Since the pilot overhead using traditional channel estimation methods in the Reconfigurable Intelligent Surface （RIS）-assisted wireless communication systems is excessively high， a block sparseness based Orthogonal Matching Pursuit （OMP） channel estimation scheme was proposed. Firstly， according to the millimeter Wave （mmWave） channel model， the cascaded channel matrix was derived and transformed into the Virtual Angular Domain （VAD） to obtain the sparse representation of the cascaded channels. Secondly， by utilizing the sparse characteristics of the cascaded channels， the channel estimation problem was transformed into the sparse matrix recovery problem， and the reconstruction algorithm based on compressive sensing was adopted to recover the sparse matrix. Finally， the special row-block sparse structure was analyzed， and the traditional OMP scheme was optimized to further reduce pilot overhead and improve estimation performance. Simulation results show that the Normalized Mean Squared Error （NMSE） of the proposed optimized OMP scheme based on the row-block sparse structure decreases about 1 dB compared with that of the conventional OMP scheme. Therefore， the proposed channel estimation scheme can effectively reduce pilot overhead and obtain better estimation performance.

The accurate flight delay prediction results can provide a great reference value for the prevention of large-scale flight delays. The flight delays prediction is a time-series prediction in a specific space， however most of the existing prediction methods are the combination of two or more algorithms， and there is a problem of fusion between algorithms. In order to solve the problem above， a Convolutional Long Short-Term Memory （Conv-LSTM） network flight delay prediction model was proposed that considers the temporal and spatial sequences comprehensively. In this model， on the basis that the temporal features were extracted by Long Short-Term Memory （LSTM） network， the input of the network and the weight matrix were convolved to extract spatial features， thereby making full use of the temporal and spatial information contained in the dataset. Experimental results show that the accuracy of the Conv-LSTM model is improved by 0.65 percentage points compared with LSTM， and it is 2.36 percentage points higher than that of the Convolutional Neural Network （CNN） model that only considers spatial information. It can be seen that with considering the temporal and spatial characteristics at the same time， more accurate prediction results can be obtained in the flight delay problem. In addition， based on the proposed model， a flight delay analysis system based on Browser/Server （B/S） architecture was designed and implemented， which can be applied to the air traffic administration flow control center.

K-anonymous algorithm makes the data reached the condition of K-anonymity by generalizing and suppressing the data. It can be seen as a special feature selection method named K-anonymous feature selection which considers both data privacy and classification performance. In K-anonymous feature selection method， the characteristics of K-anonymity and feature selection are combined to use multiple evaluation criteria to select the subset of K-anonymous features. It is difficult for the filtered K-anonymous feature selection method to search all the candidate feature subsets satisfying the K-anonymous condition， and the classification performance of the obtained feature subset cannot be guaranteed to be optimal， and the wrapper feature selection method has very high-cost calculation. Therefore， a hybrid K-anonymous feature selection method was designed by combining the characteristics of filtered feature sorting and wrapper feature selection by improving the forward search strategy in the existing methods and thereby using classification performance as the evaluation criterion to select the K-anonymous feature subset with the best classification performance. Experiments were carried out on multiple public datasets， and the results show that the proposed algorithm can outperform the existing algorithms in classification performance and has less information loss.

To seek a good balance between efficiency and fairness in emergency relief supply allocation, a high-dimensional multi-objective adaptive allocation algorithm based on two-dimensional integer encoding was developed. First of all, a high-dimensional multi-objective optimization model was constructed with the consideration of total emergency response time, panic degree of the victims, unsatisfactory degree of relief supplies, fairness of supply allocation, loss of the victims, and total cost of emergency response. Then, two-dimensional integer encoding and Adaptive Individual Repair (AIR) were adopted to resolve potential emergency resource conflicts. Finally, the shift-based density estimation and Strength Pareto Evolutionary Algorithm 2 (SPEA2) were introduced to design a high-dimensional multi-objective allocation algorithm for disaster relief supplies. Simulation results show that compared with Encoding Repair and Non-dominated Sorting based Differential Evolution algorithm (ERNS-DE) and Greedy-Search-based Multi-Objective Genetic Algorithm (GSMOGA), the proposed algorithm had coverage values increased by 34.87%, 100% and 23.59%, 100% in two emergency environments, respectively. Moreover, the hypervolume values of the proposed algorithm were much higher than those of the two comparison algorithms. Experimental results verify that the proposed model and algorithm allow decision makers to select emergency schemes according to actual emergency needs, and have better flexibility and efficiency.

