When the traditional method is used to offload mobile big data, the result is unreasonable and the shunting balance to different processors is poor. To this end, a new method for decentralized mobile big data is proposed. Given the overall structure of mobile big data distribution platform in cloud computing environment, mobile big data shunt unit is installed on each programmable switch in cloud computing environment. The problem of mobile big data shunting is transformed into the problem of solving the optimal solution, and the shunting is realized by state feedback. The optimal decomposition method is combined with Lagrange multiplication to complete the distributed computing and obtain the balanced traffic diversion model of mobile big data on the link. The switch shunt unit is used to obtain the diversion status of the network in the cloud computing environment. The partial derivative of the revenue function at this moment is obtained and the two are summed to obtain the final shunt result. Analysis of mobile big data shunt delay. The experimental results show that the proposed method can ensure the balance and rationality of diversion.