Mobile Ad-hoc and Sensor Networks

In smart grid information systems, the electricity usage data should be audited by data users, such as the market analysts to finish their tasks. Besides that, electricity company always outsources the data to the cloud server (CS) to release its data management pressure. Since the CS is untrusted and the detailed electricity usage data contains users’ privacy, the privacy concern of the data and data users’ queries is raised. Although many schemes have been proposed to achieve the encrypted data query in smart grid, they are not applied well due to the numeric attributes in electricity usage data and privacy concern in smart grid application. In this paper, we provide an efficient privacy-preserving scheme for range query in smart grid. Our scheme achieves the range query without disclosing the privacy of the data and queries. And the performance shows that our scheme can reduce the computation cost for both the data owner and data users, and shorten the response time of every query, which is great significance for smart grid application.

Multipath TCP (MPTCP) enables terminals utilizing multiple interfaces for data transmission simultaneously, which provides better performance and brings many benefits. However, using multiple paths brings some new challenges. The asymmetric parameters among different subflows may cause the out-of-order problem and load imbalance problem, especially in wireless network which has more packet loss. Thus it will significantly degrade the performance of MPTCP. In this paper, we propose a Receive Buffer Pre-division based flow control mechanism (RBP) for MPTCP. RBP divides receive buffer according to the prediction of receive buffer occupancy of each subflow, and controls the data transmission on each subflow using the divided buffer and the number of out-of-order packets, which can significantly improve the performance of MPTCP. We use the NS-3 simulations to verify the performance of our scheme, and the simulation results show that RBP algorithm can significantly increase the global throughput of MPTCP.

User behavioral analysis is expected to act as a promising technique for identity theft detection in the Internet. The performance of this paradigm extremely depends on a good individual-level user behavioral model. Such a good model for a specific behavior is often hard to obtain due to the insufficiency of data for this behavior. The insufficiency of specific data is mainly led by the prevalent sparsity of users’ collectable behavioral footprints. This work aims to address whether it is feasible to effectively detect identify thefts by jointly using multiple unreliable behavioral models from sparse individual-level records. We focus on this issue in mobile social networks (MSNs) with multiple dimensions of collectable but sparse data of user behavior, i.e., making check-ins, posing tips and forming friendships. Based on these sparse data, we build user spatial distribution model, user post interest model and user social preference model, respectively. Here, as the arguments, we validate that there is indeed a complementary effect in multi-dimensional blended behavioral analysis for identity theft detection in MSNs.

The reliability and availability of cloud computing services improved dramatically in recent years. More and more users tend to migrate their business systems to cloud environment. The compliance of SLAs (Service Level Agreement) in cloud services must be efficiently evaluated to ensure the enforcement of SLAs. Traditionally, an Evaluation Center will be established to collecting performance and reliability metrics data of SLAs. However, since the volume of data collected is huge, and the speed of data generation is fast, large bandwidth and computation capacity is needed in the Evaluation Center. This paper proposes a hierarchical architecture for monitoring and evaluating of the compliances of SLAs. In this architecture, a Data Collection Service is deployed in the intranet of each cloud service provider. Metrics data is first collected and analyzed by local Data Collection Service. When SLA violations are detected, related data is packed, signed and sent to the Evaluation Center located on the Internet. Simulations show that the use of local Data Collection Service will effectively reduce the amount of data transferred via Internet and will not cause much overhead in the construction of evaluation infrastructure.

This paper focuses on the uncapacitated k-median facility location problem, which asks to locate k facilities in a network that minimize the total routing time, taking into account the constraints of nodes that are able to serve as servers and clients, as well as the level of demand in each client node. This problem is important in a wide range of applications from operation research to mobile ad-hoc networks. Existing algorithms for this problem often lead to high computational costs when the underlying network is very large, or when the number k of required facilities is very large. We aim to improve existing algorithms by taking into considerations of the community structures of the underlying network. More specifically, we extend the strategy of local search with single swap with a community detection algorithm. As a real-world case study, we analyze in detail Auckland North Shore spatial networks with varying distance threshold and compare the algorithms on these networks. The results show that our algorithm significantly reduces running time while producing equally optimal results.

