Attention mechanism, which selectively focuses on source-side information to learn a context vector for generating target words, has been shown to be an effective method for neural machine translation (NMT). In fact, generating target words depends on not only the source-side information but also the target-side information. Although the vanilla NMT can acquire target-side information implicitly by recurrent neural networks (RNN), RNN cannot adequately capture the global relationship between target-side words. To solve this problem, this paper proposes a novel target-attention approach to capture this information, thus enhancing target word predictions in NMT. Specifically, we propose three variants of target-attention model to directly obtain the global relationship among target words: 1) a forward target-attention model that uses a target attention mechanism to incorporate previous historical target words into the prediction of the current target word; 2) a reverse target-attention model that adopts a reverse RNN model to obtain the entire reverse target words information, and then to combine with source context information to generate target sequence; 3) a bidirectional target-attention model that combines the forward target-attention model and reverse target-attention model together, which can make full use of target words to further improve the performance of NMT. Our methods can be integrated into both RNN based NMT and self-attention based NMT, and help NMT get global target-side information to improve translation performance. Experiments on the NIST Chinese-to-English and the WMT English-to-German translation tasks show that the proposed models achieve significant improvements over state-of-the-art baselines.

In this letter, we investigate the problem of feedforward timing estimation for burst-mode satellite communications. By analyzing the correlation property of frame header (FH) acquisition in the presence of sampling offset, a novel data-aided feedforward timing estimator that utilizes the correlation peaks for interpolating the fractional timing offset is proposed. Numerical results show that even under low signal-to-noise ratio (SNR) and small rolloff factor conditions, the proposed estimator can approach the modified Cramer-Rao bound (MCRB) closely. Furthermore, this estimator only requires two samples per symbol and can be implemented with low complexity with respect to conventional data-aided estimators.

A convolutional neural network (CNN) for broadband direction of arrival (DOA) estimation of far-field electromagnetic signals is presented. The proposed algorithm performs a nonlinear inverse mapping from received signal to angle of arrival. The signal model used for algorithm is based on the circular antenna array geometry, and the phase component extracted from the spatial covariance matrix is used as the input of the CNN network. A CNN model including three convolutional layers is then established to approximate the nonlinear mapping. The performance of the CNN model is evaluated in a noisy environment for various values of signal-to-noise ratio (SNR). The results demonstrate that the proposed CNN model with the phase component of the spatial covariance matrix as the input is able to achieve fast and accurate broadband DOA estimation and attains perfect performance at lower SNR values.

Small cell networks have been promoted as an enabling solution to enhance indoor coverage and improve spectral efficiency. Users usually deploy small cells on-demand and pay no attention to global profile in residential areas or offices. The reduction of cell radius leads to dense deployment which brings intractable computation complexity for resource allocation. In this paper, we develop a semi-distributed resource allocation algorithm by dividing small cell networks into clusters with limited inter-cluster interference and selecting a reference cluster for interference estimation to reduce the coordination degree. Numerical results show that the proposed algorithm can maintain similar system performance while having low complexity and reduced information exchange overheads.

Large scale first-hand tweets motivate automatic event detection on Twitter. Previous approaches model events by clustering tweets, words or segments. On the other hand, event clusters represented by tweets are easier to understand than those represented by words/segments. However, compared to words/segments, tweets are sparser and therefore makes clustering less effective. This article proposes to represent events with triple structures called frames, which are as efficient as, yet can be easier to understand than words/segments. Frames are extracted based on shallow syntactic information of tweets with an unsupervised open information extraction method, which is introduced for domain-independent relation extraction in a single pass over web scale data. This is then followed by bursty frame element extraction functions as feature selection by filtering frame elements with bursty frequency pattern via a probabilistic model. After being clustered and ranked, high-quality events are yielded and then reported by linking frame elements back to frames. Experimental results show that frame-based event detection leads to improved precision over a state-of-the-art baseline segment-based event detection method. Superior readability of frame-based events as compared with segment-based events is demonstrated in some example outputs.

