frame work for Access point selection based on Markov decision process

I need to implement the attached paper on my system. i need a help from an expert in Markov chains and Markov decision process to use it in the cellular network to predict the user location under each cell in the future and optimize network using matlab we will use Markov decision process to chose the best node in wireless network based on the arrival rate and service time for each connection, after that we calculate the net gain for each path between a specific user and different access nodes using value iteration or policy iteration algorithm. then we decide which access node is the best to serve the mobile user. i need all simulation to be done with MATLAB. Paper firmin (7) is contains the main idea of the project.