Survey of AIOps for Air Traffic Control Automated Systems

As the next generation air traffic control automated system (ATCAS) being deeply studied, how to guarantee the stability and reliability of air traffic service has gained massive momentum from both academia and aviation industry. But operations of the next generation ATCAS are confronted with tough challenges, due to extending services like general aviation or UAV control, as well as applying newly techniques like cloud computing or visualization. Conventional system monitoring methods no longer be able to support the system operation tasks. Algorithmic IT operations (AIOps) precisely and fine-grainedly depicts the system via algorithms, which is proposed to realize autonomous management and is of great significance for supporting operations of the next generation ATCAS. This paper presents a systematical review of existing work and applications of AIOps. Both necessity and technical architecture of AIOps for ATCAS are discussed. The existing research achievements in AIOps are discussed from four perspectives which are methods based on time series analysis, text mining, deep learning and newly retroactive inference learning . Finally, the future research of AIOps for ATCAS is predicted.