The recent advent of smartphones and mobile thin-clients has increased opportunities for people to download a large volume of data via wireless access networks while they on the move. In this study, we considered a scenario in which a mobile user moving from one location to another has to arrive at the destination within a certain period of time and wants to obtain as much data as possible before arriving there. In scenario like this, wireless access points on routes that a lot of people pass through become congested. In particular, if the communication range of the wireless access point is limited like wireless local access networks, the wireless bandwidth becomes more competitive. We therefore propose route instruction mechanisms for mobile users that consider longcut paths. Our mechanisms direct mobile users to another route on which wireless access points are less competitive. However, we need to consider the tradeoff between the loss caused by the additional movement and the amount of additionally downloaded data. Furthermore, when there are two or more users, this route instruction problem can be modeled as a cooperation problem because the users selfishly try to occupy the best route for themselves. We therefore came up with and compared cooperative and non-cooperative route instruction mechanisms. Computer simulation demonstrated that the former approach brings better fairness for users.