Search

Federated learning (FL) has been widely adopted to train machine learning models over massive data in edge computing. However, machine learning faces critical challenges, \eg, data imbalance, edge dynamics, and resource constraints, in edge computing.The existing FL solutions cannot well cope with data imbalance or edge dynamics, and may cause high resource cost. In this paper, we propose an adaptive asynchronous federated learning (AAFL) mechanism. To deal with edge dynamics, the parameter server will aggregate local updated models only from a certain fraction $\alpha$ of all edge nodes in each epoch. Moreover, the system can intelligently vary the number of local updated models for global model aggregation in different epochs with network situations. We then propose experience-driven algorithms based on deep reinforcement learning (DRL) to adaptively determine the optimal value of $\alpha$ in each epoch for two cases of AAFL, single learning task and multiple learning tasks, so as to achieve less completion time of training under resource constraints. Extensive experiments on the classical models and datasets show high effectiveness of the proposed algorithms. Specifically, AAFL can reduce the completion time by about 55\% and improve the learning accuracy by 18\% under resource constraints, compared with the state-of-the-art solutions.

Mobile edge computing is a new paradigm in which the computing and storage resources are placed at the edge of the Internet. Data placement and retrieval are fundamental services of mobile edge computing when a network of edge clouds collaboratively provide data services. However existing methods such as distributed hash tables (DHTs) are not enough to achieve efficient data placement and retrieval services for cooperative edge clouds. This paper presents GRED, a novel data placement and retrieval service for mobile edge computing, which is efficient in not only the load balance but also routing path lengths and forwarding table sizes. GRED utilizes the programmable switches to support a virtual-space based DHT with only one overlay hop. Data location can be easily implemented on top of the GRED. We implement GRED in a P4 prototype, which provides a simple and efficient solution. Results from theoretical analysis, simulations, and experiments show that GRED can efficiently balance the load of edge clouds, and can fast answer data queries due to its low routing stretch.

Distributed in-memory computing frameworks usually have lots of parameters (e.g., the buffer size of shuffle) to form a configuration for each execution. A well-tuned configuration can bring large improvements of performance. However, to improve resource utilization, jobs are often share the same cluster, which causes dynamic cluster load conditions. According to our observation, the variation of cluster load reduces effectiveness of configuration tuning. Besides, as a common problem of cluster computing jobs, overestimation of resources also occurs during configuration tuning. It is challenging to efficiently find the optimal configuration in a shared cluster with the consideration of memory-sparing. In this paper, we introduce MespaConfig, a job-level configuration optimizer for distributed in-memory computing jobs. Advancements of MespaConfig over previous work are features including memory-sparing and load-sensitive. We evaluate MespaConfig by 6 typical Spark programs under different load conditions. The evaluation results show that MespaConfig improves the performance of six typical programs by up to 12 compared with default configurations. MespaConfig also achieves at most 41% reduction of configuration memory usage and reduces the optimization time overhead by 10.8 compared with the state-of-the-art approach.

In this article, a reinforcement learning (RL) based controller is proposed for a multirotor-based transportation system, guaranteeing that the trained RL controller is effective in both simulation and practical experiments. The main novelty lies in that, as far as we know, this is the first attempt of combining the advantages of nonlinear and intelligent control techniques to derive a practice-oriented RL controller for the multirotor-based transportation system, where the high-dimensional complicated dynamics are fully considered in the framework. Specifically, inspired by the physical insight of the system, a new nonlinear control approach is proposed, in which the underactuated properties and the nontrivial couplings are well handled. On this basis, a RL network is proposed to parameterize the nonlinear controller, where the obtained algorithm presents the features of strong reliability and fast convergence even in complicated working conditions (e.g., model uncertainties, parameters drift, external disturbances, and so on). Subsequently, the states are proven to asymptotically converge to the equilibrium point by Lyapunov analysis and RL techniques. A series of simulation and real world experiments are implemented to verify satisfactory positioning accuracy and robustness of the proposed algorithm.

Unexpected bursty traffic brought by certain sudden events, such as news in the spotlight on a social network or discounted items on sale, can cause severe load imbalance in backend services. Migrating hot data—the standard approach to achieve load balance—meets a challenge when handling such unexpected load imbalance, because migrating data will slow down the server that is already under heavy pressure. This article proposes PostMan, an alternative approach to rapidly mitigate load imbalance for services processing small requests. Motivated by the observation that processing large packets incurs far less CPU overhead than processing small ones, PostMan deploys a number of middleboxes called helpers to assemble small packets into large ones for the heavily-loaded server. This approach essentially offloads the overhead of packet processing from the heavily-loaded server to helpers. To minimize the overhead, PostMan activates helpers on demand, only when bursty traffic is detected. The heavily-loaded server determines when clients connect/disconnect to/from helpers based on the real-time load statistics. To tolerate helper failures, PostMan can migrate connections across helpers and can ensure packet ordering despite such migration. Driven by real-world workloads, our evaluation shows that, with the help of PostMan, a Memcached server can mitigate bursty traffic within hundreds of milliseconds, while migrating data takes tens of seconds and increases the latency during migration.

