Numerous variable tap-length algorithms can be found in some literature and few strategies are derived from a basic theoretical formula. Thus, some algorithms lack of theoretical depth and their performance are unstable. In view of this point, the novel variable tap-length algorithm which is based on the mixed error cost function is presented in this letter. By analyzing the mixed expectation of the prior and the posterior error, the novel variable tap-length strategy is derived. The proposed algorithm has a more valid proximity to the optimal tap-length and a good convergence ability by the performance analysis. It can solve many deficiencies comprising large fluctuations of the tap-length, the high complexity and the weak steady-state ability. Simulation results demonstrate that the proposed algorithm equips good performance.

Pose estimation is a basic requirement for the autonomous behavior of robots. In this article we present a robust and fast visual odometry method to obtain camera poses by using RGB-D images. We first propose a motion estimation method based on sparse geometric constraint and derive the analytic Jacobian of the geometric cost function to improve the convergence performance, then we use our motion estimation method to replace the tracking thread in ORB-SLAM for improving its runtime performance. Experimental results show that our method is twice faster than ORB-SLAM while keeping the similar accuracy.

In this letter, a new semi-blind approach incorporating the bounded nature of communication sources with the distance between the equalizer outputs and the training sequence is proposed. By utilizing the sparsity property of l1-norm cost function, the proposed algorithm can outperform the semi-blind method based on higher-order statistics (HOS) criterion especially for transmitting sources with non-constant modulus. Experimental results demonstrate that the proposed method shows superior performance over the HOS based semi-blind method and the classical training-based method for QPSK and 16QAM sources equalization. While for 64QAM signal inputs, the proposed algorithm exhibits its superiority in low signal-to-noise-ratio (SNR) conditions compared with the training-based method.

In the impulse noise removal from a color image, vector filters are suitable for suppressing false color generation. However, the vector filters do not select optimal vectors to restore noise corrupted pixels. To cope with this problem, a cost function-based vector filter is proposed in this letter.

Code division multiple access (CDMA) technique is used widely since it can flexibly support multi-rate multi-media services by changing the number of orthogonal spreading codes. In this paper, we present a new adaptive code assignment algorithm, which consists of three steps: reserved-space, improved-crowded-first-space, and multi-code combination to fully use the code space. Compared with the existing algorithms, the proposed algorithm can avoid the code blocking problem and lower its total blocking probability while keeping its computational complexity relatively low. Simulation results show that increasing the free space reduces the average total blocking probability while increasing the blocking probability of high rate users.

In the problem of determining the major frequency components of a signal disturbed by noise, a model selection criterion has been proposed. In this paper, the criterion has been extended to cover a penalized cost function that yields a componentwise shrinkage estimator, and it exhibited a consistent model selection when the proposed criterion was used. Then, a simple numerical simulation was conducted, and it was found that the proposed criterion with an empirically estimated componentwise shrinkage estimator outperforms the original criterion.

The normalized least mean square (NLMS) algorithm has the drawback that the convergence speed of adaptive filter coefficients decreases when the reference signal has high auto-correlation. A technique to improve the convergence speed is to apply the decorrelated reference signal to the calculation of the gradient defined in the NLMS algorithm. So far, only the effect of the improvement is experimentally examined. The convergence property of the adaptive algorithm to which the technique is applied is not analized yet enough. This paper first defines a cost function properly representing the criterion to estimate the coefficients of adaptive filter. The name given in this paper to the adaptive algorithm exploiting the decorrelated reference signal, 'normalized least mean EE' algorithm, exactly expresses the criterion. This adaptive algorithm estimates the coefficients so as to minimize the product of E and E' that are the differences between the responses of the unknown system and the adaptive filter to the original and the decorrelated reference signals, respectively. By using the cost function, this paper second specifies the convergence condition of the normalized least mean EE' algorithm and finally presents computer simulations, which are calculated using real speech signal, to demonstrate the validity of the convergence condition.

We previously proposed a robust hybrid edge detector which relaxes the trade off between robustess against noise and accurate localization of the edges. This hybrid detector separates the tasks of localization and noise suppresion between two sub-detectors. In this paper, we present an extension to this hybrid detector to determine its optimal parameters, independently of the scene. This extension defines a probabilistic cost function using for criteria the probability of missing an edge buried in noise and the probability of detecting false edges. The optimization of this cost function allows the automatic selection of the parameters of the hybrid edge detector given the height of the minimum edge to be detected and the variance of the noise, σ2n. The results were applied to the 2D case and the performance of the adaptive hybrid detector was compared to other detectors.

The multiple registration schemes (MRSs) proposed here are classified into 3 cases by combining five registration schemes which are power up registration scheme (PURS), power down registration scheme (PDRS), zone based registration scheme (ZBRS), distance based registration scheme (DBRS), and implicit registration scheme (IRS) as follows: the first is MRS1 which covers PURS, PDRS, and ZBRS; the second is MRS2 which covers PURS, PDRS, and DBRS; the third is MRS3 which covers PURS, PDRS, IRS, and DBRS. The three proposed schemes are compared each other by analyzing their combined signaling traffic of paging and registration with considering various parameters of a mobile station behavior (unencumbered call duration, power up and down rate, velocity, etc.). Also, we derive allowable location areas from which the optimal location area is obtained. Numerical results show that MRS3 yields better performance than ZBRS, DBRS, MRS1, and MRS2 in most cases of a mobile station behavior, and it has an advantage of distributing the load of signaling traffic into every cell, which is important in personal communication system.

The most commonly used activation function in Backpropagation learning is sigmoidal while linear function is also sometimes used at the output layer with the view that choice between these activation functions does not make considerable differences in network's performance. In this letter, we show distinct performance between a network with linear output units and a similar network with sigmoid output units in terms of convergence behavior and generalization ability. We experimented with two types of cost functions, namely, sum-squared error used in standard Backpropagation and log-likelihood recently reported. We find that, with sum-squared error cost function and hidden units with nonsteep sigmoid function, use of linear units at the output layer instead of sigmoidal ones accelerates the convergence speed considerably while generalization ability is slightly degraded. Network with sigmoid output units trained by log-likelihood cost function yields even faster convergence and better generalization but does not converge at all with linear output units. It is also shown that a network with linear output units needs more hidden units for convergence.