Deep neural network (NN) has been widely accepted for enabling various AI applications, however, the limitation of computational and memory resources is a major problem on mobile devices. Quantized NN with a reduced bit precision is an effective solution, which relaxes the resource requirements, but the accuracy degradation due to its numerical approximation is another problem. We propose a novel quantized NN model employing the “dithering” technique to improve the accuracy with the minimal additional hardware requirement at the view point of the hardware-algorithm co-designing. Dithering distributes the quantization error occurring at each pixel (neuron) spatially so that the total information loss of the plane would be minimized. The experiment we conducted using the software-based accuracy evaluation and FPGA-based hardware resource estimation proved the effectiveness and efficiency of the concept of an NN model with dithering.

Most of latency-sensitive mobile applications depend on computational resources provided by a cloud computing service. The problem of relying on cloud computing is that, sometimes, the physical locations of cloud servers are distant from mobile users and the communication latency is long. As a result, the concept of distributed cloud service, called mobile edge computing (MEC), is being introduced in the 5G network. However, MEC can reduce only the communication latency. The computing latency in MEC must also be considered to satisfy the required total latency of services. In this research, we study the impact of both latencies in MEC architecture with regard to latency-sensitive services. We also consider a centralized model, in which we use a controller to manage flows between users and mobile edge resources to analyze MEC in a practical architecture. Simulations show that the interval and controller latency trigger some blocking and error in the system. However, the permissive system which relaxes latency constraints and chooses an edge server by the lowest total latency can improve the system performance impressively.

Many quality metrics have been proposed for the compliance perception to assess haptic device performance and perceived results. Perceived compliance may be influenced by factors such as object properties, experimental conditions and human perceptual habits. In this paper, analysis of softness perception was conducted to find out relevant quality metrics dominating in the compliance perception system and their correlation with perception results, by expressing these metrics by basic physical parameters that characterizing these factors. Based on three psychophysical experiments, just noticeable differences (JNDs) for perceived softness of combination of different stiffness coefficients and damping levels rendered by haptic devices were analyzed. Interaction data during the interaction process were recorded and analyzed. Preliminary experimental results show that the discrimination ability of softness perception changes with the ratio of damping to stiffness when subjects exploring at their habitual speed. Analysis results indicate that quality metrics of Rate-hardness, Extended Rate-hardness and ratio of damping to stiffness have high correlation for perceived results. Further analysis results show that parameters that reflecting object properties (stiffness, damping), experimental conditions (force bandwidth) and human perceptual habits (initial speed, maximum force change rate) lead to the change of these quality metrics, which then bring different perceptual feeling and finally result in the change of discrimination ability. Findings in this paper may provide a better understanding of softness perception and useful guidance in improvement of haptic and teleoperation devices.

This paper presents a hybrid architecture for a neural network (NN) trained by a particle swarm optimization (PSO) algorithm. The NN is implemented on the hardware side while the PSO is executed by a processor on the software side. In addition, principal component analysis (PCA) is also applied to reduce correlated information. The PCA module is implemented in hardware by the SystemVerilog programming language to increase operating speed. Experimental results showed that the proposed architecture had been successfully implemented. In addition, the hardware-based NN trained by PSO (NN-PSO) program was faster than the software-based NN trained by the PSO program. The proposed NN-PSO with PCA also obtained better recognition rates than the NN-PSO without-PCA.

Recognizing the different segments of speech belonging to the same speaker is an important speech analysis task in various applications. Recent works have shown that there was an underlying manifold on which speaker utterances live in the model-parameter space. However, most speaker clustering methods work on the Euclidean space, and hence often fail to discover the intrinsic geometrical structure of the data space and fail to use such kind of features. For this problem, we consider to convert the speaker i-vector representation of utterances in the Euclidean space into a network structure constructed based on the local (k) nearest neighbor relationship of these signals. We then propose an efficient community detection model on the speaker content network for clustering signals. The new model is based on the probabilistic community memberships, and is further refined with the idea that: if two connected nodes have a high similarity, their community membership distributions in the model should be made close. This refinement enhances the local invariance assumption, and thus better respects the structure of the underlying manifold than the existing community detection methods. Some experiments are conducted on graphs built from two Chinese speech databases and a NIST 2008 Speaker Recognition Evaluations (SREs). The results provided the insight into the structure of the speakers present in the data and also confirmed the effectiveness of the proposed new method. Our new method yields better performance compared to with the other state-of-the-art clustering algorithms. Metrics for constructing speaker content graph is also discussed.