In this study, a novel M-estimation based sparse grid quadrature filter (MSGQF) is proposed to improve the robust performance of the nonlinear system. We present a systematic formulation of the sparse grid quadrature filter (SGQF), and extend it to the discrete-time nonlinear system with abnormal measurement values. The M-estimation method is introduced in the SGQF, which uses the Huber’s cost function to update the measurement covariance. Convergence on the modified robust SGQF is established and proved. The sufficient conditions are shown to ensure stochastic stability of the MSGQF. A target tracking problem has been conducted to demonstrate the accuracy of the MSGQF. When measurement abnormal values appear, it outperforms the unscented Kalman filter (UKF), the cubature Kalman filter (CKF) and the SGQF. Theoretical analysis and simulation results prove that the MSGQF provides significant performance improvement in the robustness of the nonlinear system.

In a software defined network, the controller needs to obtain/collect traffic measurement information ( i.e. , flow statistics) from switches for different applications, such as traffic engineering. Existing solutions seldom consider the per-switch cost, which may lead to heavy statistics collection cost ( e.g. , high CPU overhead) on some switches. Due to limited computing power on most commodity switches, heavy statistics collection cost on those switches may seriously interfere with the basic rule operations, especially when some switches need to deal with many new-arrival flows or update routes of existing flows. To address this challenge, we design and implement efficient flow statistics collection (FSC) with limited interference on the basic rule operations. We formally propose a cooperative flow statistics collection with per-switch cost constraint (CP-FSC) problem. We prove that the CP-FSC problem is NP-hard and present an efficient algorithm with approximation ratio 1/2, based on dynamic programming. To reduce the time complexity, a greedy-based algorithm with approximation ratio 1/3 is also presented. We implement the proposed FSC algorithms on our SDN platform. The experimental results and the extensive simulation results show 36%-59% performance improvement compared with the existing solutions.

New capabilities enabled by quantum networks, e.g. quantum key distribution, rely on the concurrent quantum entanglements via multiple untrusted repeaters between the source-destination pairs, which is known as quantum routing. Existing quantum routing algorithms pre-assign the free qubits to specific paths, in order to avoid resource contention in the process of entanglement swapping. Since not all pre-assigned qubits can be utilized, the resource fragmentation problem caused can limit the network throughout. In this paper, a new fragmentation-aware entanglement routing (FER) algorithm is proposed. With FER, the entangled qubits are shared by all the paths, and the resource contention problem is solved by a priority connection mechanism. Simulation results show that FER consistently outperforms existing algorithms under different network settings.

The hierarchical mobile edge computing satisfies the stringent latency requirements of data access and processing for emerging edge applications. The data location service is a basic function to provide data storage and retrieval to enable these applications. However, it still lacks research of a scalable and low-latency data location service in the environment. The existing solutions, such as DNS and DHT, fail to meet the requirement of those latency-sensitive applications. Therefore, in this article, we present a low-latency hybrid data-sharing framework, HDS. The HDS divides the data location service into two parts: intra-region and inter-region. More precisely, we design a data sharing protocol called Cuckoo Summary to achieve fast data localization in intra-region. Furthermore, for the inter-region data sharing, we develop a geographic routing based scheme to achieve efficient data localization with only one overlay hop. The advantages of HDS include short response latency, low implementation overhead, and few false positives. We implement the HDS framework based on a P4 prototype. The experimental results show that, compared to the state-of-the-art solutions, our design achieves 50.21% shorter lookup paths and 92.75% fewer false positives.

We propose to study mobile object tracing, which allows a mobile system to report the shape, location, and trajectory of the mobile objects appearing in a video camera and identifies each of them with its cyber-identity (ID), even if the appearances of the objects are not known to the system. Existing tracking methods either cannot match objects with their cyber-IDs or rely on complex vision modules pre-learned from vast and well-annotated datasets including the appearances of the target objects, which may not exist in practice. We design and implement TagAttention, a vision-RFID fusion system that achieves mobile object tracing without the knowledge of the target object appearances and hence can be used in many applications that need to track arbitrary un-registered objects. TagAttention adopts the visual attention mechanism, through which RF signals can direct the visual system to detect and track target objects with unknown appearances. Experiments show TagAttention can actively discover, identify, and track the target objects while matching them with their cyber-IDs by using commercial sensing devices in complex environments with various multipath reflectors. It only requires around one second to detect and localize a new mobile target appearing in the video and keeps tracking it accurately over time.

There is an increasing trend that enterprises outsource their middlebox processing to a cloud for lower cost and easier management. However, outsourcing middleboxes brings threats to the enterprise’s private information, including the traffic and rules of middleboxes, all of which are visible within the cloud. Existing solutions for secure middlebox outsourcing either incur significant performance overhead or do not support incremental updates. In this article, we present a secure and dynamic middlebox outsourcing framework, SICS, short for Secure In-Cloud Service. SICS encrypts each packet header and uses a label for in-cloud rule matching, which enables the cloud to perform its functionalities correctly with minimum header information leakage. Evaluation results show that SICS achieves higher throughput, faster construction and update speed, and lower resource overhead at the enterprise and in the cloud when compared with existing solutions.

