Search

UAVs are widely used in different fields of social life. With the increasing use of UAV, the main direction of future UAV technology development is “intellectualization”. Facing the problems of abundant information sources, large amount of information, interaction of various equipment systems and stability of communication signals, UAV can not only rely on manual operation, so it is important to strengthen the ability of process data to UAV. Nowadays deep learning is one of the hottest topics in the field of science and technology, and there are more and more researches based on the theory of deep learning, which provides a method for realizing artificial intelligence. This paper based on deep learning technology builds a neural network framework to study UAV path planning. Compared with the traditional UAV path algorithm, the neural network model is smaller and the recognition speed is faster. A detection and recognition method suitable for the network is proposed, which is applied to the design of UAV obstacle avoidance system to realize UAV’s recognition of the surrounding environment, obstacle avoidance and to ensure UAV’s flight safety.

In the Internet of things, trust establishment plays an important role of improving cooperation and enhancing security. To evaluate the trust relationships among sensor nodes accurately, an appropriate trust estimation model should be designed so as to against attack and misbehavior effectively. In this paper, a novel quantitative model of trust value is proposed. A variety of trust factors related to the behaviors of sensor nodes is measured, including the packet forwarding capacity, the repetition rate, the consistency of the packet content, the delay, the integrity, etc. To void the impact of subjective setting, each trust factor is determined by means of the information entropy theory. In addition, the D-S theory is adopted to deduce and synthesize the trust, and the statistic factor of nodes' behavior is introduced to modify the synthesis result. Evaluation results show that our scheme performs better in defeating attacks.

In this paper, we present an analog voice transmission system based on Visible Light Communication (VLC). At the transmitter of the system, by the Voltage-Frequency (V-F) conversion circuit, the analog voice signal is transformed into the frequency signal, which is linear with the amplitude of the signals. At the receiver, the voice signal can be demodulated by the Frequency-Voltage (F-V) conversion circuit. Finally the test of the voice transmission system based on visible light communication is carried out, which shows that the input and output waveforms are consistent, the transmitted voice can be received without distortion. The test result shows that the design is correct and feasible.

A range-free visible light communication (VLC) based positioning method is proposed for indoor wireless sensor networks (WSN). Then the proposed VLC based positioning method and the conventional received signal strength indication (RSSI) based positioning methods are merged into a hybrid positioning method for indoor WSN to get more accuracy position message with lower power consumption. Computer simulations demonstrate that in a Rician fading environment, the accuracy of positioning can be effectively improved with the proposed hybrid positioning method and also longer battery life can be expected since the lower power consumption.