Computational and conditioning issues of a discrete model for cochlear sensorineural hypoacusia