Concerning the high computational complexity and communication overhead of wireless network node localization algorithm based on message passing algorithm, a ranging-based hybrid message passing node localization method with low complexity and cooperative overhead was proposed. The uncertainty of the reference nodes was taken into account to avoid error accumulation, and the messages on factor graph were restricted to be Gaussian distribution to reduce the communication overhead. Firstly, the factor graph was designed based on the system model and the Bayesian factorization. Secondly, belief propagation and mean filed methods were employed according to the linear state transition model and the nonlinear ranging model to calculate the prediction messages and the cooperation messages, respectively. Finally, in each iteration, the non-Gaussian beliefs were approximated into Gaussian distribution by Taylor expansions of the nonlinear terms. The simulation results show that the positioning accuracy of the proposed algorithm is compareable to that of Sum-Product Algorithm over a Wireless Network (SPAWN), but the information transmitted between nodes decreases from a large number of particles to mean vector and covariance matrix, and the comupational complexity is also dramatically reduced.

The goal of algorithm selection is to automatically select the best suitable algorithm for current problem from a batch of available algorithms. For this purpose, an intelligent recommendation framework based on meta-learning approach was presented. The automatic selection procedure for Genetic Algorithm (GA), Particle Swarm Optimization (PSO) and Simulated Annealing (SA) was designed according to this framework by using Multi-mode Resource-Constrained Project Scheduling Problem (MRCPSP) as the validation data set. Three hundred and seventy-eight instances of MRCPSP were randomly picked out from the Project Scheduling Problem Library (PSPLib), and the inherent and statistic features of each instance were extracted and used as the metadata, then the prediction meta-model for new examples was obtained by using Feed-forward Neural Network (FNN) algorithm. The empirical results demonstrate that the hit rate reaches 95% at most, and the average hit rate is 85% when choosing one algorithm from two ones; the best hit rate reaches 92% and 80% respectively when choosing one algorithm from three ones. The proposed intelligent recommendation framework is successful and the automatic selection for optimization algorithms is feasible.

In order to improve the effectiveness of vein feature extraction, a dorsal hand vein recognition method based on sparse coding was proposed. Firstly, during image acquisition process, acquisition system parameters were adaptively adjusted in real-time according to image quality assessment results, and the vein image with high quality could be acquired. Then concerning that the effectiveness of subjective vein feature mainly depends on experience, a feature learning mechanism based on sparse coding was proposed, thus high-quality objective vein features could be extracted. Experiments show that vein features obtained by the proposed method have good inter-class separableness and intra-class compactness, and the system using this algorithm has a high recognition rate.

Concerning the application requirements of slope stability analysis in large areas and the defects of current calculation method, a three-dimensional analysis method based on Digital Elevation Model (DEM) data was proposed. In this method, a sphere was used, instead of ellipsoid, to search slippery surface, and the slope safety factor under the three-dimensional conditions was calculated through integral operator with the results of two-dimensional analysis. Finally, the position and shape of possible landslide were determined according to the safety factor. The practical application results confirm that this method simplifies the search algorithm, ensures the accuracy of slope stability analysis, and improves the efficiency of analysis and calculation.

To increase the classification accuracy of the database classification system, this paper proposed a new classification method. Firstly, the continuous attributes were dispersed by the Fuzzy C-Mean (FCM) algorithm. Secondly, an improved fuzzy association method was proposed to mine the classification association rules. Eventually, the compatibility between the generated rules and patterns was used to construct a set of feature vectors, which were used to generate a classifier. The experimental results demonstrate that the method has high discrimination and efficiency.

QoS requirement is denoted by the range to support the flexible and heterogeneous QoS.According to the microeconomics theory and method,a Kelly/PSP model-based pricing strategy was presented to support the inter-group fairness.The ELSD(Equal Link Split Downstream) method was adopted to apportion the cost among group members,thus the intra-group fairness was provided.Based on the parallelized FCNN(Firing Coupled Neural Network),a parallel and intelligent QoS multicast routing algorithm was introduced,exploiting the inherent parallelism in FCNN fully and improving the scalability to the network size and the problem complexity significantly.Combining the above,a parallel,fair and intelligent QoS multicast routing mechanism was established.Simulation results have shown that the proposed mechanism is both effective and efficient,and the runtime efficiency of the proposed parallelized algorithm is higher than its corresponding serialized one.