To solve the problems of color distortion and local detail information loss caused by the histogram expansion of High Dynamic Range (HDR) image generated by single image, an imaging algorithm of high dynamic range image based on luminance partition fusion was proposed. Firstly, the luminance component of normal exposure color image was extracted, and the luminance was divided into two intervals according to luminance threshold. Then, the luminance ranges of images of two intervals were extended by the improved exponential function, so that the luminance of low-luminance area was increased, the luminance of high-luminance area was decreased, and the ranges of two areas were both expanded, increasing overall contrast of image, and preserving the color and detail information. Finally, the extended image and original normal exposure image were fused into a high dynamic image based on fuzzy logic. The proposed algorithm was analyzed from both subjective and objective aspects. The experimental results show that the proposed algorithm can effectively expand the luminance range of image and keep the color and detail information of scene, and the generated image has better visual effect.

Concerning the problem that low-cost Dual-antenna Attitude determination System (DAS) has low accuracy and gross error because of using direct solution, an improved algorithm based on carrier phase and pseudo-range double-difference Real-Time Kinematic (RTK) Kalman filter was proposed. Firstly, the baseline length was employed as the observation, then the precise baseline length obtained in advance was taken as the observation error. Secondly, the position of master antenna was corrected according to the epoch time interval of the slave antenna receiver and the integer ambiguity was solved by MLABMDA (Modified LABMDA) algorithm. Experimental results in static and dynamic mode show that the accuracy of the heading angle calculated by the proposed algorithm is about 1 degree and the calculated pitch angle accuracy is about 2-3 degrees in the case of baseline length 1.1 m with GPS and Beidou dual systems. The proposed algorithm improves the robustness and accuracy of the system greatly compared with the traditional dual-antenna attitude determination by direct solution.

As a large-scale distributed data processing framework, Hadoop has been widely used in industry during the past few years. Currently manual parameter optimization and experience-based parameter optimization are ineffective due to complex running process and large parameter space. In order to solve this problem, a method and an analytical framework for Hadoop parameter auto-optimization were proposed. Firstly, the operation process of a job was broken down into several microoperations and the microoperations were determined from the angle of finer granularity directly affected by variable parameters, so that the relationship between parameters and the execution time of a single microoperation was able to be analyzed. Then, by reconstructing the job operation process based on microoperations, a model of the relationship between parameters and the execution time of whole job was established. Finally, various searching optimization algorithms were applied on this model to efficiently and quickly obtain the optimized system parameters. Experiments were conducted with two types of jobs, terasort and wordcount. The experimental results show that, compared with the default parameters condition, the proposed method reduce the job execution time by at least 41% and 30% respectively. The proposed method can effectively improve the job execution efficiency of Hadoop and shorten the job execution time.

Concerning the problem that a predefined time window is usually used to mine simple association dependencies between events in the traditional frequent episode discovery, which cannot effectively handle interleaved temporal correlations between events, a concept of time-lag episode discovery was proposed. And on the basis of frequent episode discovery, Adjacent Event Matching set (AEM) based time-lag episode discovery algorithm was proposed. Firstly, a probability statistical model introduced with time-lag was introduced to realize event sequence matching and handle optional interleaved associations without a predefined time window. Then the discovery of time lag was formulated as an optimization problem which can be solved iteratively to obtain time interval distribution between time-lag episodes. Finally, the hypothesis test was used to distinguish serial and parallel time-lag episodes. The experimental results show that compared with Iterative Closest Event (ICE) algorithm which is the latest method of time-lag mining, the Kullback-Leibler (KL) divergence between true and experimental distributions discovered by AEM is 0.056 on average, which is decreased by 20.68%. AEM algorithm measures the possibility of multiple matches of events through a probability statistical model of time lag and obtains a one-to-many adjacent event matching set, which is more effective than the one-to-one matching set in ICE for simulating the actual situation.

Traditional network operation and maintenance evaluation method has two problems. First, it is too dependent on domain experts' experience in indicator selection and weight assignment, so that it is difficult to obtain accurate and comprehensive assessment results. Second, the network operation and maintenance quality involves data from multiple manufacturers or multiple devices in different formats and types, and a surge of users brings huge amounts of data. To solve the problems mentioned above, an indicator selection method based on correlation was proposed. The method focuses on the steps of indicator selection in the process of evaluation. By comparing the strength of the correlation between the data series of indicators, the original indicators could be classified into different clusters, and then the key indicators in each cluster could be selected to construct a key indicators system. The data processing methods and weight determination methods without human participation were also utilized into the network operation and maintenance quality evaluation model. In the experiments, the indicators selected by the proposed method cover 72.2% of the artificial indicators. The information overlap rate is 31% less than the manual indicators'. The proposed method can effectively reduce human involvement, and has a higher prediction accuracy for the alarm.

