空间复用系统在相关信道下的两步选择天线算法

针对采用贝尔实验室垂直分层空时(V-BLAST)检测算法的空间复用系统存在信道相关的问题,提出一种新的两步选择发射天线的准则,首先利用信道矩阵奇异值确定有效发射天线数,然后根据最低检测后信噪比确定发射天线和子数据流之间的映射关系.仿真结果表明两步选择准则能获得比已有算法更高的选择增益,此外还分析并仿真了天线间距和角度扩展对符号向量错误概率的影响.     

Power Allocation Algorithms for GMD or UCD Based Joint Transceiver Designs

In a downlink multiuser multiple-input multiple-output channel, joint transceiver designs based on geometric mean decomposition and uniform channel decomposition can provide a BER performance enhancement by making all substreams of each user have the same SINR. However, if a user power allocation (uPA) algorithm is not applied in their designs, they actually make the same SINR for all substreams in each user while causes different SINR with one another, and therefore the average BER is dominated by the user with the worst SINR. In this regards, it is possible to enhance BER performance further provided that the uPA algorithm is taken into account under a total power constraint. The objective of this paper is to maximize the worst user’s SINR by applying the uPA algorithm with the total power constraint. Because of proposed uPA algorithms, all substreams across all users can have the same SINR, which leads to minimization of average BER. Simulation results show that joint transceiver designs with proposed uPA algorithms have an SNR gain about 2 dB at the BER of \(10^{-3}\).     

基于块对角化和服务公平性的次最优用户选择方案

该文基于块对角化的思想提出了一种次最优用户选择方案。首先推导出一个和速率容量的上界公式,将其作为该文提出算法的选择准则;然后用此上界公式代替瞬时和容量公式作为自适应准则,得到了所提出算法的简化算法;最后考虑到用户之间服务公平性的因素,把该文的提出算法与比例公平调度机制相结合,使得系统既能获得一定的信道容量增益又能让所有用户均得到公平的服务。仿真结果表明,该文所提出算法具有接近最优的性能,与比例公平调度机制相结合后可以达到系统和容量与服务公平性的折衷。     

新的V-BLAST系统次优天线选择

该文提出新的基于ZF SIC检测的V-BLAST系统次优天线选择准则:最小化信道矩阵伪逆的最大行范数。基于贪婪选择思想,发射天线选择采用使得该范数增加最小的递增选择策略,接收天线选择采用使得该范数减少最大的递减选择策略。仿真表明所提出的新准则明显优于已有的最大第1检测层后处理信噪比准则,且相应的快速选择算法可以获得最优的基于最大最小准则的全搜索选择的大部分分集增益,而复杂度很低。     

Unequal error protection of SPIHT encoded image bit streams

A derivative of the set partitioning into hierarchical trees (SPIHT) image coding method, which generates substreams with different error-resilience properties, is proposed. By dividing the image bit stream into three classes, substreams with different immunity properties are obtained. The unequal protection of these substreams with different channel coding rates improves the overall performance of the method against channel errors. Simulation results show the superiority of the proposed method over some of the state-of-the-art methods     

Improving power allocation to MIMO eigenbeams under imperfect channel estimation

Eigenbeamforming exploits the spatial diversity of the MIMO channel by multiplexing data substreams along orthogonal modes. Estimation errors result in self-interference, which must be taken into account in the selection of modes for transmission and the allocation of transmitter power. Using perturbation analysis to evaluate the self-interference, a technique is proposed for determining the power allocation which achieves a specified performance, based only on the information available at the receiver.     

Closed-loop Grouped Space–Time Block Code: Encoding,Decoding and Codeword Selection

A design of closed-loop grouped space–time block codes (G-STBCs) including encoding, decoding and codeword selection is proposed for the downlink over Rayleigh flat-fading channels. In particular, at the transmitter, the antenna array is partitioned into a number of groups, each of which is encoded based on the orthogonal STBC (O-STBC). At the receiver, by exploiting the algebraic structure of orthogonal codes, a low-complexity, in recursion form, group-wise ordered successive interference cancellation (OSIC) detector is developed. Moreover, the G-STBC codeword is designed and a G-STBC codeword selection criterion that minimizes the BER performance under the constraints of a fixed spectral efficiency and total transmit power is then proposed. The selection index of the G-STBC codeword and the associated modulation type are determined at the receiver and conveyed to the transmitter with a limited feedback overhead to choose an appropriate mode for transmission. Finally, Numerical examples are used for illustrating the performance of the proposed G-STBCs, OSIC based detection and G-STBC codeword selection criterion.

