High energy cost is a big challenge faced by the current data centers, wherein computing energy and cooling energy are main contributors to such cost. Consolidating workload onto fewer servers decreases the computing energy. However, it may result in thermal hotspots which typically consume greater cooling energy. Thus the tradeoff between computing energy decreasing and cooling energy decreasing is necessary for energy saving. In this paper, we propose a minimized-total-energy virtual machine (VM for short) migration model called C2vmMap based on efficient tradeoff between computing and cooling energies, with respect to two relationships: one for between the resource utilization and computing power and the other for among the resource utilization, the inlet and outlet temperatures of servers, and the cooling power. Regarding online resolution of the above model for better scalability, we propose a VM migration algorithm called C2vmMap_heur to decrease the total energy of a data center at run-time. We evaluate C2vmMap_heur under various workload scenarios. The real server experimental results show that C2vmMap_heur reduces up to 40.43% energy compared with the non-migration load balance algorithm. This algorithm saves up to 3x energy compared with the existing VM migration algorithm.

In recent years, deep learning based approaches have substantially improved the performance of face recognition. Most existing deep learning techniques work well, but neglect effective utilization of face correlation information. The resulting performance loss is noteworthy for personal appearance variations caused by factors such as illumination, pose, occlusion, and misalignment. We believe that face correlation information should be introduced to solve this network performance problem originating from by intra-personal variations. Recently, graph deep learning approaches have emerged for representing structured graph data. A graph is a powerful tool for representing complex information of the face image. In this paper, we survey the recent research related to the graph structure of Convolutional Neural Networks and try to devise a definition of graph structure included in Compressed Sensing and Deep Learning. This paper devoted to the story explain of two properties of our graph - sparse and depth. Sparse can be advantageous since features are more likely to be linearly separable and they are more robust. The depth means that this is a multi-resolution multi-channel learning process. We think that sparse graph based deep neural network can more effectively make similar objects to attract each other, the relative, different objects mutually exclusive, similar to a better sparse multi-resolution clustering. Based on this concept, we propose a sparse graph representation based on the face correlation information that is embedded via the sparse reconstruction and deep learning within an irregular domain. The resulting classification is remarkably robust. The proposed method achieves high recognition rates of 99.61% (94.67%) on the benchmark LFW (YTF) facial evaluation database.

A term is a connected acyclic graph (unrooted unordered tree) pattern with structured variables, which are ordered lists of one or more distinct vertices. A variable of a term has a variable label and can be replaced with an arbitrary tree by hyperedge replacement according to the variable label. The dimension of a term is the maximum number of vertices in the variables of it. A term is said to be linear if each variable label in it occurs exactly once. Let T be a tree and t a linear term. In this paper, we study the graph pattern matching problem (GPMP) for T and t, which decides whether or not T is obtained from t by replacing variables in t with some trees. First we show that GPMP for T and t is NP-complete if the dimension of t is greater than or equal to 4. Next we give a polynomial time algorithm for solving GPMP for a tree of bounded degree and a linear term of bounded dimension. Finally we show that GPMP for a tree of arbitrary degree and a linear term of dimension 2 is solvable in polynomial time.

The energy of a threshold circuit C is defined to be the maximum number of gates outputting ones for an input assignment, where the maximum is taken over all the input assignments. In this paper, we study computational power of threshold circuits of energy at most two. We present several results showing that the computational power of threshold circuits of energy one and the counterpart of energy two are remarkably different. In particular, we give an explicit function which requires an exponential size for threshold circuits of energy one, but is computable by a threshold circuit of size just two and energy two. We also consider MOD functions and Generalized Inner Product functions, and show that these functions also require exponential size for threshold circuits of energy one, but are computable by threshold circuits of substantially less size and energy two.

