- EI核心期刊
Social search is closely related to numerous research fields. First, the relationship between social search and traditional search pattern is analyzed and compared, the definition of social search is elaborated. Then, the theoretical basis of social search is introduced, the current status and key technology of social search are summarized. Finally, the new challenges of social search are analyzed, and the prospective research direction of social search is proposed. A more comprehensive and clear overview of the research area is outlined.
In order to balance energy consumption of network, a non-uniform grid partition mechanism (NuGPM) based on particle swarm optimization is proposed to improve network’s performance in wireless sensor networks. In NuGPM, network is split into <em>k</em> layers, the grids of the same layer have the same length, and the grids in different layers have different width. If the layer is closer to base station, its grid has bigger width. The particle swarm optimization algorithm is adopted to search the best width-combination of grids in all layers. The mechanism is able to make grids which are closer to base station have more nodes. By adopting this method, the grids near base station will have more energy to forward packets from grids of upper layer, and the hot spot problem near base station can be improved. The simulation results show that the proposed mechanism is able to balance the energy consumption of whole network, improve the hot spot problem near base station effectively, and extend the network’s lifetime.
The existing reconstruction algorithms in compressed sensing (CS) theory commonly cost too much time. A novel reconstruction algorithm based on inner product optimization is proposed to reduce reconstruction time. And also stopping criterion is derived from theory. The proposed algorithm computes the inner product of measurement matrix and the residual only in the first iteration during the reconstruction process. In the remaining iterations, the inner product of vectors instead of matrices is calculated. Then least square calculation is done only once to reconstruct the signal after iterations stopped. Experiments show that the proposed algorithm reduces the reconstruction time largely without degrading the quality of the signal.
Current intra-domain transmission in the overlay network ignores the leader-followers game characteristic between the Internet service providers and user. This will result in the problem of not achieving the best optimization state because the intra-domain transmission in real network applications may lack a reasonable equilibrium point. To solve this problem, a hierarchical overlay network architecture based on multi-agent game was proposed. Through constructing the leader-followers stackelberg game model between Internet service provider agent and user agent. Based on this model, the existence and uniqueness of Nash equilibrium was analyzed to get the best price and transmission rate under the Nash equilibrium state. Then, a quantitative description of the network’s best running was given. Finally, the effect of network topology parameters on the best running state was analyzed.
For solving the difficulties of web service composition with hybrid quality of service (QoS) and multiple decision makers, based on multi-attribute group decision making theory, hybrid QoS-aware web service composition (GDMA_HQoS) algorithm was presented. GDMA_HQoS can support multiple decision-makers and hybrid QoS information expressed by real numbers, interval numbers, triangular fuzzy numbers, intuitionistic fuzzy numbers (I). Other contributions included a novel QoS model, an algorithm of aggregating QoS. Experimental results show the proposed algorithm can be better applied in service composition with hybrid QoS and multiple decision makers.
Billing is a key function in the commercialization of cloud service, the service can be measured effectively by billing according to the demand of users. Analyzing the dynamic nature of process combination of service modules in cloud computing environment, the process billing model of hierarchical billing system is established based on modeling and tracking of dynamic process by stochastic Petri nets. Furthermore, in order to realize the billing control of cloud service providers based on state and optimize billing policy, the billing process of this system is analyzed and designed by Petri nets. Thus, the service measurement of cloud computing based on process model is achieved, and its corresponding pricing by this model can be made to realize billing according to the demand of users. Finally, the effectiveness and maneuverability of the billing model is verified by experiment, so the model can be used to support the billing decision of cloud service provider.
In conventional multicast scheme, the total throughput of multicast group is constrained by user with the worst channel quality. In order to overcome this problem of limited throughput, a resource allocation algorithm is proposed based on reducing feedback strategy and joint coding strategy for multicast systems. The layered coding combined with Reed-Solomon (RS) coding strategy which divides the multicast data into many layers is explored and compensated for data packet loss. In order to reduce the feedback load of uplink, different feedback strategies are given for different layers data. The optimal problem for resource allocation is formulated as well. In order to reduce the computational complexity, the suboptimal proportional fairness subcarrier allocation algorithm and water-filling power allocation algorithm with quality of service guarantees are presented. To further reduce the complexity, a power allocation algorithm with increasing fixed power (IFP-Q) is also proposed. Simulation shows that the proposed reducing feedback strategy can reduce more the feedback load of uplink, and the joint coding strategy can further improve the performance of systems.
