Smart cities make full use of information technology so as to make intelligence responses to all requirements, including network and city services. This paper proposes a differential evolution back propagation (DE-BP) neural network traffic prediction model applicable for a smart cities network to predict the network traffic. The proposed approach takes the impact factor of network traffic as the input layer and the network traffic as the output layer and trains the DE-BP network with the past traffic data so as to obtain the mapping relationship between the impact factor and the network traffic and get the predicted value of the network traffic. The experimental results show that the proposed approach can accurately predict the trend of network traffic. Within the allowable error range, the predicted traffic volume is consistent with the actual traffic volume trend, and the predicted error is small. INDEX TERMS Smart city, network traffic prediction, urban computing, BP neural network, global optimization.
Retinal vascular segmentation is very important for diagnosing fundus diseases. However, the existing methods of retinal vascular segmentation have some problems, such as low microvascular segmentation and wrong segmentation of pathological information. To solve these problems, we developed a fundus retinal vessels segmentation based on the improved deep learning U-Net model. First, enhance the retinal images. Second, add the residual module in the process of designing the network structure, which solved the problem of the traditional deep learning U-Net model is not deep enough. By using the improved deep learning U-Net model, it can connect the output of the convolutional layer with the output of the deconvolution layer to avoid low-level information sharing, and solved the problem of performance degradation of deep convolutional neural networks in residual networks under extreme depth conditions. By verifying on the DRIVE (digital retinal images for vessel extraction) dataset, the segmentation accuracy, sensitivity, and specificity of the proposed method are 96.50%, 93.1%, and 98.63% respectively.
INDEX TERMSRetina, blood vessels, improved deep learning U-Net model, residual.
In traditional static wireless sensor networks (WSNs), the unbalanced communication overhead in different regions will result in premature death of some monitoring nodes. The introduction of mobile sink in WSNs can not only balance the node traffic load, but also obtain even energy consumption of nodes, thus effectively avoiding the ''hot spot'' problem and prolonging the network lifetime. However, the mobility of the sink will lead to frequent changes in the aspect of network topology, which can aggravate the overhead of the node's reorganization in hierarchical WSNs. Therefore, it is essential to obtain the optimal trajectory design of the mobile sink so as to improve the ability of data gathering. In this paper, a mobile sink-based path optimization strategy in WSNs using artificial bee colony algorithm is proposed. First, the problem of overall energy consumption in the network can be transformed into the minimization of the total hops between all subnodes and the rendezvous points of the mobile sink. The objective function and the constraint criterion should be established. Second, an improved artificial bee colony algorithm is proposed to solve the problem. On the one hand, the cumulative factor is introduced to the position update of the employed bee stage to speed up the convergence of the algorithm. On the other hand, the Cauchy mutation operator is presented to increase the diversity of the feasible solution and enhance the global search ability of the algorithm. The simulation results show that the proposed algorithm is better than the traditional methods in the aspects of energy efficiency and the real-time performance of data collection.INDEX TERMS Wireless sensor networks, mobile sink, path optimization, artificial bee colony algorithm.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.