Development and Performance Evaluation of a Connected Vehicle Application Development Platform

Abstract

Connected vehicle (CV) application developers need a development platform to build, test, and debug real-world CV applications, such as safety, mobility, and environmental applications, in edge-centric cyber-physical system (CPS). The objective of this paper is to develop and evaluate a scalable and secure CV application development platform (CVDeP) that enables application developers to build, test, and debug CV applications in real-time while meeting the functional requirements of any CV applications. The efficacy of the CVDeP was evaluated using two types of CV applications (one safety and one mobility application) and they were validated through field experiments at the South Carolina Connected Vehicle Testbed (SC-CVT). The analyses show that the CVDeP satisfies the functional requirements in relation to latency and throughput of the selected CV applications while maintaining the scalability and security of the platform and applications.

Serverless Cloud-based Speed Advisory Application for Connected Vehicles – Case Study

“…7 presents the processing time, and the end-to-end delay reported during our experiments using box charts and Table 3 provides the averages of the processing time and the end-to-end delay for each CV in the cloud for the three traffic density conditions. From Table 3, the end-to-end delay is about 452 ms (on average for all three traffic density conditions), which meets the requirement of a real-time CV mobility application, i.e., maximum allowable delay of 1000 ms [10,25]. Besides, we observe from Fig.…”

“…Each cluster contains multiple serverless CV advisory modules that process information from the CVs. To meet the latency requirement of a real-time CV mobility application, i.e., less than or equal to 1000 ms [10,25], we define the capacity of each serverless CV advisory module in terms of the maximum number of CVs to be processed, which is 50 CVs per module in our AWS implementation, and run all the CV advisory modules in parallel. The usage of parallel computing in a cluster makes our Serverless CloSA fast and scalable.…”

“…The primary contributions of this study are (i) to develop a serverless cloud computing architecture using AWS for a CV speed advisory application in a TCPS environment, which we refer to as "Serverless CloSA" in this paper, and (ii) to develop an optimization-based real-time CV speed advisory application which is deployable in the cloud in terms of end-to-end latency requirement. Similar serverless architectures can also be used to develop other types of CV mobility applications, such as queue warnings and eco-driving advisories, where the maximum end-to-end latency threshold is considered to be 1000 ms [10]. In our Serverless CloSA, CVs and connected traffic signals send their state information, i.e., basic safety messages of CVs, and traffic signal phase and timing information of connected traffic signals, into the serverless cloud computing infrastructure.…”

Implementation of a Sensor Big Data Processing System for Autonomous Vehicles in the C-ITS Environment

“…Islam et al (2020) aimed to implement a platform for developing connected car services, where the platform users could access data collected from connected cars [16]. Applications developed on the platform directly deliver messages such as BSM to a connected car.…”

“…This is the first time that Kafka and the database connector are connected, causing a delay. In addition, when using the Kafka server for both ROS to Python application and Python application to Database processes, the proposed C-ITS environment guarantees stable latency because the maximum latency for occupant safety does not exceed 200 ms, and an average speed of 14 ms is guaranteed [16].…”