The storage-carrying-forwarding of messages of the node is a way of short-distance communication in the mobile social networks, and the transmission performance is the key factor that affects the user interaction experience. If the user can transmit the message according to the interest or the community, the transmission performance can be improved. For the short-distance communication in the mobile social networks, the existing research is mainly either interest-based or community-based transmission. In order to make users to have a better interactive experience, we proposed InComT (Interest Community based Transmission) which combines the user interest with the community. We measure the interest value of a node in the mobile social networks, and the community is divided according to its interest value to determine the whole community interest value. Then the relay community and the relay node are selected by the interest value to realize the transmission of the message. The simulation results show that the scheme can get a higher transmission success rate with low transmission overhead and low average delay.

This paper proposed a lossless data hiding scheme by variable length code (VLC) mapping, which focused on embedding additional data into JPEG bitstream. The entropy-coded data in JPEG bitstream consists of a sequence of VLCs and the appended bits. Not all VLCs defined in the JPEG file header are used in the entropy-coded data and the replacement of unused-VLCs do nothing with image decompression. Hence, additional data can be embedded by mapping the unused-VLCs to the used-VLCs. To obtain higher embedding capacity, we improved the mapping rules in this paper. Employing the proposed mapping scheme, larger embedding capacity and no image distortion are both achieved while the filesize of JPEG is preserved after data embedding.

Saliency detection in images attracts much research attention for its usage in numerous multimedia applications. Beside on the detection within the single image, co-saliency has been developed rapidly by detecting the same foreground objects in different images and trying to further promote the performance of object detection. This paper we propose a co-saliency detection method based on Siamese Network. By using Siamese Network, we get the similarity matrix of each image in superpixels. Guided by the single image saliency map, each saliency value, saliency score matrix is obtained to generate the multi image saliency map. Our final saliency map is a linear combination of these two saliency maps. The experiments show that our method performs better than other state-of-arts methods.

The tremendous growth of WiFi fingerprint-based localization techniques has significantly facilitated localization services. The traditional techniques pose a threat to both client’s and server’s privacies, because it is likely to reveal sensitive information about the client and the server during providing localization services. Many existing works have proposed privacy preserving localization schemes based on homomorphic cryptographic systems. However, the state of the art homomorphic cryptographic systems turn out to bear a time-consuming process for recourse-constrained devices. Hence, preserving location privacy while guaranteeing efficiency and usability is still a challenging problem. In this paper, we propose a privacy preserving indoor localization scheme employing oblivious transfer, called OTPri, to preserve the privacy of both clients and server in the process of localization in an efficient way. Our method enables a client to efficiently compute her location locally at client side with a small amount of additional overhead compared with the non-privacy-preserving scheme. Meanwhile, we conduct comprehensive experiments, including single-floor and multi-floor scenarios in our office building. The evaluation results demonstrate the efficiency improvement and overhead reduction of our proposed scheme compared with a classical privacy-preserving indoor localization scheme.

With the availability of Smart Grid, disaggregation, i.e. decomposing a whole electricity signal into its component appliances has gotten more and more attentions. Now the solutions based on the sparse coding, i.e. the supervised learning algorithm that belongs to Non-Intrusive Load Monitoring (NILM) have developed a lot. But the accuracy and efficiency of these solutions are not very high, we propose a new efficient sparse coding-based data-mining (ESCD) scheme in this paper to achieve higher accuracy and efficiency. First, we propose a new clustering algorithm – Probability Based Double Clustering (PDBC) based on Fast Search and Find of Density Peaks Clustering (FSFDP) algorithm, which can cluster the device consumption features fast and efficiently. Second, we propose a feature matching optimization algorithm – Max-Min Pruning Matching (MMPM) algorithm which can make the feature matching process to be real-time. Third, real experiments on a publicly available energy data set REDD [1] demonstrate that our proposed scheme achieves a for energy disaggregation. The average disaggregation accuracy reaches 77% and the disaggregation time for every 20 data is about 10 s.

