Network load prediction method and device

Abstract

The embodiment of the invention provides a network load prediction method and device. The method comprises the following steps: after a specific package is popularized, calculating monthly average useflow increment of a single user according to the use proportion of each package user and the use saturation of the package; calculating the daily average total flow of the cell according to the monthly average use flow amplification of the single user; calculating the total flow of the cell during self-busy and the uplink and downlink total flow of the cell during self-busy according to the dailyaverage total flow of the cell; calculating the uplink and downlink PRB utilization rate of the cell when the cell is self-busy according to the uplink and downlink flow growth rate of the cell whenthe cell is self-busy; calculating an effective RRC connection average value when the cell is self-busy according to the increment of the effective RRC connection user number of the cell after the package is popularized; and determining a high-load cell according to the uplink and downlink PRB utilization rate when the cell is self-busy, the effective RRC connection average number when the cell isself-busy and the uplink and downlink total flow when the cell is self-busy. According to the embodiment of the invention, the high-load cell to be expanded can be accurately predicted, so that targeted expansion optimization is carried out.

Description

technical field [0001] The embodiments of the present invention relate to the technical field of communications, and in particular, to a network load forecasting method and device. Background technique [0002] With the vigorous development of 4G networks and the continuous growth of user scale, operators in various provinces and cities and many Internet companies have begun to gradually promote large data packages in order to compete for users. With the promotion of user traffic packages, the traffic is growing rapidly and the network load is increasing. Therefore, it is necessary to accurately and accurately predict high-load cells, and then expand and optimize them to avoid the impact of high-load networks on user perception. [0003] At present, the prediction of network load capacity is basically based on historical data, from the three dimensions of user number, utilization rate and traffic, to predict the future change trend, and then select the cells to be expanded a...

AI Technical Summary

Problems solved by technology

[0004] Due to the impact of marketing strategies, the current forecasting methods can no longer accurately reflect the load situation of the community. For example, in terms of traffic forecasting, the existing forecasting model only considers the natural growth rate of traffic, but the growth of user traffic is affected by marketing strategies and network usage habits. Affected by many factors, such as considering only the natural growth rate, it cannot truly reflect the growth demand of user traffic, so it cannot accurately predict the high-load community

Images