Superior human pose and shape reconstruction from monocular images depends on removing the ambiguities caused by occlusions and shape variance. Recent works succeed in regression-based methods which estimate parametric models directly through a deep neural network supervised by 3D ground truth. However, 3D ground truth is neither in abundance nor can efficiently be obtained. In this paper, we introduce body part segmentation as critical supervision. Part segmentation not only indicates the shape of each body part but helps to infer the occlusions among parts as well. To improve the reconstruction with part segmentation, we propose a part-level differentiable renderer that enables part-based models to be supervised by part segmentation in neural networks or optimization loops. We also introduce a general parametric model engaged in the rendering pipeline as an intermediate representation between skeletons and detailed shapes, which consists of primitive geometries for better interpretability. The proposed approach combines parameter regression, body model optimization, and detailed model registration altogether. Experimental results demonstrate that the proposed method achieves balanced evaluation on pose and shape, and outperforms the state-of-the-art approaches on Human3.6M, UP-3D and LSP datasets., Comment: Accepted by Pacific Graphcis 2020

Background: In Multilevel Inverters (MLI) as the number of level increases, there is a proportionate increase in the count of the semiconductor devices that are employed. Methods: This paper deals with an asymmetrical cascaded H-bridge inverter topology with half bridge cells to produce seven level output voltage waveform. Nearest Level Control (NLM) technique is used to produce the switching pulses. The operating principle of the proposed MLI and its performance abilities is verified through MATLAB/Simulink and a prototype is developed to provide the experimental results. Results: Total Harmonic Distortion (THD) is computed for proposed MLI for different types of loads in simulation environment as well as in the developed hardware prototype. Comparison between the proposed MLI and recent topologies demonstrates the advantageous features. Conclusion: The simulation and hardware results confirm the suitability of the proposed seven level MLI as the total component count, and the requirement of DC sources reduces considerably.

To effectively improve the computation efficiency in the strapdown inertial navigation system (SINS) alignment process, the authors propose a marginalised adaptive sparse-grid quadrature filter (MASGQF). The filtering method combines the virtues of marginalisation technique and accuracy level switching structure to decrease the computational burden and improve the filtering precision simultaneously. With the adaptation criterion and the predefined tolerance value, different integration rules corresponding to the different accuracy levels nested in the proposed filtering method can be adjusted autonomously. It is proved that the MASGQF is exponentially bounded by theoretical analysis. The performance comparison of different filtering methods is demonstrated by the numerical results of the simulation and offline experiment. On the basis of the data performance of the online test, the validity of the proposed MASGQF is verified in the SINS alignment application.

Machine learning has been a trending topic for which almost every research area would like to incorporate some of the technique in their studies. In this paper, we demonstrate several machine learning models using two different data sets. One data set is the thermograms time series data on a cancer study that was conducted at the University of Louisville Hospital, and the other set is from the world-renowned Framingham Heart Study. Thermograms can be used to determine a patient’s health status, yet the difficulty of analyzing such a high-dimensional dataset makes it rarely applied, especially in cancer research. Previously, Rai et al.1 proposed an approach for data reduction along with comparison between parametric method, non-parametric method (KNN), and semiparametric method (DTW-KNN) for group classification. They concluded that the performance of two-group classification is better than the three-group classification. In addition, the classifications between types of cancer are somewhat challenging. The Framingham Heart Study is a famous longitudinal dataset which includes risk factors that could potentially lead to the heart disease. Previously, Weng et al.2 and Alaa et al.3 concluded that machine learning could significantly improve the accuracy of cardiovascular risk prediction. Since the original Framingham data have been thoroughly analyzed, it would be interesting to see how machine learning models could improve prediction. In this manuscript, we further analyze both the thermogram and the Framingham Heart Study datasets with several learning models such as gradient boosting, neural network, and random forest by using SAS Visual Data Mining and Machine Learning on SAS Viya. Each method is briefly discussed along with a model comparison. Based on the Youden’s index and misclassification rate, we select the best learning model. For big data inference, SAS Visual Data Mining and Machine Learning on SAS Viya, a cloud computing and structured statistical solution, may become a choice of computing.

Channel pruning is effective in compressing the pretrained CNNs for their deployment on low-end edge devices. Most existing methods independently prune some of the original channels and need the complete original dataset to fix the performance drop after pruning. However, due to commercial protection or data privacy, users may only have access to a tiny portion of training examples, which could be insufficient for the performance recovery. In this paper, for pruning with limited data, we propose to use all original filters to directly develop new compact filters, named reborn filters, so that all useful structure priors in the original filters can be well preserved into the pruned networks, alleviating the performance drop accordingly. During training, reborn filters can be easily implemented via 1×1 convolutional layers and then be fused in the inference stage for acceleration. Based on reborn filters, the proposed channel pruning algorithm shows its effectiveness and superiority on extensive experiments.

Sequential Text Classification (STC) aims to classify a sequence of text fragments (e.g., words in a sentence or sentences in a document) into a sequence of labels. In addition to the intra-fragment text contents, considering the inter-fragment context dependencies is also important for STC. Previous sequence labeling approaches largely generate a sequence of labels in left-to-right reading order. However, the need for context information in making decisions varies across different fragments and is not strictly organized in a left-to-right order. Therefore, it is appealing to label the fragments that need less consideration of context information first before labeling the fragments that need more. In this paper, we propose a novel model that labels a sequence of fragments in jumping order. Specifically, we devise a dedicated board-game to develop a correspondence between solving STC and board-game playing. By defining proper game rules and devising a game state evaluator in which context clues are injected, at each round, each player is effectively pushed to find the optimal move without position restrictions via considering the current game state, which corresponds to producing a label for an unlabeled fragment jumpily with the consideration of the contexts clues. The final game-end state is viewed as the optimal label sequence. Extensive results on three representative datasets show that the proposed approach outperforms the state-of-the-art methods with statistical significance.

