This paper investigates the optimal design of the wireless powered communication network, where a power transmitter (PT) and multiple fog servers (FSs) are used to charge an energy-constrained wireless device (WD) by radio-frequency (RF)-based energy harvesting, and the WD's computing task is partioned into multiple parts, processed by WD itself (local computing) and FSs (fog computing) jointly, in which an orthogonal spreading coding (OSC)-based transmission protocol is employed for task offloading. For such a system, we formulate an optimization problem to maximize system computation rate by jointly optimizing the energy beamforming, the time assignment, the computing frequency and the power allocation at the WD in the presence of multiple system constraints, including the energy causality, the computation capability and the power and the storage capacities. Since the problem is non-convex, we first transform it into some equivalent forms and then design a Dinkelbach's programming-based method to solve it. Some semi-closed form results are obtained. Simulation results show that our proposed system with OSC outperforms the benchmark systems. It also confirms that with the increment of the number of FSs, the computation rate will increase, but the increasing rate is rapidly decreased if the FS number is larger than 3.