NOISE CANCELLATION DURING VOICE CALLS USING KALMAN FILTER

Abstract : Quality of the speech is one of the important concerns in communications via mobiles phones. In order to achieve uninterrupted communication, the quality of the speech signal should be clear, so that person in other side can hear properly and reply, this should happen on both side. Real time noise cancellation mechanism can be provided as important feature in mobiles phones to provide better quality speech signal at receiver end. All of us have experienced trying to make a mobile phone call from a noisy street, crowded restaurant or train station where the background noise can make it impossible to hear the incoming call as shown in Figure 1, where background noise is added with speech of the person at transmitter side. It can be worse when the person next to you in these situations is yelling into the receiver in an attempt to be heard. These noises can be eliminated by applying some noise cancellation techniques using suitable filters.

 EXISTING SYSTEM :

 • In existing system, Adaptive digital filters are used • It’s not applicable for real time process • Noise Cancellation System is available for android devices

 DISADVANTAGE :

 • Use of two microphones increases the delay • For processing it needs a higher end processor • In order to reduce the delay this Active Noise Cancellation technique is introduced

 PROPOSED SYSTEM :

  Kalman Filter minimizes the mean square error of estimated parameters during estimation prediction and update techniques are used. Kalman Filter is an optimal estimator, so it gives good results. Kalman Filter is a recursive data processing algorithm that estimates the state of noisy linear dynamic system. The methodology and terms used in Kalman Filter have their influence in the design of Kalman filter. The Kalman filter processes all available measurements to estimate the state both stationery and non-stationery process.

 ADVANTAGE :

 The performance of any speech signal processing system is degraded in the presence of noise (either additive or convolution). This is due to the acoustic mismatch between the speech features used to train and test this system and the ability of the acoustic models to describe the corrupted speech. When processing the speech signal, the quality of speech may be at risk from various sources of interference or distortions. sources of interference are: