JP7231208B2 - Phase recovery optical synchronous detector - Google Patents

Description

The present invention relates to systems and methods for receiving optical signals and detecting information including optical phase. More specifically, the present invention is a system or system capable of reproducing the optical phase information of an optical signal by causing interference in an optical signal and applying a phase recovery technique based on the observation result of the optical signal intensity after interference. Regarding the method.

Polarization diversity optical coherent receivers, which are widely used in backbone optical communication networks today, include a polarization beam splitter for separating polarization multiplexed signals, local light for synchronous detection, And an optical 90-degree hybrid for optical IQ mixing composed of complicated optical elements is required as an optical preprocessing circuit. In addition, in a large-capacity optical spatial multiplexing transmission system using multiple fiber modes, optical preprocessing circuits are required for the number of spatial channels, and an optical system for spatial channel separation is additionally required. Become.

In the next-generation optical coherent reception technology, spatial expandability for optical spatial multiplexing transmission and simplification for application to access networks are newly required.

K. Kikuchi, "Fundamentals of coherent optical fiber communications,"J. Lightwave Technol. , vol. 34, no. 1, pp. 157-179, 2016.

SUMMARY OF THE INVENTION It is an object of the present invention to provide a system and method for recovering optical phase information and enabling coherent reception with a simplified device configuration.

The present invention basically provides systems and methods based on novel coherent reception techniques based on phase retrieval techniques. This system uses a direct detector array with more elements than the number of optical signal channels and performs spatially redundant observations to recover the optical phase information lost by direct detection by means of signal processing. This enables simultaneous coherent reception of spatially multiplexed optical signals without complicated optical preprocessing.

In other words, the system of the present invention does not extract the optical electric field phase due to interference with local light (heterodyne or homodyne), but rather, for example, the optical intensity fluctuation due to spatio-temporal random interference between phase-modulated signal lights. , reproduces the optical electric field information including the phase in terms of signal processing.

One aspect described in this specification relates to a system 1 for receiving an optical signal and detecting information including optical phase.
The system 1 comprises an interference element 3 , a detector array 7 comprising a plurality of detectors 5 and processing circuitry 9 .
The interference element 3 is an element for scattering optical signals and causing interference.
The detector 5 is an element for measuring the optical intensity of the optical signal with interference caused by the interference element 3 .
The processing circuit 9 is an element for recovering the phase information of the optical signal using the information about the light intensity measured by the detector array 7 .

One aspect of the system described above is that the interference element 3 induces random interference in the optical signal.

In one aspect of the above system, where Nr is the number of detectors 5 and Nt is the number of optical signal channels, Nr is four times or more Nt.

In one aspect of the system described above, the processing circuit 9 recovers the phase information of the optical signal using factors relating to the discreteness of the optical signal.

In one aspect of the system described above, the processing circuit 9 uses the known signal to estimate a randomness component of the interference component 3, and uses the estimated randomness component of the interference component 3 to process the phase information of the optical signal. to recover.

One aspect described herein is a method for receiving an optical signal and detecting information including optical phase.
This method includes a step (S101) in which the interference element 3 scatters the optical phase-modulated signal to cause interference;
a step (S102) of measuring the optical intensity of the optical signal with interference caused by the interference element 3 by the detector array 7 including a plurality of detectors 5;
and a step (S103) in which the processing circuit 9 uses the information about the light intensity measured by the detector array 7 to recover the phase information of the optical signal.

INDUSTRIAL APPLICABILITY The present invention can provide a system and method for recovering optical phase information and enabling coherent reception with a simplified device configuration. The system of the present invention functions as a phase-recovery coherent receiver, has a simple configuration, and is excellent in spatial expandability. It is expected to accelerate the development of coherent technology.