Today, with the presence of a large number of Man-In-The-Middle (MITM) attacks, identity authentication plays an important role in computer communication network. Series of authentication protocols have been proposed to resist against MITM attacks. Due to the lack of two-way certification between the client and the server, an attack named Man-In-The-Middle-Script-In-The-Browser (MITM-SITB) still works in most protocols. In order to protect against this kind of attack, a Channel-ID based authentication protocol named Server-Invariance-with-Strong-Client-Authentication (SISCA) is put forward. This protocol can not support key update and execute inefficiently. To solve this problem, we propose a Communication-Effectiveness-of-Web-Authentication (CEWA) protocol. We design a new certification process to make the protocol support key update, thus avoiding the risk of key leaks. Simultaneously, We designed the key storage method to manage the keys. We improve the efficiency of implementation. We also analyze its security and the experimental analysis shows the better performance of the efficiency than that in SISCA protocol.

With the widespread use of mobile devices, mobile online advertising is taking more and more market share. Cost per click and cost per view are the most popular pricing modes in mobile internet advertising, which take effective clicks or displaying duration as the charging basis. However, at the same time, ad fraud, which uses illegal and invalid clicks to fraud advertisers in order to obtain unreasonable income, become a serious problem. Most of the previous studies on click fraud in website focused on network traffic data analysis. This makes them cannot solve the placement fraud problem, which use invalid placement to mislead user to click on it in mobile apps. In this paper, we propose a joint crowdsourcing and data analyzing based placement click fraud detection system. For the characteristic of placement fraud in mobile apps, automatic processing cannot cover every possible fraud. To overcome this, our report system provides a platform to find all possible placement fraud through crowdsourcing. Report system has three main services: a monitor service for monitoring user’s call; a layout service for recording the screen; a data service for recording the backend data. Because the placement fraud only appears when users use the apps, the report system based on crowdsourcing can cover every possible placement fraud. We implement our system in 10 tablets with 500 apps to evaluate its effectiveness. Experiment result shows that our approach can record enough data to analysis which app has placement fraud. What’s more, our system can figure out some special placement fraud which pop ads when user is using other apps. This placement fraud cannot be solved through automatic method in previous studies.

As the basis of vehicle ad hoc networks, the method of forwarding data is one of the most important parts which ensures the stability and efficiency of network communication. However, the high-speed mobile vehicle nodes cause frequent changes of network topology and disconnections of network links, casting a big challenge to the performance of network data delivery. Data forwarding methods based on the prior knowledge of vehicle’s trajectory are difficult to adapt to the changing vehicle trajectory in real world applications, while getting destination vehicles’ positions in broadcast way are extremely costly. To solve the above problems, we have proposed an association state based optimized data forwarding method (ASODF) with the assistance of low loaded road side units (RSU). The proposed method maps the urban road network into a directed graph, utilizes the carry-forward mechanism and decomposes the data transmission into decision-making data forwarding at intersections and data delivery on roads. The vehicles carried data combine the destination nodes locations obtained by low loaded road side units and their locations into association states, and the association state optimization problem is formalized as a Reinforcement Learning problem with Markov Decision Process (MDP). We utilized the value iteration scheme to figure out the delay-optimal policy, which is further used to forward data packets to obtain the best delay of data transmission. Experiments based on a real vehicle trajectory data set demonstrate the effectiveness of our model ASODF.

As the fast-paced market of smart phones, navigation application is becoming more popular especially when traveling to a new place. As a key function, shortest path recommendation enables a user routing efficiently in an unfamiliar place. However, the source and destination are always critical private information. They can be used to infer a user’s personal life. Sharing such information with an app may raise severe privacy concerns.In this paper, we propose a practical navigation system that preserves user’s privacy while achieving practical shortest path recommendation. The proposed system is based on graph encryption schemes that enable privacy assured approximate shortest path queries on large-scale encrypted graphs. We first leverage a data structure called a distance oracle to create sketch information, and we further add path information to the data structure and design three structured encryption schemes. The first scheme is based on oblivious storage. The second scheme takes advantage of the latest cryptographic techniques to find the minimal distance and achieves optimal communication complexity. The third scheme adopts homomorphic encryption scheme and achieves efficient communication overhead and computation overhead on the client side. We also evaluated our construction. The results show that the computation overhead and communication overhead are reasonable and practical.