Seed region selection and homogeneity criterion for doorplate image segmentation in mobile robot navigation

Focused on the seed region selection and homogeneity criterion in Seeded Region Growing （SRG）, an unsupervised seed region selection and a polynomial fitting homogeneity criterion for SRG are proposed in this paper. First of all, making use of Peer Group Filtering （PGF） techniques, an unsupervised seed region selection algorithm is presented to construct a seed region. Then based on the constructed seed region a polynomial fitting homogeneity criterion is applied to solve the concrete problem of doorplate segmentation appearing in the robot navigation along a corridor. At last, experiments are performed and the results demonstrate the effectiveness of the proposed algorithm.     

MIMO广播系统中基于有限反馈的多用户选择新算法

在基于有限反馈多输入多输出(MIMO)广播系统的多用户选择研究中,目前已有的用户信噪干扰比(SINR)估计存在较大误差,这使得整个系统的性能受到制约。该文利用推导出的上下界来联合估计用户接收到的有用信号功率,得到了一种误差较小的SINR估计式,并基于该式给出了一种用户选择新算法。仿真结果表明,新的用户选择算法可以在低信噪比和高信噪比下同时达到较优性能,且复杂度较低。     

Multiuser beamforming based on max SINR for downlink spatial multiplexing MIMO

In this article, a method based on max signal interference noise ratio (SINR) criterion is proposed, to mitigate the interuser interference for downlink multiuser spatial multiplexing multi-input multi-output (MIMO) systems.Unlike the zero forcing (ZF) scheme in which the SNR is decreased when the interference is eliminated completely, max SINR method makes a compromise between noise and interuser interference.When the number of substreams is larger than the difference between the number of base station antennas and the sum of interference mobile station antennas, the ZF is infeasible.An existing coordinated TX-RX block diagonalization (COOR BD) method uses preprocessing at the receiver to cancel the interuser interference.However, it cannot obtain more receive diversity gain because of the preprocessing.Analysis and simulation show that the max SINR scheme has better performance than the ZF method.Moreover, when the ZF is infeasible, the max SINR scheme can obtain more receive diversity gain than COOR BD in the two-user case.     

Signal constellations design for differential unitary space-time in MIMO-OFDM system over frequency-selective fading channels

In this paper, the design of signal constellations parameters is studied for Differential Unitary Space-Time Modulation （DUSTM） based on the design criterion of maximizing the diversity product. Further, noninteger searching method for the signal constellation parameters design is proposed in order to get better codes. Experimental results show that under the different Doppler spread and data transmission rate, the proposed design performs better than the previous design using integer parameters in Multiple Input Multiple Output Orthogonal Frequency Division Multiplexing （MIMO-OFDM） system over frequency-selective fading channels.     

An Efficient Multimode Quantized Precoding Technique for MIMO Wireless Systems

Multimode quantized precoding (QP) can provide full diversity gain or high capacity gain by adapting the number of substreams, as well as the precoding matrix, according to the instantaneous channel condition with low-rate feedback. Conventional multimode QP (MM-QP), however, does not consider the adaptive rate allocation among substreams; thus, it cannot have the additional gain by adaptive modulation. Furthermore, it is computationally complex since exhaustive matrix inversions are required to determine the optimal mode. In this paper, we propose an efficient MM-QP system that improves the performance of a conventional system in terms of error rate and has a lower computational complexity than the conventional system. First, we define the rate-partitioning vector as the mode and control the rate among substreams and the number of substreams according to the channel instantaneous condition. Second, to reduce the computational complexity for the receiver to determine the optimal mode, the simplified mode-selection technique using estimates of the modal metric is proposed. In the proposed mode-selection technique, the optimal mode can be obtained by several multiplication and division operations. Finally, the mode-reduction technique eliminating the less-frequently used modes is proposed, which leads to a significant reduction of the feedback information with negligible performance loss. In numerical experiments, it was verified that the proposed MM-QP system gives a better error-rate performance than the conventional system, with much less computational complexity for the same amount of feedback information.      

Generalized multiple maximum scatter difference feature extraction using QR decomposition

Multiple maximum scatter difference (MMSD) discriminant criterion is an effective feature extraction method that computes the discriminant vectors from both the range of the between-class scatter matrix and the null space of the within-class scatter matrix. However, singular value decomposition (SVD) of two times is involved in MMSD, rendering this method impractical for high dimensional data. In this paper, we propose a generalized MMSD (GMMSD) criterion for feature extraction and classification. GMMSD allows relatively-free selection of a suitable transformation matrix to reduce dimensions. Based on GMMSD criterion, we demonstrate that the same discriminant information can be extracted by QR decomposition, which is more efficient than SVD. Next, GMMSD is compared with several classical feature extraction methods to justify the validity of the proposed method. Our experiments on three face databases and two facial expression databases demonstrate that GMMSD provides favorable recognition performance with high computational efficiency.     