This paper presents a strategy together with tool support for the translation of state machines from equational theories into rewrite theories, aiming at automatically generating rewrite theory specifications. Duplicate effort can be saved on specifying state machines both in equational theories and rewrite theories, when we incorporate the theorem proving facilities of CafeOBJ with the model checking facilities of Maude. Experimental results show that efficiencies of the generated specifications by the proposed strategy are significantly improved, compared with those that are generated by three other existing translation strategies.

In this paper, a new non-uniform weight-updating scheme for adaptive digital beamforming (DBF) is proposed. The unique feature of the letter is that the effective working range of the beamformer is extended and the computational complexity is reduced by introducing the robust DBF based on worst-case performance optimization. The robust parameter for each weight updating is chosen by analyzing the changing rate of the Direction of Arrival (DOA) of desired signal in LEO satellite communication. Simulation results demonstrate the improved performance of the new Non-Uniform Weight-Updating Beamformer (NUWUB).

This paper considers a multi-user large-scale multiple-input multiple-output (MIMO) system with single cell working in full-duplex mode. Maximum ratio combining/maximum ratio transmission (MRC/MRT) is applied to maximize the output signal to noise ratio (SNR) of the receiver. Then we deduce the asymptotic uplink and downlink sum rate in full-duplex mode by using the large number theorem, also giving the comparison of traditional half-duplex and full-duplex. Besides, we analyze the influence of Doppler shift on the performance of the system. Finally, the change of the system performance on the user velocity is illustrated.

This letter presents a model with queueing theory to analyze the performance of non-saturated IEEE 802.11 DCF networks. We use the closed queueing network model and derive an approximate representation of throughput which can reveal the relationship between the throughput and the total offered load under finite traffic load conditions. The accuracy of the model is verified by extensive simulations.

The proper allocation of network resources from a common physical substrate to a set of virtual networks (VNs) is one of the key technical challenges of network virtualization. While a variety of state-of-the-art algorithms have been proposed in an attempt to address this issue from different facets, the challenge still remains in the context of large-scale networks as the existing solutions mainly perform in a centralized manner which requires maintaining the overall and up-to-date information of the underlying substrate network. This implies the restricted scalability and computational efficiency when the network scale becomes large. This paper tackles the virtual network mapping problem and proposes a novel hierarchical algorithm in conjunction with a substrate network decomposition approach. By appropriately transforming the underlying substrate network into a collection of sub-networks, the hierarchical virtual network mapping algorithm can be carried out through a global virtual network mapping algorithm (GVNMA) and a local virtual network mapping algorithm (LVNMA) operated in the network central server and within individual sub-networks respectively with their cooperation and coordination as necessary. The proposed algorithm is assessed against the centralized approaches through a set of numerical simulation experiments for a range of network scenarios. The results show that the proposed hierarchical approach can be about 5-20 times faster for VN mapping tasks than conventional centralized approaches with acceptable communication overhead between GVNCA and LVNCA for all examined networks, whilst performs almost as well as the centralized solutions.

For the reconstruction of the feedback polynomial of a synchronous scrambler placed after a convolutional encoder, the existing algorithms require the prior knowledge of a dual word of the convolutional code. To address the case of a dual word being unknown, a new algorithm for the reconstruction of the feedback polynomial based on triple correlation characteristic of an m-sequence is proposed. First, the scrambled convolutional code sequence is divided into bit blocks; the product of the scrambled bit blocks with a dual word is proven to be an m-sequence with the same period as the synchronous scrambler. Second, based on the triple correlation characteristic of the generated m-sequence, a dual word is estimated; the generator polynomial of the generated m-sequence is computed by two locations of the triple correlation peaks. Finally, the feedback polynomial is reconstructed using the generator polynomial of the generated m-sequence. As the received sequence may contain bit errors, a method for detecting triple correlation peaks based on the constant false-alarm criterion is elaborated. Experimental results show that the proposed algorithm is effective. Ulike the existing algorithms available, there is no need to know a dual word a priori and the reconstruction result is more accurate. Moreover, the proposed algorithm is robust to bit errors.