In-vehicle networks which were originally designed to operate in a closed environment without secure concerns are now being connected to external nodes/networks and providing useful services. However, communications with the external world introduce severe security threats to the vehicle. For connected vehicle, many attacks, which were only feasible with physical access to a vehicle, can now be carried out remotely over wireless networks. To overcome this problem, we propose a security protocol to protect in-vehicle networks based on current Controller Area Network specifications. First, we generate the secure in-vehicle networks by using a group key. Then, we make the gateway join the secure in-vehicle networks after authenticating it. Finally, we generate the pair-ware key to ensure the secure communication between the external node and the gateway. The security analysis and performance evaluation show that the proposed scheme is secure and practical.

With the arrival of big data era, distributed computing framework Hadoop has become the main solution to deal with big data now. People usually promote the performance of distributed computing by adding new computing nodes to cluster. With the expansion of the scale of the cluster, it produces a large amount of power consumption because of lack of reasonable management strategy. So how to make full use of computing resources in the cluster to improve the performance of the whole system and reduce the power consumption has become the main research direction of scholars and industrial circles. For the above, in order to make best use of computing resources and reduce the power consumption, this paper firstly proposes to optimize a reasonable configuration of the parameters provided by Hadoop. Comparing with the default configuration of Hadoop. It shows we can get better performance by parameter tuning. This paper proposes a task scheduling mechanism based on memory usage prediction. In this task schedule, it predicts the future use status of memory in the computing nodes by analyzing the use status before. The task scheduling mechanism can reduce the memory pressure by reducing the allocation of tasks when the computing node is under memory pressure. The task scheduling mechanism can be more flexible by setting the threshold of memory usage. This mechanism based on predicting memory usage can improve the performance of the system by making full use of the computing resources.

Recently there is an increasing need for online community analysis on large scale graphs. Community search (CS), which can retrieve communities efficiently on a query request, has received significant research attention. However, existing CS methods leave edge direction and vertex attributes out of consideration, which results in poor performance of community accuracy and cohesiveness. In this paper, we propose DACQ (directed attribute community query), a novel framework of retrieving effective communities in directed attributed graphs. DACQ first supplements attributes according to the topological structure and generate attribute combinations, after which DACQ finds the strongly connected k-cores (k-SCS) with attributes in the directed graph. Finally, DACQ retrieves effective communities, which are cohesive in terms of the structure and attributes. Extensive experiments demonstrate the efficiency and effectiveness of our proposed algorithms in large scale directed attributed graphs.

Road users are now able to retrieve safety information, computing task results and subscribing content through various vehicular ad hoc network (VANET) services. Most commonly used services are safety beacon, cloud computation, and content subscription. Road users concern more about data security than ever. Privacy preserving authentication (PPA) is one main mechanism to secure inter-vehicle messages. However, for historical reasons, PPA for three services are different and therefore hard to be unified and not lightweight enough. To improve the flexibility and efficiency of PPA for various VANET services, it is necessary to securely authenticate messages preserving privacy for individual service, but also to unify PPA processes of various services in one vehicle. Here we propose an Efficient Pseudonymous-based Inter-Vehicle Authentication Framework for various VANET services. Our novel framework employs three methods. Method No. 1 consists of a decentralized certificate authority (CA), which allows vehicles to communicate only if vehicles registering themselves. Method No. 2 adopts a three-stage mutual authenticating process, which adapts to different communicating models in various services. Method No. 3 we design a universal basic module that requires only lightweight hashing and MAC operations to accomplish the signing and verifying processes. To analyze the security performance of our EPAF, we use automated tool under symbolic approach. Our results strongly suggest that EPAF is secure, robust and adaptable in vehicular safety, as well as in content and cloud computation services. To analyze the performance of EPAF, we calculate benchmarks and simulate the network. Our results strongly suggest that EPAF reduces computation cost by 370–3500 times, decreases communication overhead by 45.98%–75.53% and CA need not to manage CRL compared with classical schemes. In conclusion our framework provides insights into how data privacy can be simultaneously protected using our EPAF, while also improving communication and computing speed even in high traffic density.

