In this paper, we focus on an information gathering system where a reader continuously collects information from mobile nodes in its access area, such as environmental information cameras and sensors. We assume that a mobile node is relatively tiny and does not have a high-precision antenna to sense carriers emitted by other nodes. Although a random access method like ALOHA can be easily used, it has disadvantages of transmission efficiency and energy consumption. To tackle these problems, we propose a novel method that is a combination of random and selective accesses. At first, a reader sends an ID request to all nodes. Then, each node replies its ID to the reader at a response probability involved in the request. Finally, the reader selectively gathers information from nodes according to the obtained ID lists. In our method, non-registered nodes and non-deleted nodes affect the system performance. The non-registered node is a node that is in the access area but its ID is not registered to the reader. The non-deleted node is a node that leaves the area but its ID is still registered to the reader. In this paper, we first derive their numbers by an analysis using the Inversion Formula of Palm Calculus. Then, we conduct simulation experiments to verify the analysis. Simulation results show that the proposed method performs well in a wide range of mobility by appropriately controlling the response probability.