We provide the first solution to an important question, “how a physical-layer authentication method can defend against signal replay attacks”. It was believed that if an attacker can replay the exact same reply signal of a legitimate authentication object (such as an RFID tag), any physical-layer authentication method will fail. This paper presents Hu-Fu, the first physical layer RFID authentication protocol that is resilient to the major attacks including tag counterfeiting, signal replay, signal compensation, and brute-force feature reply. Hu-Fu is built on two fundamental ideas, namely inductive coupling of two tags and signal randomization. Hu-Fu does not require any hardware or protocol modification on COTS passive tags and can be implemented with COTS devices. We implement a prototype of Hu-Fu and demonstrate that it is accurate and robust to device diversity and environmental changes, including locations, distance, and temperature. Hu-Fu provides a new direction of battery-free/low-power device authentication that enables numerous IoT applications.

The existing methods of generating adversarial texts usually change the original meanings of texts significantly and even generate the unreadable texts. These less readable adversarial texts can misclassify the machine classifier successfully, but they cannot deceive the human observers very well. In this paper, we propose a novel method that generates readable adversarial texts with some perturbations that can also confuse human observers successfully. Based on the continuous bag-of-words (CBOW) model, the proposed method looks for the appropriate perturbations to generate the adversarial texts through controlling the perturbation direction vectors. Meanwhile, we apply adversarial training to regularize the classification model and extend it to semi-supervised tasks with virtual adversarial training. Experiments are conducted to show that the generated adversaries are interpretable and confused to humans and the virtual adversarial training effectively improves the robustness of the model.

Network Function Virtualization (NFV) is a new paradigm to enable service innovation through virtualizing traditional network functions. To construct a new NFV-enabled network, there are two critical requirements: minimizing server deployment cost and satisfying switch resource constraints . However, prior work mostly focuses on the server deployment cost, while ignoring the switch resource constraints ( e.g. , switch’s flow-table size). It thus results in a large number of rules on switches and leads to massive control overhead. To address this challenge, we propose an incremental server deployment (INSD) problem for construction of scalable NFV-enabled networks. We prove that the INSD problem is NP-Hard, and there is no polynomial-time algorithm with approximation ratio of $(1-\epsilon) \cdot \ln m$ , where $\epsilon $ is an arbitrarily small value and $m$ is the number of requests in the network. We then present an efficient algorithm with an approximation ratio of $2\cdot H(q\cdot p)$ , where $q$ is the number of VNF’s categories and $p$ is the maximum number of requests through a switch. We evaluate the performance of our algorithm with experiments on physical platform (Pica8), Open vSwitches, and large-scale simulations. Both experimental results and simulation results show high scalability of the proposed algorithm. For example, our solution can reduce the control and rule overhead by about 88% with about 5% additional server deployment, compared with the existing solutions.

Passive Radio Frequency Identification (RFID) tags have been widely applied in many applications, such as logistics, retailing, and warehousing. In many situations, the order of objects is more important than their absolute locations. However, state-of-art ordering methods need a continuing movement of tags and readers, which limit the application domain and scalability. In this paper, we propose a 2-dimension ordering approach for passive tags that requires no device movement. Instead, our method utilizes signal changes caused by arbitrary movement of human beings around tags, who carry no device for horizontal dimension ordering. Hence, our method is called Human Movement based Ordering (HMO). The basic idea of HMO is that when people pass between the reader antenna and tags, the received signal strength will change. By observing the time-series RSS changes of tags, HMO can obtain the order of tags along with a specific horizontal direction. For vertical dimension, we employ a linear programming method that is tolerant of tiny errors in practice. We implement HMO with commodity off-the-shelf RFID devices. The experimental results show that HMO can achieve up to 88.71 and 90.86 percent average accuracies in the signal- and multi-person cases, respectively.

Public-key certificate validation is an important building block for various security protocols for IoT devices, such as secure channel establishment, handshaking, and verifying sensing data authenticity from cloud storage. However, certification validation incurs non-trivial overhead on resource-constrained IoT devices, because it either brings long latency or large cache space. This work proposes to utilize the power of distributed caching and explores the feasibility of using the cache spaces on all IoT devices as a large pool to store validated certificates. We design a Collaborative Certificate Validation (CCV) protocol including a memory-efficient and fast locator for certificate holders, a trust model to evaluate the trustworthiness of devices, and a protocol suite for dynamic update and certificate revocation. Evaluation results show that CCV only uses less than 25% validation time and reduces >90% decryption operations on each device, compared to a recent method. Malicious devices that conduct dishonest validations can be detected by the network using the proposed trust model.