With the rapid development of wearable sensor technology, Human Activity Recognition (HAR) based on sensor data has attracted more and more attentions. The Hidden Markov Model (HMM) with perfect performance in speech recognition has a good effect on HAR. However, almost all these techniques train multiple Hidden Markov Models for different classes of activity. For a given activity sequence with multiple activities, the activity corresponding to the HMM with the maximum generating probability is selected as the recognition result, which is not suitable for continuous HAR with multiple activities. For this problem, we propose an improved Hidden Markov activity recognition algorithm where discriminative model and generative model are utilized. The discriminative model SVM is used to produce the observation sequence of HMM, and the generative model HMM is used to generate the final result. Compared with the traditional Hidden Markov HAR model, our proposal has good performance in terms of precision, recall and F1 score.

For flow prediction in intelligent traffic system, one certain model cannot get excellent performance under different environments. Predicting models should also be updated according to data stream. In order to resolve these problems, a prediction method based on complex event processing was proposed. With fuzzy ontology to model historical event context and context clustering to partition events, this method could learn Bayesian network models according to different data during complex event processing. Appropriate Bayesian network model or combination of Bayesian network models could be provided by this method for real-time prediction and analysis of current context of events. The experimental result shows that this method can process events stream of Cyber Physical System (CPS) effectively and has favorable prediction performance.

Vehicular network can provide Internet connectivity for mobile vehicle by handover mechanism. However, existing handover schemes still face poor handover performance when they are applied in vehicular network. Software Defined Network (SDN) is a new architecture which can be used to optimize vehicular network by making network devices to be programmable. In this paper, we propose a new fast handover scheme for SDN based vehicular network to improve handover performance. SDN controllers of our scheme predict movement of vehicles by detecting port status of SDN switches, and then they start to perform the proactive handover procedure based on prediction results. Evaluation results show that the handover delay and packet loss of our scheme are lower than the contrast schemes. Simulation results prove that our handover scheme is more fit for delay sensitive vehicular network.

In unattended wireless sensor networks (UWSNs), static sensor nodes monitor environment, store sensing data in memory temporally. Mobile sink patrols and collects the sensors’ data itinerantly. Mobile sink is granted with more permissions than static sensor nodes, rendering it more attractive to the adversary. By compromising the mobile sinks, the adversary can not only seek the sensing data, but it also can steel all kinds of keys and access permissions, which may be abused to undermine other benign sensor nodes, even worse to upset the whole network. Currently, many related works focus on key management, permission management to restrict the compromised mobile sink or authentication to guarantee data reliability. However, the issue of compromised mobile sinks attracts little attention, and gradually become one obstacle to the application of UWSNs.In this paper, we proposed a secret-sharing method for detecting compromised mobile sink in UWSNs. Before the sensing data are collected by the mobile sink, every sensor node splits the digest of its data into shares by using a polynomial secret sharing algorithm, and dispatches these secret shares to randomly chosen neighbor nodes, which thereafter send to the base-station through different routes. After enough shares are gathered, the base-station recovers the original data digest, which will be used to validate the sensing data submitted by the mobile sink. If the validation fails, it reveals a compromised mobile sink. Theoretical analysis and evaluation indicate the effectiveness and efficiency of our method. Also, we proposed two types of attacking model of the mobile adversary, and obtained the respective detection probability.

With its continuous development, location information acquisition technology is able to collect more and more trajectory data, and the rich information contained therein is gradually attracting attention from researchers. Trajectory data involves complex relationships among moving objects, time, space, which are hard to understand and be used directly. Nowadays, visual analysis of trajectory data is mainly focus on its representation and interaction, but fails to address the complex correlation contained in trajectory data. Hence, we propose TrajHIN, a heterogeneous information network model built on trajectory data, measure the meta path-based similarity and centrality, and use a visual analytics method to deeply understand trajectory data. The example of visual analysis of real trajectory data has been interpreted and given feedback from domain experts, which proves effectiveness of TrajHIN and feasibility of mining implicit semantic information from trajectory data.

