Supervised Neural Fuzzy Schemes in Video Transmission over Bluetooth

Abstract

This paper is concerned with an intelligent application of Moving Picture Expert Group (MPEG) video transmission in a Bluetooth link. MPEG Variable Bit Rate (VBR) video is data hungry and results in excessive data loss and time delay over wireless networks. In a Bluetooth channel transmission rate is unpredictable as result of interferences by channel noises or other close by wireless devices. Therefore, it is difficult to transmit MPEG VBR video in Bluetooth without data loss and/or time delay. A buffer entitled ‘traffic regulating buffer’ is introduced before the Bluetooth protocol stack to prevent excessive overflow of MPEG video data over the network. Two novel neural fuzzy schemes are presented in this paper to monitor the traffic regulating buffer input and output rates, in order to ensure the video stream conforms to the traffic conditions of Bluetooth. The computer simulation results demonstrate that the two supervised neural fuzzy schemes reduce standard deviation and data loss, when compared with a traditional MPEG VBR video transmission over Bluetooth.KeywordsVideo TransmissionInternal Model ControlGeneral Packet Radio ServiceAdaptive Neural Fuzzy Inference SystemMove Picture Expert GroupThese keywords were added by machine and not by the authors. This process is experimental and the keywords may be updated as the learning algorithm improves.