Focusing on the issues that large fluctuation of Received Signal Strength Indication (RSSI), complex clustering of fingerprint database and large positioning error in traditional iBeacon fingerprinting, a new Bluetooth localization algorithm based on sort feature matching and distance weighting was proposed. In the off-line stage, the RSSI vector size was used to generate the sorting characteristic code. The generated code combined with the information of the position coordinates constituted the fingerprint information, to form the fingerprint library. While in the online positioning stage, the RSSI was firstly weighted by sliding window. Then, indoor pedestrian positioning was achieved by using the sort eigenvector fingerprint matching positioning algorithm and distance-based optimal Weighted K Nearest Neighbors (WKNN). In the localization simulation experiments, the feature codes were used for automatical clustering to reduce the complexity of clustering with maximum error of 0.952 m of indoor pedestrian localization.

The traditional speech-based Parkinson's Disease (PD) prediction method is to predict the motor Unified Parkinson's Disease Rating Scale (motor-UPDRS) and the total Unified Parkinson's Disease Rating Scale (total-UPDRS) respectively. In order to solve the problem that the traditional method could not use the shared information between tasks and the poor prediction performance in the process of single task prediction, a multi-task regression method based on model filtering was proposed to predict the motor-UPDRS and total-UPDRS of Parkinson's disease patients. Firstly, considering the different effects of the subtask speech features on the predicted motor-UPDRS and total-UPDRS, an L1 regularization term was added for feature selection. Secondly, according to different Parkinson's patient objects distributed in different domains, a filtering mechanism was added to improve the prediction accuracy. In the simulation experiments of remote Parkinson data set, the Mean Absolute Error (MAE) of motor-UPDRS is 67.2% higher than that of the Least Squares (LS) method. Compared with the Classification And Regression Tree (CART) in the single task, the motor value increased by 64% and the total value increased by 78.4%. The results of experiment show that multi-task regression based on model filtering is superior to the single task regression algorithm for UPDRS prediction.

According to the new application scenario with both historical data and stream features, an online feature selection based on group feature selection algorithm and streaming features was proposed. To compensate for the shortcomings of single clustering algorithm, the idea of clustering ensemble was introduced in the group feature selection of historical data. Firstly, a cluster set was obtained by multiple clustering using k-means method, and the final result was obtained by integrating hierarchical clustering algorithm in the integration stage. In the online feature selection phase of the stream feature data, the feature group generated by the group structure was updated by exploring the correlation among the features, and finally the feature subset was obtained by group transformation. The experimental results show that the proposed algorithm can effectively deal with the online feature selection problem in the new scenario, and has good classification performance.

The traditional feature selection methods always remove redundant features by using correlation measures, and it is not considered that there is a large amount of noise in a high-dimensional correlation matrix, which seriously affects the feature selection result. In order to solve the problem, a feature selection method based on Random Matrix Theory (RMT) was proposed. Firstly, the singular values of a correlation matrix which met the random matrix prediction were removed, thereby the denoised correlation matrix and the number of selected features were obtained. Then, the singular value decomposition was performed on the denoised correlation matrix, and the correlation between feature and class was obtained by decomposed matrix. Finally, the feature selection was accomplished according to the correlation between feature and class and the redundancy between features. In addition, a feature selection optimization method was proposed, which furtherly optimize the result by comparing the difference between singular value vector and original singular value vector and setting each feature as a random variable in turn. The classification experimental results show that the proposed method can effectively improve the classification accuracy and reduce the training data scale.

Due to the volume, complex and fast-changing characteristics of big data, traditional machine learning platforms are not applicable. Therefore, designing an efficient and general machine learning platform for big data has become an important research issue. By introducing and analyzing the characteristics of machine learning algorithms and the data and model parallelization for large-scale machine learning, some common parallel computing models were presented. Bulk Synchronous Parallel (BSP), Stale Synchronous Parallel (SSP) computing models and the differences between BSP, SSP, and Asynchronous Parallel model (AP) were introduced. Then the typical machine learning platforms based on these parallel models and the advantages and disadvantages of these platforms were mainly introduced, and what kind of big data each typical machine learning platform was best suited for was pointed out. Finally, the typical machine learning platforms were summarized from the aspects of abstract data structure, parallel computing model and fault tolerance mechanism. Some suggestions and prospects were put forward.