FIG. 1 is a block diagram showing a basic configuration example of the system of the present invention. FIG. 2 is a diagram for explaining the principle of the present invention. FIG. 3(a) is a conceptual diagram showing a configuration example of the system in the embodiment. FIG. 3(b) is an enlarged view of the 32-pixel two-dimensional PD array portion. FIG. 4 is a graph in place of a drawing showing an example of channel response estimation. FIG. 5 shows the number of iterations N of the weighted steepest descent phase recovery algorithm (DRAF) considering the signal discreteness versus the bit error rate (BER) performance for different N _r for a single transmission channel. It is a graph instead of a drawing. FIG. 6 is an alternative graph showing bit error rate (BER ₎ characteristics versus the number of DRAF iterations N for different Nr for two-channel spatial multiplexing transmission.

DESCRIPTION OF THE PREFERRED EMBODIMENTS Embodiments for carrying out the present invention will be described below with reference to the drawings. The present invention is not limited to the embodiments described below, and includes appropriate modifications within the scope obvious to those skilled in the art from the following embodiments.

FIG. 1 is a block diagram showing a basic configuration example of the system of the present invention. As shown in FIG. 1, the system 1 includes an interference element 3, a detector array 7 including a plurality of detectors 5, and processing circuitry 9. In FIG. This system 1 relates to a system for receiving an optical signal and detecting information contained in the optical signal, including the optical phase.

An example of an optical signal is a phase-modulated signal carrying phase information. Examples of specific optical signals are optical QAM (Quadrature Amplitude Modulation) signals, optical PSK (Phase Shift Keying) signals, and optical OFDM (Orthogonal Frequency Division Multiplexing) modulated signals. Examples of optical PSK (phase shift keying) signals are optical binary phase shift keying (BPSK) signals, optical quadrature phase shift keying (QPSK) signals, and optical amplitude phase shift keying (APSK) signals. An example of information that includes optical phase is a modulation signal carried on these optical signals. Preferably, the optical signal has a high data transmission rate. Examples of the data transmission rate are 1 Gbps or more and 100 Tbps or less, 10 Gbps or more and 100 Tbps or less, or 50 Gbps or more and 50 Tbps or less. The optical signal preferably passes through an optical fiber or free space in order to support high-speed data communication.

The interference element 3 is an element for scattering optical signals and causing interference. The interference element 3 is not particularly limited as long as it can scatter optical signals and cause interference. The interfering element 3 causes the optical signals to interfere randomly, for example temporally and spatially. The interference element 3 may be an optical element or an element containing an optical element. Examples of interferometric elements 3 are multimode fibers, spatial light modulators, holographic plates, diffusers and optical integrated circuits. If the optical signal channel itself causes interference, the channel itself may act as the interfering element 3 . Interference element 3 preferably produces random interference in the optical signal. Scattering means that the optical signal travels in multiple directions. Also, random interference means that not all optical signals have the same interference.

The detector 5 is an element for measuring the optical intensity of the optical signal with interference caused by the interference element 3 . A plurality of detectors 5 receive temporal and spatial light intensity fluctuations due to the interference elements 3 . An example of the detector 5 is a high-speed photodetector (PD). The photodetector is preferably integrated. Assuming that the number of detectors 5 is Nr and the number of optical signal channels is Nt, Nr is preferably four times Nt or more, and Nr may be four times Nt+1 or more. It may be 6 times or more of Nt, and Nr may be 8 times or more of Nt. However, if the number of detectors becomes enormous, the apparatus becomes complicated, so Nr is preferably 50 times or less, 10 times or less, or 8 times or less than Nt. In this way, by performing spatially redundant observations using a direct detector array consisting of a large number of detector quintuplets, the optical phase information lost by direct detection can be recovered by means of signal processing. As a result, simultaneous coherent reception of spatially multiplexed signals can be realized without complicated optical preprocessing.

