In this paper, a new single soft switched forward converter with a self driven synchronous rectification (SDSR) is introduced. In the proposed converter, a soft switching condition (ZCS turn on and ZVS turn off) is provided for the switch, by an auxiliary circuit without any extra switch. In additional, this auxiliary circuit does not impose high voltage or current stresses on the converter. Since the proposed converter uses SDSR to reduce conductive loss of output rectifier, the rectifier switches are switched under soft switching condition. So, the conductive and switching losses on the converter reduce considerably. Also, implementing control circuit of this converter is very simple, due to the self-driven method employed in driving synchronous rectification and the converter is controlled by pulse width modulation (PWM). The experimental results of the proposed converter are presented to confirm the theoretical analysis.

The effect of annealing on the field emission characteristics of a field emitter comprising diamond micropowder was investigated. The threshold voltage Vth at which the emission current begins to flow decreased as the annealing temperature increased, and a minimum Vth was obtained at an annealing temperature of 1345K. The reduction in threshold voltage was due to a reduction in the work function with annealing.

We present a method for recognition of acoustic events in conversation scenarios where speech usually overlaps with other acoustic events. While speech is usually considered the most informative acoustic event in a conversation scene, it does not always contain all the information. Non-speech events, such as a door knock, steps, or a keyboard typing can reveal aspects of the scene that speakers miss or avoid to mention. Moreover, being able to robustly detect these events could further support speech enhancement and recognition systems by providing useful information cues about the surrounding scenarios and noise. In acoustic event detection, state-of-the-art techniques are typically based on derived features (e.g. MFCC, or Mel-filter-banks) which have successfully parameterized the spectrogram of speech but reduce resolution and detail when we are targeting other kinds of events. In this paper, we propose a method that learns features in an unsupervised manner from high-resolution spectrogram patches (considering a patch as a certain number of consecutive frame features stacked together), and integrates within the deep neural network framework to detect and classify acoustic events. Superiority over both previous works in the field, and similar approaches based on derived features, has been assessed by statical measures and evaluation with CHIL2007 corpus, an annotated database of seminar recordings.

Integral image is the sum of input image pixel values. It is mainly used to speed up the process of a box filter operation, such as Haar-like features. However, large memory capacity for integral image data can be an obstacle in an embedded environment with limited hardware. In a previous research, [5] reduced the size of integral image memory using 2×2 block structure with additional calculations. It can be easily extended to n×n block structure for further reduction, but it requires more additional calculations. In this paper, we propose a new block structure for the integral image by modifying the location of the reference pixel in the block. It results in much less additional calculations by reducing the number of memory accesses, while keeping the same amount of memory as the original block structure.

By investigating the properties that the offsets should satisfy, this letter presents a brief proof of general QAM Golay complementary sequences (GCSs) in Cases I-III constructions. Our aim is to provide a brief, clear, and intelligible derivation so that it is easy for the reader to understand the known Cases I-III constructions of general QAM GCSs.

Millimeter-wave (mmw) frequency bands, especially 60GHz unlicensed band, are considered as a promising solution for gigabit short range wireless communication systems. IEEE standard 802.11ad, also known as WiGig, is standardized for the usage of the 60GHz unlicensed band for wireless local area networks (WLANs). By using this mmw WLAN, multi-Gbps rate can be achieved to support bandwidth-intensive multimedia applications. Exhaustive search along with beamforming (BF) is usually used to overcome 60GHz channel propagation loss and accomplish data transmissions in such mmw WLANs. Because of its short range transmission with a high susceptibility to path blocking, multiple number of mmw access points (APs) should be used to fully cover a typical target environment for future high capacity multi-Gbps WLANs. Therefore, coordination among mmw APs is highly needed to overcome packet collisions resulting from un-coordinated exhaustive search BF and to increase total capacity of mmw WLANs. In this paper, we firstly give the current status of mmw WLANs with our developed WiGig AP prototype. Then, we highlight the great need for coordinated transmissions among mmw APs as a key enabler for future high capacity mmw WLANs. Two different types of coordinated mmw WLAN architecture are introduced. One is distributed antenna type architecture to realize centralized coordination, while the other is autonomous coordination with the assistance of legacy Wi-Fi signaling. Moreover, two heterogeneous network (HetNet) architectures are also introduced to efficiently extend the coordinated mmw WLANs to be used for future 5th Generation (5G) cellular networks.

