This paper introduces a method of predicting LTE network traffic in time dimension

including O+B domain data to build users’ portrait

and establishes the LTE network prediction model based on big data. By the method

we can predict the cell users’ number and traffic volume and other data in the future

which will support the market development

network planning and capacity evaluation

meanwhile

local short-term changes of the network can be controlled with rapid optimization and adjustment. It also introduces capacity expansion criteria in different scenarios for network resources estimation based on users’ perception.