At virtually every public venue, announcements for visitors are made via a public addressing system. It is important that announcements transmitted by means of a public addressing system are not only audible, but also well understood by the general public. This thesis covers one of three subsystems required to make
a product that is capable of improving intelligibility of speech based on noise statistics in a room. Moreover, these statistics allow for an automatic volume control in order for listeners to experience an improved audio level. Therefore, this thesis aims at estimating the noise statistics in real time, while prior knowledge on the distorting announcement signal is available. Consequently, the concept of adaptive filtering deems suiting and is extensively studied. In order to meet the real time processing constraint, members of least mean squares algorithms are examined. Therefore, the method of steepest decent is covered, leading to the expounding of the traditional least mean squares (LMS) algorithm and the normalized LMS (NLMS) algorithm.
By assessment of the algorithms regarding different step sizes and filter lengths, the NLMS algorithm is shown to be superior in terms of faster convergence speed and better stability characteristics considering similar conditions for both algorithms. Subsequently, the results show that the NLMS is able to estimate the
noise in noise-to-signal ratio’s higher than -15dB. Also, its low complexity allows it to be suitable for real time applications, hence meeting the requirements.

Several algorithms to enhance the intelligibility of speech in near-end noise were analyzed and implemented. The algorithms considered were assessed based on the intrusive instrumental intelligibility metric SIIB_Gauss. An implementation based on the direct optimization for this metric is assessed, as well as an implementation based on human induced speech modifications, including increased sound intensity, flattening of the spectral tilt, increased vowel duration and increased consonant-vowel ratio. Another implemented algorithm is the amplification of the transient component of speech. Results show that for increased vowel duration a decrease in intelligibility was found in SIIB_Gauss value as well as in informal listening tests. The other implementations did show an increase in intelligibility according to SIIB_Gauss at SNRs between -4 dB and 6 dB in both stationary and fluctuating noise, under a power constraint. Finally, the implementations were combined into a system that automatically selects the optimal algorithm to use under the given noise conditions. It is shown that this combined system is able to increase intelligibility of speech in the presence of non-fluctuating noise, fluctuating noise, speech shaped noise, and competing speaker noise.