This document sets out a detailed technical study into the coexistence of 4G mobile phones and hearing aids and associated accessories.
This work has been carried out ahead of the award of new airwaves for mobile broadband in the 2.3 GHz band (2350-2390 MHz). This band sits near the licence exempt 2.4 GHz band, which is used by some hearing aids, radio aids, and sound processors used for cochlear implants - together referred to as assistive listening devices (ALDs).
We also ran tests using a 4G signal in the 10 MHz ‘guard band’ between the 2.3 GHz frequencies we will be auctioning and the spectrum band used by ALDs. Even though this block will not be part of our upcoming award in the UK it may be used in other countries.
Compatibility of 2.3 GHz 4G mobile with Assistive Listening Devices (PDF, 1.6 MB)
Executive summary in Welsh (PDF, 88.8 KB)
To investigate the impact of 2.3 GHz 4G mobile phones on ALDs, we focused on evaluating any influence on the audio quality of the speech file transmitted over the radio link of the ALD systems. During the testing, over 200 audio files were recorded and assessed.
The following ten example files are published as a complement to test results and analysis presented in the report.
File group A contains five audio samples referred to as outliers in sections 4 and section 5, representative of some variations in the audio quality potentially due to one of several reasons including 4G interference. To highlight the effect of interference, each sample file was edited by concatenating 10 seconds of the baseline file (recorded without the 4G signal) with 10 seconds of the test file (recorded with the 4G signal signal).
These samples are examples of a wide range of audio qualities observed in our test programme. Each sample files lasts for 10 seconds.
File group C has only one sample file, denoting the largest degree of variation in the audio quality between the left and the right ear for a baseline file, for which POLQA gave a delta MOS of 0.94. The audio samples last for 10 seconds as well.