Card-based cryptographic protocols provide secure multi-party computations using a deck of physical cards. The most important primitive of those protocols is the shuffling operation, and most of the existing protocols rely on uniform cyclic shuffles (such as the random cut and random bisection cut) in which each possible outcome is equally likely and all possible outcomes constitute a cyclic subgroup. However, a couple of protocols with non-uniform and/or non-cyclic shuffles were proposed by Koch, Walzer, and Härtel at Asiacrypt 2015. Compared to the previous protocols, their protocols require fewer cards to securely produce a hidden AND value, although to implement of such unconventional shuffles appearing in their protocols remains an open problem. This paper introduces “pile-shifting scramble,” which can be a secure implementation of those shuffles. To implement such unconventional shuffles, we utilize physical cases that can store piles of cards, such as boxes and envelopes. Therefore, humans are able to perform the shuffles using these everyday objects. Furthermore, we show that a certain class of non-uniform and/or non-cyclic shuffles having two possible outcomes can be implemented by the pile-shifting scramble. This also implies that we can improve upon the known COPY protocol using three card cases so that the number of cases required can be reduced to two.

Generalized approximate message passing (GAMP) can be applied to compressive phase retrieval (CPR) with excellent phase-transition behavior. In this paper, we introduced the cartoon-texture model into the denoising-based phase retrieval GAMP(D-prGAMP), and proposed a cartoon-texture model based D-prGAMP (C-T D-prGAMP) algorithm. Then, based on experiments and analyses on the variations of the performance of D-PrGAMP algorithms with iterations, we proposed a 2-stage D-prGAMP algorithm, which makes tradeoffs between the C-T D-prGAMP algorithm and general D-prGAMP algorithms. Finally, facing the non-convergence issues of D-prGAMP, we incorporated adaptive damping to 2-stage D-prGAMP, and proposed the adaptively damped 2-stage D-prGAMP (2-stage ADD-prGAMP) algorithm. Simulation results show that, runtime of 2-stage D-prGAMP is relatively equivalent to that of BM3D-prGAMP, but 2-stage D-prGAMP can achieve higher image reconstruction quality than BM3D-prGAMP. 2-stage ADD-prGAMP spends more reconstruction time than 2-stage D-prGAMP and BM3D-prGAMP. But, 2-stage ADD-prGAMP can achieve PSNRs 0.2∼3dB higher than those of 2-stage D-prGAMP and 0.3∼3.1dB higher than those of BM3D-prGAMP.

We consider a device-to-device (D2D) underlaid cellular network where D2D communications are allowed to share the same radio spectrum with cellular uplink communications for improving spectral efficiency. However, to protect the cellular uplink communications, the interference level received at a base station (BS) from the D2D communications needs to be carefully maintained below a certain threshold, and thus the BS coordinates the transmit power of the D2D links. In this paper, we investigate on-off power control for the D2D links, which is known as a simple but effective technique due to its low signaling overhead. We first investigate the optimal on-off power control algorithm to maximize the sum-rate of the D2D links, while satisfying the interference constraint imposed by the BS. The computational complexity of the optimal algorithm drastically increases with D2D link number. Thus, we also propose an on-off power control algorithm to significantly reduce the computational complexity, compared to the optimal on-off power control algorithm. Extensive simulations validate that the proposed algorithm significantly reduces the computational complexity with a marginal sum-rate offset from the optimal algorithm.

Usowan is one of Nikoli's pencil puzzles. We study the computational complexity of Usowan puzzles. It is shown that deciding whether a given instance of the Usowan puzzle has a solution is NP-complete.

To learn the working situation of rod-pumped wells under ground, we always need to analyze dynamometer diagrams, which are generated by the load sensor and displacement sensor. Rod-pumped wells are usually located in the places with extreme weather, and these sensors are installed on some special oil equipments in the open air. As time goes by, sensors are prone to generating unstable and incorrect data. Unfortunately, load sensors are too expensive to frequently reinstall. Therefore, the resulting dynamometer diagrams sometimes cannot make an accurate diagnosis. Instead, as an absolutely necessary equipment of the rod-pumped well, the electric motor has much longer life and cannot be easily impacted by the weather. The electric power curve during a swabbing period can also reflect the working situation under ground, but is much harder to explain than the dynamometer diagram. This letter presented a novel deep learning architecture, which can transform the electric power curve into the dimensionless dynamometer diagram image. We conduct our experiments on a real-world dataset, and the results show that our method can get an impressive transformation accuracy.