The processing circuit 9 is an element for recovering the phase information of the optical signal using the information about the light intensity measured by the detector array 7 . The processing circuitry may be implemented using only hardware for faster processing, or may be implemented using both hardware and software. The software may be a program for performing predetermined processing. A processing circuit 9 receives information about a set of optical electric field intensity information from a plurality of detectors 5, and reproduces the phase information of the optical signal in terms of signal processing, for example, by a phase retrieval technique. Methods for causing interference in signals and reproducing (recovering) the phase information of the original optical signal from a plurality of interference signals are known in this way. Therefore, the processing circuit 9 may implement a known phase recovery algorithm to recover the phase information of the optical signal. Specific phase retrieval algorithms include iterative error correction methods such as the Gerchberg-Saxton method, positive semidefinite programming methods such as PhaseLift, non-convex optimization methods such as Wirtinger flow, alternating direction method of multipliers (ADMM) and approximation. Convex optimization techniques such as message propagation (AMP).

At this time, the optical signal is discretely modulated, for example, binary (0, π), tetravalent (0, π/2, π, 3π/2). For this reason, the processing circuit 9 preferably recovers the phase information of the optical signal using factors relating to the discreteness of the optical signal. When reproducing the optical phase information, it is desirable to receive information about the modulation format of the optical signal, estimate the phase according to the modulation format, and then recover the phase information. For example, if the optical signal is a binary (0, π) phase modulated signal, the phase of the optical signal is either 0 or π. With knowledge, the accuracy and robustness of phase information recovery can be improved. In this way, by recovering the phase information of the optical signal using the elements related to the discreteness of the optical signal, the amount of signal processing can be greatly reduced, and the phase information of the optical signal can be recovered quickly.

The processing circuit 9 preferably uses the randomness component of the interference component 3 to recover the phase information of the optical signal. For the randomness of the interference element 3, the known signal is used to obtain the randomness of the interference element 3, and the obtained element is used as an estimated value.

Scattering and interference by the interference element 3 are not always constant and change. Therefore, the influence (randomness) of the interference element 3 on the optical signal changes. For this reason, using a known signal, the interference element 3 causes diffusion and interference, the detector array 7 detects a set of interference light intensities, and the processing circuit 9 recovers the phase information of the optical signal. Since the phase information of is known, the influence (randomness) of the interference element 3 can be obtained. Although the randomness of the interference element 3 may vary between when using a known signal and when using the actual signal light, it is related to the randomness of the interference element 3 obtained using the known signal. By reproducing the phase information of the signal light using the element, the phase information can be recovered more quickly.

Next, a method for detecting information including optical phase using the above system will be described.
An optical signal travels to the interference element 3 via the optical transmission line. Note that the optical transmission line itself may function as an interference element. The interference element 3 scatters the optical phase modulated signal, and the scattered optical signal causes interference (S101).
A detector array 7 including a plurality of detectors 5 measures the optical intensity of the optical signal interfered with by the interference elements 3 (S102). Specifically, the intensity of the optical signal received by each of the plurality of detectors 5 is measured. The detector array 7 outputs intensity information of the optical signals from the plurality of detectors 5 to the processing circuit 9 .
The processing circuit 9 recovers the phase information of the optical signal using the information on the multiple light intensities measured by the detector array 7 (S103). Using this recovered phase information, the information carried on the optical signal may be analyzed and output as appropriate.

が，Ｎ_ｒ個のＰＤ素子により検波されたとする。すると，環境雑音がないと仮定した場合の，時間ｔにおける第ｊ番目のＰＤの出力振幅

は，式（１）で与えられる。

ここで，

は，第ｉ番目の送信器と第ｊ番目のＰＤとの間のチャネルのインパルス応答を示す。ここで，Ｄは，チャネル遅延広がり（ｃｈａｎｎｅｌ ｄｅｌａｙ ｓｐｒｅａｄ）を示し，

である。 Theoretical Theory An example of the principle of the phase recovery method of the present invention will be described below. The invention is not limited to the following examples. FIG. 2 is a conceptual diagram showing a system example of the present invention. In Fig. 2, N _t- channel optical spatially multiplexed signals

is detected by N _r PD elements. Then, the output amplitude of the j-th PD at time t, assuming that there is no ambient noise, is

is given by equation (1).

here,

denotes the impulse response of the channel between the i-th transmitter and the j-th PD. where D denotes the channel delay spread,

is.

