Cost Efficient Application Placement for Smart Public Transportation

Abstract

By pushing applications and their service data to cloudlets close to end users, mobile edge computing paradigm significantly reduces network latency and communication cost. However, it may bring additional operation cost for each application replica and induce migration cost due to users’ movements. Specially, considering the scenario of public transportation with movement in citywide range, we formulate the cost efficient application placement problem with the goal of reducing the total costs for application deployment, while preserving users’ experienced QoS. By solving it with optimization solver, we evaluate the performance of the formulated optimization model. Preliminary evaluations exhibit that the proposed model could explicitly result in cost saving for application deployments targeting the public transportation.