Electric load forecasting has become crucial to the safe operation of power grids and cost reduction in the production of power. Although numerous electric load forecasting models have been proposed, most of them are still limited by poor effectiveness in the model training and a sensitivity to outliers. The limitations of current methods may lead to extra operational costs of a power system or even disrupt its power distribution and network safety. To this end, we propose a new hybrid load-forecasting model, which is based on a robust extreme-learning machine and an improved whale optimization algorithm. Specifically, Huber loss, which is insensitive to outliers, is proposed as the objective function in extreme learning machine (ELM) training. In addition, an improved whale optimization algorithm is designed for the robust ELM training, in which a cellular automaton mechanism is used to enhance the local search. To verify our improved whale optimization algorithm, some experiments were then conducted based on seven benchmark test functions. Due to the enhancement of the local search, the improved optimizer was around 7% superior to the basic. Finally, our proposed hybrid forecasting model was validated by two real electric load datasets (Nanjing and New South Wales), and the experimental results confirmed that the proposed hybrid load-forecasting model could achieve satisfying improvements in both datasets.

In this paper, firstly we propose a new protocol for quantum bidirectional teleportation of two particles in a six-qubit entangled channel, which is based on GHZ measurement and unitary operation. Then, also based on the six-qubit entangled state channel, a protocol for quantum bidirectional teleportation of two-three particles is proposed. The protocol utilizes Toffoli gate to convert three-qubit states into two-qubit states and a single qubit, and the rest of the teleportation is the same as the first protocol. Finally, we introduce an auxiliary bit to restore the state of the sender. Both these proposed protocols are more efficient than others in the same situation, which will be compared at the end of the paper.

In this letter, we propose a suboptimal resource allocation scheme for the uplink vehicular communications networks. Our aim is to maximize the minimum throughput of all vehicle-to-vehicle (V2V) links under the constraints of the minimum vehicle-to-infrastructure (V2I) link throughput requirements. The optimization problem has a non-linear and non-convex objective function and we cannot obtain the optimal solutions directly. To solve this problem, we first propose a V2V partitioning method from the view point of game theory. Then, an efficient power allocation algorithm is developed to obtain the optimal power allocation solutions by exploiting the D.C. (difference of convex functions) structure of the objective function. Simulation results demonstrate the effectiveness of the proposed resource allocation algorithms.

Diversity analysis and taxonomic profiles can be generated from marker-gene sequence data with the help of many available computational tools. The Quantitative Insights into Microbial Ecology Version 2 (QIIME2) has been widely used for 16S rRNA data analysis. While many articles have demonstrated the use of QIIME2 with suitable datasets, the application to pre-clinical data has rarely been talked about. The issues involved in the pre-clinical data include the low-quality score and small sample size that should be addressed properly during analysis. In addition, there are few articles that discuss the detailed statistical methods behind those alpha and beta diversity significance tests that researchers are eager to find. Running the program without knowing the logic behind it is extremely risky. In this article, we first provide a guideline for analyzing 16S rRNA data using QIIME2. Then we will talk about issues in pre-clinical data, and how they could impact the outcome. Finally, we provide brief explanations of statistical methods such as group significance tests and sample size calculation.

Allowing a device to verify the digital certificate of another device is an essential requirement and key building block of many security protocols for emerging and future IoT systems that involve device-to-device communication. However, on-device certificate verification is challenging for current devices, mainly because the certificate revocation (CR) checking step costs too much resource on IoT devices and the synchronization of CR status to devices yields a long latency. This paper presents an on-device CR checking system called TinyCR, which achieves 100% accuracy, memory and computation efficiency, low synchronization latency, and low network bandwidth, while being compatible with the current certificate standard. We design a new compact and dynamic data structure called DASS to store and query global CR status on a device in TinyCR. Our implementation shows that TinyCR only costs each device 1.7 MB of memory to track 100 million IoT certificates with 1% revocation rate. Checking the CR status of one certificate spends less than 1 microsecond on a Raspberry Pi 3. TinyCR can also be updated instantly when there are new certificates added or revoked.

Three-dimensional (3D) imaging technology has been widely applied in various fields, such as intelligent manufacturing, online inspection, reverse engineering, cultural relic protection, etc. In this work, we present a high-accuracy real-time omnidirectional 3D scanning and inspection system based on fringe projection profilometry. Firstly, a multi-camera system based on geometric constraints is constructed to perform stereo phase unfolding without additional auxiliary projection images to ensure high-accuracy 3D data acquisition in real time. Then, we propose a rapid 3D point cloud registration approach combining simultaneous localization and mapping (SLAM) with iterative closest point (ICP) techniques to achieve alignment of point cloud slices with accuracy of up to 100 microns. Finally, a cycle-positioning-based registration scheme is developed to allow for real-time 360 degree 3D surface defect inspection. The experimental results show that our system is capable of real-time omnidirectional 3D modelling and real-time 360° defect detection.

Currently, video semantic segmentation mainly faces two challenges: 1) the demand of temporal consistency; 2) the balance between segmentation accuracy and inference efficiency. For the first challenge, existing methods usually use optical flow to capture the temporal relation in consecutive frames and maintain the temporal consistency, but the low inference speed by means of optical flow limits the real-time applications. For the second challenge, flow based key frame warping is one mainstream solution. However, the unbalanced inference latency of flow-based key frame warping makes it unsatisfactory for real-time applications. Considering the segmentation accuracy and inference efficiency, we propose a novel Sparse Temporal Transformer (STT) to bridge temporal relation among video frames adaptively, which is also equipped with query selection and key selection. The key selection and query selection strategies are separately applied to filter out temporal and spatial redundancy in our temporal transformer. Specifically, our STT can reduce the time complexity of temporal transformer by a large margin without harming the segmentation accuracy and temporal consistency. Experiments on two benchmark datasets, Cityscapes and Camvid, demonstrate that our method achieves the state-of-the-art segmentation accuracy and temporal consistency with comparable inference speed.

