The stellar decomposition of Gaussian quantum states
Arsalan Motamedi, Yuan Yao, Kasper Nielsen, Ulysse Chabaud, J. Eli Bourassa, Rafael N. Alexander, Filippo M. Miatto · 2025
Read The stellar decomposition of Gaussian quantum states by Arsalan Motamedi, Yuan Yao, Kasper Nielsen, Ulysse Chabaud, J. Eli Bourassa, Rafael N. Alexander, Filippo M. Miatto (2025) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2504-10455.txt.
Introduction
Arsalan Motamedi and six co-authors introduce the stellar decomposition of Gaussian quantum states in two thousand twenty-five. Photon counting on Gaussian states makes non-Gaussian states of light. Their method splits such a state into a Gaussian core plus a transform. The transform acts on part of the modes and ignores the measurement.
The core has an exact, finite form in the photon number basis. For pure states a valid pure pair always exists, they prove. Mixed states get sharp conditions for when the split stays Gaussian. Bounds on state quality for photonic circuits follow, with GKP states in view.
Catalog record
| CREATOR | Arsalan Motamedi, Yuan Yao, Kasper Nielsen, Ulysse Chabaud, J. Eli Bourassa, Rafael N. Alexander, Filippo M. Miatto |
|---|---|
| YEAR | 2025 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 45 KB |
| VERSION | 1.0 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2504-10455
Browse the full library · About the Tower of Babel
Full text
The opening of the work, from the library's hosted text:
The stellar decomposition of Gaussian quantum states Arsalan Motamedi, Yuan Yao, Kasper Nielsen, Ulysse Chabaud, J. Eli Bourassa, Rafael N. Alexander, Filippo M. Miatto Abstract We introduce the stellar decomposition , a novel method for characterizing non-Gaussian states produced by photon-counting measurements on Gaussian states. Given an ( m + n ) (m+n) -mode Gaussian state G G , we express it as an ( m + n ) (m+n) -mode “Gaussian core state” G core G_{\mathrm{core}} followed by an m m -mode Gaussian transformation T T that only acts on the first m m modes. The defining property of the Gaussian core state G core G_{\mathrm{core}} is that measuring the last n n of its modes in the photon-number basis leaves the first m m modes on a finite Fock support, i.e. a core state. Since T T is measurement-independent and G core G_{\mathrm{core}} has an exact and finite Fock representation, this decomposition exactly describes all non-Gaussian states obtainable by projecting n n modes of G G onto the Fock basis. For pure states we prove that a physical pair ( G core , T ) (G_{\mathrm{core}},T) always exists with G core G_{\mathrm{core}} pure and T T unitary. For mixed states, we establish necessary and sufficient conditions for ( G core , T ) (G_{\mathrm{core}},T) to be a Gaussian mixed state and a Gaussian channel. We also develop a semidefinite program to extract the ‘largest’ possible Gaussian channel when these conditions fail. Finally, we present a formal stellar decomposition for generic operators, which is useful in simulations where the only requirement is that the two parts contract back to the original operator. The stellar decomposition leads to practical bounds on achievable state quality in photonic circuits and for GKP state generation in particular. Our results are based on a new characterization of Gaussian completely positive maps in the Bargmann picture, which may be of independent interest. keywords quantum optics, quantum information 1 Introduction Quantum information processing promises a radical paradigm shift compared with its classical counterpart [ 1 , 2 , 3 ] . This second quantum revolution [ 4 ] is based on several physical platforms, among which photonics stands out for its potential for deterministic entanglement generation, room-temperature operation, high clock speeds, and long coherence times [ 5 , 6 , 7 , 8 , 9 ] . Quantum states of light—and by extension, quantum operations—are usually separated into two broad classes, Gaussian [ 10 ] and non-Gaussian [ 11 ] , which share mathematical similarities with the Clifford/non-Clifford dichotomy in qubit systems [ 12 ] . Gaussian states and operations are generated by quadratic Hamiltonians in the bosonic canonical operators, and are often regarded as easier to handle from a theoretical point of view. However, they