Base station location determination model based on 5G network coverage

Based on the rapid development of 5G networks, the wider the bandwidth, the more limited the coverage. The problem of site selection is becoming more and more prominent. According to the coverage of the existing 5G network, this problem gives the weak coverage area of the existing network. It requires selecting a certain number of locations so that after the new base stations are built at these points, the weak coverage of the existing network can be solved. Network coverage issues in the coverage area. This paper establishes a related multi-objective programming model, and a heuristic algorithm is used to solve it. We first cleaned the data, judged the Euclidean distance between the existing base station and the weak coverage point according to the required minimum threshold of 10, and eliminated the points that did not meet the requirements. Due to the vast amount of data, the business volume was huge. Small dots are blurred. Set the coverage area of the base station as a sector. By adjusting the relationship between the calculation coverage radius and the angle, the angle between the three main directions is added as a constraint, and the planning is carried out again based on problem one location selection. The angle change is used as the decision variable, and the greedy algorithm is used to solve the problem. Moreover finally, the coordinates of each base station and the optimal sector angle that satisfy more than 90% of the traffic volume are: (232.6573, -111.668, - 29.7508). Using the k-means clustering algorithm, the weak coverage points are clustered according to the requirements of less than 20, and the frequency and percentage of clusters are analyzed. The time complexity is calculated by testing the DVI index and the number of iterations. On this basis, Optimize our model. Finally, the evaluation and promotion of our model are carried out, and the detailed distribution map and clustering effect map of various base stations are given in the appendix.