ここで，

であり，

は，

からなるブロックテプリッツ行列であり，

である。 For simplification, Expression (1) can be expressed as Expression (2) below by using vectors and matrices.

here,

and

teeth,

is a block Toeplitz matrix consisting of

is.

のように表記する。 Here, for simplicity,

Notation like

Coherent detection of the optical signal x is equivalent to solving the simultaneous quadratic equations of Equation (2). This kind of problem is known as a (generalized) phase retrieval problem in many fields of science and engineering, such as X-ray analysis and scattering imaging (Y. Shechtman et al., IEEE Signal Process. Mag. , 32(3), 87 (2015).

個の強度情報により

を正確に再構成するのに十分であることが指摘されている（［Ａ． Ｃｏｎｃａ ｅｔ ａｌ．， Ａｐｐｌ． Ｃｏｍｐｕｔ． Ｈａｒｍｏｎ． Ａｎａｌ．， ３８（２）， ３４６ （２０１５））。ゲルヒベルク・ザクストンやＰｈａｓｅＬｉｆｔ法（Ｋ． Ｊａｇａｎａｔｈａｎ， ｉｎ； Ｏｐｔｉｃａｌ Ｃｏｍｐｒｅｓｓｉｖｅ Ｉｍａｇｉｎｇ， ＣＲＣ， ２６３ （２０１５）．）といった，従来のＰＲアルゴリズムの多くは，計算量が大きく，高速通信システムへの応用には適さない。一方，近年，低要求演算量のアルゴリズムとして，非凸最適化手法に基づく手法が提案された（Ｅ． Ｊ． Ｃａｎｄｅｓ ｅｔ ａｌ， ＩＥＥＥ Ｔｒａｎｓ． Ｉｎｆ． Ｔｈｅｏｒｙ， ６１（４）， １９８５ （２０１５）． Ｇ． Ｗａｎｇ ｅｔ ａｌ．， ＩＥＥＥ Ｔｒａｎｓ． Ｓｉｇｎａｌ Ｐｒｏｃｅｓｓ．， ６６（１１）， ２８１８ （２０１８）．）この手法においては，式 （２）を直接解くのではなく，下記式（３）のように，損失関数を最小化することを考える。 In those fields:

according to individual strength information

([A. Conca et al., Appl. Comput. Harmon. Anal., 38(2), 346 (2015)). Many conventional PR algorithms, such as the Gerchberg-Saxton method and the PhaseLift method (K. Jaganathan, in; Optical Compressive Imaging, CRC, 263 (2015).), have a large computational complexity and are not suitable for application to high-speed communication systems. . On the other hand, in recent years, a method based on a non-convex optimization method has been proposed as an algorithm with a low computational requirement (EJ Candes et al, IEEE Trans. Inf. Theory, 61(4), 1985 (2015). G. Wang et al., IEEE Trans. Signal Process., 66(11), 2818 (2018).) In this method, instead of directly solving equation (2), as in equation (3) below, Consider minimizing the loss function.

ここで，

は，重みベクトルであり，｜｜・｜｜は，ユークリッドノルムを示し，

は， 要素ごとの積（アダマール積）を示す。式（３）で示される損失関数は，凸関数ではなく，また滑らかでもないが，ｗを適切に選ぶことによって広義の勾配を導き出すことができる（Ｅ． Ｊ． Ｃａｎｄｅｓ ｅｔ ａｌ， ＩＥＥＥ Ｔｒａｎｓ． Ｉｎｆ． Ｔｈｅｏｒｙ， ６１（４）， １９８５ （２０１５），Ｇ． Ｗａｎｇ ｅｔ ａｌ．， ＩＥＥＥ Ｔｒａｎｓ． Ｓｉｇｎａｌ Ｐｒｏｃｅｓｓ．， ６６（１１）， ２８１８ （２０１８）．）。

