Quantization of Multiple Sources Using Modified Optimum Nonnegative Integer Adaptive Bit Allocation

The optimum bit allocation (OBA) problem was first developed by Huang and Schultheiss in 1963. In 1966 Fox proposed an algorithm for solving the optimum nonnegative integer bit allocation (ONIBA) problem. In 2006 Farber and Zeger proposed another ONIBA algorithm with less complexity. This algorithm is not valid if variances of some samples of the signal are zero. Also this algorithm is proposed for a static bit allocation and the algorithm does not work adaptively. No simulation was proposed for this algorithm. In this paper we first extend the algorithm for the case that variances of some samples are zero. Then we make this algorithm adaptive using signal power as an adaptive estimation for variances of samples of signal. Finally we simulate the algorithm and compare the performance of the algorithm with the PCM method. Also we found an error in Farber and Zeger algorithm during the simulation and corrected it.

کلیدواژه ها:

optimum nonnegative integer bit allocation ، adaptive bit allocation