Deep Learning-based Joint Optimization and Application Research of Urban Network Communication System and Landscape Design

With the accelerated development of urbanization, urban neighborhood landscape design is faced with many challenges such as the shortage of urban space and the loss of living environment. Therefore, in this study, the application of artificial neural network model in architectural landscape design is carried out, which provides new ideas and methods for realizing more scientific and accurate design assistance. To explore the feasibility of an artificial neural network model in architectural landscape design, this study takes the streetscape data as the basis and extracts three kinds of visual landscape elements of urban outdoor environment, namely, green visibility, sky visibility, and building visibility, and measures six perceptual indexes of urban neighborhood residents, namely, safety, vitality, beauty, affluence, depression, and boredom. Finally, multiple linear regression analysis was used to reveal the relationship between visual landscape elements and residents' spiritual perceptions. The results show that green space, blue space, and gray space in the city shows an aggregated distribution pattern according to the functional characteristics of the city: the overall green visual elements are more scarce; there is obvious spatial differentiation between positive and negative perceptions; and the green visibility rate has a significant positive effect on the residents' comprehensive perceptions. In addition, the optimized schematic diagram shows that the contour density and coverage of communication facilities in the optimized landscape network as well as the rationality of landscape optimization have been greatly improved and enhanced.