Analysis of market data with Noir: DEBS grand challenge

BSTRACTThe 2022 DEBS Grand Challenge targets the analysis of real-world market data to define a trading strategy that triggers buy/sell advice based on specific temporal patterns. This paper presents a solution to this problem based on Noir, a distributed data processing framework developed at Politecnico di Milano that aims to provide high-level programming abstractions with minimal overhead. Noir abstracts all concerns related to deployment, concurrency, and synchronization, allowing developers to concentrate on the logic of the processing task. It simplifies the definition of a solution and minimizes the time to develop it and make it operational, while at the same time it does not sacrifice absolute performance. Our experiments show that Noir can analyze millions of events per second and deliver answers with a latency of few milliseconds.