In order to solve the problem of communication interference in Transform Domain Communication System (TDCS) under complex electromagnetic environment, a TDCS model based on store and forward mechanism was proposed. Based on the analysis of the electromagnetic spectrum environment, the spectrum sensing cases of transmitter and receiver were classified according to their respective perceptions. With the analysis of consistent spectrum conditions, the TDCS model for long distance aviation communication system was established based on store and forward modules in inconsistent spectrum conditions, and the specific communication schemes were designed, which can effectively improve the system performance. The simulation results show that the interference model and spectrum setting are reasonable, the performance of TDCS is close to the ideal bit error rate under the condition of the same spectrum. Under the condition of different frequency spectrum with more concentrated comb spectrum interference, the bit error rate is reduced by about 24.48%; moreover, with the increase of Signal-to-Noise Ratio (SNR), the performance improvement is more obvious, and the performance improvement is about 1dB under the same bit error rate.

In order to solve the link prediction problem of knowledge graph, a shared variable network model named LCPE (Local Combination Projection Embedding) was proposed, which realized the prediction of links by embedding entities and relationships into vector space. By analyzing the Unstructured Model, it was derived that the distance between related entities' embedding was shorter in the vector space, in other words, similar entities were more likely to be related. In LCPE model, ProjE model was used based on similarity between two entities to judge whether the two entities were related and the relation type between them. The experiment shows that with the same number of parameters, the LCPE improves Mean Rank by 11 and lifts Hit@10 0.2 percentage points in dataset WN18 while improves Mean Rank 7.5 and lifts Hit@10 3.05 percentage points in dataset FB15k, which proves that the similarity between entities, as auxiliary information, can improve predictive capability of the ProjE model.

For the unique environment of the Underwater Wireless Sensor Network (UWSN) in river, the model was built by method of fluid dynamics to obtain the real-time position of sensor nodes and simulate the movement law of sensor nodes in real river environment. Furthermore, on the problem of data transmission in UWSN, a Routing Algorithm based on Layered Mechanism (RALM) was proposed for river environment. The topology information was calculated and updated by each node periodically based on the receiving speed from sink. The node to transmit data would choose the neighbor node in upper layer which has the most residual energy to be the next hop. If the node has no neighbor node in upper layer, the next hop would be the neighbor node in the same layer which has the most residual energy. The simulation results show that, compared with DBR (Depth-Based Routing) and Layered-DBR (Layered-Depth Based Routing), RALM algorithm can effectively reduce the network redundancy and packet loss rate, and the network life cycle is raised by 71% and 45%.

Focusing on the fact that the order of locations and time play a pivotal role in understanding user mobility patterns for spatio-temporal trajectories, a novel deep representation model for trajectories was proposed. The model considered the characteristics of spatio-temporal trajectories: 1) different orders of locations indicate different user mobility patterns; 2) trajectories tend to be cyclical and change over time. First, two time-ordered locations were combined in location sequence; second, the sequence and its corresponding time bin were combined in the temporal location sequence, which was the basic unit of describing the features of a trajectory; finally, the deep representation model was utilized to train the feature vector for each sequence. To verify the effectiveness of the deep representation model, experiments were designed to apply the temporal location sequence vectors to user mobility patterns mining, and empirical studies were performed on a real check-in dataset of Gowalla. The experimental results confirm that the proposed method is able to discover explicit movement patterns (e.g., working, shopping) and Word2Vec is difficult to discover the valuable patterns.

Due to the disadvantages such as inefficiency of getting high-quality samples, bad impact of bootstrap to the whole learning-efficiency and final classifier performance in the negative samples bootstrap process of matrix-structural learning of cascade classifier algorithm. This paper proposed a fast learning algorithm-matrix-structural fast learning of cascaded classifier for negative sample inheritance. The negative sample bootstrap process of this algorithm combined sample inheritance and gradation bootstrap, which inherited helpful samples from the negative sample set used by last training stage firstly, and then got insufficient part of sample set from the negative image set. Sample inheritance reduced the bootstrap range of useful samples, which accelerated bootstrap. And sample pre-screening, during bootstrap process, increased sample complexity and promoted final classifier performance. The experiment results show that the proposed algorithm saves 20h in training time and improves 1 percentage point in detection performance, compared with matrix-structural learning of cascaded classifier algorithm. Besides, compared with other 17 human detection algorithms, the proposed algorithm achieves good performance too. The proposed algorithm gets great improvement in training efficiency and detection performance compared with matrix-structural learning of cascaded classifier algorithm.

