Cochlear implants are devices designed to provide a measure of hearing to the deaf. It has been very successful in restoring partial hearing to profoundly deaf people. Most deaf individuals have lost the ability to translate sound into patterns of electric activity normally present on the 30,000 fibers of auditory nerve. Because these patterns are input to the brain that result in sound sensation, cochlear implant deliver electric stimuli to these fibers in an attempt to artificially elicit patterns of spike activity that mimic the pattern present in a normal-hearing ear. Due to the variability in performance among cochlear implant patients, it's becoming increasingly important to find ways to compare the response of implanted patient with normal response. This paper presents method of speech comparison by utilizing chaotic signal processing techniques for control and synchronization of patient response with actual response. Phase space features and Lyapunov exponents are used for this purpose, so that chaos synchronization could be attained using nonlinear nature of speech signal.