垂直分层空时码的MAP检测器及其神经网络实现

该文基于最大后验概率准则,给出了垂直分层空时码的最佳检测箅法。算法对接收到的符号矢量的符号分量进行逐个地判决以使误符号率最小。同时还详细说明了径向基函数神经网络可以完全等价于最大后验概率检测算法,由于神经网络的硬件可实现性,本文提出的算法是可能在实际中应用的。     

Multicarrier CDMA with adaptive frequency hopping for mobile radiosystems

A modified multicarrier (MC) direct-sequence code-division multiple-access (DS-CDMA) system has been proposed for use over slow multipath fading channels with frequency selectivity in the reverse link transmission of a cellular network. Instead of transmitting data substreams uniformly through subchannels, data substreams hop over subchannels with the hopping patterns adaptively adjusted to the channel fading characteristics. The problem of determining the optimal hopping pattern is formulated as a multiobjective optimization problem, for which an efficient algorithm, based on the water-filling (WF) principle, is designed to solve the problem practically. Simulation results show that the performance in terms of the average bit-error probability (BEP) (over all users) is better than that of single carrier RAKE receiver systems, conventional MC CDMA systems applying moderate error protection, or diversity systems with different combining techniques     

采用故意时延的V-BLAST系统的最佳迫零检测

Shao(2007)提出了一种故意时延的垂直贝尔实验室分层空时码结构(V-BLAST)系统,但是其中提出的迫零检测算法并不是满足迫零准则下的最优算法。本文直接从接收天线处的未采样连续信号数学模型分析入手,利用泛函分析的方法推导出了该系统最优的迫零检测算法,理论和仿真都表明该算法优于Shao提出的迫零算法。本文的工作对进一步研究这种新型的故意时延的V-BLAST系统具有较为重要的意义。     

基于加权多核学习的FLDA在人脸识别中的应用

模式识别技术的应用及研究表明,单核学习在人脸识别中的应用已经很成熟,但是,在单核学习中分类的效果并不是很好。基于此,提出了一种多核构造方法,即基于加权多核学习的FLDA方法(WMKL-FLDA),通过一系列带有权值约束的基本内核线性组合构建内核,并且利用权值优化迭代方案对最大边缘准则(MMC)进行优化。在FERET及CMU PIE人脸数据库上的实验表明,与以往的单核FLDA方法相比,提出的多核学习方法不仅实现了更高的识别性能,在构造内核方面也放松了参数的选择要求。     

Least squares-based receive antenna selection for MIMO spatial multiplexing systems with linear receivers

In this letter, a novel receive antenna selection technique is proposed for multiple input multiple output (MIMO) spatial multiplexing systems with linear receivers in the presence of unknown interference. This antenna selection technique is directly implemented based on training sample sequence under the least squares (LS) criterion. It avoids the channel estimation and retain the diversity benefit by antenna selection in the presence of unknown multiple access interference (MAI). In addition, practical implementation with manageable complexity is made possible by extending the fast backward greedy algorithm (BGA) into the proposed antenna selection algorithm.     

Receive antenna selection for MIMO-SM systems with linear MMSE receivers in the presence of unknown interference

In this paper, a novel receive antenna subset selection technique is proposed for multiple input multiple output spatial multiplexing (MIMO-SM) systems with linear minimum mean square error (MMSE) receivers in the presence of unknown co-channel interference. This antenna selection technique is directly implemented based on training sample sequence by utilizing backward greedy algorithm (BGA) under the least squares (LS) criterion. In the case of practical implementation. Diagonal loading (DL) technique is incorporated into the selection process to insure robustness of antenna selection given limited training sample support. Simulation results show that the proposed antenna selection technique is able to retain the diversity order of a full complexity system in the presence of unknown multiple access interference (MAI)     

车联网中基于NOMA-MEC的卸载策略研究

随着车联网(IoV)的迅猛发展,请求进行任务卸载的汽车终端用户也逐渐增长,而基于移动边缘计算(MEC)的通信网络能够有效地解决任务卸载在上行传输时延较高的挑战,但是该网络模型同时也面临着信道资源不足的问题。该文引入的非正交多址(NOMA)技术相较于正交多址(OMA)能够在相同的信道资源条件下为更多的用户提供任务卸载,同时考虑到任务卸载过程中多方面的影响因子,提出了混合NOMA-MEC卸载策略。该文设计了一种基于深度学习网络(DQN)的博弈算法,帮助车辆用户进行信道选择,并通过神经网络多次迭代学习,为用户提供最优的功率分配策略。仿真结果表明,该文所提出的混合NOMA-MEC卸载策略能够有效地优化多用户卸载的时延以及能耗,最大限度保证用户效益。