Traffic flow often contains massive amounts of information that is related to location and shows some regularity. And the traffic flow analysis based on trajectory data has become one of the most popular research topics in recent years. With the wide application of deep learning and for its higher accuracy than other approaches, methods such as convolution neural network and deep residual network have been introduced in traffic flow research and achieve good results. However, these methods usually require the training of a large number of parameters, which leads to some problems. For example, frequent manual adjustment is needed, and some parameters cannot be dynamically adjusted with the training process. We find that learning rate plays a crucial role in all parameters, which has important influence on the training speed of the residual network. In other words, the soundness of traffic flow predication results depends on the learning rate. Hence, we propose G4 algorithm to automatically determine the learning rate. It can be adjusted automatically in the process of trajectory data mining, and therefore solve the traffic flow prediction problem. Experiments on real data sets show that our method is effective and superior over some traditional optimizing methods of traffic flow analysis.

Ad hoc networks due to its open and dynamic nature are susceptible to a variety of security threats. Recently trust management emerged as the promising candidate to provide less computational and secure solutions. This paper based on novel idea i.e. trust mechanism based AODV (Ad hoc On-demand Distance Vector) routing protocol in the network layer of MANET to enhance network security. In our proposed scheme trust among nodes is represented by opinions, which is an item derived from three Valued subjective logic (3VSL). The opinions are dynamic, therefore judged, combined and updated frequently. In 3VSL the theoretical capabilities are based on Dirichlet distribution by considering prior and posterior uncertainty of the said event. Meanwhile, using advised and personal trust, a node can make relative judgment for forward node authenticity. The simulated results validate the accuracy of our proposed scheme.

Mobile edge computing (MEC) is a technology that transfers resource to the edge of network, which spares more attention to giving users easier access to network and computation resources. Due to the large amount of data computation needed for devices in Internet of Things, MEC technology is applied to improve computing efficiency. Though MEC can be applied to the Internet of Things, it needs further consideration on how to efficiently and reasonably allocate computing resources, and how to minimize the computing time of all users. This paper proposes a computing resources allocation scheme based on hybrid quantum-behaved particle swarm optimization. Simulation experiments with the network environment based on the Internet of Things is carried out. The results show that this algorithm can accelerate the whole computing process and reduce the number of iterations.

In modern life, software plays an increasingly important role and ensuring the reliability of software is of particular importance. In space, a Single Event Upset occurs because of the strong radiation effects of cosmic rays, which can lead to errors in software. In order to guarantee the reliability of software, many software-based fault tolerance methods have been proposed. The majority of them are based on data redundancy, which duplicates all data to prevent data corruption during the software execution. But this fault tolerant approach will make the data redundant and increase memory overhead and time overhead. Duplicating critical variables only can significantly reduce the memory and performance overheads, while still guaranteeing very high reliable results in terms of fault-tolerance improvement. In this paper, we propose an analysis model, named CDM (Critical Data Model), which can compute the critical of variables in the programs and achieve the purpose of reducing redundancy for the reliable program. According to the experimental results, the model proposed in this paper can enhance the reliability of the software, reduce the time and memory cost, and improve the efficiency of the reliable program.

With the development of sensors on smart devices, many applications usually learn an accurate model based on the collected sensors’ data to provide new services for users. However, the collection of data from users presents obvious privacy issues. Once the companies gather the data, they will keep it forever and the users from whom the data is collected can neither delete it nor control how it will be used.In this paper, we design, implement, and evaluate a practical privacy-preserving deep learning model that enables multiple participants to jointly learn an accurate model for a given objective. We introduce a light-weight data sanitized mechanism based on differential privacy to perturb participant’s local training data. After that, the service provider will collect all participants’ sanitized data to learn a global accurate model. This offers an attractive point: participants preserve the privacy of their respective data while still benefitting from other participants’ data. Finally, we theoretically prove that our APDL can achieves the $$\varepsilon $$ε-differential privacy and the evaluation results over a real-word dataset demonstrate that our APDL can perturb participant data effectively.