Aiming to solve the change of the extrinsic parameters between the two cameras in free binocular stereo vision system caused by the rotation of the cameras, a method for acquiring the dynamic extrinsic parameters based on calibration of rotation axis was proposed. Multiple rotation and translation matrixes were obtained by the calibration at different positions, then the parameters of rotation axis could be calculated by using least square method. Combined with the intrinsic and extrinsic parameters at initial position and rotation angle, the dynamic extrinsic parameters between the two cameras could be calculated in real time. The chessboard corners were reconstructed with the dynamic extrinsic parameters calculated by the proposed method, the result showed that the average error was 0.241mm and the standard deviation was 0.156mm. Compared with the calibration method based on multiple-plane calibration target, the proposed method is easier to implement and has higher precision, where dynamic extrinsic parameters can be acquired without real-time calibration.

Aiming at the problems existing in point-to-point communication model of Wireless Local Area Network (WLAN) protocol via Near Field Communication (NFC) authentication, such as plaintext transferring, user's anonymous access, data being easily tapped and tampered, a security design of WLAN protocol via NFC was put forward. The security tunnel was built using Diffie-Hellman key exchange algorithm and second generation Secure Hash Algorithm (SHA) to transfer the random information, and the user's anonymity was eliminated using Elliptic Curve Digital Signature Algorithm (ECDSA). A prototype implementation on computer was given from requirement analysis, architecture design and sequence steps of the protocol. The experimental results by using Colored Petri Net (CPN) modeling show that the proposed protocol can execute stably and deal with the unauthorized access and eavesdropping problems of WLAN.

At present, the accuracy of image retrieval is a difficult problem to study, the main reason is the method of feature extraction. In order to improve the precision of image retrieval, a new image retrieval method based on multi-feature called CAUC (Comprehensive Analysis based on the Underlying Characteristics) was presented. First, based on YUV color space, the mean value and the standard deviation were used to extract the global feature from an image that depicted the global characteristics of the image, and the image bitmap was introduced to describe the local characteristics of the image. Secondly, the compactness and Krawtchouk moment were extracted to describe the shape features. Then, the texture features were described by the improved four-pixel co-occurrence matrix. Finally, the similarity between images was computed based on multi-feature fusion, and the images with high similarity were returned.On Corel-1000 image set, the comparative experiments with method which only considered four-pixel co-occurrence matrix showed that the retrieval time of CAUC was greatly reduced without significantly reducing the precision and recall. In addition, compared with the other two kinds of retrieval methods based on multi-feature fusion, CAUC improved the precision and recall with high retrieval speed. The experimental results demonstrate that CAUC method is effective to extract the image feature, and improve retrieval efficiency.

In order to deal with the channel fading in Underwater Wireless Sensor Networks (UWSN) changing randomly in time-space-frequency domain, underwater cooperative communication model with relays was proposed in this paper to improve reliability and obtain diversity gain of the communication system. Based on the new model, a relay selection algorithm for UWSN was proposed. The new relay selection algorithm used new evaluation criteria to select the best relay node by considering two indicators: channel gain and long delay. With the selected relay node, source node and relay nodes could adjust their sending power by the power allocation algorithm which was based on the principle of minimizing the bit error rate. In a typical scenario, by comparing with the traditional relay selecting algorithm and equal power allocation algorithm, the new algorithm reduces the delay by 16.7% and lowers bit error rate by 1.81dB.

In order to improve the speed and accuracy of image retrieval, the drawbacks of image retrieval based on a variety of clustering algorithms were analyzed, then a new partition clustering method for image retrieval was presented in this paper. First, based on the asymmetrical quantization of the color in HSV model, color feature of image was extracted by color coherence vectors. Then, global shape feature of image was extracted based on improved Hu invariant moment. Finally,images were clustered based on contribution according to color and shape features, and image feature index library was established. The methods described above were used for image retrieval based Corel image library. The experimental results show that compared with image retrieval algorithms based on improved K-means algorithms, precision ratio and recall ratio of the proposed algorithm are improved greatly.

