Ph.D. Defense: Jing Geng
Tuesday, December 7, 2021
301 405 3681
ANNOUNCEMENT: Ph.D. Defense
Name: Jing Geng
Professor Shuvra Bhattacharyya (Chair)
Professor Richard La
Professor Manoj Franklin
Dr. Richard Candell (Special member)
Professor Jeffery Herrmann (Dean’s representative)
Date/time: Tuesday, December 7, 2021 at 11am-1pm
Location: AVW 2224
Title: Methods and Tools for Simulation of Industrial Wireless Networks
In recent years, there has been significantly increasing interest in integrating wireless communication capability within factory automation (FA) systems. This interest has arisen due to the important potential advantages brought about by wireless communications technology, including lower cost and increased operational efficiency. Such integration motivates the study of interactions among the physical layout of factory workcells, wireless communication, and improving the overall factory system performance. However, design of such cyber-physical systems with wireless networks involves stringent constraints on real-time performance and reliability, and requires the assessment of performance for complex interactions among process control, factory topology construction, and wireless communication.
Due to the highly complex design spaces associated with wireless-integrated factory systems, efficient co-simulation methods as well as tools for modeling and estimating the interactions between physical process flows and wireless communication are needed. In this thesis, a new integrated platform, called the Experimentation Platform for Advanced Industrial wireless Systems (EPAIS), is proposed. EPAIS integrates a collection of novel software tools and modeling libraries to support assessment and experimentation in the design of next-generation cyber-physical systems integrated with wireless networks.
Within EPAIS, four key tools are introduced. This thesis introduces each of these tools along with extensive experimental results that validate the utility of the tools. The first tool is a model-based co-simulation framework to support assessment and experimentation in the design of next-generation FA systems with wireless communication networks. The framework is novel in its model-based approach, where different modeling techniques are employed to efficiently and flexibly integrate the design perspectives of FA topologies (system-level modeling of networks of machinery and control units), FA nodes (processing within the individual networked nodes), and wireless communications (protocol functionality and channel characteristics).
In addition to the encapsulated co-simulation framework, EPAIS contains a companion tool that supports more streamlined processes for simulation model development; provides synthesis of detailed, low-level simulation code from high-level, user-friendly representations; provides for more efficient modeling and simulation of complex simulation scenarios; and enables advanced design space exploration methods.
In order to more accurately reflect the harsh communication conditions in industrial environments, two link layer simulation plug-ins are also developed for EPAIS. These plug-ins, which represent the third and fourth tools developed as part of EPAIS, generate a modeling subsystem called an EPAIS channel model library. Such channel model libraries are constructed using a new method for channel model construction based on link-layer simulation. A distinguishing characteristic of the method is that it systematically incorporates field measurements of wireless communication channels from industrial networks. The generated library components capture channel characteristics in the form of lookup tables, which can be flexibly integrated into system-level simulators or network simulation to increase the efficiency as well as the accuracy of simulation.