Abhijit Mitra

Enhanced production of amylase by optimization of nutritional constituents using response surface methodology

Abstract Response surface methodology was employed to study the cumulative interactive effect of the macronutrients of the media and to optimize their concentration to enhance the production of maltooligosaccharide-forming amylase from Bacillus circulans GRS 313. A 23 factorial central composite experimental design was used to study the combined effect of the nutritional constituents like soybean meal, yeast extract and wheat bran. The p-value of the coefficient for linear effect of soybean meal concentration was found to be 0.081, suggesting that this was the principal experimental variable having the greatest effect on the production of maltooligosaccharide-forming amylase. The optimal combination of the media constituents for amylase production from the contour plots were: soybean meal =4.84 g /100 ml , yeast extract =1.58 g /100 ml , wheat bran =2.84 g /100 ml . The optimization of the media increased the amylase yield by 1.25 times.

Core Networks in the Flexgrid Era

This paper reviews the status of research into elastic optical networks and flexgrid, illustrating the benefits of new flexible technologies to allow higher network capacities and support superchannel flows. The paper assesses the potential for the various forms of elasticity and describes open questions of a currently active research area.

A block floating-point treatment to the LMS algorithm: efficient realization and a roundoff error analysis

An efficient scheme is presented for implementing the LMS-based transversal adaptive filter in block floating-point (BFP) format, which permits processing of data over a wide dynamic range, at temporal and hardware complexities significantly less than that of a floating-point processor. Appropriate BFP formats for both the data and the filter coefficients are adopted, taking care so that they remain invariant to interblock transition and weight updating operation, respectively. Care is also taken to prevent overflow during filtering, as well as weight updating processes jointly, by using a dynamic scaling of the data and a slightly reduced range for the step size, with the latter having only marginal effect on convergence speed. Extensions of the proposed scheme to the sign-sign LMS and the signed regressor LMS algorithms are taken up next, in order to reduce the processing time further. Finally, a roundoff error analysis of the proposed scheme under finite precision is carried out. It is shown that in the steady state, the quantization noise component in the output mean-square error depends on the step size both linearly and inversely. An optimum step size that minimizes this error is also found out.

A block floating-point realization of the gradient adaptive lattice filter

We present a novel scheme to implement the gradient adaptive lattice (GAL) algorithm using block floating point (BFP) arithmetic that permits processing of data over a wide dynamic range at a cost significantly less than that of a floating point (FP) processor. Appropriate formats for the input data, the prediction errors, and the reflection coefficients are adopted, taking care so that for the prediction errors and the reflection coefficients, they remain invariant to the respective order and time update processes. Care is also taken to prevent overflow during prediction error computation and reflection coefficient updating by using an appropriate exponent assignment algorithm and an upper bound on the step-size mantissa.

The NLMS algorithm in block floating-point format

We present a novel scheme to implement the normalized least mean square algorithm in block floating-point (BFP) format, which permits processing of data over a wide dynamic range, at a cost significantly less than that of a floating-point processor. Appropriate BFP formats for both the data and the filter coefficients are adopted. Care is taken so that the chosen formats remain invariant to interblock transition and weight-updating operation, respectively. Care is also taken to prevent overflow during filtering, as well as weight-updating processes, by using a dynamic scaling of the data and a slightly reduced range for the step size control parameter, with the latter having negligible effect on convergence speed.

Multiple-input–multiple-output channel modelling using multi-layer perceptron with finite impulse response and infinite impulse response synapses

Multiple-input-multiple-output (MIMO) wireless technology is a viable option likely to be able to meet the demands of the ever-expanding mobile networks. For MIMO system, channel estimation is still a challenging area due to several difficulties. Soft-computational approaches can be additions to the list of traditional methods of MIMO channel modelling primarily because these tools, for their ability to learn, are better placed to use channel side information for improved performance. One of the viable means of such innovative channel estimation is the use of the artificial neural network (ANN) in a feedforward format known as multi-layer perceptron (MLP). But as these ANNs prove to be suitable for static and slowly varying cases, time-varying MIMO channels are modelled using modified MLP with temporal attributes developed using finite impulse response (FIR) and infinite impulse response (IIR) blocks in place of the synaptic links. Six sub-classes of each of the FIR-MLP and the IIR-MLP are formulated, which show better performance than the conventional MLP in modelling the MIMO channels.

A Low Power Architecture of Digital Sinusoid Generator using Cubic Spline Interpolation

Evaluation of basic functions like sinusoids with hardware/firmware combination is important in many applications for proper implementation of systems. Such a scheme with low power approach is proposed in this paper with cubic spline interpolation. In particular, two different approaches are proposed to find the interpolating polynomial of sin (x) within the range [–π/2, π/2]–one with only one fixed data point and the other with two. The architectures for these are also proposed, which are shown to exhibit flexibility of implementation with low power requirement.

Recurrent Fuzzy-Neural MIMO Channel Modeling

Abstract. Fuzzy systems and artificial neural networks (ANN), as important components of soft-computation, can be applied together to model uncertainty. A composite block of the fuzzy system and the ANN shares a mutually beneficial association resulting in enhanced performance with smaller networks. It makes them suitable for application with time-varying multi-input multi-output (MIMO) channel modeling enabling such a system to track minute variations in propagation conditions. Here we propose a fuzzy neural system (FNS) using a fuzzy time delay fully recurrent neural network (FTDFRNN) that has the capability to tackle time-varying inputs in fuzzified form and is used to model MIMO channels. The inference engine is constituted by novel FTDFRNN blocks which determine the decision boundaries and tracks in-phase and quadrature components of input signals encompassing stochastic behavior of the MIMO channel. The system shows significant improvement in performance compared to statistical and ANN approaches in terms of faster processing time, lower bit error rate (BER) margins and better precision while carrying out symbol recovery of transmitted data through severely faded MIMO channels.

An efficient block floating point implementation of the LMS algorithm

An efficient scheme is presented for implementing the LMS-based transversal adaptive filter in block floating point (BFP) format which permits processing of data over a wide dynamic range at a processor cost marginally higher than that of a fixed point processor. Appropriate BFP formats for both the data and the filter coefficients have been adopted and adjustments made in filtering as well as weight updating operations in order to sustain the adopted format and also to prevent overflow in both these operations jointly. For the presented method to work properly, the algorithm step size is to be chosen below an upper limit, which is, however, not very restrictive when compared with the upper bound for convergence, thereby having marginal effect on convergence speed.

Level Crossing Rate in Land Mobile Satellite Channel with a Modified Nakagami-Lognormal Distribution

A modified Nakagami-lognormal distribution has been proposed to represent statistical nature of land mobile satellite (LMS) channel. The normal or Gaussian variable associated with lognormal distribution is taken to be correlated with its time derivative. It does not change the probability density function but modifies the expression of level crossing rate (LCR), which is derived in this paper. The proposed model takes the conventional Nakagami-lognormal model as its special case which can be achieved by setting the correlation coefficient to zero. This correlation coefficient provides an extra degree of freedom in modeling the LMS channel. The effect of this correlation coefficient and other important parameters on LCR, computed from derived expression, has been depicted with LCR curves.