To address the problem that traditional PhotoVoltaic （PV） power prediction models are affected by random power fluctuation and tend to ignore important information， resulting in low prediction accuracy， ADDPG and ARDPG models were proposed by combining the attention mechanism with Deep Deterministic Policy Gradient （DDPG） and Recurrent Deterministic Policy Gradient （RDPG）， respectively， and a PV power prediction framework was proposed on this basis. Firstly， the original PV power data and meteorological data were normalized， and the PV power prediction problem was modeled as a Markov Decision Process （MDP）， where the historical power data and current meteorological data were used as the states of MDP. Then the attention mechanism was added to the Actor networks of DDPG and RDPG， giving different weights to different components of the state to highlight important and critical information， and learning critical information in the data through the interaction of Deep Reinforcement Learning （DRL） agents and historical data. Finally， the MDP problem was solved to obtain the optimal strategy and make accurate prediction. Experimental results on DKASC and Alice Springs PV system data show that ADDPG and ARDPG achieve the best results in Root Mean Square Error （RMSE）， Mean Absolute Error （MAE） and R². It can be seen that the proposed models can effectively improve the prediction accuracy of PV power， and can also be extended to other prediction fields such as grid prediction and wind power generation prediction.

For Wireless Sensor Network (WSN), many applications rely on the coordination of synchronized clock nodes. However, the crystal oscillator of the node is affected by itself and the external environment, so that the clock skew and clock offset change and then lead to the clocks falling out of synchronization. Consequently, a new clock synchronization algorithm based on distributed Kalman filter and consistency compensation, namely DKFCC, was proposed. First of all, the optimal estimation of the clock skew and offset was obtained by two-way message exchange mechanism and distributed Kalman filter. Then consistency compensation method based on the optimal estimation value of clock parameters was adopted to achieve clock synchronization. The experimental results show that compared with the Asynchronous Consensus-based time synchronization (AC) algorithm in the WSN with 100 randomly-deployed nodes, the Synchronous Root Mean Square Error (SRAMSE) of the DKFCC synchronization algorithm with virtual global consistency is reduced by about 95%, which means DKFCC synchronization algorithm has higher synchronization accuracy. At the same time, the proposed algorithm achieves synchronization from the clock parameter level without operating clock synchronization frequently, thus it has better energy efficiency compared to AC synchronization algorithm.

Focusing on the information security and implementation of RFID systems, this paper analyzed the chaotic dynamics of the linear feedback shift register and the piecewise Logistic mapping, then presented a RFID encryption algorithm combined with chaotic dynamical disturbance. The information was encrypted by the key generated by chaotic sequence generator before being transmitted, then the ciphertext feedback and dynamical disturbance would generate the next key. The analysis and simulation shows that the algorithm is safe, easy to be implemented and so on.

The existing speckle reduction algorithms of Synthetic Aperture Radar (SAR) image can efficiently reduce the speckle effects but unfortunately smear edges and details. A new method, based on morphological Haar wavelet, was proposed. In this method, the SAR image was firstly decomposed by 2-D morphological Haar wavelet. Thus, the edges, details and textures were well preserved in low-frequency sub-band. The speckle noise was mainly distributed in high-frequency sub-bands. Then, the average filtering and median filtering were run on the corresponding high frequency sub-bands according to the noise features. Finally, 2-D morphological Haar wavelet inverse transform was carried on to low-frequency sub-band coefficients and filtered high-frequency sub-bands coefficients to reconstruct SAR image accurately. The experimental results show that the proposed method can not only filter the speckle noise efficiently, but well preserve the image textures and details of SAR image. The proposed method is better than the traditional Lee filtering, Frost filtering, Kuan filtering and wavelet soft-threshold overall.