To address the issues of insufficient long-range motion modeling and over-parameterization caused by channel redundancy in video steganography tasks under small-sample conditions， an enhanced video steganographic network — GAB3D-SEVSN was proposed by integrating 3D Global Attention Block （GAB-3D） and Squeeze-and-Excitation （SE） channel attention. In the model， through the optimized GAB-3D module， key motion trajectories in the 3D spatio-temporal domain were focused on adaptively， thereby enhancing the capability of modeling long-range dependencies. Meanwhile， by embedding the SE module into the reversible architecture， the channel-level adaptive calibration was achieved， which suppressed redundant parameters and alleviated over-parameterization effectively. Experimental results on the UCF101 dataset （13K video samples） demonstrate that， compared to the LF-VSN baseline model， the proposed model achieves improvements of 0.5 dB in Peak Signal-to-Noise Ratio （PSNR） and 2.06% in Structural SIMilarity （SSIM）. Ablation experimental results verify the effectiveness and synergistic effect of various modules. Test results on high-dynamic scene subsets and videos with different attributes show that the model outperforms baseline models in PSNR and SSIM significantly， demonstrating excellent robustness and generalization ability.

Adopting embedded framework and network communication mode, a kind of multi-axis embedded motion controller hardware platform design for Panasonic A5N drive was proposed, which was based on the modular control core (ARM + FPGA) and could adapt to the new real-time network communication named RTEX. It was transplanted in μC/OS-Ⅱ. The processes of controller's functional design, hardware design and software design were explained in detail. Up to now, the motion controller's hardware platform had been completed and verified by communication experiments and position and velocity control experiments based on SCARA robot. The experimental results show that the controller has good communication function and stable performance, and can well accomplish the functions of servo control.

The basic ray-based method is time consuming and only uses the information of triangle facets. An extended ray-based method was proposed based on the principle of non-intersecting pencil of planes. The key points of this method were as follows: firstly, a group of rays was scattered evenly from the center of the 3D model to intersect with triangle facets, and the non-intersecting pencil of planes determined by the rays was used to get the intersection points; secondly, the retrieval model was established to improve the 3D model retrieval effectiveness, according to the distances from the center to those intersection points and the vertices of the 3D model. Applying this method on ten categories of 3D models in PSB (Princeton Shape Benchmark), the results show that this approach not only reduces the processing time, but improves the retrieval accuracy.

AutoRegression (AR) model and Clarke model (CLARKE R H. A statistical theory of mobile-radio reception. Bell System Technology Journal, 1968, 47(6): 957-1000) are typically used for the simulation of Rayleigh fading channels. But the results show that AR model has unavoidable numerical difficulties, hence it cannot generate the desired correlations of Rayleigh fading channels. Based on the Clarke model, an improved Sum-Of-Sinusoids (SOS) model was proposed, which reduced the distribution of the angel of arrivals and ensured its random distribution. Therefore, the new model could achieve better statistical properties of generated channels. Judging the variance of the time average statistical properties from their ideal ensemble averages, the new model has better performance and its statistical convergence to the desired correlations is faster than Clarke model.

The authentication and key agreement protocol adopted by 3rd Generation Partnership Project (3GPP) System Architecture Evolution (SAE) Release 8 standard was analyzed in contrast with 3G, and several security defects in SAE protocol were pointed out, then an improved 3GPP SAE authentication and key agreement protocol was put forward based on public key cryptosystem. In the new protocol, user’s identity information and authentication vector in network domain were encrypted based on public key cryptosystem, public parent key adopted in local authentication was generated by random data. The security and efficiency of the proposed new scheme was analyzed at last. The analysis results show that the proposal can effectively solve the problems mentioned above and improve the security of protocol with less cost.

A new adaptive blind equalization technology was proposed for Direct Sequence-Ultra WideBand (DS-UWB) and Time-Hopping-Ultra WideBand (TH-UWB) wireless communication systems. This method can effectively track the variation of Ultra WideBand (UWB) channels by using the variable forgetting factor to the adaptive algorithm and thus can compensate the interference due to the properties of channels. When receiving the signals, it can adjust the coefficients of equalizers without using the training sequence and then obtain the estimation of the sent signals. Furthermore, this method has relatively rapid convergence rate and good stability. Simulation results show that the proposed method has better performance in tracking UWB fading channels and it also help obtain lower Bit Error Rate (BER).

The data processed by semantic grid are often semi-structured data. Therefore, how to model and retrieve these semi-structured data is a key issue of semantic Web. The presenting model for semi-structured data base Resource Description Framework （RDF） and its information retrieval mechanism were proposed in this paper. Then a semantic grid infrastructure was described, which implemented knowledge sharing and retrieval in chemical engineering domain based on a chemical computing grid platform.

The analysis system of subscriber churn in wireless carriers was constructed based on data mining. It used Bayesian network to build the subscriber churn model from existing business data. So the market people could make corresponding policy. Experiments prove that the priority knowledge can be corrected continually by inference information of Bayesian network, and help to solve the problem of subscriber churn.