To improve the performance of the cubature kalman filter (CKF) in nonlinear non-Gaussian filtering system, the Gaussian sum cubature kalman filter (GSCKF) is proposed. A formulation of the CKF for the nonlinear system with non-Gaussian process noise and non-Gaussian measurement noise is presented. The GSCKF uses the Gaussian sum theory to divide the non-Gaussian system into several Gaussian sub-systems. And each subsystem is conducted the filtering process by the CKF. A target tracking problem with non-Gaussian noise is used as a simulation application to compare the filtering performance of the GSCKF. The simulation results show that the GSCKF outperforms the unscented Kalman filter (UKF) and CKF under non-Gaussian conditions. It proves that for the nonlinear systems with non-Gaussian noise, the filtering performance of the GSCKF is significantly improved.

For now, the comparator is an important structure in the Analog to Digital Converter. Advanced comparators are widely used in integrated circuits. The paper addresses several kinds of comparator design to compare the difference between various comparators based on different parameters. The different circuits are designed mathematically and suitably for emphasizing the advantages and disadvantages. The simulation is also used to find out the delay and energy cost.

The present studies on battery electric vehicles (BEVs) has mainly focused on the single-objective or weighted multi-objective optimization based on energy management, which can not manifest the coupling relationship among the vehicle performance objectives essentially. To optimize the handling stability, ride comfort and economy of BEV, this paper built the stability dynamics analysis model, ride comfort simulation half-car model and power consumption calculation model of BEV, as well as two-point virtual random excitation model on Level B road and proposed related evaluation indexes, including vehicle handling stability factor, weighted acceleration root-mean-square (RMS) value of vertical vibration at the driver’s seat and power consumption per 100 m at a constant speed. The Pareto optimum principle–based multi-objective evolutionary algorithm (MOEA) of BEV was also designed, which was encoded with real numbers and obtained the target values of all optional schemes via MATLAB/Simulink simulation software. The merits and demerits of alternative schemes could be judged according to the Pareto dominance principle, so that alternative schemes obtained after optimization were realizable. The results of simulation experiment suggest that the proposed algorithm can perform the multi-objective optimization on BEV, and obtain a group of Pareto optimum solutions featured by high handling stability, favorable ride comfort and low energy consumption for the decision-makers.

Clock drawing test is a classic and reliable clinical assessment of cognitive functions. However, it requires manual inspection and scoring. In this paper, via several computer vision and artificial intelligence techniques, we successfully design and implement an intelligent system with the ability to score the test in an automatic manner. As a mobile application, our work allows users to take a clock drawing test anytime and anywhere.

Underwater acoustic environment (UAE) prediction is essential for accurately estimating sonar performance and making credible tactical decisions. Modular Ocean Data Assimilation System (MODAS) provides an efficient way to reconstruct real-time three-dimensional ocean sound speed field and has potential use in UAE prediction. However, the effect of MODAS on UAE prediction is not clear, which prevents MODAS from being used in practice. Taking into account that convergence zone (CZ) is a typical feature of deep-sea UAE, this paper studies the effect of MODAS on CZ distance prediction. A method is proposed to describe the accuracy of CZ distance prediction quantitatively. The method takes the CZ distance as a random variable and uses Monte Carlo sampling method to obtain the probability density function (PDF) of CZ distance. The performance of MODAS on CZ distance prediction is numerically simulated in a range-independent and a mesoscale eddy waveguides. It is shown that: (1) In the range-independent waveguide, if the sound speed disturbance of internal waves with respect to the background sound speed profile is less than the RMSE of MODAS, the prediction of CZ distance for local measured sound velocity profile is more accurate. (2) In the mesoscale eddy waveguide, the performance of MODAS is better and the performance improves as the RMSE of MODAS decreases.

Intent-Based Networking (IBN) has been increasingly deployed in production enterprise networks. Automated network configuration in IBN lets operators focus on intents- i.e., the end to end business objectives-rather than spelling out details of the configurations that implement these objectives. Automation brings its own concerns as the administrators cannot rely on traditional network troubleshooting tools. This situation is further exacerbated in the case of stateful Network Functions (NFs) whose packet processing behavior depends on previously observed traffic patterns. To ensure that the network configuration and state derived from network automation matches the administrator’s specified intent, we propose, Epinoia, a network intent checker for stateful networks. Epinoia relies on a unified model for NFs by leveraging the causal precedence relationships that exist between NF packet I/Os and states. Scalability of Epinoia is achieved by decomposing intents into sub-checking tasks and maintaining a causality graph between checked invariants. Epinoia checks for network-wide intent violations incrementally to reduce overhead in the event of network changes. Our evaluation results using real-world network topologies show that Epinoia can perform comprehensive checking within a few seconds per network with intent updates.

Wi-Fi-based indoor localization gained a lot of attention over recent years due to low cost and open access properties. However, existing schemes might not be applicable in the real environment if their robustness is low. This paper presents CLRS, a novel distributed Indoor Positioning System (IPS) with high robustness which uses Wi-Fi signals to divide the space twice based on Angle of Arrival (AoA) and Effective Channel State Information (ECSI). The proposed scheme trade the redundancy of Access Point (AP) quantity to improve the tolerance of data measurement error. We performed simulations as well as real-world experiments, in which simulation results proved that the theoretical average error is the least when the routers are placed vertically in our localization method while the real-world experiments proved the high accuracy and robustness of CLRS.