In this letter, a polarization filtering based transmission (PFBT) scheme is proposed to enhance the spectrum efficiency in wireless communications. In such scheme, the information is divided into several parts and each is conveyed by a polarized signal with a unique polarization state (PS). Then, the polarized signals are added up and transmitted by the dual-polarized antenna. At the receiver side, the oblique projection polarization filters (OPPFs) are adopted to separate each polarized signal. Thus, they can be demodulated separately. We mainly focus on the construction methods of the OPPF matrix when the number of the separate parts is 2 and 3 and evaluate the performance in terms of the capacity and the bit error rate. In addition, we also discuss the probability of the signal separation when the number of the separate parts is equal or greater than 4. Theoretical results and simulation results demonstrate the performance of the proposed scheme.

In order to solve the influence of scale change on target tracking using the drone, a multi-scale target tracking algorithm is proposed which based on the color feature tracking algorithm. The algorithm realized adaptive scale tracking by training position and scale correlation filters. It can first obtain the target center position of next frame by computing the maximum of the response, where the position correlation filter is learned by the least squares classifier and the dimensionality reduction for color features is analyzed by principal component analysis. The scale correlation filter is obtained by color characteristics at 33 rectangular areas which is set by the scale factor around the central location and is reduced dimensions by orthogonal triangle decomposition. Finally, the location and size of the target are updated by the maximum of the response. By testing 13 challenging video sequences taken by the drone, the results show that the algorithm has adaptability to the changes in the target scale and its robustness along with many other performance indicators are both better than the most state-of-the-art methods in illumination Variation, fast motion, motion blur and other complex situations.

The distinctive characteristics of unmanned aerial vehicle networks (UAVNs), including highly dynamic network topology, high mobility, and open-air wireless environments, may make UAVNs vulnerable to attacks and threats. Due to the special security requirements, researching in the high reliability of the power and communication network in drone monitoring system become special important. The reliability of the communication network and power in the drone monitoring system has been studied. In order to assess the reliability of the system power supply in the drone emergency monitoring system, the accelerated life tests under constant stress were presented based on the exponential distribution. Through a comparative analysis of lots of factors, the temperature was chosen as the constant accelerated stress parameter. With regard to the data statistical analysis, the type-I censoring sample method was put forward. The mathematical model of the drone monitoring power supply was established and the average life expectancy curve was obtained under different temperatures through the analysis of experimental data. The results demonstrated that the mathematical model and the average life expectancy curve were fit for the actual very well. With overall consideration of the communication network topology structure and network capacity the improved EED-SDP method was put forward in drone monitoring. It is concluded that reliability analysis of power and communication network in drone monitoring system is remarkably important to improve the reliability of drone monitoring system.

Fault prediction aims to identify whether a software module is defect-prone or not according to metrics that are mined from software projects. These metric values, also known as features, may involve irrelevance and redundancy, which hurt the performance of fault prediction models. In order to filter out irrelevant and redundant features, a Hybrid Feature Selection (abbreviated as HFS) method for software fault prediction is proposed. The proposed HFS method consists of two major stages. First, HFS groups features with hierarchical agglomerative clustering; second, HFS selects the most valuable features from each cluster to remove irrelevant and redundant ones based on two wrapper based strategies. The empirical evaluation was conducted on 11 widely-studied NASA projects, using three different classifiers with four performance metrics (precision, recall, F-measure, and AUC). Comparison with six filter-based feature selection methods demonstrates that HFS achieves higher average F-measure and AUC values. Compared with two classic wrapper feature selection methods, HFS can obtain a competitive prediction performance in terms of average AUC while significantly reducing the computation cost of the wrapper process.

We propose two secret sharing schemes realizing general access structures, which are based on unauthorized subsets. In the proposed schemes, shares are generated by Tassa's (k,n)-hierarchical threshold scheme instead of Shamir's (k,n)-threshold scheme. Consequently, the proposed schemes can reduce the number of shares distributed to each participant.

