shielded box.

An exception to this rule is DUT 6. As detailed in Section 7.6.1 the link between this device and its companion was found to be “weak” even under line‐of‐sight conditions.

7.7.2Results

Overall, 188 measurements were conducted. In 136 cases no degradation of the HAS audio signal was observed.

92 measurements were conducted with “strong” victim links. In eight cases (related to DUTs 8 and 9) degradation of the audio signal quality was observed.

96 measurements were conducted with “weak” victim links. In 44 cases degradation of the audio signal quality was observed.

Severe degradations occurred only when audio signals were transmitted over “weak” victim links. There was one exception, DUT 8 which reacted considerably more sensitive to the presence of interference than other DUTs. It is therefore quite possible that the system operated at minimum sensitivity and the link actually was “weak”.

A similar observation was made for minor degradations. These, too, occurred only when audio signals were transmitted over “weak” victim links. Again, there was one exception, in this case DUT 9, where audio signals recorded in the presence of interference showed several noise bursts. Although barely perceptible, these were taken into account as minor degradations.

The calculated separation distances (line‐of‐sight) for different combinations of interfering signals and LTE UE transmit power levels which resulted in degradation of the audio signal are provided in Table 5 and Table 613.

12 Mean power during transmission. 13 The centre frequencies are 2390 MHz for TD‐LTE and 2505 MHz for FDD‐LTE. The channel widths are 20 MHz for TD‐LTE and 10 MHz for FDD‐LTE.

40

If an LTE UE transmits at low output power (‐ 9 dBm in this case) a minor degradation of the HAS audio quality can occur if the UE is located within a short distance from the HAS (Table 5). When more than one interferer is active separation distances increase slightly but not in all cases.

For LTE UE transmitting at high output power (+20 dBm) a minor degradation of the HAS audio quality can occur if the UE is located within several metres from the HAS (Table 5). When more than one interferer is active separation distances increase in some but not all cases.

Audio degradation Minor

UE transmit power Minimum Maximum

Interferer

Equivalent separation distance [m] Equivalent separation distance [m]

TD‐LTE UE FDD‐LTE UE TD‐LTE UE FDD‐LTE UE

Range Median Range Median Range Median Range Median

TD‐LTE only 0.02 ‐ 0.03 0.03 0.6 ‐ 0.9 0.71

TD‐LTE + Wi‐Fi 0.22 ‐ 0.45 0.33 6.3 ‐ 12.6 9.43

FDD‐LTE only 0.03 ‐ 0.29 0.2 5.7 5.7

FDD‐LTE + Wi‐Fi 0.03 ‐ 0.23 0.13 0.8 ‐ 6.4 3.62

TD‐LTE + FDD‐LTE 0.03 ‐ 0.45 0.11 0.16 ‐ 0.41 0.38 0.7 ‐ 12.6 3.16 4.6 ‐ 11.4 10.83

TD‐LTE + FDD‐LTE + Wi‐Fi 0.03 ‐ 0.45 0.20 0.03 ‐ 0.51 0.23 0.7 ‐ 12.6 5.61 0.8 ‐ 14.4 6.44

Table 5: Equivalent separation distances for different levels of LTE UE Tx power and minor audio quality degradation

For severe degradation of audio signal quality to occur LTE UE transmitting at low output power (‐ 9 dBm) have to be located within a very short distance from the HAS (Table 6). When more than one interferer is active separation distances increase insignificantly.

For LTE UE transmitting at high output power (+20 dBm) severe degradation of the HAS audio quality can occur if the UE is located within approximately one metre from the HAS (Table 6). When more than one interferer is active separation distances can increase to a few metres.