We propose an analytical model for IEEE 802.11 wireless local area networks (WLANs). The analytical model uses macroscopic descriptions of the distributed coordination function (DCF): the backoff process is described by a few macroscopic states (medium-idle, transmission, and medium-busy), which obviates the need to track the specific backoff counter/backoff stages. We further assume that the transitions between the macroscopic states can be characterized as a continuous-time Markov chain under the assumption that state persistent times are exponentially distributed. This macroscopic description of DCF allows us to utilize a two-dimensional continuous-time Markov chain for simplifying DCF performance analysis and queueing processes. By comparison with simulation results, we show that the proposed model accurately estimates the throughput performance and average queue length under light, heavy, or asymmetric traffic.

This work is conducted to solve the current problem in the attribute-based keyword search (ABKS) scheme about how to securely and efficiently delegate the search rights to other users when the authorized user is not online. We first combine proxy re-encryption (PRE) with the ABKS technology and propose a scheme called attribute-based keyword search with proxy re-encryption (PABKS). The scheme not only realizes the functions of data search and fine-grained access control, but also supports search function sharing. In addition, we randomly blind the user's private key to the server, which ensures the confidentiality and security of the private key. Then, we also prove that the scheme is selective access structure and chosen keyword attack (IND-sAS-CKA) secured in the random oracle model. A performance analysis and security proof show that the proposed scheme can achieve efficient and secure data search in the cloud.

A triangular decomposition-based multipath super-resolution method is proposed to improve the range resolution of small unmanned aerial vehicle (UAV) radar altimeters that use a single channel with continuous direct spread waveform. In the engineering applications of small UAV radar altimeter, multipath scenarios are quite common. When the conventional matched filtering process is used under these environments, it is difficult to identify multiple targets in the same range cell due to the overlap between echoes. To improve the performance, we decompose the overlapped peaks yielded by matched filtering into a series of basic triangular waveforms to identify various targets with different time-shifted correlations of the pseudo-noise (PN) sequence. Shifting the time scale enables targets in the same range resolution unit to be identified. Both theoretical analysis and experiments show that the range resolution can be improved significantly, as it outperforms traditional matched filtering processes.

In recent years, deep learning based image recognition, speech recognition, text translation and other related applications have brought great convenience to people's lives. With the advent of the era of internet of everything, how to run a computationally intensive deep learning algorithm on a limited resources edge device is a major challenge. For an edge oriented computing vector processor, combined with a specific neural network model, a new data layout method for putting the input feature maps in DDR, rearrangement of the convolutional kernel parameters in the nuclear memory bank is proposed. Aiming at the difficulty of parallelism of two-dimensional matrix convolution, a method of parallelizing the matrix convolution calculation in the third dimension is proposed, by setting the vector register with zero as the initial value of the max pooling to fuse the rectified linear unit (ReLU) activation function and pooling operations to reduce the repeated access to intermediate data. On the basis of single core implementation, a multi-core implementation scheme of Inception structure is proposed. Finally, based on the proposed vectorization method, we realize five kinds of neural network models, namely, AlexNet, VGG16, VGG19, GoogLeNet, ResNet18, and performance statistics and analysis based on CPU, gtx1080TI and FT2000 are presented. Experimental results show that the vector processor has better computing advantages than CPU and GPU, and can calculate large-scale neural network model in real time.

A new method for adaptive digital beamforming technique with compressed sensing (CS) for sparse receiving arrays with gain/phase uncertainties is presented. Because of the sparsity of the arriving signals, CS theory can be adopted to sample and recover receiving signals with less data. But due to the existence of the gain/phase uncertainties, the sparse representation of the signal is not optimal. In order to eliminating the influence of the gain/phase uncertainties to the sparse representation, most present study focus on calibrating the gain/phase uncertainties first. To overcome the effect of the gain/phase uncertainties, a new dictionary optimization method based on the total least squares (TLS) algorithm is proposed in this paper. We transfer the array signal receiving model with the gain/phase uncertainties into an EIV model, treating the gain/phase uncertainties effect as an additive error matrix. The method we proposed in this paper reconstructs the data by estimating the sparse coefficients using CS signal reconstruction algorithm and using TLS method toupdate error matrix with gain/phase uncertainties. Simulation results show that the sparse regularized total least squares algorithm can recover the receiving signals better with the effect of gain/phase uncertainties. Then adaptive digital beamforming algorithms are adopted to form antenna beam using the recovered data.

