Application Development for Cyber-Physical Systems: Programming Language Concepts and Case Studies

Sookoor, Tamim, Computer Science - School of Engineering and Applied Science, University of Virginia
Whitehouse, Cameron, Department of Computer Science, University of Virginia

Cyber-Physical Systems (CPSs) combine low-power radios with tiny embedded processors in or-der to simultaneously cover large geographic areas and provide high-resolution sensing/actuation. However, CPSs are extremely difficult to program. This issue is addressed by presenting a macro programming framework called MacroLab that offers a vector programming abstraction similar to Matlab for Cyber-Physical Systems. The user writes a single program for the entire network using Matlab-like operations such as addition, find, and max. The framework executes these operations across the network in a distributed fashion, a centralized fashion, or something between the two – whichever is most efficient for the target deployment. This approach is called deployment-specific code decomposition (DSCD). MacroLab programs can be executed on mote-class hardware such as the Telos motes. The results indicate that MacroLab introduces almost no additional overhead in terms of message cost, power consumption, memory footprint, or CPU cycles over TinyOS programs.

As a crucial component of the application development cycle, debugging CPSs is addressed by MDB, the first system to support the debugging of macroprograms. MDB allows the user to set breakpoints and step through a macroprogram using a source-level debugging interface similar to GDB, a process called macrodebugging. A key challenge of MDB is to step through a macroprogram in sequential order even though it executes on the network in a distributed, asynchronous manner. Besides allowing the user to view distributed state, MDB also provides the ability to search for bugs over the entire history of distributed states. Finally, MDB allows the user to make hypothetical changes to a macroprogram and to see the effect on distributed state without the need to redeploy, execute, and test the new code. Macrodebugging is both easy and efficient: MDB consumes few system resources and requires few user commands to find the cause of bugs. A lightweight version of MDB, called MDB Lite, is also provided. It can be used during the deployment phase to reduce re-source consumption while still eliminating the possibility of heisenbugs: changes in the manifestation of bugs caused by enabling or disabling the debugger.

While a large number of CPS applications have been developed over the last few years, very few, if any, have been developed using a macroprogramming language. This is in spite of the vast array of macroprogramming languages, and abstractions, that are available. In an attempt to understand why macroprogramming languages are not being used for CPS application development and in order to evaluate the effectiveness of MacroLab in implementing a real-world CPS, a series of case studies involving the occupancy-based control of a Heating, Ventilation, and Air Conditioning (HVAC) system is implemented. Occupant-oriented HVAC control was selected due to it being a complete cyber-physical system involving the interaction between the real world and computation. Unlike many sensor network applications that involve primarily data collection, occupant-oriented HVAC control requires all three constituents of a complete CPS: sensing, computation, and actuation. The computation involved in these case studies range from making control decisions to predictions.

The case studies can be written in MacroLab in fewer lines of code than their implementations in Python and MacroLab can optimize their implementations depending on the topology of the network and capabilities of the sensors. Yet, the freedom for optimization decreases as the complexity of computation increases, or the size of the network decreases. Thus, for future cyber-physical systems the abstractions provided by macroprogramming systems, such as MacroLab presenting all sensor values as vectors and actuations as function calls, would be more beneficial than any implementation optimizations they may afford.

PHD (Doctor of Philosophy)
cyber-physical systems, embedded networks, macroprogramming, programming abstractions, source-level debugging, wireless embedded networks, building energy, wireless sensor networks, sensing
All rights reserved (no additional license for public reuse)
Issued Date: