AN EFFICIENT FINITE DIFFERENCE METHOD FOR STOCHASTIC LINEAR SECOND-ORDER BOUNDARY-VALUE PROBLEMS DRIVEN BY ADDITIVE WHITE NOISES

In this paper,we develop and analyze a finite difference method for linear second-order stochastic boundary-value problems(SBVPs)driven by additive white noises.First we regularize the noise by the Wong-Zakai approximation and introduce a sequence of linear second-order SBVPs.We prove that the solution of the SBVP with regularized noise converges to the solution of the original SBVP with convergence order O(h)in the meansquare sense.To obtain a numerical solution,we apply the finite difference method to the stochastic BVP whose noise is piecewise constant approximation of the original noise.The approximate SBVP with regularized noise is shown to have better regularity than the original problem,which facilitates the convergence proof for the proposed scheme.Convergence analysis is presented based on the standard finite difference method for deterministic problems.More specifically,we prove that the finite difference solution converges at O(h)in the mean-square sense,when the second-order accurate three-point formulas to approximate the first and second derivatives are used.Finally,we present several numerical examples to validate the efficiency and accuracy of the proposed scheme.

MAXIMUM LIKELIHOOD SOURCE SEPARATION FOR FINITE IMPULSE RESPONSE MULTIPLE INPUT-MULTIPLE OUTPUT CHANNELS IN THE PRESENCE OF ADDITIVE NOISE

Blind identification-blind equalization for Finite Impulse Response (FIR) Multiple Input-Multiple Output (MIMO) channels can be reformulated as the problem of blind sources separation. It has been shown that blind identification via decorrelating sub-channels method could recover the input sources. The Blind Identification via Decorrelating Sub-channels(BIDS)algorithm first constructs a set of decorrelators, which decorrelate the output signals of subchannels, and then estimates the channel matrix using the transfer functions of the decorrelators and finally recovers the input signal using the estimated channel matrix. In this paper, a new approximation of the input source for FIR-MIMO channels based on the maximum likelihood source separation method is proposed. The proposed method outperforms BIDS in the presence of additive white Gaussian noise.

Recovery of Transient Signals in Noise by OptimalThresholding in Wavelet Domain

：研究用离散子波变换复原被加性高斯白噪声污染的瞬态信号．在子波域中提出了一种最佳门限方法，该方法涉及到对于波系数的假设检验，利用似然比、奈曼，皮尔逊准则和最小均方差设计该门限，计算机仿真证明，该方法在较低信噪比下复原信号的有效性．

Construction of Regular Rate-Compatible LDPC Convolutional Codes

In this paper, we propose a new method to derive a family of regular rate-compatible low-density parity-check（RC-LDPC） convolutional codes from RC-LDPC block codes. In the RC-LDPC convolutional family, each extended sub-matrix of each extended code is obtained by choosing specified elements from two fixed matrices HE1K and HE1K, which are derived by modifying the extended matrices HE1 and HE2 of a systematic RC-LDPC block code. The proposed method which is based on graph extension simplifies the design, and prevent the defects caused by the puncturing method. It can be used to generate both regular and irregular RC-LDPC convolutional codes. All resulted codes in the family are systematic which simplify the encoder structure and have maximum encoding memories which ensure the property. Simulation results show the family collectively offer a steady improvement in performance with code compatibility over binary-input additive white Gaussian noise channel（BI-AWGNC）.

SYMMETRIC INFORMATION RATE OF CPM AND OPTIMAL DETECTOR DESIGN

This paper first calculates numerically the Symmetric Information Rate (SIR) of Cotinuous Phase Modulation (CPM) schemes over Additive White Gaussian Noise (AWGN) channel, modeling CPM and channel as a Memory-less Modulator (MM) plus a Markov Finite State Channel (FSMC), then proposes an optimal detector, which is of BCJR based Decision Feedback Detector (DFD) but non-iterative. It can achieve the SIR as the sequence length approaches infinity while the complexity is extremely low. Finally, both theoretic proof and numerical simulation are performed to show the op- timality. Simulation results show that it nearly achieves the theoretic bound and outperforms the Viterbi Detector (VD) with 2dB in low Eb/N0.

ASYMPTOTIC BEHAVIOR FOR GENERALIZED GINZBURG-LANDAU POPULATION EQUATION WITH STOCHASTIC PERTURBATION

In this paper,we are devoted to the asymptotic behavior for a nonlinear parabolic type equation of higher order with additive white noise.We focus on the Ginzburg-Landau population equation perturbed with additive noise.Firstly,we show that the stochastic Ginzburg-Landau equation with additive noise can be recast as a random dynamical system.And then,it is proved that under some growth conditions on the nonlinear term,this stochastic equation has a compact random attractor,which has a finite Hausdorff dimension.

Multiple Input Multiple Output CDSK Chaotic Communication System and Its Performance Analysis

This paper propose a novel noncoherent chaotic com- munication scheme named multiple-input multiple-output correlation-delay-shill- keying （MIMO-CDSK）. In this scheme, multiple antennas are employed to strengthen the robustness in transmission, and to get more diversity gain. The bit error rate （BER） of the MIMO-CDSK is studied analytically in AWGN channel model and multipath fading channel model. The theory and simulation results show that, the performance gain can be obtained with multiple antennas allocated in the transmitter and receiver. Moreover, it is observed that MIMO-CDSK system can effectively reduce the multipath interference.

A Novel Minimum Output Variance Estimator for Carrier Frequency Offset in OFDM Systems

In orthogonal frequency-division multiplexing （OFDM） systems, the carrier frequency offset （CFO） destroys the orthogonality among subcarriers which degrades system performance. Various CFO estimation methods have been developed to compensate for the CFO at the receiver. This paper describes a novel minimum output variance method for OFDM systems with CFO in additive white Gaussian noise channels. This method utilizes the phase and the amplitude of the received signal and reduces the mean square error of the CFO by about 3 dB compared with the original minimum output variance method.