A Two-step Keyword Spotting Method Using Fuzzy Search Algorithm

For many practical applications, input signal is a spontaneous conversation while the acoustic model was trained with read speech because of data availability. The performance of Keyword spotting system will degrade significantly. To solve this problem, we presented a two- step keyword spotting strategy. A fuzzy search algorithm was adopted to improve the retrieval performance. We generated confusion matrix from confusion network. It gives more freedom in the representation so as to alleviate the effect phoneme mlsrecognition. Furthermore a hybrid confidence measure was applied to reject false accepts. Experiments show that the proposed algorithms significantly reduce Equal error rate （EER） on the telephone conversational task.     

Fast Fuzzy Keyword Spotting Using Syllable Confusion Network Indexing

This paper presents a fast fuzzy search algorithm to extract keyword candidates from Syllable confusion networks （SCNs） in Mandarin spontaneous speech. Since the recognition accuracy of spontaneous speech is quite poor, Syllable confusion matrix （SCM） is applied to compensate for the recognition errors and to improve recall. In order to scale up to large collections and support quick query response, an efficient vocabulary-independent index structure is designed, which selects individual arcs of syllable confusion network as indexing unit. An inverted search algorithm that use syllable confusion matrix to calculate relevance score and search in this index structure is proposed. In experiments performed on a telephone conversational task, the Equal error rate （EER） was reduced by about 33% relative over the baseline where keywords are directly extracted from phoneme lattices. Additionally, it only took computer one or two seconds to search 100 keywords in one hour speech data.     

Top-K Query Framework in Wireless Sensor Networks for Smart Grid

The smart grid has caught great attentions in recent years, which is poised to transform a centralized, producer-controlled network to a decentralized, consumer- interactive network that＇s supported by fine-grained monitoring. Large-scale WSNs （Wireless Sensor Networks） have been considered one of the very promising technologies to support the implementation of smart grid. WSNs are applied in almost every aspect of smart grid, including power generation, power transmission, power distribution, power utilization and power dispatch, and the data query processing of ＇WSNs in power grid＇ become an hotspot issue due to the amount of data of power grid is very large and the requirement of response time is very high. To meet the demands, top-k query processing is a good choice, which performs the cooperative query by aggregating the database objects＇ degree of match for each different query predicate and returning the best k matching objects. In this paper, a framework that can effectively apply top-k query to wireless sensor network in smart grid is proposed, which is based on the cluster-topology sensor network. In the new method, local indices are used to optimize the necessary query routing and process intermediate results inside the cluster to cut down the data traffic, and the hierarchical join query is executed based on the local results.Besides, top-k query results are verified by the clean-up process, and two schemes are taken to deal with the problem of node＇s dynamicity, which further reduce communication cost. Case studies and experimental results show that our algorithm has outperformed the current existing one with higher quality results and better efficiently.     

A Novel Ship Detection Approach for Polarimetric SAR Images Based on a Foreground/Background Separation Framework

In this paper, we present a novel ship detection approach for Polarimetric synthetic aperture radar （PolSAR） images based on a Foreground/background separation （FBS） framework, which exploits the statistical dissimilarity of PoISAR data to separate desired targets from the background clutter. Since the FBS framework takes the spatial relations between pixels into consideration and the separation process exploits the inherent characteristics of PolSAR data, the proposed detector is stable to speckle. And the detection process is filter-free, thus it can preserve the edge and shape information of the extracted targets. Experimental results and comparisons with the standard polarimetric detector show that the proposed method is promising. Factors that affect the performance of the proposed detector are also analyzed and discussed in this paper.     

Self-Organization Approaches for Optimization in Cognitive Radio Networks

Cognitive radio （CR） is regarded as a promising technology for providing a high spectral efficiency to mobile users by using heterogeneous wireless network architectures and dynamic spectrum access techniques. However, cognitive radio networks （CRNs） may also impose some challenges due to the ever increasing complexity of network architecture, the increasing complexity with configuration and management of large-scale networks, fluctuating nature of the available spectrum, diverse Quality-of-Service （QoS） requirements of various applications, and the intensifying difficulties of centralized control, etc. Spectrum management functions with self-organization features can be used to address these challenges and realize this new network paradigm. In this paper, fundamentals of CR, including spectrum sensing, spectrum management, spectrum mobility and spectrum sharing, have been surveyed, with their paradigms of self-organization being emphasized. Variant aspects of self- organization paradigms in CRNs, including critical functionalities of Media Access Control （MAC）- and network-layer operations, are surveyed and compared. Furthermore, new directions and open problems in CRNs are also identified in this survey.     