Audio degradation Severe

UE transmit power Minimum Maximum

Interferer

Equivalent separation distance [m] Equivalent separation distance [m]

TD‐LTE UE FDD‐LTE UE TD‐LTE UE FDD‐LTE UE

Range Median Range Median Range Median Range Median

TD‐LTE only 0.03 0.03 0.71 0.71

TD‐LTE + Wi‐Fi 0.03 ‐ 0.04 0.03 0.71 ‐ 1.26 0.71

FDD‐LTE only 0.03 0.03 0.81 0.81

FDD‐LTE + Wi‐Fi 0.03 0.03 0.81 0.81

TD‐LTE + FDD‐LTE 0.03 0.03 0.71 0.03 0.03 0.71 0.81 0.81

TD‐LTE + FDD‐LTE + Wi‐Fi 0.03 ‐ 1.58 0.12 0.03 0.03 0.71 ‐ 44.6 3.43 0.81 0.81

Table 6: Equivalent separation distances for different levels of LTE UE Tx power and severe audio quality degradation

The Friis Free space equation is valid only in the far field region of the transmitting antenna. The very small separation distances calculated above should therefore be considered qualitative indications.

41

8.SummaryandconclusionsIn this study we examine the effects of adjacent‐band LTE signals on the quality of audio signals received by ALDs and hearing aids. For this purpose we conducted measurements with 21 devices from six major manufacturers in 23 different test configurations. Overall, 192 individual measurements were made. We focused on the effect of transmissions from LTE User Equipment (UE) operating in proximity of hearing aid systems.

We observed that when HAS receiver and transmitter were operating at a distance from each other that is representative of typical operating conditions almost all systems proved to be very robust against interference. Even in the presence of multiple high‐power in‐band interferers the HAS which all appeared to employ adaptive frequency hopping managed to maintain stable connections and provide distortion‐free audio. It could actually be observed how the (relatively) narrow‐band HAS signals moved to less‐interfered or unoccupied parts of the 2.4 GHz spectrum after an interfering signal had been activated. As the bit rate of digital hearing aids typically is around 64 kbits/s a few carriers appear to be sufficient to maintain a sufficient quality of service.

When HAS were operating near the receiver sensitivity level, i.e. when their RF signals were highly attenuated, the presence of strong adjacent‐band LTE signals resulted in degradation of the audio signals in a number of cases. Adding in‐band Wi‐Fi signals generally worsened the situation.

In combination, TD‐LTE and FDD‐LTE degraded victim signal quality slightly more than individually.

For adjacent‐band LTE signals to cause degradation of a HAS audio signal a number of conditions must be fulfilled:

The quality of the RF link between HAS transmitter and receiver is poor, i.e. the signal‐to‐noise ratio (SNR) at the receiver is low.

There is a nearby LTE UE transmitting continuously, e.g. during the upload of a large file to a remote base station.

The LTE UE is located close to the HAS receiver. Depending on the model of LTE UE the distance at which the audio signal is impaired can be between a few centimetres to 1 metre for severe degradation, and up to 11 metres for minor degradation. These values were calculated for free‐space / line‐of‐sight conditions.

LTE is operating at the band edges, i.e. 2390 MHz for TD‐LTE and 2505 MHz for FDD‐LTE14.

We also noted that the RF emissions from certain HAS models can severely degrade Wi‐Fi performance.

Overall, our findings are perfectly in line with those of the various previous studies. We conclude that while HAS audio signal quality can be impaired by adjacent‐band TD‐LTE signals the combination of prerequisites for this to happen makes the overall risk appear low. Furthermore, we conclude that the additional presence of FDD‐LTE UE signals in the 2.5 GHz band does not significantly increase the degradation of HAS audio quality.

14 Due to time constraints the impact of LTE signals further removed from the 2.4 GHz band edges could not be assessed. While previous studies considered only TD‐LTE and frequencies up to 2390 MHz the conditions created in this study correspond to worst‐case scenarios.

42

AppendixA–Listofmeasurementequipment

Tx antenna 1 Schwarzbeck BBHA 9120EGain at 2.4 GHz: 15 dBi / 13 dBi (4 m / 1 m)

Tx antenna 2 Schwarzbeck BBHA 9120DGain at 2.5 GHz: 10 dBi /10 dBi (4 m / 1 m)

Monitoring antenna Schwarzbeck ESLP 9145Gain at 2.4 ‐2.5 GHz: 7.2 dBi

Signal generator 1 Rohde & Schwarz SMU200AMax. output power: +20 dBm

Signal generator 2 Rohde & Schwarz SMBV100AMax. output power: +30 dBm

Spectrum analyser Rohde & Schwarz FSV

Audio DAC FocusRite Scarlett 2i2

Appen

The gratested A

ndixB–A

phs below dALDs and hea

Audiorec

depict the saring aids.