Vehicular network has the following characteristics such as nodes moving fast, topology changing rapidly. The direct use of Global Positioning System (GPS) devices causes large positioning error and low routing connectivity rate. Therefore, the packet delivery rate of the existing location-based routing algorithm is not high enough to provide reliable routing. A geographic routing algorithm based on anchor node in vehicle networks named Geographic Routing based on Anchor Nodes (GRAN) was proposed. Using street lamps as anchor nodes, a vehicle could locate itself through the anchor nodes. Combined with the road gateway and the central data, GRAN established a hierarchical routing structure, thus removing the steps of route discovery and the whole network broadcast. Thus, the routing overhead was reduced and the routing efficiency and the packet delivery rate were improved. By using the NS-2 software and selecting a realistic urban scene, a simulation was conducted on Greedy Perimeter Stateless Routing (GPSR), Graphic Source Routing (GSR) and GRAN. The experimental results show that GRAN can provide a lower average delay, higher packet delivery ratio and throughput at a lower load, compared with several typical location-based routing protocols.

In view of the low delivery ratio of conventional probabilistic routing in opportunistic networks, an improved routing algorithm based on History Meeting Predictability Routing (HMPR) was put forward. The algorithm was primarily based on the contact duration and the meeting frequency of history information of nodes, and predicted the utility of packets successfully delivered to the destination. Through comparing the utility value, nodes could determine packets whether to be forwarded from them to next hop nodes. The simulation results show that, compared with traditional epidemic routing and probabilistic routing, the proposed routing scheme has better performance in the delivery ratio of packets, the average delay time and the average buffer time.

A dynamic Retransmission Timeout (RTO) algorithm: DRTO (Dynamic RTO) for solving the TCP packets out-of-order caused by LTE network handover was proposed. The essence of DRTO was to use the TCP packet sequence number to distinguish the old and the new packets. Hence the multiplicative factor calculated in the past traditional RTO could be replaced by the difference of the serial number. The algorithm did not need to modify the handover mechanism, which can solve the packet out-of-order between the first part of packets (the packets transferred from source eNB to target eNB) which was transferred before handover process and the second part of the packets (the packets sent by server) which was transferred after handover process. Finally, the DRTO algorithm was compared with the traditional RTO algorithm on NS-2 simulation platform. The simulation results show the DRTO algorithm is better than the traditional RTO algorithm in terms of throughput, the number of retransmission packets and latency.

By discussing the shortages of access control policy, and analyzing the complementarity and completeness between access control and state control, the necessity of state control was proposed. A formal description about state control policy was defined, and the policy's description rules based on XML were regulated. At the same time, according to different control goal and control object, some application patterns for state control policy were provided. In addition, the complexity of state control policy was discussed, and some solutions were provided.

In the cognitive radio networks, the main functions of MAC protocols include channel sensing, channel selection and access, the detection time and transmission time has the important influence on network performance. In dynamic wireless network environment, how to reasonably distribute detection time and transmission time is a challenge. A two radios multi-channel distributed MAC(TM-MAC) protocol are proposed, which is not necessary to detect channels before node transmit data. Node can detect the spectrum resource while other nodes transmit data. Then, the node can transmit data use the idle spectrum resource. A model is built to analyze the throughput of TM-MAC under the saturation condition. The analysis results show that TM-MAC improves the throughput in cognitive networks.

The unsuitable iterative number of the Regularized Orthogonal Matching Pursuit (ROMP) algorithm in the framework of Compressive Sensing (CS) may reduce the quality of image reconstruction greatly. In this paper, an algorithm was proposed to determine the proper iterative number. In order to guarantee the quality of image reconstruction, the best iterative number of every image block with various sparsity was obtained by using the prior knowledge which resulted from last iteration. The experimental results show that the method can get better reconstruction performance than the ROMP algorithm with deterministic iterative number.

The performance of DiffServ in wireless ad hoc networks was evaluated by simulation. The simulation results show that bandwidths obtained by high and low priority traffics were not consistent with their WRR(Weighted Round Robin) weight ratios. Combined with simulation trace, this phenomenon is caused by MAC mechanism. We conclude that it is impossible to proportionally distribute resource only using DiffServ at network layer in wireless ad hoc networks.

The integration of MPLS and traditional mobile IP has many limitations. In order to improve its performance, many researches on integrating MPLS into micro-mobility were carried out. Based on these researches, especially HMPLS and MMPLS, this paper analysed their problems and then proposed a solution. Through LSP extension within a domain, this scheme overcomes some defects of current schemes of micro-mobility integrated with MPLS,supports fast and smooth handover and has lower signalling overhead.