here,

is the weight vector, ||・|| indicates the Euclidean norm,

indicates the element-wise product (Hadamard product). Although the loss function given by equation (3) is neither convex nor smooth, a broad-sense gradient can be derived by appropriately choosing w (EJ Candes et al, IEEE Trans. Inf. Theory, 61(4), 1985 (2015), G. Wang et al., IEEE Trans. Signal Process., 66(11), 2818 (2018).).

式（４）を用いることにより，式（３）を

（ここでμ_ｎはステップサイズである。）で示される最急降下法により，効率的に最小化できる。このアルゴリズムは，Ｒｅｗｅｉｇｈｔｅｄ Ａｍｐｌｉｔｕｄｅ Ｆｌｏｗ（ＲＡＦ）として知られている． That is,

By using equation (4), equation (3) is reduced to

(where _μn is the step size) can be efficiently minimized by the steepest descent method. This algorithm is known as Reweighted Amplitude Flow (RAF).

を送信することを考える。すると式（１）及び（２）は，以下のＮ_ｒ個の並列した位相回復問題として表現できる。 In the derivation of the phase retrieval algorithm described above, it was implicitly assumed that the observation matrix H was known. However, in real systems, H also needs to be estimated from amplitude information only. K independent pilot sequences

Consider sending a Equations (1) and (2) can then be expressed as the following N _r parallel phase retrieval problems.

ここで，

である。 That is,

here,

is.

Therefore, by solving each phase retrieval problem with the known detection matrix P, the channel _hj for each detector can be estimated. Although there is phase uncertainty between h _i and h _j (i≠j), it does not affect the loss function l(x;w) used for signal detection.

On the other hand, many of the existing phase retrieval algorithms, including the RAF mentioned above, have been proposed on the premise of random observation. As a result, the observation matrix H often has a specific structure. As described above, in optical spatial multiplexing transmission, H is modeled as a Toeplitz matrix representing the response of a linear convolution channel. It is known that such a phase retrieval problem with a specific structure of the observation matrix has stricter requirements for the required number of observations and retrieval algorithms than the phase retrieval problem based on random observations. ([Q. Qu et al., arXiv: 1712.00716 (2017).). Therefore, in order to solve this problem, it is beneficial to use known information about the modulation scheme of the optical signal. Here, we propose a RAF considering the discreteness of the elements of the signal vector x by using the sum of absolute value (SOAV) optimization method (M. Nagahara, IEEE Signal Process. Lett., 22(10), 1575 (2015).).

を変調信号の組とすると，離散性を考慮した損失関数

は以下の式（５）のように表現できる。

損失関数の和

は近接分離法によって効率的に最小化することができる。ここでは，高速反復収縮しきい値アルゴリズム（ＦＩＳＴＡ）（Ａ． Ｂｅｃｋ， Ｍ． Ｔｅｂｏｕｌｌｅ， ＳＩＡＭ Ｊ． Ｉｍａｇｉｎｇ Ｓｃｉ．， ２， １８３ （２００９））によって高速化した，離散性を考慮したＲＡＦ（ＤＲＡＦ）アルゴリズムを表１に要約する。

is a set of modulated signals, the discrete loss function

can be expressed as in the following equation (5).

sum of loss functions

can be efficiently minimized by the proximity separation method. Here, a discreteness-aware RAF (DRAF) accelerated by the Fast Iterative Shrinkage Threshold Algorithm (FISTA) (A. Beck, M. Teboulle, SIAM J. Imaging Sci., 2, 183 (2009)) The algorithms are summarized in Table 1.

Here, the proximity map of g(x) is given below, for example (R. Hayakawa and K. Hayashi, IEEE Access, 6, 66499 (2018)).

