Search

An unbiased weighted least square (UWLS) adaptive algorithm is developed and proposed to cooperate with a second-order linear predictor to estimate the frequency of a single tone embed in noise. A new objective function and updating equations are introduced to improve the performance of previous WLS algorithm. Extensive simulations are drawn to verify the performance of the proposed estimator.

Another scheme of a binary frequency phase shift keying (BFSK) signal demodulation has been presented in this paper. The proposed demodulated scheme utilizes an ANC-based adaptive signal processing. The adaptive algorithm which adapts the coefficients of the ANC system will be constrained in adaptation sequence corresponding to each symbol interval of the received BFSK signal. By using this scheme, the system will provide demodulated behavior similar to an integrated and dump correlator of an optimum receiver. The analytical and simulation results have shown that the proposed system can demodulate the received BFSK signal in AWGN channel. Furthermore, the system provides the same performance as the optimum receiver.

A constrained adaptive IIR notch filter (c-ANF) using weighted least square (WLS) algorithm (c-ANF-WLS) is introduced for frequency estimation of a single real tone embed in Gaussian noise. As compared with some previous techniques, the proposed algorithm exhibits better performance in terms of both convergence rate and steady state mean square error (MSE). Computer simulation results are provided to assert the claim.

This paper presents a modified phase based frequency adaptive algorithm (MPBAA). We have modified the original PBAA by using i) an all-pole bandpass filter (BPF) instead of zero-pole BPF ii) only one first-order autoregressive processes (AR(1)) instead of two ARs(1). Moreover, the process of de-composing for real and imaginary parts is not required. Therefore, MPBAA requires less computation than PBAA while maintaining almost the same performances.

Adaptive Sinusoidal interference cancellation (ASIC) based on recursive least square (RLS) algorithm is proposed. RLS algorithm allows us to use a recursion instead of inversion of the autocorrelation matrix to evaluate the optimum weights. The main advantage of RLS algorithm is that it exhibits extremely high rate of convergence. Computer simulations are performed to demonstrate the performance of the proposed algorithm.

This paper has proposed the technique in order to remove salt-and-pepper noise from the corrupted image. The proposed scheme consists of two steps. The first, the adaptive median filter is applied to noisy image for detecting noisy and noise-free pixels. Finally, the output image of previous step will be divided into many non-overlapping windows of which the noise-free pixel values are used to compute the new pixel value by the Shepard method. The experiment results of our approach had shown the better performance than the existing method.

This paper presents a simple lattice-structure complex adaptive linear predictor (LP) to estimate and track an unknown frequency of a noisy complex sinusoid. Instead of a widely used LMS-based algorithm, a weighted least squared (WLS) algorithm is proposed to improve the convergence rate of the predictor. Analytic results for asymptotic bias is derived in closed form and demonstrated that the WLS is asymptotic unbiased estimator.

At present, mobile stations have high data rates and high-speed of mobile stations (MS) such as high-speed trains. Therefore is the cause of frequent handover and handover fails. This makes communication interrupted a challenge to solve. The speed of the mobile station is also an important factor in starting the handover (HO) that should not be ignored. To handle delivery delays, this article proposes a new HO scheme. This scheme uses dynamic HO criteria that use techniques. Link-Going-Down (LGD) to determine the HO process. Starting this technique depends on the delivery time and the speed of the mobile device. Therefore, the HO process begins at the right time and the use of network resources has been improved to be more efficient and reduce interrupt time.

Currently, a mobile network environment has communication high data rate and high-speed of the mobile station (MS) such as a high-speed train. Therefore, it causes frequent handover. Consequently, the handover fails or increases latency in the system. If the system has latency more than 42 ms, it will affect the real-time application. This causes interruptions in communication to be a challenge in solving problems. The speed of the mobile station is also an important factor in getting started handover. To handle the problems of handover latency time, this paper proposes a new handover method. This method applies the dynamic handover threshold to define the starting handover process. This technique is based on the measured received signals strength (RSS) by the Doppler Effect technique, the mobile speed and handover latency time. Consequently, the initial handover process at the right time will reduce the handover time and the efficient use of network resources.

Sign-algorithm (SA) for a constrainted adaptive IIR notch filter (ANF) is very attractive because it is simple to implement. Although the slow convergence speed and biased estimate are two main drawbacks of SA, in some applications, it provides satisfactory results. In this work we propose the bias removal technique to mitigate the biased estimate of SA. Computer simulation has been drawn to show the performance of the proposed technique.

A fast adaptive method to eliminate a single sinusoidal interference of known frequency is proposed. The coupled-oscillator is used to generate the reference signal instead of look-up table. Unlike the least mean square (LMS) based adaptive algorithm, the weighted least squared error criterion is employed instead to evaluate the system parameters. It has been shown that the proposed method can quickly eliminate unwanted interference as compared to previous methods, Moreover, the extensive simulations have been done and revealed that proposed algorithm yields a constant output signal to interference ratio (SIR) of about 20 dB regardless of input SIR.

In this paper, the inverse distance weighting (IDW) is utilized in order to restore contaminated images which are corrupted by the salt-and-pepper noise (SPN). Our method consists of three steps. The first step, the noise candidate pixels are identified by the adaptive median filter. For the second step, the noisy image is divided into many sub-windows. The last step, for each sub-window from previous step, the IDW is applied to calculate the new pixel values that use the neighbour uncorrupted pixels in each sub-region. The results of experiments demonstrate that the proposed technique is better performance than the existing methods.

This paper proposes the high efficient method to remove salt-and-pepper noise from the noisy image. In this paper, we propose two step process which contain the noisy pixels identify and noise restoration. In the first step, the noisy pixels are detected by using the minimal and maximal value of the contaminated image. Finally, the Shepard interpolation method is operated to restore the corrupted pixels on each sliding block. In performance evaluation, it demonstrates the efficiency of the proposed method is better than the existing methods.

The LMS-based indirect frequency estimation algorithm (IFE) is reformulated using weighted least-square error criterion. Theoretical analyses for steady state bias and mean square error (MSE) are addressed. It has been shown that the proposed algorithm outperforms the conventional LMS-based algorithms in terms of convergence speed at the same value of MSE.

In this paper, the salt-and-pepper noise is removed by Shepard interpolation method. Our method has three steps. The first step, the noise location of the noisy image is detected by applying the adaptive median filter. The second step, the corrupted image is split into sub-regions. The window size of each sub-region is defined by the noise density. Finally, the new pixel value of each noisy pixel which locates in each sub-regions is calculated by Shepard method. The experiment results demonstrate the good performance some recent methods.