Support creating noise model

[king-p3nguin]

How to create a noise model

According to [1], the type of errors in LOQC can be grouped into detection errors, source errors, and circuit errors.

Detection errors

The detector may count fewer or more photons than were actually present. The former is called photon loss, and the latter is called dark count. Adding the effect of photon loss is not difficult because we can add a loss channel to Perceval processors, which is a non-unitary component that simulates photon loss.
https://perceval.quandela.net/docs/components.html#loss-channel

For Dark count, [2] may be helpful.

Source errors

Perceval has already implemented an error model for single-photon source:
https://perceval.quandela.net/docs/reference/source.html

[3] describes the details of the error model.

Circuit errors

There are many possible causes for errors that occur during calculations.

• Mode mismatching
  This is considered to be the main error in the circuit error [1]. It is due to the indistinguishability of the photons, and it can be configured in perceval.Source.
• Absorption and scattering of photons
  Linear optical components may absorb photons or scatter photons back to the source, which results in a loss of photons. This can be implemented as photon loss.
• Feed-forward process
  An error can occur due to the read-out error of a photon detector and the malfunction of an optical switch. So far, we do not need to consider the error of the feed-forward process because Perceval does not support the feed-forward process.
• Losing photon in a fiber
  Storing a photon for 100 µs in a fiber has a loss probability of p ≈ 0.54 [1].

Parameters of the noise model

Creating a realistic error model is another issue. Need to find literature on the basis for determining the parameters of the error model.

[1] KOK, Pieter, et al. Review article: Linear optical quantum computing. arXiv preprint quant-ph/0512071, 2005, 29. https://arxiv.org/abs/quant-ph/0512071
[2] KOK, Pieter, et al. Review article: Linear optical quantum computing. arXiv preprint quant-ph/0512071, 2005, 29. https://arxiv.org/abs/quant-ph/0512071
[3] PONT, Mathias, et al. High-fidelity generation of four-photon GHZ states on-chip. arXiv preprint arXiv:2211.15626, 2022. https://arxiv.org/abs/2211.15626

List of sub-tasks

• Support detection noise model
• Support source noise model
• Support circuit noise model
• Add an example to create a realistic noise model