Computational microscopy, as a subfield of computational imaging, combines optical manipulation and image algorithmic reconstruction to recover multi-dimensional microscopic images or information of micro-objects. In recent years, the revolution in light-emitting diodes (LEDs), low-cost consumer image sensors, modern digital computers, and smartphones provide fertile opportunities for the rapid development of computational microscopy. Consequently, diverse forms of computational microscopy have been invented, including digital holographic microscopy (DHM), transport of intensity equation (TIE), differential phase contrast (DPC) microscopy, lens-free on-chip holography, and Fourier ptychographic microscopy (FPM). These computational microscopy techniques not only provide high-resolution, label-free, quantitative phase imaging capability but also decipher new and advanced biomedical research and industrial applications. Nevertheless, most computational microscopy techniques are still at an early stage of "proof of concept" or "proof of prototype" (based on commercially available microscope platforms). Translating those concepts to stand-alone optical instruments for practical use is an essential step for the promotion and adoption of computational microscopy by the wider bio-medicine, industry, and education community. In this paper, we present four smart computational light microscopes (SCLMs) developed by our laboratory, i.e., smart computational imaging laboratory (SCILab) of Nanjing University of Science and Technology (NJUST), China. These microscopes are empowered by advanced computational microscopy techniques, including digital holography, TIE, DPC, lensless holography, and FPM, which not only enables multi-modal contrast-enhanced observations for unstained specimens, but also can recover their three-dimensional profiles quantitatively. We introduce their basic principles, hardware configurations, reconstruction algorithms, and software design, quantify their imaging performance, and illustrate their typical applications for cell analysis, medical diagnosis, and microlens characterization.

We present a new application direction named Pareidolia Face Reenactment, which is defined as animating a static illusory face to move in tandem with a human face in the video. For the large differences between pareidolia face reenactment and traditional human face reenactment, two main challenges are introduced, i.e., shape variance and texture variance. In this work, we propose a novel Parametric Unsupervised Reenactment Algorithm to tackle these two challenges. Specifically, we propose to decompose the reenactment into three catenate processes: shape modeling, motion transfer and texture synthesis. With the decomposition, we introduce three crucial components, i.e., Parametric Shape Modeling, Expansionary Motion Transfer and Unsupervised Texture Synthesizer, to overcome the problems brought by the remarkably variances on pareidolia faces. Extensive experiments show the superior performance of our method both qualitatively and quantitatively. Code, model and data are available on our project page1.

Today's large-scale enterprise networks, data center networks, and wide area networks can be decomposed into multiple administrative or geographical domains. Domains may be owned by different administrative units or organizations. Hence protecting domain information is an important concern. Existing general-purpose Secure Multi-Party Computation (SMPC) methods that preserves privacy for domains are extremely slow for cross-domain routing problems. In this paper we present PYCRO, a cryptographic protocol specifically designed for privacy-preserving cross-domain routing optimization in Software Defined Networking (SDN) environments. PYCRO provides two fundamental routing functions, policy-compliant shortest path computing and bandwidth allocation, while ensuring strong protection for the private information of domains. We rigorously prove the privacy guarantee of our protocol. To improve time efficiency we design the QuIck Pathing (QIP) technique. QIP only requires one-time offline preprocessing and very fast online computation. We have implemented a prototype system that runs PYCRO and QIP on servers in a campus network. Experimental results using real ISP network topologies show that PYCRO and QIP are very efficient in computation and communication costs.

Doors as densely-deployed natural landmarks play an important role in improving indoor positioning systems. However, the state-of-the-art door event detection works are based on either vision or infrastructure, thus incurring non-trivial device or management cost. To address these problems, we present a Light-weight Magnetic-based Door Event Detection method, called LMDD. It leverages built-in magnetic sensors of common smartphones to achieve infrastructure-free door event detection. After analyzing the special features of sensors’ readings changes caused by the door, we design LMDD scheme with three main components, including data acquisition, events identification and events denoising. Moreover, an improved and robust door event detection framework based on a majority-voting model is proposed to fuse multiple-dimensional sensing data from non-magnetic built-in sensors. We have implemented a prototype of LMDD on Android-based platform. Experimental results show that LMDD with only magnetic sensor achieves door event detection accuracy of around 80 percent on average, ranging from 70 to 87 percent in various typical indoor environments. The enhanced LMDD based on the fusion of heterogeneous sensors can achieve a much higher door event detection accuracy of 90 percent on average.

The Challenge for the flight control of flapping wing vehicles has been baffling the control scientists and engineers for decades. Based on this observation, we propose a quaternion-based hybrid attitude control strategy to implement stable hovering and forward flight for a flapping wing vehicle. Specifically, two hysteretic functions are adopted to switch between two control strategies, namely regulating to the desired attitude and pulling back to the stable flight region, so as to deal with a specific geometric constraint, and retain good robustness. Meanwhile, considering the inherent underactuation property of the studied vehicle, the reduced quaternion is introduced to construct feedback signals which are convenient and practical to use. These innovations contribute to a reliable control strategy for the flapping wing vehicles, even with underactuation in attitude. The equilibrium set is proved to be globally asymptotically stable with the help of hybrid control theory. Finally, the proposed controller is implemented in our self-made flapping wing vehicle, with the obtained experimental results clearly showing its superior performance.

