Abstract: In the context of large-scale distributed resources participating in cooperative regulation of distribution grids, the limited communication and sensing resources pose great challenges in satisfying the demands of low-latency transmission and processing of massive device data. An integrated communication-sensing resource allocation architecture is constructed to improve the distribution grid data-section consistency. Secondly, aiming to minimize the network average end-to-end queuing delay and delay-section deviation, the optimization problem is decomposed into compressed sensing mode selection, device scheduling optimization subproblem, and time allocation optimization subproblem based on Lyapunov optimization theory. Finally, a two-stage communication-sensing resource allocation method based on enhanced swap matching is proposed. In the first stage, the delay-section deviation and queuing delay deficit are introduced to improve the matching benefit function. Subsequently, the optimal compressed sensing mode selection and device scheduling strategy are solved through iterative swap matching. In the second stage, the optimal time allocation strategy is solved based on convex optimization. Simulation results demonstrate that the proposed algorithm exhibits excellent performance in reducing end-to-end queuing delay and improving the consistency of the data-section.