In this letter, we propose Dynamic Linear Bellman Combination that allows us to composite optimal policies of Kullback-Leibler control for similar tasks with different terminal and non-terminal costs. Simulation results demonstrate the effectiveness of our method.

Sparsity-based methods have been recently applied to abnormal event detection and have achieved impressive results. However, most such methods suffer from the problem of dimensionality curse; furthermore, they also take no consideration of the relationship among coefficient vectors. In this paper, we propose a novel method called consistent sparse representation (CSR) to overcome the drawbacks. We first reconstruct each feature in the space spanned by the clustering centers of training features so as to reduce the dimensionality of features and preserve the neighboring structure. Then, the consistent regularization is added to the sparse representation model, which explicitly considers the relationship of coefficient vectors. Our method is verified on two challenging databases (UCSD Ped1 database and Subway batabase), and the experimental results demonstrate that our method obtains better results than previous methods in abnormal event detection.

In this paper, we first propose ten new discrimination features of SAR images in the moving and stationary target acquisition and recognition (MSTAR) database. The Ada_MCBoost algorithm is then proposed to classify multiclass SAR targets. In the new algorithm, we introduce a novel large-margin loss function to design a multiclass classifier directly instead of decomposing the multiclass problem into a set of binary ones through the error-correcting output codes (ECOC) method. Finally, experiments show that the new features are helpful for SAR targets discrimination; the new algorithm had better recognition performance than three other contrast methods.

Most research projects with respect to energy saving are trying to improve power efficiency and are using software to manage the power systems in the power on mode; but in our design, we modify the original Suspend to RAM mode-S3 state, which is the 3rd system state as defined by the ACPI specification, in order to reduce power consumption. We've redesigned the control circuit to save power while a PC is in the standby mode. First, we re-examine the entire circuit in the standby mode, and clarify which chip is used both to wake up the system and to turn off all unnecessary standby power previously used by the chips. Secondly, we redesign the power sequence and use an additional chip to control the system power supply, to allow a PC's normal system's operation to turn off the unnecessary power control chips. Third, in order to save power supply in the standby mode, we have simplified the multiple remote wake-up mechanism to control the remote boot device. The improvement shows that our design reduced power consumption to 0.21W from the original 0.56W while all the remote wake-up functions are disabled; and consumes 0.42W when using multiple remote wake-up functions. We implement the above modification from the legacy S3 state, and obtain lower power consumption. In order to distinguish the standby states, we name the modified S3 state as Deep S3 state.

A low loss intelligent power module (IPM) that specifically designed for high performance frequency-alterable air conditioner applications is proposed. This IPM utilizes 600 V trench gate field stop insulated gate bipolar transistors (TFS-IGBTs) as the main switching devices to deliver extremely low conduction and switching losses. In addition, 600 V SiC schottky barrier diodes (SBDs) are employed as the freewheeling diodes. Compared to conventional silicon fast recovery diodes (FRDs) SiC SBDs exhibit practically no reverse recovery loss, hence can further reduce the power loss of the IPM. Experimental results reveal that the power loss of the proposed IPM is between 3.5∼21.7 W at different compressor frequencies from 10 to 70 Hz, which achieving up to 12.5%∼25.5% improvement when compared to the state-of-the-art conventional Si-based IGBT IPM.

