Since the first days of telecommunication networks, service providers wanted to know beforehand if their planned network reaches the requirement specification. It turned out that these requirement specifications were heavily related to the consumer needs, and that the characteristics of the telecommunication system have to be adapted to the users' behavior in the communication situation. The quality the user experiences when using the system can ultimately only be investigated by observing the user behavior and asking for the user opinion.
In principle, there are two types of models for predicting speech communication quality of modern telecommunication systems. On the one hand, there are planning models like the E-model or SUBMOD, which can be used for transmission planning in order to have sufficient end-to-end transmission quality for the users. They allow network planners to estimate the overall quality of a telephone call done within a future network. The models will come up with a single value and allow the identification of quality constraints.
On the other hand, signal based models like PEAQ, PEMO-Q, or PESQ are employed to instrumentally predict the auditory speech quality for communication channels. Signal based models estimate speech quality by comparing the degraded output signal with the clean input signal. They are mainly used for comparative measurements. Such comparative measures are very effective when new codecs (speech coding/decoding algorithm) or other new algorithms have to be tested and compared to standards whose performances are known, and when auditory tests are too time-consuming or expensive. For example, they will be used to find out the optimum parameter settings for a new codec, and to get an impression of the achievable quality in comparison to competing products already on the market.
So far, all of these models have been set up and optimized with normal hearing test persons. Our goal here is to extend one parameter based and one signal based model towards hearing impaired persons. These models and the work to be done to extend the models will help to set up new normative criteria and guidelines. They will help to improve standard and special telephone equipment and telephone transmission systems for the usage by hearing impaired persons.