Maximizing the lifetime of barrier coverage is a critical issue in randomly deployment sensor networks. In this paper, we study the barrier coverage lifetime maximization problem in a bistatic radar network, where the radar nodes follow a uniform deployment. We first construct a coverage graph to describe the relationship among different bistatic radar pairs. We then propose a solution to maximize the barrier lifetime: An algorithm is first proposed to find all barriers based on coverage graph and then determines the operation time for each barrier by using linear programming method. We also propose two heuristic algorithms called greedy algorithm and random algorithm for large-scale networks. Simulation results validate the effectiveness of the proposed algorithms.

Satellite communication system is expected to play an important role in wireless networks because of its appealing contributions to ubiquitous coverage, content multicast and caching, reducing user expenditure, and so on. However, due to the inherent broadcasting nature and serious channel conditions, satellite communication system is highly vulnerable to eavesdropping attacks. As an initial step towards this end, this paper focuses on the physical layer security technique and explores the secrecy performance of a multibeam satellite system, where multiple legitimate users are served and each user is exposed to an eavesdropper located in the same beam. With perfect channel state information at the satellite and adopting the complete zero-forcing approach for signal processing, we first derive the optimal beamforming vectors to maximize the achievable secrecy rate. Based on this, we further calculate the secrecy outage probabilities of an individual user and the whole system, respectively. Finally, simulation and numerical results are provided to show the secrecy performance of the multibeam satellite system.

Gestures recognition as the main technology of human-computer interaction draws a great amount attention of researchers. Comparing to existing methods, the RFID-based passive gesture recognition requires no specialized equipment which makes it much easier to be used. To achieve the goal, we build a priori gesture database according to signal features caused by perturbation of different gestures. Then, the modified dynamic time warping (DTW) algorithm has been used to match with the priori fingerprint database. Besides, we propose a wireless phase calibration algorithm by utilizing the theory that the noise subspace and the signal subspace is orthogonal in multiple signal classification (MUSIC) algorithm to estimate and remove phase errors that may caused by equipment differences so that we can ensure the accuracy of angle of arrival (AoA) estimation. To evaluate the effectiveness of our gesture recognition system, the experiments in a real scene were carried out. And the experimental results show that we can achieve about 92% accuracy.

Integrating GPU with CPU on the same chip is increasingly common in current processor architectures for high performance. CPU and GPU share on-chip network, last level cache, memory. Do not need to copy data back and forth that a discrete GPU requires. Shared virtual memory, memory coherence, and system-wide atomics are introduced to heterogeneous architectures and programming models to enable fine-grained CPU and GPU collaboration. Programming model such as OpenCL 2.0, CUDA 8.0, and C++ AMP support these heterogeneous architecture features. Data partition is one of the collaboration patterns. It is essential for improving performance and energy-efficiency to balance the data processed between CPU and GPU. In this paper, we first demonstrate that the optimal allocation of data to the CPU and GPU can provide 20% higher performance than fixed ratio of 20% for one application. Second, we evaluate another 5 heterogeneous applications covering the latest architecture features, found the relation of the data partitioning with performance.

Energy Internet is considered as a promising approach to solve the problems of energy crisis and carbon emission. It needs to collect user’s real-time data for optimizing the energy utilization. Edge nodes like GWs (gateway) are used for data aggregation to improve the efficiency of the system. Due to a large number of GWs are widely distributed and difficult to be managed, which brings potential security threats for the Energy Internet. Existing data aggregation schemes fails in preventing the adversary from controlling or destroying GWs. In this paper, we propose an IBE-based Device Traceable Privacy-Preserving Aggregation Scheme, named IBE-DTPPA. Increasing the RA (Residential Area) users’ data aggregation integrity verification by BGN Cryptosystem; using IBE Cryptosystem to encrypt aggregation data, calculating ciphertext based on GW’s dynamic ID, realizing the target GW traceability; choosing CC (Control Center) dynamic identity information as public key to realize CC authentication, preventing adversary from using CC’s identity fraudulently. Through extensive analysis, we demonstrate that IBE-DTPPA resists various security threats, and can trace target GW efficiently.

