|Dimensions||2500μm x 2500μm|
|Power||0.045 mW, 1.2V|
One of the major problems of modern hearing aids is the poor speech intelligibility in noisy environments. First key point to increase the actual Signal-to-Noise Ratio (SNR) is the choice of an application-oriented algorithm. Unfortunately, many proposed Digital Noise Reduction (DNR) and DSE (Digital Speech Enhancement) approaches tend to generate some kind of distortion, are excessively complicated to be implemented in hearing aids, or introduce a large latency. All this can not be accepted by hearing-impaired people.
The most effective way to improve the signal quality is still the separation of the speech from the noise prior to the main signal processing. This can be achieved by means of adaptive directional microphones, exploiting the spatial diversity of signal and noise.
The practical idea implemented during this semester thesis project is to combine the signals from two electro-acoustic transducers such, as to obtain an overall directional characteristic that attenuates the disturbing noise.
Thanks to a careful optimization across different abstraction levels, the final design is very compact and attains outstanding performance in terms of SNR degradation and power consumption.