The Impact of QoT on Estimation and Control
pdf
Inexpensive computation and ubiquitous embedded sensing, actuation, and communication provide tremendous opportunities for societal impact, but also great challenges in the design of networked control systems, because the traditional unity feedback loop that operates in continuous time or at a fixed sampling rate is not adequate when sensor data arrives from multiple sources, asynchronously, delayed, possibly corrupted, and — especially important for this project — the different entities that participate in the control system do not share a common clock.
Significant performance gains can be gained by allowing controllers to use timestamps that specify the time at which a particular sensor measured the state (or a portion of it). However, in practice, three types of errors afflict timestamps:
- Delays between the time at which the hardware collects the data and the time at which the timestamp is generated (often by software);
- Mismatches between the local clock of the sensor that timestamps the data and an absolute time reference;
- Mismatches between the local clock of a controller node that processes the data and the absolute time reference.
These errors limit the system’s ability to have an accurate and unified notion of time and we use the term Quality of Time (QoT) to generically denote a characterization of how well one can localize time events with respect to an absolute time reference.
A key goal of this project is to investigate the impact of QoT in the stability and performance of a networked control system. When considering time-stamping errors, we expect the stability conditions to also permit flexibility in terms of tradeoffs between control performance, message rates (related to communication bandwidth), network delays, and QoT. This enables synergetic system designs, in which control applications make effective use of information about QoT by adapting to it. An additional project goal is to investigate the robustness of distributed control and estimation algorithms to attacks focused on compromising the QoT of the network.
Submitted by Joao Hespanha
on
Inexpensive computation and ubiquitous embedded sensing, actuation, and communication provide tremendous opportunities for societal impact, but also great challenges in the design of networked control systems, because the traditional unity feedback loop that operates in continuous time or at a fixed sampling rate is not adequate when sensor data arrives from multiple sources, asynchronously, delayed, possibly corrupted, and — especially important for this project — the different entities that participate in the control system do not share a common clock.
Significant performance gains can be gained by allowing controllers to use timestamps that specify the time at which a particular sensor measured the state (or a portion of it). However, in practice, three types of errors afflict timestamps:
- Delays between the time at which the hardware collects the data and the time at which the timestamp is generated (often by software);
- Mismatches between the local clock of the sensor that timestamps the data and an absolute time reference;
- Mismatches between the local clock of a controller node that processes the data and the absolute time reference.
These errors limit the system’s ability to have an accurate and unified notion of time and we use the term Quality of Time (QoT) to generically denote a characterization of how well one can localize time events with respect to an absolute time reference.
A key goal of this project is to investigate the impact of QoT in the stability and performance of a networked control system. When considering time-stamping errors, we expect the stability conditions to also permit flexibility in terms of tradeoffs between control performance, message rates (related to communication bandwidth), network delays, and QoT. This enables synergetic system designs, in which control applications make effective use of information about QoT by adapting to it. An additional project goal is to investigate the robustness of distributed control and estimation algorithms to attacks focused on compromising the QoT of the network.