Problems of planning the execution of a set of jobs arising during the design and operation of backbone networks and transport and logistics centers are considered. This problem is formalized as the problem of finding the best schedule in a multi-agent system. Various options of problems of job planning and correction of deadlines depending on available throughput and processing capabilities are considered (all jobs are non-preemptive; all jobs allow interruptions and switching from one machine to another; part of jobs allows interruptions and switching, and part does not allow; in addition to machines non-renewable resource is also used). The mathematical solution is based on the reduction of incoming and outgoing flows to transport and logistics centers with a minimum cost in the backbone network. The areas of using the obtained results in the design systems of transport and logistics centers are proposed