Pre-processing of Speech Signal serves various purposes in any speech processing application. It includes Noise Removal, Endpoint Detection, Pre-emphasis, Framing, Windowing, Echo Canceling etc. Out of these, silence portion removal along with endpoint detection is the fundamental step for applications like Speech and Speaker Recognition. The proposed method uses multi-layerperceptronalong with hierarchal mixturemodel for classification of voiced part of a speech from silence/unvoiced part. The work shows better end point detection as well as silence removal.The study is based on timing-based traffic analysis attacks thatcan be used to reconstruct the communication on end-to-end voip systems by taking advantage of the reduction or suppression of the generation of traffic whenever the sender detects a voice inactivity. Removal of the silence in between the inter packet time in order to obtain a clear communication. The proposed attacks can detect speakers of encrypted speech communications with high probabilities. With the help of different codecs we are going to detect speakers of speech communications In comparison with traditional traffic analysis attacks, the proposed traffic analysis attacks do not require simultaneous accesses to one traffic flow of interest at both sides.