Entropy, Information, and Reconciliation: A Quantitative Perspective on Sensor Networks
Roberto Nieto Linares · 2025
This paper explores Shannon's information-theoretic entropy as a key performance indicator (KPI) for data reconciliation in sensor networks, reinforcing the importance of entropy as a measure of uncertainty and methodology that accurately estimates unmeasured streams and identifies the equalization of entropies at one measurement, enhancing the robustness and applicability of the approach. Using a series flow network with 3 nodes and 4 streams, we demonstrate that entropy has a direct influence on the variance that entropy before reconciliation increases with more measurements, while entropy after reconciliation decreases, both entropies equalizing at one measurement. The methodology aligns with recent entropybased reconciliation techniques, offering insights for sensor network design, fault detection, and process monitoring.