Shigenori Kinjo

A design of FIR filter using CSD with minimum number of registers

This paper proposes an algorithm to be used for the VLSI design of a FIR filter using a canonical signed digit (CSD) representation with a minimum number of registers. The coefficients represented by CSD have a common digit pattern, so that they can reduce the number of adders to calculate same digit patterns. The proposed method takes advantage of this technique without using more registers than that of the transfer function.

Time domain channel estimation schemes for OFDM systems with multiple antenna transmissions

In this report, time domain channel estimation schemes for OFDM systems are studied. It is known that the time domain maximum likelihood estimation (MLE) scheme out-performs the conventional frequency domain channel estimation; however, the performance degradation becomes noticeable at low SNRs. Although we proposed a singular-value truncation approach for the MLE to improve the performance, we have been required more performance improvement in noisy environment. A minimum mean square error (MMSE) estimation is a promising technique to improve the performance in the low SNRs. However, we need considerable efforts to carry out the time domain MMSE channel estimation because it requires statistical information such as power delay profiles of MIMO channels in advance. Concerning this issue, we propose to construct the covariance matrix of the channels by using the estimated power delay profiles given by the MLE in every frames, and solve a normal equation by using the conjugate gradient method (CGM). Simulation results will show some packet error rate performances of an MIMO-OFDM system when we apply these channel estimation schemes.

A new cost function allowing parallel processing of adaptive filters

An alias free parallel structure for adaptive digital filters (ADFs) is considered. We propose a new cost function for parallel ADFs. The limiting value analysis of system identification using proposed cost function is given in stochastic sense. The cost function is defined in each bin; accordingly, it enables the parallel processing of ADFs.<<ETX>>

Linear programming design of two-channel perfect-reconstruction biorthogonal filter banks-linear phase and low-delay

This paper discusses the use of a linear programming technique for the design of two-channel perfect reconstruction (PR) biorthogonal FIR filter banks with linear phase or low-delay. Both the objective function and the design constraints for the PR filter banks are formulated as a set of linear equations and inequalities. The method is simple and good stop-band attenuation can be obtained. Two design examples are given to illustrate the effectiveness of the proposed method.

A subband adaptive filter with the optimum analysis filter bank

Conventional subband ADFs (adaptive digital filters) using filter banks have shown degradation in performance because of the non-ideal nature of filters. For this problem, we propose a new type of subband ADF incorporating two types of analysis filter bank. In this paper, we show that we can design the optimum filter bank which minimizes the LMSE (least mean squared error). In other words, we can design a subband ADF with less MSE than that of conventional subband ADFs.

A subband adaptive filter with two types of analysis filter bank

Conventional subband ADFs (adaptive digital filters) using filter banks have shown degradation in performance because of the non-ideal nature of filters. For this problem, we propose a new type of subband ADF incorporating two types of analyses filter bank. We show that we can design the optimum filter bank which minimizes the LMSE (least mean squared error). In other words, we can design a subband ADF with less MSE than that of conventional subband ADFs.<<ETX>>

A new MMSE channel estimation algorithm for OFDM systems

In this report, we propose a new MMSE channel estimation algorithm for OFDM systems. It is well-known that time domain maximum likelihood estimators (MSEs) show highly accurate impulse response estimation by using the time domain long preamble of the OFDM frame. On the other hand, the impulse response estimation based on the minimum mean square error (MMSE) criterion achieves superior channel estimation in low SNR conditions; however, it requires prior statistical information such as delay profiles of channels. In addition, the computational complexity becomes large because of the inverse matrix calculation. In order to overcome these issues, we propose to estimate the power delay profile of the channel by using the MLE, and to suppress the increase of the computational complexity for the matrix inverse calculation by applying the conjugate gradient method without degradation of the PER performance.

An over-sampling subband adaptive filter with the optimal real filter bank

The over-sampling subband ADFs are in high demand because of shortening the convergence of adaptation algorithms and being able to avoid aliasing. However, the conventional methods must use the complex coefficient filter banks, which require a lot of computation compared to the real filter banks. The aim of this correspondence is to develop a new oversampling subband ADF which does not require complex coefficient filter banks but real filter banks. The filter banks are optimized such that the mean square error can be minimized. It turns out that the proposed scheme can minimize the aliasing effect due to decimation.

A study on reduced MLD utilizing the QR decomposition for MIMO communication systems

Reduced maximum likelihood detection (R-MLD) has been proposed as an MIMO decoder. It was reported that the R-MLD showed equivalent performance with a full MLD under the symbol level detection for 2×2 MIMO systems with reduced computational complexity. However, the computational load becomes extremely large when we apply the idea directly to MIMO systems with three transmission antennas or more. In this paper, we study to apply the idea of the R-MLD to MIMO systems with three transmission antennas or more, and show an example of 4×4 MIMO system. We apply the QR decomposition to a 4×4 MIMO channel matrix, and divide the upper triangle matrix into four 2×2 matrices including one zero matrix so that we can apply the R-MLD scheme hierarchically. Numerical simulations show that allowable packet error rate are achieved with less computational complexity.

A new two-dimensional parallel block adaptive filter with reduced computational complexity

Two-dimensional (2-D) adaptive digital filters (ADFs) for 2-D signal processing have become a fascinating area of the adaptive signal processing. However, conventional 2-D FIR ADFs require a lot of computations. For example, the TDLMS requires 2N/sup 2/ multiplications per pixel. We propose a new 2-D adaptive filter using FFTs. The proposed adaptive filter carries out fast convolution using the overlap-save method, and has a parallel structure. Thus, we can reduce the computational complexity to O(log/sub 2/N) per pixel.