ABSTRACT
The models of time-varying network have a profound impact on the study of virus spreading on the networks. On the basis of an activity-driven memory evolution model, a time-varying spatial memory model (TSM) is proposed. In the TSM model, the cumulative number of connections between nodes is recorded, and the spatiality of nodes is considered at the same time. Therefore, the active nodes tend to connect the nodes with high intimacy and close proximity. Then, the TSM model is applied to epidemic spreading, and the epidemic spreading on different models is compared. To verify the universality of the TSM model, this model is also applied to rumor spreading, and it is proved that it can also play a good inhibiting effect. We find that, in the TSM network, the introduction of spatiality and memory can slow down the propagation speed and narrow the propagation scope of disease or rumor, and memory is more important. We then explore the impact of different prevention and control methods on pandemic spreading to provide reference for COVID-19 management control and find when the activity of node is restricted, the spreading will be controlled. As floating population has been acknowledged as a key parameter that affects the situation of COVID-19 after work resumption, the factor of population mobility is introduced to calculate the interregional population interaction rate, and the time-varying interregional epidemic model is established. Finally, our results of infectious disease parameters based on daily cases are in good agreement with the real data, and the effectiveness of different control measures is evaluated.
ABSTRACT
Nowadays people are mostly focused on their work while ignoring their health which in turn is creating a drastic effect on their health in the long run. Remote health monitoring through telemedicine can help people discover potential health threats in time. In the COVID-19 pandemic, remote health monitoring can help obtain and analyze biomedical signals including human body temperature without direct body contact. This technique is of great significance to achieve safe and efficient health monitoring in the COVID-19 pandemic. Existing remote biomedical signal monitoring methods cannot effectively analyze the time series data. This paper designs a remote biomedical signal monitoring framework combining the Internet of Things (IoT), 5G communication and artificial intelligence techniques. In the constructed framework, IoT devices are used to collect biomedical signals at the perception layer. Subsequently, the biomedical signals are transmitted through the 5G network to the cloud server where the GRU-AE deep learning model is deployed. It is noteworthy that the proposed GRU-AE model can analyze multi-dimensional biomedical signals in time series. Finally, this paper conducts a 24-week monitoring experiment for 2000 subjects of different ages to obtain real data. Compared with the traditional biomedical signal monitoring method based on the AutoEncoder model, the GRU-AE model has better performance. The research has an important role in promoting the development of biomedical signal monitoring techniques, which can be effectively applied to some kinds of remote health monitoring scenario.