As a new future Internet architecture, information-centric networking (ICN) emerges as the time goes and is widely used in the field of Internet of Things. Caching technology is an important feature of ICN, and has an important impact on the content transmission performance in ICN-IoT. Due to the characteristics of data updating frequently and the strict requirements for data freshness in ICN-IoT, the traditional network caching technology of ICN is facing challenges. This study proposes a distributed caching strategy based on content popularity and network topology, considering the freshness of content. To maximize the caching efficiency, each caching node first caches the content with higher popularity and closer to users. In order to adapt to the frequent updating of content in IoT, a content caching reward prediction method based on grey prediction is proposed, which is convenient to obtain the caching reward of new content quickly. Simulation results show that, compared with the traditional caching strategy, the proposed scheme can effectively improve the caching efficiency and hit ratio, reduce the access delay, and improve the user experience.