In previous studies, we determined that workloads often contain many input-output (IO) concentrations. Such concentrations are aggregations of IO accesses. They appear in narrow regions of a storage volume and continue for durations of up to about an hour. These narrow regions occupy a small percentage of the logical unit number capacity, include most IO accesses, and appear at unpredictable logical block addresses. We investigated these workloads by focusing on page-level regularity and found that they often include few regularities. This means that simple caching may not reduce the response time for these workloads sufficiently because the cache migration algorithm uses page-level regularity. We previously developed an on-the-fly automated storage tiering (OTF-AST) system consisting of an SSD and an HDD. The migration algorithm identifies IO concentrations with moderately long durations and migrates them from the HDD to the SSD. This means that there is little or no reduction in the response time when the workload includes few such concentrations. We have now developed a hybrid storage system consisting of a cache drive with an SSD and HDD and a multi-tier SSD that uses OTF-AST, called “OTF-AST with caching.” The OTF-AST scheme handles the IO accesses that produce moderately long duration IO concentrations while the caching scheme handles the remaining IO accesses. Experiments showed that the average response time for our system was 45% that of Facebook FlashCache on a Microsoft Research Cambridge workload.

Recently, Network Function Virtualization (NFV) has drawn attentions of many network researchers with great deal of flexibilities, and various network service chains can be used in an SDN/NFV environment. With the flexibility of virtual middlebox placement, how to place virtual middleboxes in order to optimize the performance of service chains becomes essential. Some past studies focused on placement problem of consolidated middleboxes which combine multiple functions into a virtual middlebox. However, when a virtual middlebox providing only a single function is considered, the placement problem becomes much more complex. In this paper, we propose a new heuristic method, the gradual switch clustering based virtual middlebox placement method, in order to improve the performance of service chains, with the constraints of end-to-end delay, bandwidth, and operation cost of deploying a virtual middlebox on a switch. The proposed method gradually finds candidate places for each type of virtual middlebox along with the sequential order of service chains, by clustering candidate switches which satisfy the constraints. Finally, among candidate places for each type of virtual middlebox, the best places are selected in order to minimize the end-to-end delays of service chains. The evaluation results, which are obtained through Mininet based extensive emulations, show that the proposed method outperforms than other methods, and specifically it achieves around 25% less end-to-end delay than other methods.

With the acceptance of social sharing, public bike sharing services have become popular worldwide. One of the most important tasks in operating a bike sharing system is managing the bike supply at each station to avoid either running out of bicycles or docks to park them. This requires the system operator to redistribute bicycles from overcrowded stations to under-supplied ones. Trip demand prediction plays a crucial role in improving redistribution strategies. Predicting trip demand is a highly challenging problem because it is influenced by multiple levels of factors, both environmental and individual, e.g., weather and user characteristics. Although several existing studies successfully address either of them in isolation, no framework exists that can consider all factors simultaneously. This paper starts by analyzing trip data from real-world bike-sharing systems. The analysis reveals the interplay of the multiple levels of the factors. Based on the analysis results, we develop a novel form of the point process; it jointly incorporates multiple levels of factors to predict trip demand, i.e., predicting the pick-up and drop-off levels in the future and when over-demand is likely to occur. Our extensive experiments on real-world bike sharing systems demonstrate the superiority of our trip demand prediction method over five existing methods.

Various systems that share remote co-worker's awareness information have been proposed for realizing efficient collaborative work among distributed offices. In this study, we implemented an interruptibility sharing system in a university laboratory and assessed the observation behavior for the displayed information. Observation behavior for each target member was detected using an eye tracker to discuss the usage and effect of the system in a quantitative manner, along with the considerations of workers' job positions and relationships. The results suggested that participants observed interruptibility information approximately once an hour while at their desks. Observations were frequent during break-times rather than when the participants wanted to communicate with others. The most frequently observed targets were the participants themselves. The participants gazed the laboratory members not only in a close work relationship but also in a weak relationship. Results suggested that sharing of interruptibility information assists worker's self-reflection and contributes to the establishment of horizontal connection in an organization including members in weak work relationship.

In recent years, multi-party computation (MPC) frameworks based on replicated secret sharing schemes (RSSS) have attracted the attention as a method to achieve high efficiency among known MPCs. However, the RSSS-based MPCs are still inefficient for several heavy computations like algebraic operations, as they require a large amount and number of communication proportional to the number of multiplications in the operations (which is not the case with other secret sharing-based MPCs). In this paper, we propose RSSS-based three-party computation protocols for modular exponentiation, which is one of the most popular algebraic operations, on the case where the base is public and the exponent is private. Our proposed schemes are simple and efficient in both of the asymptotic and practical sense. On the asymptotic efficiency, the proposed schemes require O(n)-bit communication and O(1) rounds,where n is the secret-value size, in the best setting, whereas the previous scheme requires O(n2)-bit communication and O(n) rounds. On the practical efficiency, we show the performance of our protocol by experiments on the scenario for distributed signatures, which is useful for secure key management on the distributed environment (e.g., distributed ledgers). As one of the cases, our implementation performs a modular exponentiation on a 3,072-bit discrete-log group and 256-bit exponent with roughly 300ms, which is an acceptable parameter for 128-bit security, even in the WAN setting.