Cloud computing is an Information Technology (IT) model that provides convenient, on-demand network access to a shared pool of configurable computing resources (e.g., networks, servers, storage, applications, and services), which can be rapidly provisioned and released with minimal management effort and service provider interaction. Infrastructure as a Service (IaaS) is a new trend setter in the field of cloud computing which recently emerged as a new architype for hosting and delivering services on the internet. This study will discuss the characteristics and benefits of operating Virtual Web-Hosting together with Infrastructure as a Service (IaaS) model of cloud computing. Moreover, this study will also highlight the architectural principles, main concepts, and state of the art implementation and challenges of virtual web-hosting on Infrastructure as a service (IaaS).

“E-photo” is to solve the inconvenient problem when users want some photos. It is a self-service model with the help of internet crowdsourcing. Users download “E-photo” and register as “E-photo” members, then they can get the task and earn the corresponding reward as incentives. The pricing task is the core problem in “E-photo”. If the pricing is not reasonable, some tasks will not be cared, thus involving the failure of commodity inspection. In this paper, we fuse the logistic regression model and cohesive hierarchical model, and propose a better pricing method for “E-photo”.

Indoor planting can purify the air, beautify the environment, satisfy people by closing to the nature and farming. However, the plants are easy to stop growing or even to die due to the lack of proper environment situations, such as lack of water or sunlight. This paper leverages the Internet of Things (IoT) and cloud computing technology to monitor the light intensity, air temperature and soil humidity of indoor plants. The plant growth condition and environment situation are also reflected to the user’s smartphone and stored in the cloud. Outdoor users can also control the water pump to irrigate the plants and LED to add light supply via their smartphones. With our prototype, our system accurately monitors the environment and intelligently controls the plant growth.

Amplification attack, as a new kind of DDoS attack, is more destructive than traditional DDoS attack. Under the existing Internet architecture, it is difficult to find effective measures to deal with amplification attack. In this paper, we propose a two-phase reference detecting scheme by utilizing Software Defined Infrastructure capabilities: switch side is volume-based and controller side is feature-based. The proposed scheme is protocol-independent and lightweight, unlike most of the existing strategies. It can also detect amplification attack in the request phase for a small price, before these attacks cause actual harm. Upon the architecture, we design detection algorithms and a prototype system. Experimental results with both online and offline data sets show that the detection scheme is effective and efficient.

The amount of research done in the field of mobile ad hoc networks is extraordinarily large. Evaluation of protocols designed for ad hoc networks is challenging as the cost of node deployment in terms of resources required is high, hence, most of the researchers use simulations for performance evaluation. In this paper we address the pitfalls of simulation studies in ad hoc routing protocols published in recent years. We have conducted a survey to evaluate the current state of simulation studies published in top conference/journals of the communication domain. In majority of the published papers (the way simulation results are reported) we have found design flaws, unrealistic assumptions, are non-reproducible, and statistically invalid results. We also propose a standardizing architecture for automating the reporting and replication process for network simulators. This platform independent architecture alleviates the challenge of simulation parameter reporting and facilitates in designing better network simulation experiments.

Mobile cloud computing (MCC) can significantly enhance computation capability and save energy of smart mobile devices (SMDs) by offloading remoteable tasks from resources-constrained SMDs onto the resource-rich cloud. However, it remains a challenge issue how to appropriately partition applications and select the suitable cloud to offload the task under the constraints of execution cost including completion time of the application and energy consumption of SMDs. To address such a challenge, in this paper, we first formulate the partitioning and cloud selection problem into execution cost minimization problem. To solve the optimization problem, we then propose a system framework for adaptive partitioning and dynamic selective offloading. Based on the framework, we design an optimal cloud selection algorithm with execution cost minimization which consists of offloading judgement and cloud selection. Finally, our experimental results in a real testbed demonstrate that our framework can effectively reduce the execution cost compared with other frameworks.