Abstract
A lot of research has been done on the problem of finding the k nearest neighbor to a query point. Existing studies are usually intended to work on static data. Even the minimal number of existing work done on dynamic objects has not solved the problems caused by their dynamic nature. The problem with KNN algorithms is how to keep the results fresh and avoid unnecessary computation cost each time the object changes position. This type of algorithm is in fact very used in many applications. In this document, a new challenge has been accepted to solve a complex problem. We propose a new approach to look for KNNs on continuously moving objects while guaranteeing a freshness of the results during a safety period during which the results of the query are always valid even if the object changes continuously its position. In order to take advantage of this type of algorithm in difficult situations such as the emergency decision-making process, we propose a new efficient algorithm to determine the K closest resources that are circulating in the same area of the query point. Our approach is progressive and relies on the Safe Region pruning method. As long as the object remains in its respective safe region, the new expensive computation is not necessary. The result of deep-seated experiments on our approach, validates its efficiency in terms of communication and calculation cost through a search restriction area method.