Energy harvesting networks is a new form of computer networks.It can convert the environmental energy into usable electric energy

and supply the electrical energy as a primary or secondary power source to the electronic device for network communication.However

most of the energy harvesting networks use the analytical probability distribution function to describe the energy acquisition process

which can not accurately simulate the actual situation because the lack of authenticity.We propose an energy harvesting networks simulation method based on scenario generation in this paper.Firstly

based on the historical data of the harvested energy

the method does not need to set a probability distribution function in advance

and uses optimal scenario reduction technology to generate representative scenarios in single period.Secondly

it uses homogeneous simulated annealing algorithm to generate daily scenario sequences

so that can accurately simulate the random characteristics in energy harvesting networks.Taking the actual wind power data as an example

the accuracy and stability of the method are verified by comparison with the real data.Then we cite an instance to optimize network throughput

the optimal solution and data analysis showed the method based on scenario generation was feasible and effective in energy harvesting networks.