In order to solve the problem that the traditional deep learning based remote sensing image classification algorithms cannot effectively fuse multiple deep learning features and their classifiers have poor performance, an improved high-resolution remote sensing image classification algorithm based on deep learning was proposed. Firstly, a seven-layer convolutional neural network was designed and constructed. Secondly, the high-resolution remote sensing images were input into the network to train it, and the last two fully connected layer outputs were taken as two different high-level features for the remote sensing images. Thirdly, Principal Component Analysis (PCA) was applied to the output of the fifth pooling layer in the network, and the obtained dimensionality reduction result was taken as the third high-level features for the remote sensing images. Fourthly, the above three kinds of features were concatenated to get an effective deep learning based remote sensing image feature. Finally, a logical regression based classifier was designed for remote sensing image classification. Compared with the traditional deep learning algorithms, the accuracy of the proposed algorithm was increased. The experimental results show that the proposed algorithm performs excellent in terms of classification accuracy, misclassification rate and Kappa coefficient, and achieves good classification results.

Concerning the security problem of local and global attacks on the Integrity Report Protocol (IRP), the StatVerif syntax was extended by adding constructors and destructors associated with the integrity measurement. The security of the Platform Configuration Attestation (PCA) was analyzed and the local and global attacks were found, including tampering the platform configuration register or revising stored measurement log by running unauthorized commands. The abilities of attackers were modeled, and how attackers accumulated knowledge and tampered PCA protocol by using constructors and destructors was introduced. Finally, the existence of attacking sequence was proved theoretically when PCA does not satisfy the correspondence property; and several propositions that PCA can meet the local reliability and gloabal reliability were given, which were proved by the formal verification tool Proverif.

Instant messaging is fundamental to various mobile Internet applications; however, it is still an open problem to implement secure instant messaging in untrusted Internet environment. An approach for secure instant messaging of mobile intelligent terminal was presented, and a protocol for Trusted Session Key Agreement (TSKA) was designed and implemented. Theoretical analysis shows that the proposed TSKA can ensure the authenticity, freshness and confidentiality of the negotiated session key, even in the condition that both of the instant messaging server and the communication channel are not trusted. After TSKA, instant audio/video messages can be sent to the other side in a confidential and complete way. Experimental results in real Internet environment show that the proposed approach is efficient and secure, the session key can be negotiated within 1-2 seconds, and attackers cannot obtain any plaintext of instant messages.

Concerning the problem of spatial-temporal uncertainty caused by low bandwidth and high latency as well as insufficient network state observations in UnderWater Acoustic Sensor Network (UWASN), a medium access control protocol based on Partially Observable Markov Decision Process (POMDP) for UWASN was presented in this paper. Firstly, the link quality and residual energy of each node were divided into multiple discrete levels for expressing nodes' state information. After that, channel access probability was predicted by receivers through the history information of channel state observation and channel access actions, and then the optimal channel scheduling strategy for senders was acquired. Senders communicate with receivers and transmit data packets in their time slots according to the assigned sequence from the optimal channel scheduling strategy. When the communication was completed, the states of next time slot of the related nodes were predicted based on the statistics of the network transfer probabilities. Simulation results show that the proposed protocol can improve the data packet transmission rate as well as the network throughput, and decrease the energy consumption.

In order to decrease the influence caused by low bandwidth and high latency on Media Access Control (MAC) layer in Underwater Acoustic Sensor Network (UWASN), an Evolutionary Game Theory based MAC (EGT-MAC) protocol was proposed. In EGT-MAC, each sensor node adopted two strategies including spatial multiplexing and temporal multiplexing. With the replication kinetics equation, each strategy got an evolutionary stable strategy and reached stable equilibrium of evolution. In this way, it improved channel utilization rate and data transmission efficiency to achieve performance optimization for MAC protocol. The simulation results show that EGT-MAC can improve the network throughput as well as the transmission rate of data packet.

The security problem of the data stored in the cloud is a challenge for cloud computing. Encryption is the main method to solve the problem of data storage security in cloud computing. The confidentiality issue of the encryption is the key management. The attribute-based encryption and re-encryption scheme in cloud computing was proposed. It combined attribute-based encryption and re-encryption technology. The cloud server could re-encrypt for different users, and re-encrypt for a group of users at a time. Thereby the scheme reduced the number of re-encryption keys. The data owner could generate and send the keys of re-encryption for a group of users and the data requester could use the key corresponding to the attribute set to decrypt the data of several data owners, which reduced the amount of keys’ transmission. The scheme reduced the difficulty of key management and improved the efficiency of the scheme. In the end, the security and efficiency of the scheme were discussed.

Considering the multi-sourcing and heterogeneous data streams in three dimensional underwater acoustic sensor networks (3D UWASNs), this paper gives formalized definition and modeling for describing the characters and attributes of burst and presents an evolutionary game theory based burst detection algorithm for reducing processing time and improving detection performance through optimized choosing the size of slide window. We demonstrate through simulations that our burst detection algorithm consumes less processing time than traditional algorithms in the same condition of data distribution, burst probability or maximum slide window size.

Owing to the insufficiency of the training sets, the performance of the initial classifier is not satisfactory and can not track the users' needs dynamically. Concerning the defect, an incremental learning method of Bayesian classifier combined with feedback information was proposed. To reduce the redundancy between features effectively and improve representative ability of feedback feature subset, an improved feature selection method based on Genetic Algorithm (GA) was used to choose the best features from feedback sets to amend classifier. The experimental results show that the algorithm optimizes classification significantly and has good overall stability.