Recent advances in wireless data networking, satellite services and cell connections allow us to take advantage of mobile computing systems. Here, a problem is the mobility of computing process which needs to get information from databases. In fact, system movement during the transaction interrupts the mobile network connection to server wireless systems, and the disconnection causes failing in the running transaction so the mobile system has to run the transaction again. One way to avoid transaction failure is doing prediction process until connecting to the next wireless network. The objective of this study is to use particle swarm optimization (PSO) algorithm in order to provide a solution for prediction of the information sent to the mobile system for using the memory. Generally, PSO algorithms result in the best or an acceptable solution for optimized selection and design in many scientific and technological problems. Here, the objective of PSO is to select the best information to send to the mobile system when the connection is interrupted in order to prevent transaction failures. This study focuses only on information sent to the mobile system.