Integrating Data Distribution Service – DDS – in an existing software architecture : exploring the effects of performance in scaling

Sammanfattning: Data Distribution Service (DDS) is a flexible, decentralized, peer-to-peer communication middle-ware. This thesis presents a performance analysis of using DDS in the Toyota Smartness platform that is used inside Toyota’s Autonomous Guided Vehicles (AGVs). The purpose of this thesis is to investigate if DDS is suitable for internal communication between the modules that reside within the Smartness platform.  Several different approaches of how to integrate DDS into the Smartness platform are explored, tests are carried out in regards to how increasing the number of interfaces affect the execution time and system utilization of these implementations. The results of these tests are presented in a multitude of ways to show the effects of DDS, this includes, results of how the performance scales, measurements of how the performance changes over time, and analytical results using linear regression.  Results show that DDS can be utilized in the future development of Smartness, but adds additional overhead to the system which negatively affects Smartness’ ability to keep up with the desired cycle time.