Distributed Sensing: Entanglement Delivery and Protocols
Distributed sensing can use entanglement between remote sensors to estimate a spatially distributed parameter. This page asks the network question: can the required probe be delivered and synchronized often enough that its information gain survives distribution cost? The sensor-network assessment owns measurement objectives and field-sensing applications; this page owns delivery, scheduling and protocol feasibility.
Required background. Distributed quantum sensing develops weighted parameters and matched separable bounds.
Information per unit time is the relevant currency
Section titled “Information per unit time is the relevant currency”For repeated independent interrogation cycles, a useful diagnostic is Fisher information per total cycle time,
Use classical Fisher information for the actual measurement; quantum Fisher information is an upper bound unless an attainable measurement is supplied. Random preparation time and failed distribution attempts must be averaged consistently. A highly sensitive accepted probe can lose to independent sensors when it is produced rarely.
The target may be a weighted sum of local phases rather than every phase separately. Distributing one large GHZ state, several smaller entangled clusters or continuous-variable correlations gives different loss and synchronization tradeoffs.
Evidence and limitations
Section titled “Evidence and limitations”Zhang and Zhuang review distributed sensing tasks and demonstrations. Zhuang, Zhang and Shapiro analyze a continuous-variable entangled sensing scheme and show explicitly why loss changes ideal scaling. Established: entanglement can improve specified distributed estimation tasks under stated resources. Active: retaining that benefit after lossy distribution and realistic duty cycles. Conjectural: universal improvement for arbitrary spatial fields or unknown nuisance parameters. Review, primary analysis.
A common phase reference and synchronized interrogation can be expensive shared resources. Give the separable comparator access to the same classical synchronization and total sensing energy.
Open questions
Section titled “Open questions”When should a network abandon an aged multipartite probe and restart? Can adaptive cluster size beat a fixed global state under outages? Which entanglement graph best matches a weighted field functional? How can unknown phase-reference drift be separated from the signal without consuming the expected gain?
Graduate project: choose a cluster size under loss
Section titled “Graduate project: choose a cluster size under loss”Model a small network estimating a common phase with product probes, one GHZ probe and several GHZ clusters. Include independent loss, dephasing and a stochastic entanglement-preparation delay. Compute the information from the actual readout distribution and compare at equal particle-time and elapsed time. A laptop can enumerate small systems.
Deliver the optimal cluster size versus loss and preparation rate, with analytic no-noise checks. Failure tests add correlated dephasing, an unknown phase offset and unequal link rates. If the advantage exists only when distribution time is ignored, record that as the result. Research begins with a delivery policy or probe design that remains useful under a measured network channel and the application’s nuisance parameters.
Entry points and dates
Section titled “Entry points and dates”Read the review for the task, then the primary calculation for a loss-sensitive example. Sources checked on 11 October 2026; selected evidence spans 2018–2021 rather than a current sensor-network performance census.