Search

We present a classification scheme based on a new kind of RBFNN (radial basis function neural network) whose structure is similar to that of AWNN (adaptive wavelet neural network). To be more suitable for HRRP (high resolution range profile) classification, this kind of RBFNN substitutes wavelet basis functions in AWNN with Gaussian basis functions. In addition, we also devise an RBFNN initialization method of clear physical significance, and propose a decision rule based on average output vectors of RBFNNs. The new scheme is applied to HRRP classification of six aircraft at different SNR levels, and the results are compared with that obtained by MCCM (maximum correlation coefficient method). It is indicated that the RBFNN-based classification method has the potential in complex target classification and is promising to develop more practical HRRP classifiers.

The SVM (support vector machine) has good generalization property and is proven to be a powerful classification method to a wide variety of applications. In this paper, the SVM classification method is applies to the HRRP (high resolution range profile) classification problem. Furthermore, we propose a preprocessing method and present a new SVM model selection method, the combined leave-one-out cross-validation scheme with VC bound method. Accordingly, the SVM classifiers for five aircraft are designed. Their anti-noise performance and tolerance to aspect variation are compared with that of MCM (maximum correlation method). The experimental results indicate that the preprocessing procedure and model selection method are effective and the SVM method has potential in radar complex target classification.

ATM networks oriented connections provide pure QoS (quality of service) for diversified services through a series of traffic management mechanisms, the ABR (available bit rate) flow control is especially important among these approaches. In the binary flow control scheme, the cell rate and queue length may oscillate with great magnitude to reduce link utilization, so that the EFCI (explicit forward congestion indication) algorithm is regard as ineffective, however its simplicity is attractive to high performance switch design. In this paper, the EFCI algorithm is analyzed based on classical control theory. It is found that the nonlinear hysteresis determines the congestion, and is the dominant reason that causes the oscillation. Then, a probability congestion detection approach, called p-EFCI is put forward. Numerical results show that the improved algorithm deeply constrains the oscillation magnitude, and the queue length is controlled in limited scope to guarantee cell zero loss.

With the increase in demand for personal wireless data communications within the past decade, there are growing needs for both higher bit rate data transmission and multiple access. The BER of the OFDM-CDMA system with parallel interference cancellation (PIC) detector is analyzed and simulated in the Rayleigh fading channel. It is shown that the proposed multiuser detector could reduce the multiple access interference effectively and affords important performance gains over conventional single-user systems.