To meet the network information interaction requirements for effective mission execution in unmanned swarm systems, the challenges of resource allocation rigidity and limited protocol flexibility arising from planning and design in network organization were investigated. A dynamic resource allocation optimization protocol was then proposed, which can adapt to task variations. The protocol incorporates mechanisms such as low-overhead distributed resource reservation, resource priority lists, and temporary member resource allocation. Specifically, numerical simulations were conducted to analyze key parameters affecting protocol performance, including the maximum proportion of dynamic resource demands, the number of members requesting dynamic resources, and the number of time slots with the highest priority allocated to members. The simulation results demonstrate that the proposed dynamic resource allocation optimization protocol can effectively adapt to changing resource demands from tasks with varying priorities, meet the flexible resource requirements of members, and improve the utilization rate of dynamic resources.