Abstract: With the rapid development of mobile internet technology and daily increase in wireless user equipment as well as demand of mobile applications, the required quality of service (QoS) by mobile users for wireless communication networks is increasing, and the pressure of backhaul network is also increasing. The emerging cloud radio access network (C-RAN) can effectively improve the network capacity and user service quality. Meanwhile, Passive Optical Network (PON) was used for backhaul network that offers a large bandwidth, high reliability, and low latency backhaul. When the demand of mobile applications is ever-changing and backhaul network resource is limited, an efficient backhaul network resource scheduling strategy is of great importance. This can improve the backhaul network resource utilization and waiting-transmitting delay in C-RAN. To save backhaul network resource, improve the load balance of wavelengths and the utilization of network resource, a downlink resource scheduling strategy was proposed. According to the network demand of wireless users in hot spots, three optimization objectives including the number of used wavelengths, load balance, and well-distribution of real-time traffic were considered in the optimization of adaptive-weighted parallel genetic algorithm. Thus, the wavelength resources were dynamically allocated and resource utilization was improved. The results demonstrate that the proposed downlink resource scheduling strategy can effectively improve the load balance and resource utilization and reduce the waiting-transmitting delay of real-time traffic.