Robust Sensor Placement for Resilient Network Observability

15 concepts8 papers

In the design of monitoring systems—whether for power grids, robot navigation, or air traffic control—the traditional goal has been efficiency. Engineers typically seek the minimum number of sensors required to make a system "observable," meaning the internal state of the entire system can be inferred from the available measurements.

However, a growing body of research suggests that this drive for minimality creates a dangerous fragility. When a sensor network is designed with zero redundancy, it possesses "critical measurements": single points of failure where the loss of one sensor or communication link renders the entire system unobservable.

The Fragility of Minimal Coverage

The core problem is a trade-off between cost and resilience. In power systems, Li et al. observe that while basic observability ensures a mathematical solution exists, it often relies on these critical measurements. If a single Phasor Measurement Unit (PMU) fails or a communication line is disrupted, the grid operator effectively goes blind.

This vulnerability is not unique to power grids. In crowdsourced aircraft surveillance networks like OpenSky, Strohmeier et al. highlight that large geographic areas often depend on a single receiver. In these "single-sensor coverage" zones, even low-power interference or simple hardware failure can cause a total loss of situational awareness. Similarly, Komatsu et al. note that robot camera systems designed to minimize visual overlap to save processing power are left without any backup if a single camera fails.

Redundancy as a Design Constraint

To move beyond this fragility, researchers propose treating redundancy not as a luxury, but as a fundamental constraint in the optimization process. The goal shifts from finding the minimum set of sensors to finding the minimum robust set.

Liu et al. frame this challenge as a set cover problem. By identifying all possible single-failure scenarios (sensor or link failures), they use polynomial-complexity algorithms to select a sensor suite that maintains structural observability across every scenario. This approach ensures that every state is "covered" by enough independent information sources that no single failure can break the chain of inference.

In the context of transportation networks, Rinaldi et al. argue that incorporating overlapping route data—which might seem redundant in a traditional model—is essential for resilience. They demonstrate that while non-overlapping designs minimize initial sensor counts, they suffer "catastrophic information loss" when faced with even minor budget fluctuations or hardware failures.

Mapping Observational Equivalence

If a sensor does fail, how does a system recover? The answer lies in understanding the network's topology. Doostmohammadian et al. and the authors introduce the concept of observational equivalence and "contractions" in complex networks.

This framework identifies sets of nodes that provide identical information for monitoring based on their position in the network. If a sensor at one location fails, the system doesn't necessarily need a new design; it needs a replacement sensor at an "equivalent" site. By mapping these equivalent sets, researchers can create a roadmap for restoring visibility with minimal effort.

Moving Targets and Adversarial Resilience

Beyond accidental failure, redundancy is a critical tool for security. Sengupta et al. propose a Moving Target Defense (MTD) for power grids. Instead of relying on one fixed set of sensors, the system dynamically switches between different redundant sets of PMUs. Because each set is capable of full observability, the system remains functional, but the "shifting surface" makes it much harder for an attacker to know which sensors to compromise to blind the operator.

Ultimately, this research direction suggests that the most efficient sensor network is not the one with the fewest sensors, but the one that maintains its ability to "see" even when its components fail. By mathematically formalizing redundancy, researchers are building systems that are robust by design rather than by accident.

Key Foundational Work

Minimum Robust Sensor Placement for Large Scale Linear Time-Invariant Systems: A Structured Systems Approach (2013) — contributed 5 concept elements to this direction.
A Framework for Phasor Measurement Placement in Hybrid State Estimation Via Gauss–Newton (2014) — contributed 3 concept elements to this direction.
History & citationVersion 1 · 2026-10-04

Current · Permanent link

View frozen version 1

Cite this direction

Latest version (a living page):

@misc{lacuna_single_sensor_coverage_dependency_076b5535,
  title = {single-sensor coverage dependency},
  author = {{Lacuna}},
  howpublished = {Lacuna research direction},
  year = {2026},
  url = {https://lacuna.tiptreesystems.com/direction/single-sensor-coverage-dependency/rdr_076b5535cd9f4a8c83d9b66df9d0ba79},
  note = {Latest version of a living research direction; cite a version for a fixed text. Accessed 2026-11-19},
}

This exact version (version 1, frozen text):

@misc{lacuna_single_sensor_coverage_dependency_076b5535_v1,
  title = {single-sensor coverage dependency},
  author = {{Lacuna}},
  howpublished = {Lacuna research direction},
  year = {2026},
  url = {https://lacuna.tiptreesystems.com/direction/single-sensor-coverage-dependency/rdr_076b5535cd9f4a8c83d9b66df9d0ba79/v/1},
  note = {Version 1, frozen 2026-10-04. Accessed 2026-11-19},
}

See something wrong? Report a mistake