Manifestation of Rank-Tuned Weak Measurements Towards Featured State Generation

TL;DR

This paper introduces a measurement-based method to generate and analyze multipartite entangled states, revealing how measurement rank influences the type and robustness of entanglement produced.

Contribution

It classifies unsharp measurement protocols into biased and unbiased types and explores how measurement rank affects the generation of different entangled states and their properties.

Findings

Rank-2 measurements produce GHZ-class states.

Rank-4 measurements produce W-class states.

Average GME increases as measurement rank decreases.

Abstract

We propose that an unsharp measurement-based process to generate genuine multipartite entanglement from an entangled initial state with a fewer number of qubits can be classified in two ways -- biased and unbiased inflation protocols. In the biased case, genuine multipartite entanglement (GME) of the resulting state obtained after a single measurement outcome is optimized, thereby creating a possibility of states with high GME while in the unbiased case, average GME is optimized over all possible outcomes. Interestingly, we show that the set of two-qubit unsharp measurements can generate multipartite states having different features according to GME measure, generalized geometric measure, the monogamy-based entanglement measure, tangle and robustness against particle loss quantified via persistency depending on the rank of the unsharp measurement operators. Specifically, in the process of producing three-qubit pure states, we prove that rank-$2$ measurements can create only Greenberger Horne Zeilinger (GHZ)-class states while only W-class states are produced with rank-$4$ measurements although rank-$3$ measurements are capable to generate both. In the case of multipartite states with an arbitrary number of qubits, we report that the average content of genuine multipartite entanglement increases with the decrease of the rank in the measurement operators although the persistency decreases with the rank, both in the biased as well as unbiased protocols.