An evolutionary game scheduling algorithm for cloud computing is proposed based on differentiated service. In this algorithm, cloud computing tasks are competed for resources by means of kinds of preferences, and virtual machine resources competed for tasks by their quality assessments such as calculation type, storage type, bandwidth type, all of them can form a mixed game. The algorithm can improve score of virtual machine resources and their populations according to task scheduling information and users feedback score. Finally, the equilibrium of the game can be achieved. Experiments indicate that the algorithm is available and efficient. It can allocate the virtual machine resources with different characteristics according to the types of tasks and guarantee that different kinds of users can have a better quality of service by users feedback score.
Combined opportunistic routing and network coding, a new any-path coding-aware opportunistic routing scheme is proposed. For reasonably choosing candidate forwarding nodes and assigning the nodes forwarding priority, the expected coding-aware transmission metric is presented. This routing scheme takes advantage of the inter-flow network coding to reduce the transmission numbers, and then improves the transmission efficiency. As the simulation is showed, it can improve forwarding efficiency and throughput of the network.
Establishing the loops mathematic module, the phase tracking error equation of the second-order frequency locking loop(FLL) assisted third-order phase locking loop(PLL) is derived. In the module, the tracking error of FLL is mixed into PLL. Two points of view are given: firstly, the phase dynamic tracking error of frequency locking loop assisted PLL(FPLL) is equal to zero in case of jerk; the phase thermal noise tracking error includes the thermal noise of PLL and the noise component of FLL, secondly, a new optimization method is proposed to optimizing both bandwidths of PLL and FLL. The optimization method minimizes the total phase tracking error to meet the need of constraint of FLLs frequency tracking error being within the PLLs locking bandwidth. Simulation shows that the analysis of phase tracking error is accurate; the optimization method is able to improve the tracking accuracy while being stable. This research will help the later extra design of FPLL.
The visual display unit of a personal computer(PC) emits electromagnetic waves that contain the information from the PC display. The information signals can be received and reconstructed secretly in the far field. To choose the best frequency range for receiving the compromising emanations effectively and quickly, a method based on the common equipments is proposed. Firstly, the frequency spectrum of 10 MHz to 1 GHz are measured about 10 times respectively when the PC is displaying special image with white-black vertical strips and power off. The frequency range which may contain the display information can be obtained through frequency subtraction. And then, the special image signal is searched around the dot frequency and harmonics. If the special signal disappears when displaying the total black image, it can be judged to be the real special signal. Finally, the best receiving frequency range can be chosen by comprehensively considering the background electromagnetic noise and the peak value of the special image signal in the frequency point. Experiment shows when intercepting the computer display, the optimal image could be restored within the chosen frequency range.
To solve the problem of low precision on information propagation prediction caused by frustrated link prediction in social networks, a new approach, based on time-series analysis, combined with information novelty, is proposed with no requiring the knowledge of social network.When the node is infected, the global influence function of a node is estimated. The overall propagation volume of information is predicted. Simulations on dataset about news articles and blogs show that the proposal can accurately and reliably predict the temporal variations of information diffusion.
To improve the adaptability in the complex pipe environment, the structure of a flexible squirm pipe robot is optimally designed and the force the pipe acts on the robot is analyzed. The robot is made up of flexible axl, front/back body, guiding head and supporting wheels. Analyzing the motion character of the pipe robot in straight and L-type pipe, a mechanical model of the pipe robot is established. The moving condition of the pipe robot is also inferred. A new experiment system is set up to verify the validity of the design.
A model of asymmetric digital subscriber line (ADSL) on-off-line behavior based on non-homogeneous poisson process is presented. First, a model of ADSL on-off-line behavior is established by formula deduction. Then, the model is used to analyze and calculate the user average probability of logout based on real data from ADSL network. The value is predicted by using the model and analyzed by contrasting with real data. Finally, the change of average probability of logout and continued online in different user group are given by the model. The analysis result shows that the model could give a quantitative description of On-off-line behavior of network users effectively, and will make great effort to anomaly detection for ADSL on-off-line behavior and public opinion survey.
Trust network is a social network and is constructed by trust relationships between agents. Different contributions devoted by two dimensions of trust are analyzed, and definitions of service trust and recommendation trust are given. The social characteristics of the trust network is studied and a trust prediction method based on agent’s role-based trust and reputation is proposed. The influences of interactions number, time, reputation, service trust and recommendation trust are comprehensively considered which concord with the cognitive semantics of trust computing and propagation in trust network. The experiment verifies the differences of the role-based trust values and the effectiveness of trust prediction.
