In the fields of industrial production or safety monitoring, wireless sensor networks are often content with unreliable and time-varying channels that are susceptible to interference. Consequently, ensuring both transmission reliability and data accuracy has garnered substantial attention in recent years. Although multipath routing-based schemes can provide transmission reliability for wireless sensor networks, achieving high data accuracy simultaneously remains challenging. To address this issue, an Energy-efficient Multipath Routing algorithm balancing data Accuracy and transmission Reliability (EMRAR) is proposed to balance the reliability and accuracy of data transmission. The multipath routing problem is formulated into a multi-objective programming problem aimed at optimizing both reliability and power consumption while adhering to data accuracy constraints. To obtain the solution of the multi-objective programming, an adaptive artificial immune algorithm is employed, in which the antibody initialization method, antibody incentive calculation method, and immune operation are improved, especially for the multipath routing scheme. Simulation results show that the EMRAR algorithm effectively balances data accuracy and transmission reliability while also saving energy when compared to existing algorithms.

Aiming at the problems of early weak fault feature extraction of bearings in rotating machinery, an improved stochastic resonance (SR) is proposed combined with the advantage of SR to enhance weak characteristic signals with noise energy. Firstly, according to the characteristics of the large parameters of the actual fault signal, the amplitude transform coefficient and frequency transform coefficient are introduced to convert the large parameter signal into small parameter signal which can be processed by SR, and the relationship of second-order parameters are introduced. Secondly, a comprehensive evaluation index (CEI) consisted of power spectrum kurtosis, correlation coefficient, structural similarity, root mean square error, and approximate entropy, is constructed through BP neural network. Moreover, this CEI is adopted as fitness function to search the optimal damping coefficient and amplitude transform coefficient with adaptive weight particle swarm optimization (PSO). Finally, according to the improved optimal SR system, the weak fault feature can be extracted. The simulation and experiment verify the effectiveness of the proposed method compared with traditional second-order general scale transform adaptive SR.

Connecting the educational resources of specific disciplines to form a knowledge network is a key step in the intellectualization of education and research. Based on a knowledge graph of power grid technology education resources of 42510 literature and information in China National Knowledge Infrastructure (CNKI) in our former research, a recommendation algorithm is proposed in this paper. The algorithm aims at the problem of data sparsity and cold start of traditional recommendation algorithm, which adopts hybrid similarity fusion to quickly match core resources and improve the pertinence and efficiency of education and research. The work in this paper show that the integration of knowledge graph technology and academic literature database can provide scientific and intelligent data analysis methods for discipline education and research.

As the Vehicle to Vehicle or Vehicle to Infrastructure communication in Vehicular Ad Hoc Network (VANET) is usually based on opportunity network, the transmission delay will be high with the low vehicle density or far transmission distance. As the result, the Quality of Service requirements of safety message dissemination cannot be satisfied. The paper proposed a safety message broadcast strategy based on VANET-cellular architecture, and it does not rely too much on traffic density or additional deployment of Road Side Unit. Meanwhile, the only relay node is selected from the candidate relay node subset via the acknowledgement mechanism of short control packet. The strategy can solve the problem of competitive channel and redundant data in VANET.

Underwater Acoustic Sensor Networks (USANs) which adopt acoustic communication have been widely used in underwater, such as oceans and rivers, monitoring applications. However, the USANs designed for urban lake monitoring are facing challenges and seldom regarded. Based on the analysis of the features of the urban lakes and the monitoring requirements, a USAN system is designed in this paper. The node deployment strategy is discussed to balance the node density and transmission bandwidth. Then a hybrid topology is proposed to meet the monitoring requirements, in which the nodes on the water surface construct a two-dimensional ring topology, and the nodes underwater construct a three-dimensional grid structure. Base on this topology, a MAC protocol i-HALO is designed which improves the existing handshaking mechanism with a dynamic adaptive back off algorithm to avoid unnecessary backoff. As the result, high delay problem is addressed and the “one handshake, multiple transmission” is realized. The channel access delay is analyzed theoretically and the formula of waiting time for a frame before sending is given. The performance of throughput and transmission delay under different node densities and frame lengths are simulated. The results indicate that a node having about 13 neighbors will have the best performance on throughput, and when the frame length is longer than 100 bytes, the throughput and delay will have fluctuation. In comparison with existing protocols, i-HALO has better performance on throughput, delay, frame reception rate and fairness.

The Agricultural Internet of Things (Ag-IoT) is one of the most important technical supports for precision agriculture and smart farming, which enables automatic and continuous agriculture monitoring systems. The energy-constrained batteries used in the Ag-IoT can be charged by multi-unmanned aerial vehicle (UAV)-assisted wireless rechargeable sensor networks (WRSNs). However, it faces the problem of optimizing the hovering location, the charging task assignment, and the charging path for UAVs. To solve these problems, this article presents a reinforcement-learning-based energy charging strategy for the Ag-IoT. First, it designs a charging cell generation algorithm to optimize the hovering location for UAVs. Second, it puts forward a dynamic charging task assignment algorithm to assign charging tasks for different UAVs dynamically. Third, it proposes a reinforcement-learning-based charging path planning algorithm to adapt to dynamic changes in the state of UAV-assisted WRSN and get the optimal charging paths for UAVs. Extensive simulations show that the presented approach has significant advantages in terms of joint charging efficiency, response ratio, charging coverage ratio, and average charging delay.