A rigorous and systematical analysis of the performance of the power inversion adaptive array is shown where the method of moments is applied to analyze the antenna system and the electric field intensity is used as the descriptive parameter of the incident signals. This method can be easily extended to analyze adaptive arrays with arbitrary array structure and with wire antenna elements other than monopoles and dipoles. To show the adavantages of the method, the performance of adaptive arrays with mutual coupling effect among the array elements is considered and the inverted-F antennas (IFA) are used as the elements. The steering vector in the presence of mutual coupling which differs from that in the absence of the mutual coupling is derived. The output signal-to-interference-plus-noise ratio (SINR) of two kinds of adaptive arrays composed of IFA elements is computed and the importance of properly selecting steering vector is demostrated by means of the quiescent array patterns and the output SINR performance.

Several subband array methods have been proposed as useful means to perform joint spatio-temporal equalization in digital mobile communications. These methods can be applied to mitigate problems caused by the inter-symbol interference (ISI) and co-channel interference (CCI). The subband array methods proposed so far can be classified into two major schemes: (1) a centralized feedback scheme and (2) a localized feedback scheme. In this paper, we propose subband arrays with partial feedback scheme, which generalize the above two feedback schemes. The main contribution of this paper is to derive the steady-state mean square error (MSE) performance of subband arrays implementing these three different feedback schemes. Unlike the centralized feedback scheme which can be designed to provide the optimum equalization performance, the subband arrays with localized and partial feedback schemes are in general suboptimal. The performance of these two suboptimal feedback schemes depends on the channel characteristics, the filter banks employed, and the number of subbands.

In this paper, we propose a face representation approach using multi-orientation Log-Gabor local binary pattern (MOLGLBP) for realizing face recognition under facial expressions, illuminations and partial occlusions. Log-Gabor filters with different scales (frequencies) and orientations are applied on Y, I, and Q channel image in the YIQ color space respectively. Then Log-Gabor images of different orientations at the same scale are combined to form a multi-orientation Log-Gabor image (MOLGI) and two LBP operators are applied to it. For face recognition, histogram intersection metric is utilized to measure the similarity of faces. The proposed approach is evaluated on the CurtinFaces database and experiments demonstrate that the proposed approach is effectiveness against two simultaneous variations: expression & illumination, and illumination & occlusion.

Many recent studies have focused on leveraging rich information types to increase useful information for improving fault localization effectiveness. However, they rarely investigate the impact of information richness on fault localization to give guidance on how to enrich information for improving localization effectiveness. This paper presents the first systematic study to fill this void. Our study chooses four representative information types and investigates the relationship between their richness and the localization effectiveness. The results show that information richness related to frequency execution count involves a high risk of degrading the localization effectiveness, and backward slice is effective in improving localization effectiveness.

The capacity of optical transport networks has been increasing steadily and the networks are becoming more dynamic, complex, and transparent. Though it is common to use worst case assumptions for estimating the quality of transmission (QoT) in the physical layer, over provisioning results in high margin requirements. Accurate estimation on the QoT for to-be-established lightpaths is crucial for reducing provisioning margins. Machine learning (ML) is regarded as one of the most powerful methodological approaches to perform network data analysis and enable automated network self-configuration. In this paper, an artificial neural network (ANN) framework, a branch of ML, to estimate the optical signal-to-noise ratio (OSNR) of to-be-established lightpaths is proposed. It takes account of both nonlinear interference between spectrum neighboring channels and optical monitoring uncertainties. The link information vector of the lightpath is used as input and the OSNR of the lightpath is the target for output of the ANN. The nonlinear interference impact of the number of neighboring channels on the estimation accuracy is considered. Extensive simulation results show that the proposed OSNR estimation scheme can work with any RWA algorithm. High estimation accuracy of over 98% with estimation errors of less than 0.5dB can be achieved given enough training data. ANN model with R=4 neighboring channels should be used to achieve more accurate OSNR estimates. Based on the results, it is expected that the proposed ANN-based OSNR estimation for new lightpath provisioning can be a promising tool for margin reduction and low-cost operation of future optical transport networks.