This letter deals with direction-of-arrival (DOA) estimate problem based on gravitational search algorithm (GSA) with multiple signal classification (MUSIC) criterion for code-division multiple access (CDMA) signals. It has been shown that the estimate accuracy of the searching-based MUSIC estimator strictly depends on the number of search grids used during the search process, which is time consuming and the required number of search grids is not clear to determine. In conjunction with the GSA-based optimization, the high resolution DOA estimation can be obtained; meanwhile the searching grid size is no need to know previously. In this letter, we firstly present a GSA-based DOA estimator with MUSIC criterion under high interferer-to-noise ratio circumstances. Second, for the purpose to increase the estimation accuracy, we also propose an improved GSA with adaptive multiple accelerations, which depend on Newton-Raphson method. Several computer simulations are provided for illustration and comparison.

In this letter, we present a meet-in-the-middle attack on the 5-round reduced Kuznyechik cipher which has been recently chosen to be standardized by the Russian federation. Our attack is based on the differential enumeration approach. However, the application of the exact approach is not successful on Kuznyechik due to its optimal round diffusion properties. Accordingly, we adopt an equivalent representation for the last round where we can efficiently filter ciphertext pairs and launch the attack in the chosen ciphertext setting. We also utilize partial sequence matching which further reduces the memory and time complexities. For the 5-round reduced cipher, the 256-bit master key is recovered with an online time complexity of 2140.3, a memory complexity of 2153.3, and a data complexity of 2113.

With aggressive device scaling, timing failures have become more prevalent due to manufacturing defects and process variations. When timing failure occurs, it is important to take corrective actions immediately. Therefore, an efficient and fast diagnosis method is essential. In this paper, we propose a new diagnostic method using timing information. Our method approximately estimates all the segment delays of measured paths in a design, using inequality-constrained least squares methods. Then, the proposed method ranks the possible locations of delay defects based on the difference between estimated segment delays and the expected values of segment delays. The method works well for multiple delay defects as well as single delay defects. Experiment results show that our method yields good diagnostic resolution. With the proposed method, the average first hit rank (FHR), was within 7 for single delay defect and within 8 for multiple delay defects.

Let v=p1m1p2m2…ptmt be the canonical prime factorization of v. In this paper, we give a construction of optimal ((s+1)×v,s+1,1) two-dimensional optical orthogonal codes with both at most one-pulse per wavelength and at most one-pulse per time slot, where s | gcd(p1-1,p2-1,...,pt-1). The method is much simpler than that in [1]. Optimal (m×v,k,1) two-dimensional optical orthogonal codes are also constructed based on the Steiner system S[2,k,m].

Given a graph G, a set of spanning trees of G are completely independent if for any vertices x and y, the paths connecting them on these trees have neither vertex nor edge in common, except x and y. In this paper, we prove that for graphs of order n, with n ≥ 6, if the minimum degree is at least n-2, then there are at least ⌊n/3⌋ completely independent spanning trees.

For an odd prime p and a positive integer r, new classes of binary sequences with period pr+1 are proposed from Euler quotients in this letter, which include several known classes of binary sequences derived from Fermat quotients and Euler quotients as special cases. The advantage of the new constructions is that they allow one to choose their support sets freely. Furthermore, with some constrains on the support set, the new sequences are proved to possess large linear complexities under the assumption of 2p-1 ≢ 1 mod p2.

A high energy-efficiency and area-reduction switching scheme for a low-power successive approximation register (SAR) analog-to-digital converter (ADC) is presented. Based on the sequence initialization, monotonic capacitor switching procedure and multiple reference voltages, the average switching energy and total capacitance of the proposed scheme are reduced by 99.4% and 87.5% respectively, compared to the conventional architecture.

We propose efficient algorithms for Sorting k-Sets in Bins. The Sorting k-Sets in Bins problem can be described as follows. We are given numbered n bins with k balls in each bin. Balls in the i-th bin are numbered n-i+1. We can only swap balls between adjacent bins. Our task is to move all of the balls to the same numbered bins. For this problem, we give an efficient greedy algorithm with $rac{k+1}{4}n^2+O(k+n)$ swaps and provide a detailed analysis for k=3. In addition, we give a more efficient recursive algorithm using $rac{15}{16}n^2+O(n)$ swaps for k=3.