A Probability Driven Quadri-Stage Contention Protocol for Mobile Ad Hoc Networks

The paper proposes a new contention based Time division multiple access （TDMA） Medium access control （MAC） protocol in wireless ad hoc networks, which can be termed Quadri-stage contention protocol （QSCP）. The QSCP protocol is free of both the ＂hidden terminal＂ and the ＂exposed terminal＂ problems. It is distributive and arbitrarily scalable. In the dedicated contention stage, nodes exchange their ＂Normalized accumulation and dissipation rate of payload （NADROP）＂ messages in 2-hop neighborhood, and then run a contention probability calculation algorithm to decide the probability of sending a contention frame. The algorithm guaran- tees that the number of time slot reserved by a node is proportional to its NADROP value, which can remarkably improve the end-to-end throughput when network load is heavy, and is more efficient in energy consumption. The simulation results demonstrate that QSCP performs better than the existing protocols.     

Best Effort Spectrum Allocation Scheme for Femtocell Networks in Dense Deployment

In orthogonal frequency division multiple access（OFDMA） based femtocell networks,the co-tier interference among femto base stations（FBS） becomes important in multiuser and densely deployed environment.In order to mitigate the co-tier interference and enhance the system total throughput,this paper proposed a best effort spectrum allocation scheme based on the extension of graph theory.In the scheme,a controller was proposed to collect the channel state information（CSI）of all femtocell user equipments（FUEs） in a certain range.Then,the controller evaluated the signal-to-interference Ratio（SIR） of each FUE and determined the set of its interference neighbors.By calculating the received power matrix（RPM） among FUEs and building interference graph matrix（IGM）,different spectrum resource blocks（RBs） were assigned to the users with interference relation,while users without interference relation shared the same RBs,which could increase the spectrum efficiency.Simulation results show that the proposed algorithm can significantly improve the RB usage efficiency compared with the basic graph coloring theory,and more than 80% improvement can be acquired in dense deployment scenario.Besides,the throughput of both cell edge macro user equipments（MUEs） and cell edge FUEs is guaranteed on the premise of low interference.     

A Scalable Peer-to-Peer Web Services Overlay for Semantic Discovery

The growing number of web services advocates distributed discovery infrastructures which are semantics-enabled. We introduce a novel approach for semantic discovery of web services in structured P2P-based （Peer-to-Peer） registries. We partition concepts into different CGs （Concept groups） and index these ordered concept sets in a two-phase semantic routing mechanism. Each node and query has one concept set as its unique ID. For each concept in the query, current node first route the query to a node with nodeID that shares a prefix as long as with the query, but a longer CG. Second, the query will be routed to a node sharing a longer prefix than that of the current node. Combining with semantic similarity calculation, our proposal guarantees that most existing services matching a query will be found with bounded costs in terms of number of nodes involved. The scalability and reliability of our approach are also confirmed through emulation tests.     

Action-Based Access Control Modelt

Access control is one of the powerful and generalized approaches for restricted resource access. The environmental state is introduced and the term “action” is defined based on roles, temporal states and environmental states. Actions can be used to capture security-relevant aspects of roles, environmental and temporal states in different information systems. Then, the action hierarchy, environmental hierarchy, temporal hierarchy and Action- based access control （ABAC） model are presented. ABAC is compared with the existing models and the result shows that the ABAC model can solve the problem of access control in information systems with mobile computation for its convenient and flexible designs. An application example of ABAC model is described at the end of the paper.     

Novel Active Contour Model for Image Segmentation Based on Local Fuzzy Gaussian Distribution Fitting

A novel active contour model is proposed, which incorporates local information distributions in a fuzzy energy function to effectively deal with the intensity inhomogeneity. Moreover, the proposed model is convex with respect to the variable which is used for extracting the contour. This makes the model independent on the initial condition and suitable for an automatic segmentation. Furthermore, the energy function is minimized in a computationally efficient way by calculating the fuzzy energy alterations directly. Experiments are carried out to prove the performance of the proposed model over some existing methods. The obtained results confirm the efficiency of the method.     

On the Design and Temporal Petri Net Verification of Grid Commerce Architecture

The emergence of Grid Computing settings lays a solid foundation for designing and constructing a next generation E-commerce environment. This work presents a new E-commerce architecture based on a Grid Computing environment, and models and analyzes the architecture using Temporal Petri nets （TPNs）. It can make sufficiently use of the Grid resources to implement various electronic bargains between the geographically dispersed users. By means of its TPN model, some main temporal properties are represented and verified formally.     

PTCP Incast in Data Center Networks

In recent years, dual-homed topologies have appeared in data centers in order to offer higher aggregate bandwidth by using multiple paths simultaneously. Multipath TCP （MPTCP） has been proposed as a replacement for TCP in those topologies as it can efficiently offer improved throughput and better fairness. However, we have found that MPTCP has a problem in terms of incast collapse where the receiver suffers a drastic goodput drop when it simultaneously requests data over multiple servers. In this paper, we investigate why the goodput collapses even if MPTCP is able to actively relieve hot spots. In order to address the problem, we propose an equally-weighted congestion control algorithm for MPTCP, namely EW-MPTCP, without need for centralized control, additional infrastructure and a hardware upgrade. In our scheme, in addition to the coupled congestion control performed on each subflow of an MPTCP connection, we allow each subflow to perform an additional congestion control operation by weighting the congestion window in reverse proportion to the number of servers. The goal is to mitigate incast collapse by allowing multiple MPTCP subflows to compete fairly with a single-TCP flow at the shared bottleneck. The simulation results show that our solution mitigates the incast problem and noticeably improves goodput in data centers.     