cordings

ignificant di

43

s

fferences in frequency response beetween somme of the

44

ListofTables

Table 1: List of tested devices ............................................................................................................... 15 Table 2: Observed artefacts and attributes of streamed audio signals [26] ........................................ 17 Table 3: Mean Opinion Score (MOS) .................................................................................................... 18 Table 4: Interfering signal characteristics ............................................................................................. 18 Table 5: Equivalent separation distances for different levels of LTE UE Tx power and minor audio quality degradation ............................................................................................................................... 40 Table 6: Equivalent separation distances for different levels of LTE UE Tx power and severe audio quality degradation ............................................................................................................................... 40

45

ListofFigures

Figure 1: Hearing aid manufacturers‘ global market shares (2014) [6] ................................................ 11 Figure 2: Wireless audio transmission to Gateway device ................................................................... 12 Figure 3: Direct wireless audio transmission to Hearing Aids .............................................................. 12 Figure 4: LTE and HAS frequency allocations between 2300 and 2690 MHz (Europe) ........................ 13 Figure 5: Proposed TD‐LTE frequency arrangement in the 2300‐2400 MHz band [23] ....................... 13 Figure 6: Frequency arrangement within the 2500‐2690 MHz band ................................................... 14 Figure 7: COLHA head ........................................................................................................................... 19 Figure 8: Initial HAS measurement setup inside the JRC’s anechoic chamber ..................................... 20 Figure 9: Initial HAS measurement setup (Setup 1) .............................................................................. 20 Figure 10: Final HAS measurement setup (Setups 2‐8) ........................................................................ 21 Figure 11: DUT 1 (Setup 1) .................................................................................................................... 22 Figure 12: DUT 6 (Setup 2) .................................................................................................................... 25 Figure 13: Distortion of the DUT 6 audio signal in the presence of Wi‐Fi and TD‐LTE (setup 2) .......... 26 Figure 14: DUT 6 (Setup 3) .................................................................................................................... 28 Figure 15: DUTs 7 and 8 (Setup 3)......................................................................................................... 29 Figure 16: DUT 9 (Setup 3) .................................................................................................................... 30 Figure 17: DUTs 9‐13 (Setup 4) ............................................................................................................. 31 Figure 18: DUT 14 (Setup 5) .................................................................................................................. 34 Figure 19: DUT 15 (Setup 6) .................................................................................................................. 35 Figure 20: DUTs 16‐19, 22 (Setup 7) ..................................................................................................... 36 Figure 21: DUTs 20, 21, and 23 (Setup 8) .............................................................................................. 37

46

Bibliography

[1] Copsey Communications Consultants, “ALD compatibility testing against 2.3GHz LTE, TDD signals at UK Ofcom Baldock June 2014,” 2014.

[2] Multiple Access Communications Ltd, “The Effect of TDD LTE Signals in the 2.3 to 2.4 GHz band on Bluetooth Equipment Operating in the 2.4 GHz ISM band,” May 2013.

[3] Cambridge Silicon Radio Limited, “Bluetooth Performance with 2.3 GHz LTE Interference, Report, Issue 5,” 2015.

[4] European Hearing Instrument Manufacturers Association (EHIMA), “Effects of LTE interference on Hearing Aids using Bluetooth Low Energy,” 20 October 2016.

[5] World Health Organization, “Deafness and hearing loss,” March 2015. [Online]. Available: http://www.who.int/mediacentre/factsheets/fs300/en/.

[6] GN Resound, “The Case 2015 IBCC,” 2015.

[7] World Health Organization, “1.1 billion people at risk of hearing loss,” 27 February 2015. [Online]. Available: http://www.who.int/mediacentre/news/releases/2015/ear‐care/en/.

[8] S. Hochheiser, “The History of Hearing Aids,” 22 July 2013. [Online]. Available: http://theinstitute.ieee.org/tech‐history/technology‐history/the‐history‐of‐hearing‐aids.

