In this study the Non-Linear Ultrasonic Sideband Peak Count-Index (SPC-I) technique is used as the foundation for anovel approach towards acoustic source localization (ASL) in orthotropic composite plates. The SPC-I based technique proposed here does not require the signal attenuation information or any knowledge on the time of arrival of the signal. It should be noted that since individual sensors can have varying sensitivities, the signal attenuation measured from the recorded signal amplitude is not very reliable. In addition, it is not necessary to have any prior knowledge of the mechanical properties of the composite plate material. All these are achievable by attaching 25 sensors that are well-scattered on the surface of the plate. The signals that are generated by an acoustic source are recorded by these 25 sensors. The recorded signals are then analyzed to derive the SPC-I value for each signal. The calculated SPC-I values are run through an optimization algorithm to predict the acoustic source location. Such localization is possible because the composite plate is inherently a non-linear material. Hence, as the signal travels longer distances through a composite plate, the recorded signal should show increasing level of distortion due to material non-linearity and dispersion. This phenomenon manifests itself primarily as a consequence of signal scattering and frequency modulation. Because of this, the phenomena of increasing distortion in the signal with increasing propagation distance can be exploited and utilized to predict the location of the acoustic source by solely utilizing the SPC-I values. This acoustic source localization technique is experimentally verified on a Carbon Fiber Reinforced (CFR) composite plate of dimension 500 mm x 500 mm with a thickness of 1 mm. The experimental results confirmed the feasibility of the proposed technique.