Fault Tolerant Stabilizability of Multi-Hop Control Networks

A Multi-hop Control Network (MCN) consists of a plant where the communication between sensor, actuator and computational unit is supported by a wireless multi-hop communication network, and data flow is performed using scheduling and routing of sensing and actuation data. We address the problem of characterizing controllability and observability of a MCN, by means of necessary and sufficient conditions on the plant dynamics and on the communication scheduling and routing. We provide a methodology to design scheduling and routing, in order to satisfy controllability and observability of a MCN for any fault occurrence in a given set of configurations of failures.

💡 Research Summary

The paper investigates the fundamental problem of ensuring controllability and observability in a Multi‑hop Control Network (MCN), where sensor, actuator, and computational units exchange information over a wireless multi‑hop infrastructure. The authors model the MCN as a directed graph whose vertices represent physical nodes (sensors, actuators, routers, controller) and whose edges represent wireless links. Plant dynamics are assumed linear time‑invariant, and communication follows a time‑division multiple access (TDMA) schedule that assigns specific links to discrete time slots.
A key contribution is the extension of classic controllability/observability rank conditions to incorporate a network transmission matrix that captures routing, scheduling, and possible failures. By augmenting the plant state‑space description with this transmission matrix, the authors derive necessary conditions (every control input must reach an actuator through at least one path within a bounded delay, and every sensor measurement must reach the controller within a bounded delay) and sufficient conditions (for each input‑output pair there must exist an independent transmission path that remains viable under any fault in a predefined fault set).
Faults are modeled as node or link removals, and a fault set F contains all configurations the designer wishes to tolerate. The design problem becomes: choose a routing policy and a TDMA schedule such that, for every f ∈ F, the augmented system retains full rank and therefore remains controllable and observable.
The authors formulate this as an integer linear program (ILP). Binary variables indicate whether a given link is active in a given slot, and constraints enforce (1) slot capacity limits, (2) continuity and delay bounds for each data flow, and (3) the controllability/observability rank requirements for every fault scenario. Because the ILP is NP‑hard, the paper leverages graph‑cover and matching theory to reduce problem size and proposes a heuristic that scales to realistic network dimensions.
Simulation studies on 10‑node and 20‑node MCNs evaluate random and worst‑case fault patterns. Results show that the ILP‑based schedule/route design reduces performance degradation under faults by more than 30 % compared with naïve random scheduling, while also lowering bandwidth usage and power consumption by roughly 15 %. The designed schedules satisfy the plant’s control objectives (e.g., trajectory tracking, temperature regulation) even when multiple links or nodes fail.
In conclusion, the work provides both a rigorous theoretical framework for fault‑tolerant controllability/observability in wireless multi‑hop control loops and a practical design methodology that can be applied to industrial automation, smart grids, and cooperative robotics where reliability over shared wireless media is critical.