Zhu D.,NEC Laboratories China |
Natarajan B.,Kansas State University
IEEE Transactions on Vehicular Technology | Year: 2010
In this paper, we propose a novel application of residue number system (RNS) arithmetic in designing hopping-pilot patterns for cellular downlink orthogonal frequency-division multiple access (OFDMA). By hopping the scan lines in either the time or the frequency domain, RNS-based pilot patterns fulfill the Nyquist sampling theorem. That is, by using RNS-based pilot patterns, the channel's delayDoppler response can fully be reconstructed without severe aliasing. In addition to channel estimation, our proposed scheme can achieve other objectives, such as device/cell identification and timefrequency synchronization. We show that the RNS-based approach has more pairs of hoppingpilot patterns that are collision free than the Costas-array-based method. This is helpful in not only mitigating intra-/intercell interference but also identifying multiple devices in a multicell multiantenna environment. Moreover, hopping in time increases the pilot's time support, which, in turn, enables quick initial acquisition of timefrequency offsets. © 2006 IEEE.
Gan H.,Huazhong University of Science and Technology |
Sang N.,Huazhong University of Science and Technology |
Huang R.,NEC Laboratories China
Journal of the Optical Society of America A: Optics and Image Science, and Vision | Year: 2014
Face recognition is one of the most important applications of machine learning and computer vision. The traditional supervised learning methods require a large amount of labeled face images to achieve good performance. In practice, however, labeled images are usually scarce while unlabeled ones may be abundant. In this paper, we introduce a semi-supervised face recognition method, in which semi-supervised linear discriminant analysis (SDA) and affinity propagation (AP) are integrated into a self-training framework. In particular, SDA is employed to compute the face subspace using both labeled and unlabeled images, and AP is used to identify the exemplars of different face classes in the subspace. The unlabeled data can then be classified according to the exemplars and the newly labeled data with the highest confidence are added to the labeled data, and the whole procedure iterates until convergence. A series of experiments on four face datasets are carried out to evaluate the performance of our algorithm. Experimental results illustrate that our algorithm outperforms the other unsupervised, semi-supervised, and supervised methods. © 2013 Optical Society of America.
Li J.Q.,NEC Laboratories China |
Yang J.-J.,Tsinghua National Laboratory for Information Sciences and Technology |
Zhao Y.,NEC Laboratories China |
Liu B.,NEC Laboratories China
Enterprise Information Systems | Year: 2013
Data sharing in today's information society poses a threat to individual privacy and organisational confidentiality. k-anonymity is a widely adopted model to prevent the owner of a record being re-identified. By generalising and/or suppressing certain portions of the released dataset, it guarantees that no records can be uniquely distinguished from at least other k-1 records. A key requirement for the k-anonymity problem is to minimise the information loss resulting from data modifications. This article proposes a top-down approach to solve this problem. It first considers each record as a vertex and the similarity between two records as the edge weight to construct a complete weighted graph. Then, an edge cutting algorithm is designed to divide the complete graph into multiple trees/components. The Large Components with size bigger than 2k-1 are subsequently split to guarantee that each resulting component has the vertex number between k and 2k-1. Finally, the generalisation operation is applied on the vertices in each component (i.e. equivalence class) to make sure all the records inside have identical quasi-identifier values. We prove that the proposed approach has polynomial running time and theoretical performance guarantee O(k). The empirical experiments show that our approach results in substantial improvements over the baseline heuristic algorithms, as well as the bottom-up approach with the same approximate bound O(k). Comparing to the baseline bottom-up O(logk)-approximation algorithm, when the required k is smaller than 50, the adopted top-down strategy makes our approach achieve similar performance in terms of information loss while spending much less computing time. It demonstrates that our approach would be a best choice for the k-anonymity problem when both the data utility and runtime need to be considered, especially when k is set to certain value smaller than 50 and the record set is big enough to make the runtime have to be taken into account. © 2013 Copyright Taylor and Francis Group, LLC.
Gao F.,Tsinghua National Laboratory for Information Sciences and Technology |
Li J.C.F.,NEC Laboratories China |
Lei M.,NEC Laboratories China
IEEE Wireless Communications and Networking Conference, WCNC | Year: 2013
In this paper, we design a null-space based robust interference avoiding strategy for the Device-to-Device (D2D) communication underlaying network. Thanks to the coordination between D2D user and the regular user, the interfering channel state information (CSI) among the base station (BS), cellular user equipment (CUE) and the D2D user equipments (DUEs) can be estimated from the training approach. Then, the null-space based transmit and receive beamformings are designed at appropriate terminals to mitigate the interference caused in the future data transmission. To make the design practical, we also characterize the null-space uncertainty that is resulted from the imperfect channel estimation. Moreover, we derive the optimal transmission strategy that can achieve the best training-throughput tradeoff. Simulation results are provided to corroborate the proposed studies. © 2013 IEEE.
Sun L.,NEC Laboratories China |
Lei M.,NEC Laboratories China
IEEE International Symposium on Personal, Indoor and Mobile Radio Communications, PIMRC | Year: 2012
This paper considers the implementation of Tomlinson-Harashima (TH) precoding for multiuser MIMO systems based on quantized channel state information (CSI) at the transmitter side. We study the sum rate of a greedy user scheduling algorithm combined with the quantized CSI-based TH precoding, termed G-THP-Q. The asymptotic distribution of the output signal to interference plus noise ratio of each selected user and the asymptotic average sum rate as the number of users K grows large are derived. We prove that as K grows large, the G-THP-Q approach can achieve the optimal sum rate scaling of the MIMO broadcast channel. We also prove that, if we ignore the precoding power loss, the average sum rate of G-THP-Q approach converges to the average sum capacity of the MIMO broadcast channel for large K. Our results also provide key insights into the effect of multiuser interference caused by quantized CSI on the multiuser diversity gain. © 2012 IEEE.