An overview on conflict detection and resolution methods in air traffic management using multi agent systems

Abstract

So far a number of methods have been presented in order to automate air traffic management (ATM) using multi-agent systems (MASs) technology. However, there has been little discussion about the advantages and disadvantages of these methods. In addition, there has not been created a comprehensive reference to compare these methods. Furthermore, researchers studied ATM from a specific viewpoint and used different experimental data (real or artificial) and different performance metrics for evaluation of their proposed models. Obviously, to compare different proposed methods, we should use identical and explicit criterions that indicate the utility of presented models, then using these criterions we could examine the efficiency of proposed models. In this paper, our major goal is to present an explicit framework to categorization and comparing various models for conflict detection and resolution (CDR) in ATM based on MAS technology. Also we try to present a review of different proposed models (twenty two models) by using this framework. Some of these models have been implemented and currently are in use. The others are under evaluations or currently there isn't any available technology for building them. Our classification method is based on criterions such as: agent selection (the entity which selected as Agent), agent's actions, agents' interaction method in the process of CDR, the strategy used in agents' implementation, type of the multi-agent system (pure multi-agent system or combined) conflict detection method, conflict resolution method, and management the multiple aircrafts conflict. Then, using these features that defined in a specified framework, we evaluate different models that so far presented.