[9] Sivantos, Inc., “History of Siemens Hearing Aids & Accessories ‐ Siemens Hearing Aids,” 2017. [Online]. Available: https://usa.bestsoundtechnology.com/history/.

[10] ETSI, “ETSI TR 102 791 V1.2.1 (2013‐08) Electromagnetic compatibility and Radio spectrum Matters (ERM); System Reference Document; Short Range Devices (SRD); Technical characteristics of wireless aids for hearing impaired people operating in the VHF and UHF freq,” 2013.

[11] GN Hearing A/S, “Hearing aid systems,” September 2016.

[12] Sonic Innovations Inc., “Canal Styles (CIC, ITC, ITE),” 02 10 2015. [Online]. Available: http://www.sonici.com/Sonic/About‐hearing‐and‐hearing‐loss/About‐hearing‐aids/Types‐of‐hearing‐aids/Canal‐styles.aspx.

[13] Hearing Aid Know, “Hearing Aid Types, the different types or styles and their pros & cons,” 14 12 2015. [Online]. Available: https://www.hearingaidknow.com/hearing‐aid‐types.

[14] Springer Handbook of Auditory Research, Hearing Aids, vol. 56, B. C. M. R. R. F. A. N. P. Gerald R. Popelka, Ed., Springer International Publishing, 2016.

[15] Grand View Research, Inc., “Hearing Aids Market Size & Share | Global Industry Trends Report, 2024,” September 2014. [Online]. Available: http://www.grandviewresearch.com/industry‐analysis/hearing‐aids‐market.

[16] MarketsandMarkets, “Hearing Aids Market worth 8,373.9 Million USD by 2020,” [Online]. Available: http://www.marketsandmarkets.com/PressReleases/hearing‐aids.asp.

47

[17] C. W. (. Semiconductor), “Advanced Hearing Aids: Wearables You’ll Want to Wear,” Medical Design Technology (MDT), 7 April 2016. [Online]. Available: https://www.mdtmag.com/article/2016/04/advanced‐hearing‐aids‐wearables‐youll‐want‐wear.

[18] William Demant Holding, “Trends and directions in the hearing healthcare market,” 2013.

[19] The Hearing Review, “US Hearing Aid Unit Sales Grow by 10% in Q2 of 2016,” The Hearing Review, 13 July 2016. [Online]. Available: http://www.hearingreview.com/2016/07/us‐hearing‐aid‐unit‐sales‐grow‐10‐q2‐2016/.

[20] WiFore Wireless Consulting, “The Market for Hearable Devices 2016‐2020,” November 2016.

[21] European Hearing Instrument Manufacturer’s Association (EHIMA), “EHIMA/HIA TR 101 V2.2 (2010–03–04) Draft. Electromagnetic compatibility and radio spectrum matters (ERM); Hearing instrument RF interference analysis.,” 2010.

[22] GSA Global mobile Suppliers Association, “Evolution to LTE report,” 25 January 2016. [Online]. Available: http://gsacom.com/wp‐content/uploads/2016/01/160127‐Snapshot_LTE‐TDD_extract_GSA_Evolution_to_LTE_report.pdf.

[23] CEPT, “CEPT Report 55,” 28 November 2014.

[24] ETSI, “ETSI TS 136 101 V13.3.0 (2016‐05),” 2016.

[25] Electronic Communications Committee (ECC), “ECC Decision (05)05 Harmonised utilization of spectrum for Mobile/Fixed Communications Networks (MFCN) operating within the band 2500‐2690 MHz,” 2015.

[26] J. Ramsgaard, “Sound Quality in Hearing Aid Wireless Streaming Technologies,” The Hearing Review, 16 July 2016. [Online]. Available: http://www.hearingreview.com/2016/07/sound‐quality‐hearing‐aid‐wireless‐streaming‐technologies/.

[27] S. F. M. V. &. B. K. (. Inga Holube, “Development and analysis of an International Speech Test Signal (ISTS),” International Journal of Audiology, pp. 891‐903, 2010.

[28] European Commission ‐ Joint Research Centre, “JRC Study on Coexistence between 2.3 GHz TD‐LTE and 2.4 GHz Wi‐Fi – Final Report,” 2016.