Considering the limitation of communication capacity and energy for nodes in wireless sensor networks, a beamforming scheme based on immune genetic algorithm is proposed. First,the impact of the transmission coefficients and the number of selected nodes whose arrival phase have difference on the network energy consumption is analyzed. Then,the nodes participating in transmission are selected according to their residual energy and arrival phase combined with rotated factor and their transmission coefficients are adjusted using immune genetic algorithm. Simulation shows that the energy consumption among the nodes can be balanced and the network lifetime can be prolonged effectively.
The decoding of multi-hop Alamouti amplify and forward (AAF-MH) cooperation scheme usually needs channel state information, which is difficult to acquire in practical systems. To solve this problem, a parallel factor (PARAFAC)-based blind signal detection algorithm is proposed. This algorithm transforms received signals into a PARAFAC model which contains channel and signal information, and uses the bilinear alternating least square for the global convergence. Compared with constant modulus algorithm, the proposed algorithm is with better performance, such as more stable fitting results, and can realize uniqueness of parameter estimation. Simulation is given to support the analysis.
With help of a new cost-aware cloud service request scheduling algorithm based on dynamic reuse, according to current system load and divisible character of cloud service requests, the virtual resources can be rent and reused on demand to achieve optimal scheduling of dynamic requests in reasonable time.The rental cost of the overall infrastructure for increasing cloud service providers’ profits can be minimized when meeting service level agreement constraints. Simulation indicates that our proposed algorithm shows better performance compared with other revenue-aware algorithms in terms of resource utilization and operation profit.
A Netcoding-based Efficient Broadcast Transmission scheme (NEBT) is proposed for Delay Tolerant Mobile Sensor Network (DTMSN). In NEBT, the original data packages are coded and batched transmitted by Base Station (BS) and sensor nodes that have all the original data. Because the packages are encoded, the correlation of data between sensor nodes is low which is helpful in reducing the delay of broadcasting. At the same time, sensor nodes can compute their own broadcasting gains and detect the moving trends to their neighbors based on their signal strength. Based on the moving trends between nodes, sensor nodes can decide the time to transmit their packages, which can decrease the cost of broadcasting. Simulation results show that the proposed NEBT scheme has lower broadcast delay and broadcasting overhead than NBT, Flood et al.
The metric of topology tolerance in wireless sensor networks is the foundation of the topology tolerance research. Considering that the connectivity and coverage properties can reflect the network monitoring quality, the measuring parameters of the topology connectivity and coverage is proposed according to the usability and validity of the network. A measuring approach of topology tolerance-tolerance demand is provided by using the comprehensive service demand of the network connectivity and coverage. Simulation proves the difference reflection of the topology tolerance with the new measuring approach under different destruction strategy of random, rich-rich and poor-poor. It is shown that the measure can metric the diversification of topology tolerance accurately and effectively.
Diagnosis for wireless sensor network (WSN) is difficult, for the constraint of sensor node hardware and limit of network resources. Many existing approaches mainly aim at gathering a large amount of status information from individual sensor nodes, which brings additional communication on the network. Diagnosis based on sensing data (DSD) is designed and implemented,a passive diagnosis method using sensing date in WSN. GreenOrbs is a long-term deployment WSN system in the forest, which releases a large number of actual monitoring data on the Internet (http: //www.greenorbs.org). By studying those public sensing data and network status, a knowledge library to store those related regular patterns is established. Based on this self-improvement system, failures and traffic overhead can both be deduced. The experiment carried out in Testbed and realistic environment show that DSD can improve diagnosis efficiency, decrease the burden of network traffic, and have the flexibility with the arrangement of large-scale WSN.
For the blindness of cluster head election, the imbalance of energy consumption in the cluster and the substantial energy consumption in communication with a hop in low energy adaptive clustering hierarchy (LEACH) protocol, an improved algorithm of the LEACH protocol based on "game of life" is proposed. First the cluster head election mechanism based on the estimate of the node residual energy is optimized. Then the sleeping scheduling model based on "game of life" and communication of multi-hops which used its neighbor nodes as forwarding nodes are proposed. Network simulator version 2 simulation shows that the improved protocol effectively prolongs the survival time of the wireless sensor networks and increass the amount of data transmission.
A voting mechanism based broadcast authentication scheme for the mobile sensor networks is presented. The identity of the broadcast is verified and voted by nodes within the communication range to be authenticated. If the number of pros surpasses threshold <em>T<sub><</sub>/em>1, the node passes the authentication and joins the network. When the node moves, the mentioned above is used to assist authenticating. If the number of pros surpasses threshold <em>T<sub><</sub>/em>2, the node passes the authentication. Simulation shows that as long as security is assured during the authentication process, the proposed program requires less communication traffic thus consumes less energy than the three compared methods, the proposed method performs well on storage consumption with low-consumption.