This paper proposes a location-aided flooding mechanism to disseminate data in community-based IoT networks. Inspired by two heuristics obtained from the analysis of optimal flooding problem, the proposed mechanism allows wireless nodes to cancel a duplicate packet transmission in a distributed way when all of their neighbors have received that packet in advance. Extensive evaluations have been done in two different scenarios, i.e., a random uniform distribution of wireless nodes and a real distribution of wireless nodes in the Sumida ward of Tokyo. It has been validated that the proposed mechanism is not only able to increase the success ratio of data delivery, but also capable of reducing the delay of data delivery significantly, e.g., in the best case, the proposed mechanism improves the success ratio of conventional mechanisms by 47.3% and reduces the delivery delay by 92.0%.