[29] ETSI, ETSI TS 136 141 V13.2.0 (2016‐01), 2016.

[30] ETSI , “Evolved Universal Terrestrial Radio Access (E‐UTRA); Physical channels and modulation (3GPP TS 36.211 version 13.1.0 release 13),” 2016.

[31] Sevana Oü, “AQuA ‐ Sevana,” Sevana Oü, [Online]. Available: http://sevana.biz/products/aqua/.

[32] T. Higgins, “SmallNetBuilder's Router Market Share Report ‐ Q1 2015,” 2015 April 01. [Online]. Available: http://www.smallnetbuilder.com/lanwan/lanwan‐features/32666‐smallnetbuilders‐router‐market‐share‐report‐q1‐2015.

[33] E. O. R. S. S. Japertas, “Investigation of WI‐Fi indoor signals under LOS and NLOS conditions,” International Journal of Digital Information and Wireless Communications (IJDIWC), vol. 1, no. 2, pp. 26‐32, 2012.

48

[34] T. R. S. B. R. Jadhavar, “2.4 GHz Propagation Prediction Models for Indoor Wireless Communications within Building,” International Journal of Soft Computing and Engineering (IJSCE), vol. 2, no. 3, pp. 108‐113, July 2012.

[35] Aegis Systems Limited, “In‐home propagation Final report,” Ofcom, 30th June 2011.

[36] Cisco Systems, Inc., “Cisco Wireless Mesh Access Points, Design and Deployment Guide, Release 7.4,” 30 June 2013. [Online]. Available: http://www.cisco.com/c/en/us/td/docs/wireless/technology/mesh/7‐4/design/guide/mesh74.pdf.

[37] Y. Solahuddin and R. Mardeni, “Indoor empirical path loss prediction model for 2.4 GHz 802.11n network,” in Control System, Computing and Engineering (ICCSCE), 2011 IEEE International Conference on, 25‐27 Nov. 2011.

[38] L. (. A. Wilhelmsson, “Possible Indoor Channel Models for HEW System Simulations,” IEEE 802.11 HEW Study Group, 18 March 2014.

[39] Eurostat, “Housing conditions ‐ Statistics Explained,” Eurostat, 23 October 2015. [Online]. Available: http://ec.europa.eu/eurostat/statistics‐explained/index.php/Housing_conditions.

[40] Cisco Systems, Inc., “Site Survey Guidelines for WLAN Deployment,” 10 April 2013. [Online]. Available: http://www.cisco.com/c/en/us/support/docs/wireless/5500‐series‐wireless‐controllers/116057‐site‐survey‐guidelines‐wlan‐00.html#anc24.

[41] Ubee Interactive, “Understanding Technology Options for Deploying Wi‐Fi,” Ubee Interactive.

[42] R. T. a. C. R. Lucas DiCioccio, “Measuring Home Networks with HomeNet Profiler,” in Proceedings of the 14th international conference on Passive and Active Measurement (PAM'13), 2013.

[43] S. G. S. B. Ashish Patro, “Observing home wireless experience through WiFi APs,” in Proceedings of the 19th annual international conference on Mobile computing & networking (MobiCom '13), 2013.

[44] Aruba Networks, “RF Basics ‐ Part 1,” 2007. [Online]. Available: http://community.arubanetworks.com/aruba/attachments/aruba/tkb@tkb/121/1/RF‐Basics_Part1.pdf.

[45] RF Micro Devices, Inc. (RFMD), RF5755 2.4GHz TO 2.5GHz 802.11b/g/n WiFi FRONT END MODULE.

[46] IEEE, Part 11: Wireless LAN Medium Access Control (MAC) and Physical Layer (PHY) Specifications IEEE Std 802.11™‐2012, IEEE, 2012.

[47] Texas Instruments, Understanding and Enhancing Sensitivity in Receivers for Wireless Applications, 1999.

[48] D. A. W. David D. Coleman, “WLAN Trouble Shooting and Design,” John Wiley & Sons, Inc., Indianapolis, Indiana, 2014.

[49] iperf.fr, “iPerf ‐ The network bandwidth measurement tool,” [Online]. Available:

49

https://iperf.fr/iperf‐doc.php.

