Vector codebook design using gravitational search algorithm

In this paper Gravitational search algorithm (GSA) is efficiently used to design vector codebook process known as vector quantization along with Linde-Buzo-Gray (LBG) algorithm which provides initial codebook to it. The result of this proposed technique has been tested and compared with well known vector quantization methods based on particle swarm optimization (PSO) and firefly (FF) algorithm. Here efficiency has been valued in terms of parameters like peak signal to noise ratio (PSNR) and computational time needed by CPU. The empirical results clearly shows that LBG-GSA performs better comparison to LBG-PSO in terms of both the parameters. Even if there is a little improvement of PSNR as well as a little reduction of computational time observed in LBG-GSA comparison to LBG-FF gives an ample importance to this proposed LBG-GSA over others.