As the amount of avaliable resources of detecting and attacking enemies is increased in the defense domain, the approaches of servicizing the weaponry resources and making them dynamically composed together to support mission plans have been developed. The core part of these approaches includes the service management, representation of the functionalities and capabilities of weaponry resources, and dynamic service reconfiguration based on changes in mission requirements and/or mission environments. Distributed service coordination is an essential issue in dynamic service reconfiguration because services should be managed at runtime and assigned to appropriate resources dynamically in a mission environment. The centralized service coordination approaches have the single point failure problem and may cause the centralized server to become a bottleneck as the number of requests increases. To overcome these limitations, we propose a mission-oriented and distributed service coordination approach that supports a collaborative execution of services based on connected resource-nodes in a tactical network. We developed a dynamic resource allocation algorithm for minimizing the communication overhead between gateways based on the graph clustering algorithm. We showed the effectiveness of the proposed approach by performing a simulation. The result shows that the communication overhead can be reduced by 10 to 20% in comparison to the greedy and random allocation algorithms.