**1. Introduction**

The recognition and processing of speech signal plays a key role in any real time implementation. The plethora of approaches has been followed to analyze the continuous speech spectrum namely Perceptual Linear Prediction (PLP) statistical approach, Relative Spectral (RASTA) method, Mel-frequency cepstral analysis (MFCC), hidden Markov models (HMM), artificial neural networks (ANN), etc. The speech signal processing can be further enhanced by filtering out the noisy environments using single channel or multi-channel methods.

Radek et al., [1, 2] have incorporated the multi-channel methods using least mean square algorithm (LMS) and independent component analysis (ICA) to provide the mathematical calculations to avoid additive noise in speech signal. For smart home implementation, the LabVIEW SW tool is used for visualization, speech recognition, virtual cable connection to the sound card, and the actual mathematical calculations within additive noise canceling. The speech recognition engine will provide communication between LabVIEW and the recognizer. Its function is to convert input command into stream of text outputs. The output will be compared with the recognized command and the appropriate command will be given to ON/OFF the devices like Fan, Television, Washing Machine, Vacuum Cleaner, Dish Washer, etc. in the home. The Virtual Audio Cable (VA – Cable) is a software used for communicating between the devices using the audio streams.

LabVIEW is capable of executing/implementing many parallel loops on FPGA and real-time controllers by using a set of complex math functions. To make it more powerful, it has been integrated with LabVIEW SoftMotion Module and Kollmorgen AKD drives and motors to control the smart machines in all aspects. It also helps in the maintenance of smart machines after their deployment in the world by implementing centralized resource management, diagnostics, and remote updates [3].