We present vacuum filters, a type of data structures to support approximate membership queries. Vacuum filters cost the smallest space among all known AMQ data structures and provide higher insertion and lookup throughput in most situations. Hence they can be used as the replacement of the widely used Bloom filters and cuckoo filters. Similar to cuckoo filters, vacuum filters also store item fingerprints in a table. The memory-efficiency and throughput improvements are from the innovation of a table insertion and fingerprint eviction strategy that achieves both high load factor and data locality without any restriction of the table size. In addition, we propose a new update framework to resolve two difficult problems for AMQ structures under dynamics, namely duplicate insertions and set resizing. The experiments show that vacuum filters can achieve 25% less space in average and similar throughput compared to cuckoo filters, and 15% less space and >10x throughput compared to Bloom filters, with same false positive rates. AMQ data structures are widely used in various layers of computer systems and networks and are usually hosted in platforms where memory is limited and precious. Hence the improvements brought by vacuum filters can be considered significant.

Ad hoc connection/pairing among IoT devices will be pervasive in the Next Generation Networking (NGN). However, due to its de-centralized nature, it faces more security challenges. In this paper, we propose a secure key pairing protocol for NGN, VariSecure, to enable confidential communication for smartphones. The idea is to leverage a natural human interaction mode: face-to-face interaction. Two users raise up their smartphones to establish a secure channel based on their similar observations from the view of the smartphones’ cameras. One user acts as the initiator and uses the front-facing camera to snap his/her face. The other user captures the initiator’s face using his/her rear camera. Utilizing the facial appearance as a secret substitution, the initiator delivers a randomly generated key to the other user, thus establishing a confidential communication channel between their devices. VariSecure achieves secure device pairing without the complex operations needed in prior works. We implemented VariSecure on the commercial off-the-shelf (COTS) smart devices. The experimental results show that VariSecure can establish a face-to-face secure pairing in less than 3 seconds.

With the development of 4G networks and the popularity of mobile terminals, the public’s way of expressing emotions and expressing opinions has been transferred from traditional information media to multimedia Internet platforms, which has promoted the dissemination of public information on multimedia networks into public view. The public through the APP software, smart phones, various types of mobile terminals portable, and quickly participate in the network public opinion information carried by Weibo, WeChat, forum and other media. This trend can be represented by various attributes, and those with objectivity and irrefutability provide the strongest evidence of the reality of an event. At present, scholars at home and abroad mostly use the quantitative characteristics of online public opinion texts as the basis for calculation to judge the situation and emotions of online public opinion. In this paper, information theory is introduced to the field of public opinion analysis, and S-info is defined and extended to E-info; the sum of E-info of an event shows the event trend, representing the emotional intensity and tendency and their changes over time. Based on this, a Weibo situational perception model is constructed. The dictionary method is used to calculate the emotion of the text, and is then combined with the Weibo situational perception model to analyze the sentimental trend. The model is used on a dataset of 1,448,188 Weibos, and the events were separately statistically analyzed. The sum of the E-info are collected and compared with the number of Weibos. From the comparison, key time points and emotional burst points of the event can be picked up clearly, revealing the novel characteristics of this model. Duplications and advertisements can be filtered out and the reality can be revealed, uncovering the objective and irrefutable attributes.

The design of distributed control plane is an essential part of SDN. While there is an urgent need for verifying the control plane, little, however, is known about how to validate that the control plane offers assurable performance, especially across various failures. Such validation is hard due to two fundamental challenges. First, the number of potential failure scenarios could be exponential or even non-enumerable. Second, it is still an open problem to model the performance change when the control plane employs different failure recovery strategies. In this paper, we first characterize the validation of the distributed control plane as a robust optimization problem and further propose a robust validation framework to verify whether a control plane provides assurable performance across various failure scenarios and multiple failure recovery strategies. Then, we prove that identifying an optimal recovery strategy is NP-hard after developing an optimization model of failure recovery. Accordingly, we design two efficient failure recovery strategies, which can well approximate the optimal strategy and further exhibit good performance against potential failures. Furthermore, we design the capacity augmentation scheme when the control plane fails to accommodate the worst failure scenario even with the optimal failure recovery strategy. We have conducted extensive evaluations based on an SDN test bed and large-scale simulations over real network topologies. The evaluation results show the efficiency and effectiveness of the proposed validation framework.

Smart packaging adds sensing abilities to traditional packages. This paper investigates the possibility of using RF signals to test the internal status of packages and detect abnormal internal changes. Towards this goal, we design and implement a nondestructive package testing and verification system using commodity passive RFID systems, called Echoscope. Echoscope extracts unique features from the backscatter signals penetrating the internal space of a package and compares them with the previously collected features during the check-in phase. The use of backscatter signals guarantees that there is no difference in RF sources and the features reflecting the internal status will not be affected. Compared to other nondestructive testing methods such as X-ray and ultrasound, Echoscope is much cheaper and provides ubiquitous usage. Our experiments in practical environments show that Echoscope can achieve very high accuracy and is very sensitive to various types abnormal changes.