A New Key Management Protocol to MANET

MANET （Mobile ad-hoc networks） are networks in which nodes are mobile and link connectivity might be changing all the time. The wireless and dis- tributed nature of MANET pose a great challenge to system security. Key management is an important and complex problem to the authentication and system security. The paper proposes a new method to the key management with high efficiency. Which,making use of the ideas of self-organization and local driven by intercross authenti-cation of CAs and threshold encryption. It＇s proved that the new method can reduce the communications delay and time delay in the key management process compared to other methods. So, it is a contribution to the field of key management in MANET.     

An Effective Gated Scheduling with Pre-defined QoS Cross-layer Design for Wireless Networks

The proposed M-gated scheduling algorithm rules the server in wireless networks stays with a station for at most M times gated services. By the method of embedded Markov chain and probability generating function we developed a model with vacations and acquired the theoretical value of average queue length and packet delay. Like E-limlted scheduling algorithm, its performance is close to that of the optimal polling scheme, exhaustive scheduling, but the inherent unfairness problem for the latter is avoided. And it also does not require the server to know the status of stations, queues information. Compared with E-limited, M-gated has less jitter and is more robust. We then use the derived quality of service parameter as the basis for cross-layer design. Simulations results show the network is more stable and can keep the queue length and the packets delay under QoS level. When putting E- limited and M-gated scheduling in the same cross-layer design scheme, the latter has less overhead especially when the system is heavily loaded.     

Joint Optimization of Subchannel Reassignment and Power Adaptation for Multicarrier Cooperative Systems

An amplify-and-forward multicarrier cooperative system was proposed in this paper. The optimal resource-allocation problem was investigated in terms of improving data rate. An effective resource-allocation algorithm was proposed based on three techniques, adaptive modulation, subchannel reassignment design and transmit power adaptation. The designed subchannel reassignment not only increases the average data rate, but also raises the threshold of ＂water-filling＂ in optimizing power allocation that can further increase data rate. Since the power allocation is based on the designed subchannel reassignment, the complexity of the proposed algorithm is reduced greatly. Simulation results show that the proposed algorithm effectively increases the system＇s average data rate.     

A Novel Surface Treatment Method for Improving the Antenna System Performance of Mobile Phones with Stainless Steel Frame

A novel surface treatment method of plat-ing Cu＋PPS film/coating on a mobile phone＇s stainless steel frame for improving the antenna system efficiencies is proposed. The mobile phone was measured in free space, in a silicon cover, and in the hand and cover si- multaneously. It＇s found that with this surface treat-ment, the total efficiency of the antenna system can be improved in all the four cases respectively by 14.22%, 1.38%, 15.19% and 1.72% at 940MHz （GSM900：880- 960MHz）, 2.59%, 3.21%, 4.81% and 1.43% at 1720MHz （DCS：1710-1880MHz） and 6.34%, 2.85%, 9.83% and 2.32% at 2100 MHz （WCDMA：1920-2170MHz）. This low- cost surface treatment method is an important break- through to improve antenna system performance of mobile phones especially for those with a stainless steel frame, and suitable for mass production.     

Spaceborne/Airborne Hybrid Bistatic SAR Frequency Domain Imaging Processing Algorithm

An imaging algorithm mainly processed in the frequency domain for the hybrid spaceborne/airborne BSAR is presented. The key point of deriving the algo- rithm is the analytical evaluation of the system transfer function＇s 2-D spectrum. To overcome the difficulty of resolving analytical solution for the stationary phase point, the spectrum＇s phase is approximated by two-order Taylor expanding around the point, which is in the neighborhood of the system＇s corresponding stationary phase point and can be obtained analytically. Thus the approximated analytical spectrum is pretty close to the actual one. In the imaging algorithm, both range-dependent range cell migration and azimuth-dependent range cell migration are compensated in two steps： Scaled inverse Fourier trans- form （SIFT）, which can be realized through the chirp z-transform, and phase multiplication. The validity of the algorithm is demonstrated by simulation with both the point target and area target.     

Labeling Algorithm for Face Detection Using Skin and Hair Characteristics

This research presents an algorithm for face detection based on color images using three main components： skin color characteristics, hair color characteristics, and a decision structure which converts the obtained information from skin and hair regions to labels for identifying the object dependencies and rejecting many of the incorrect decisions. Here we use face color characteristics that have a good resistance against the face rotations and expressions. This algorithm is also capable of being combined with other methods of face recognition in each stage to improve the detection.