As the capabilities and costs of Artificial Intelligence (AI) and of sensors (IoT) continue to improve, the concept of a “control system” can evolve beyond the operation of a discrete technical system based on numerical information and enter the realm of large-scale systems with both technical and social characteristics based on both numerical and unstructured information. This evolution has particular significance for applying the principles of Autonomous Decentralised Systems (ADS) [1]. This article considers the possible roles for ADS in complex technical and social systems extending up to global scales.

A composite system consists of many subsystems, which have interconnections with other subsystems. For such a system, in general, we utilize decentralized control, where each subsystem is controlled by a local controller. On the other hand, event-triggered control is one of useful approaches to reduce the amount of communications between a controller and a plant. In the event-triggered control, an event triggering mechanism (ETM) monitors the information of the plant, and determines the time to transmit the data. In this paper, we propose a design of ETMs for the decentralized event-triggered control of nonlinear composite systems using an M-matrix. We consider the composite system where there is an ETM for each subsystem, and ETMs monitor local states of the corresponding subsystems. Each ETM is designed so that the composite system is stabilized. Moreover, we deal with the case of linear systems. Finally, we perform simulation to show that the proposed triggering rules are useful for decentralized control.

We prove generalization error bounds of classes of low-rank matrices with some norm constraints for collaborative filtering tasks. Our bounds are tighter, compared to known bounds using rank or the related quantity only, by taking the additional L1 and L∞ constraints into account. Also, we show that our bounds on the Rademacher complexity of the classes are optimal.

The k-error linear complexity of a sequence is a fundamental concept for assessing the stability of the linear complexity. After computing the k-error linear complexity of a sequence, those bits that cause the linear complexity reduced also need to be determined. For binary sequences with period 2pn, where p is an odd prime and 2 is a primitive root modulo p2, we present an algorithm which computes the minimum number k such that the k-error linear complexity is not greater than a given constant c. The corresponding error sequence is also obtained.

Although Wolf Pack Algorithm (WPA) is a novel optimal algorithm with good performance, there is still room for improvement with respect to its convergence. In order to speed up its convergence and strengthen the search ability, we improve WPA with the Differential Evolution (DE) elite set strategy. The new proposed algorithm is called the WPADEES for short. WPADEES is faster than WPA in convergence, and it has a more feasible adaptability for various optimizations. Six standard benchmark functions are applied to verify the effects of these improvements. Our experiments show that the performance of WPADEES is superior to the standard WPA and other intelligence optimal algorithms, such as GA, DE, PSO, and ABC, in several situations.

Behaviors of an object-oriented system can be visualized as reverse-engineered sequence diagrams from execution traces. This approach is a valuable tool for program comprehension tasks. However, owing to the massiveness of information contained in an execution trace, a reverse-engineered sequence diagram is often afflicted by a scalability issue. To address this issue, many trace summarization techniques have been proposed. Most of the previous techniques focused on reducing the vertical size of the diagram. To cope with the scalability issue, decreasing the horizontal size of the diagram is also very important. Nonetheless, few studies have addressed this point; thus, there is a lot of needs for further development of horizontal summarization techniques. We present in this paper a method for identifying core objects for trace summarization by analyzing reference relations and dynamic properties. Visualizing only interactions related to core objects, we can obtain a horizontally compactified reverse-engineered sequence diagram that contains system's key behaviors. To identify core objects, first, we detect and eliminate temporary objects that are trivial for a system by analyzing reference relations and lifetimes of objects. Then, estimating the importance of each non-trivial object based on their dynamic properties, we identify highly important ones (i.e., core objects). We implemented our technique in our tool and evaluated it by using traces from various open-source software systems. The results showed that our technique was much more effective in terms of the horizontal reduction of a reverse-engineered sequence diagram, compared with the state-of-the-art trace summarization technique. The horizontal compression ratio of our technique was 134.6 on average, whereas that of the state-of-the-art technique was 11.5. The runtime overhead imposed by our technique was 167.6% on average. This overhead is relatively small compared with recent scalable dynamic analysis techniques, which shows the practicality of our technique. Overall, our technique can achieve a significant reduction of the horizontal size of a reverse-engineered sequence diagram with a small overhead and is expected to be a valuable tool for program comprehension.