To overcome the limitation of the assumptions of topic model-word independence and topic independence, a topic model which inosculated the spatial relationship of visual words was proposed, namely Markov Topic Random Field (MTRF). In addition, it was discussed that the "topic" of topic model represented the part of object in image processing. There is a high probability of the neighbor visual words generated from the same topic, and whether the visual words were generated from the same topic determined the topic was generated from Markov Random Field (MRF) or multinomial distribution of topic model. Meanwhile, both theoretical analysis and experimental results prove that "topic" of topic model appeared as mid-level feature to represent the parts of objects rather than the instances of objects. In experiments of image classification, the average accuracy of MTRF was 3.91% higher than that of Latent Dirichlet Allocation (LDA) on Caltech101 dataset, and the mean Average Precision (mAP) of MTRF was 2.03% higher than that of LDA on VOC2007 dataset. Furthermore, MTRF assigned topics to visual words more accurately and got the mid-level features which represented the parts of objects more effectively than LDA. The experimental results show that MTRF makes use of the spatial information effectively and improves the accuracy of the model.

For the vital arc problem of maximum dynamic flow in time-capacitated network, the classic Ford-Fulkerson maximum dynamic flow algorithm was analyzed and simplified. Thus an improved algorithm based on minimum cost augmenting path to find the vital arc of the maximum dynamic flow was proposed. The shared minimum augmenting paths were retained when computing maximum dynamic flow in new network and the unnecessary computation was removed in the algorithm. Finally, the improved algorithm was compared with the original algorithm and natural algorithm. The numerical analysis shows that the improved algorithm is more efficient than the natural algorithm

Two-stage cooperative spectrum sensing that used both soft combining and hard combining need to send much unnecessary information to fusion center in the second stage. Concerning this problem, this paper proposed an improved algorithm by estimating Signal-to-Noise Ratio (SNR). The improved algorithm selected traditional two-stage scheme to improve the performance of spectrum when SNR was low and selected hard combining scheme to decrease the number of sending data when SNR is high. The simulation results show that the proposed algorithm decreases much unnecessary soft information being sent to fusion center at the cost of a little loss in performance of detection as compared to an existing two-stage cooperative spectrum sensing algorithm.

In order to deal with the poor or invalid tracking performance caused by the color sensitivity of Continuously adaptive Mean Shift (CamShift) algorithm, a new CamShift tracking algorithm based on local feature matching was proposed. The new algorithm used the method of Speeded-Up Robust Feature (SURF) to extract the local feature points containing the image information from the target and searched areas of multi-channel images, and then matched the feature points by the method of approximate nearest neighbor searching. The location, scale and orientation information of the feature points were obtained utilizing the purified matching results, therefore the CamShift method was constrained and updated to improve the accuracy and stability of tracking. The experimental results show that the new algorithm can outperform the classic CamShift algorithm and the similar improved algorithms for rotating and zooming objects against complex backgrounds in real-time tracking.

Evolutionary MultiObjective Optimization (EMO) has become a very popular topic in the last few years. However, designing an efficient EMO algorithm for finding wellconverged and welldistributed approximate optimal set is a challenging task. In this paper, an improved IBEA algorithm was proposed to solve Multiobjective Optimization Problems (MOPs) efficiently. The proposed approach introduced a diversity promotion mechanism based on grid in objective space to ensure the approximate optimal set has good distribution. To make the algorithm converge faster, the new approach employed oppositionbased learning mechanism to evaluate the current solutions and their opposite solutions simultaneously in order to find a group of better solutions. The experiments on six benchmark problems show that the new approach is able to obtain a set of welldistributed solutions approximating the true Paretofront.

The traditional multi-objective algorithm named NSGA-Ⅱ and the multi-objective algorithm based model named RM-MEDA were analyzed. Meanwhile, the deficiencies of these two algorithms were pointed out. On the basis of that, a hybrid multi-objective algorithm based on probabilistic model was proposed and the corresponding model metric for mixing the two algorithms was designed. The proposed algorithm could take advantage of the mentioned two algorithms. The algorithm was contrasted with NSGA-Ⅱ and RM-MEDA on 10 test functions. The experimental results show that the proposed algorithm has a good performance on global convergence and diversity.

A new method for matching feature points of images from two cameras of a stereo vision system in the game engine was presented. The main issue is how to meet the request of real-time processing and accuracy of the system. Our method is to calibrate the epipolar constrains and calculate the offset of two image along with the constraints of homography. Then we used a fast method to match the feature points in a local area. This method is proved to be real-time and accurate enough to fulfill the request of our game engine.

The reliability of the large-scale cluster storage system is one important research aspect in the related domain. A high-availability data objects placement algorithm was proposed. It groups objected into redundancy sets using RAID at the algorithm level. The redundancy allowed us to reconstruct any corrupted data objects and storage nodes when it failed and assured efficiently the high availability of storage system. The availability of storage system was quantitatively analyzed by using Markov reward model, and the computing results indicate the algorithm is efficient.

A printing module based on COM was designed and realized. It could print 2D bar code of high resolution and microform effectively in appointed geometric area and form. The principle of COM technology was introduced, and the elementary method of using ATL to realize the COM module in VC was described.