suffer from severe limitations for quantum information processing [ 13 , 14 , 15 , 16 , 17 , 18 , 19 ] : in particular, Gaussian computations can be simulated efficiently by classical computers [ 20 ] . This has led to the understanding of non-Gaussian states and operations as resources for many quantum information processing tasks [ 21 , 22 , 23 ] , and to the introduction of associated resource measures [ 11 ] . One such measure, the stellar rank [ 24 , 25 ] , provides a natural way of bookkeeping the “degree of non-Gaussianity" of quantum states in situations where one does not have access to Hamiltonians of degree higher than quadratic, or couplings between bosonic operators and those of two-level systems such as quantum dots, nor measurements of observables that are beyond quadratic. Indeed, the stellar rank of a non-Gaussian quantum state is non-increasing under Gaussian operations and lower bounds the number of photonic non-Gaussian operations (photon-addition or photon-subtraction) necessary to engineer a non-Gaussian state from a Gaussian one [ 24 , 25 ] . From an experimental standpoint, Gaussian states and operations are also simpler to engineer in photonic platforms than non-Gaussian ones [ 26 ] : currently, the standard technological toolkit in quantum photonics consists of the ability to generate deterministic Gaussian states, to implement Gaussian transformations deterministically, and to perform photon-number-resolving (PNR) measurements. Using this toolkit, a practical method to produce non-Gaussian resource states is to start with a multimode Gaussian state and then measure a subset of its modes in the photon number basis (Fock basis) using PNR detectors. This projects the state of the leftover modes onto a non-Gaussian quantum state based on the heralded pattern, with its stellar rank depending on the number of detected photons [ 27 ] . There is a rich history of heralding non-Gaussian states from Gaussian states [ 26 , 11 ] , including the generation of single-photon [ 28 , 29 ] , two-photon [ 30 , 31 ] and three-photon states [ 32 ] , photon-added coherent states [ 33 ] and photon-added thermal states [ 34 ] , displaced single-photon states [ 35 ] , finite Fock superpositions [ 36 , 37 ] , photon-subtracted squeezed states [ 38 , 39 , 40 , 41 ] , cat states [ 42 , 43 , 44 , 45 , 46 , 47 , 48 , 49 ] , multimode cat states [ 50 , 51 , 52 ] , better cat states [ 53 , 54 ] and Gottesman–Kitaev–Preskill (GKP) states [ 55 , 56 ] using Gaussian breeding protocols on top of non-Gaussian heralding [ 57 , 58 , 59 , 60 , 8 ] . Such quantum states are crucial building blocks for existing and future photonic devices [ 61 , 62 , 60 , 8 ] . However, the intrinsic complexity of heralded non-Gaussian states generated by large-scale devices makes tracking such states challenging, even though it is an essential task for device characterization, architecture design, and benchmarking. In this context, state decompositions [ 63 , 64 , 65 , 66 ] are particularly useful, as they allow complex non-Gaussian states to be expressed in terms of simpler, well-understood components. In this work, we present a new mathematical decomposition for Gaussian states, which we call the stellar decomposition due to its link with the stellar rank, that dramatically simplifies the analysis of non-Gaussian state generation protocols based on heralding photons from multimode Gaussian states. Specifically, we show that a multimode Gaussian state G on m+n modes can be expressed as a “Gaussian core state” G_{\mathrm{core}} on m+n modes, followed by a Gaussian channel T on m modes. G_{\mathrm{core}} has the special property that if one counts N photons across the n heralding modes, then the cutoff in the Fock basis to represent the state of the remaining m modes before T will be at most 2N . In general, without the stellar decomposition, the Fock cutoff required to express the output state to high numerical accuracy is much greater than 2N . Moreover, we find the channel T is independent of which photon pattern is heralded, which allows us to obtain general bounds on the levels of noise in the multimode Gaussian state G that can be tolerated in non-Gaussian state preparation protocols. Our results are based on a new characterization of Gaussian CPTP maps in the