On Two Sensors Scheduling for Remote State Estimation With a Shared Memory Channel in a Cyber-Physical System...

Area of Science:

• Cyber-Physical Systems (CPSs)
• Control Theory
• Estimation Theory

Background:

• Remote state estimation in CPSs relies on sensors transmitting data over communication channels.
• Shared memory channels introduce packet reception correlations, complicating optimal scheduling.
• Minimizing estimation errors is critical for effective CPS operation.

Purpose of the Study:

• To develop an optimal scheduling policy for two sensors sharing a memory channel in CPSs.
• To minimize total estimation errors at the remote estimator.
• To reduce computational overhead associated with the scheduling policy.

Main Methods:

• Formulation of the sensor scheduling problem as a Markov decision process (MDP).
• Derivation of the optimal policy and its threshold structure.
• Application of the Whittle index policy after proving system indexability.

Main Results:

• An optimal scheduling policy was derived using MDP.
• A threshold structure for the optimal policy was identified, reducing computation.
• The Whittle index policy was shown to be effective in further reducing computational load.
• Numerical simulations validated the theoretical findings.

Conclusions:

• The proposed MDP-based and Whittle index policies effectively minimize estimation errors in CPSs.
• The threshold structure and Whittle index approach offer significant computational advantages.
• This work provides a framework for efficient sensor scheduling in networked CPS applications.