[50] ABI Research, “Installed Base of 802.15.4‐enabled Devices to Exceed 2 Billion by 2019; ZigBee to Face Increasing Competition,” 13 November 2013. [Online]. Available: https://www.abiresearch.com/press/installed‐base‐of‐802154‐enabled‐devices‐to‐exceed/.

[51] ABI Research, “With an Installed Base of 10 Billion Devices Expected in 2018, Bluetooth will be an Essential Tool for Building the Internet of Everything,” 6 August 2013. [Online]. Available: https://www.abiresearch.com/press/with‐an‐installed‐base‐of‐10‐billion‐devices‐expec/.

[52] Wi‐Fi Alliance, “Wi‐Fi® device shipments to surpass 15 billion by end of 2016,” [Online]. Available: http://www.wi‐fi.org/news‐events/newsroom/wi‐fi‐device‐shipments‐to‐surpass‐15‐billion‐by‐end‐of‐2016.

[53] eco Research, “Verbreitung und Nutzbarkeit von WLAN, WLAN‐Zugangspunkten sowie öffentlicher Hotspots in Deutschland,” November 2014.

[54] Ofcom UK, “Ofcom Technology Tracker Wave 1 2015,” 2015.

[55] IHS Inc., “Nine in 10 Global Broadband Households to Have Service Provider Wi‐Fi by 2019, IHS Says,” 25 June 2015. [Online]. Available: http://press.ihs.com/press‐release/technology/nine‐10‐global‐broadband‐households‐have‐service‐provider‐wi‐fi‐2019‐ihs‐sa.

[56] Ofcom UK, “The Communications Market 2015 Telecoms and networks,” 2015.

[57] IHS ‐ Infonetics Research, “Reliable WiFi for Multiscreen TV Is Powering Explosive Growth in WiFi,” 11 June 2015. [Online]. Available: http://www.infonetics.com/pr/2015/2H14‐Home‐Networking‐Devices‐Market.asp.

[58] Bundesministerium für Wirtschaft und Energie, “Freie Routerwahl,” 2016. [Online]. Available: http://www.bmwi.de/DE/Themen/Digitale‐Welt/Netzpolitik/freie‐routerwahl.html.

[59] Ofcom UK, “Public Sector Spectrum Release (PSSR): Technical coexistence issues for the 2.3 and 3.4 GHz award,” 2014. [Online]. Available: http://stakeholders.ofcom.org.uk/consultations/pssr‐2014/.

[60] R. S. Eldada Perahia, Next Generation Wireless LANs; Throughput, Robustness, and Reliability in 802.11n, Cambridge University Press, 2008.

[61] Cisco, “Dynamic Channel Assignment (DCA),” [Online]. Available: http://www.cisco.com/c/en/us/td/docs/wireless/controller/technotes/8‐3/b_RRM_White_Paper/b_RRM_White_Paper_chapter_0100.pdf.

[62] Cisco, “Understanding the Changing Mobile User: Gain Insights from Cisco’s Mobile Consumer Research,” 28 January 2014. [Online]. Available: https://communities.cisco.com/docs/DOC‐50412.

[63] The Hearing Review, “Hearing Aid Sales Increase by 7.2% in 2015 after Strong Q4 by Private Sector,” The Hearing Review, 25 January 2016. [Online]. Available: http://www.hearingreview.com/2016/01/hearing‐aid‐sales‐increase‐7‐2‐2015‐strong‐q4‐private‐sector/.

50

[64] ETSI, “ETSI TR 102 791 V1.2.1 (2013‐08),” 2013.

[65] Audicus, “Hearing Aid Timeline: Visual History,” [Online]. Available: https://www.audicus.com/hearing‐aid‐timeline‐visual‐history/.

[66] [Online].

[67] European Commission, “COMMISSION DECISION of 13 June 2008 on the harmonisation of the 2 500‐2 690 MHz frequency band for terrestrial systems capable of providing electronic communications services in the Community,” 2008.

PUBLIC DOCUMENT - Europa...more receivers, where each receiver can have a wired or wireless (inductive) connection to a hearing aid or be an integral part of a hearing aid. Hearing

Documents