We propose a simple framework for evaluating the performance of pairing-based cryptographic schemes for various types of curves and parameter settings. The framework, which we call ‘Opcount’, enables the selection of an appropriate curve and parameters by estimating the performance of a cryptographic scheme from a pseudo-code describing the cryptographic scheme and an implementation-information database that records the performance of basic operations in curves targeted for evaluation. We apply Opcount to evaluate and compare the computational efficiency of several structure-preserving signature schemes that involve tens of pairing products in their signature verification. In addition to showing the usefulness of Opcount, our experiments also reveal the overlooked importance of taking account of the properties of underlying curves when optimizing computations and demonstrate the impact of tight security reductions.

This paper proposes an anomaly prediction method based on k-means clustering that assumes embedded devices with memory constraints. With this method, by checking control system behavior in detail using k-means clustering, it is possible to predict anomalies. However, continuing clustering is difficult because data accumulate in memory similar to existing k-means clustering method, which is problematic for embedded devices with low memory capacity. Therefore, we also propose k-means clustering to continue clustering for infinite stream data. The proposed k-means clustering method is based on online k-means clustering of sequential processing. The proposed k-means clustering method only stores data required for anomaly prediction and releases other data from memory. Due to these characteristics, the proposed k-means clustering realizes that anomaly prediction is performed by reducing memory consumption. Experiments were performed with actual data of control system for anomaly prediction. Experimental results show that the proposed anomaly prediction method can predict anomaly, and the proposed k-means clustering can predict anomalies similar to standard k-means clustering while reducing memory consumption. Moreover, the proposed k-means clustering demonstrates better results of anomaly prediction than existing online k-means clustering.

Epidemic cyber incidents are caused by malicious websites using exploit kits. The exploit kit facilitate attackers to perform the drive-by download (DBD) attack. However, it is reported that malicious websites using an exploit kit have similarity in their website structure (WS)-trees. Hence, malicious website identification techniques leveraging WS-trees have been studied, where the WS-trees can be estimated from HTTP traffic data. Nevertheless, the defensive component of the exploit kit prevents us from capturing the WS-tree perfectly. This paper shows, hence, a new WS-tree construction procedure by using the fact that a DBD attack happens in a certain duration. This paper proposes, moreover, a new malicious website identification technique by clustering the WS-tree of the exploit kits. Experiment results assuming the D3M dataset verify that the proposed technique identifies exploit kits with a reasonable accuracy even when HTTP traffic from the malicious sites are partially lost.

Forecasting commodity futures price is a challenging task. We present an algorithm to predict the trend of commodity futures price based on a type of structuring data and back propagation neural network. The random volatility of futures can be filtered out in the structuring data. Moreover, it is not restricted by the type of futures contract. Experiments show the algorithm can achieve 80% accuracy in predicting price trends.

Dynamic sharing of limited radio spectrum resources is expected to satisfy the increasing demand for spectrum resources in the upcoming 5th generation mobile communication system (5G) era and beyond. Distributed real-time spectrum sensing is a key enabler of dynamic spectrum sharing, but the costs incurred in observed-data transmission are a critical problem, especially when massive numbers of spectrum sensors are deployed. To cope with this issue, the proposed spectrum sensors learn the ambient radio environment in real-time and create a time-spectral model whose parameters are shared with servers operating in the edge-computing layer. This process makes it possible to significantly reduce the communication cost of the sensors because frequent data transmission is no longer needed while enabling the edge servers to keep up on the current status of the radio environment. On the basis of the created time-spectral model, sharable spectrum resources are dynamically harvested and allocated in terms of geospatial, temporal, and frequency-spectral domains when accepting an application for secondary-spectrum use. A web-based prototype spectrum management system has been implemented using ten servers and dozens of sensors. Measured results show that the proposed approach can reduce data traffic between the sensors and servers by 97%, achieving an average data rate of 10 kilobits per second (kbps). In addition, the basic operation flow of the prototype has been verified through a field experiment conducted at a manufacturing facility and a proof-of-concept experiment of dynamic-spectrum sharing using wireless local-area-network equipment.