重みｗ_ｎには種々の選択肢が考えられるが，以下では，簡単のため，

とする。
ここで，

である。またβ_ｎ及びρ_ｎは，それぞれ重み及びサンプルの取捨選択に関するパラメータである。

Various options can be considered for the weight w _n , but for the sake of simplicity,

and
here,

is. Also, β _n and ρ _n are parameters relating to selection of weights and samples, respectively.

As a proof of concept, phase recovery coherent detection of 10Gbaud spatially multiplexed QPSK signals was demonstrated (Fig. 3(a)). In the transmitter, a narrow line width laser with a wavelength of 1550.72 nm and a line width of 0.1 kHz was modulated in a general QPSK format by an optical IQ modulator. IQ modulator driven by 10 GSa/s Arbitrary Waveform Generator (AWG) for single-channel transmission, polarization with 10 GSa/s Pulse Pattern Generator (PPG) for 2-channel spatial multiplexing A multiplexed IQ modulator was used. The carrier wave component in the modulated signal was appropriately suppressed, and no reference light such as that used for heterodyne or Kramers-Kronig detection was used. The quality of the modulated signal was error-free. The QPSK signal was amplified to 19 dBm through an erbium-doped fiber amplifier (EDFA), such a high amplitude due to the lack of an amplifier built into the prototype 2-D PDA module. The output of the EDFA was input to a GI50 OM3 MMF at 1.6 km.Since the transmission system consisted of single-mode devices, SC connectors were used to connect to the MMF. Therefore, unlike typical SDM transmissions that use orthogonal fiber modes, the fiber modes are excited almost randomly and interfere with each other along the fiber. In other words, the transmission line itself is used as a scrambler.Therefore, the receiver consists only of a two-dimensional PD array and lenses.The PD array consists of 32 elements and has an average bandwidth of 11 GHz.Pixel The size was 30 μm×30 μm and the pitch was 44 μm (T. Umezawa et al., JLT, 36(7), 3684 (2018).).

In this experiment, only 12 out of 32 PDs were used due to equipment limitations. . To effectively increase the number of observations, we used a 20 GSa/s 12-channel digital storage oscilloscope (DSO), oversampled the PD output at twice the transmission rate, and used fractionally spaced samples to virtually divide the number of measurements by 2. It was doubled, ie, N _r =12×2=24. Signal processing was performed off-line. For DRAF, we set M = 64, D = 12 (at symbol rate), μ _n = 6, β _n = 5, ρ _n = 0.4. A weighted maximum correlation initialization method was used for initialization (G. Wang et al., IEEE Trans. Signal Process., 66(11), 2818 (2018)).

FIG. 4 is an example of a channel impulse response estimated by RAF in the case of a single channel. As shown in FIG. 4, mode coupling and dispersion in MMF cause random interference between modes. Conventional coherent reception methods require a complex mode diversity configuration to remove this interference, while phase recovery coherent receivers utilize this distortion to recover the optical phase.

5 and 6 show the bit error rate (BER) characteristics versus the number of DRAF iterations N for different _Nr , respectively. BER is the average of 200 trials with different fiber conditions. For a given _Nr , the _Nr channels are selected in descending order of standard deviation and phase recovery is performed.

For the single transmission channel case (Fig ^. 5), the BER below the FEC (Forward Error Correction) threshold with 7% redundancy, _i . achieved for. When using 20%-FEC (BER threshold 2.0×10 ⁻² ), error-free transmission was possible for N _r >10 (5 PD) and N>400. On the other hand, when the conventional RAF, that is, the phase retrieval method that does not consider the discreteness, is used, even if all 12 PDs (N _r = 24) are used, error-free transmission cannot be achieved. This indicates that the phase retrieval problem following convolutional observations is difficult. It should be noted that the BER characteristics shown in the present embodiment are obtained after performing a phase noise (PN) elimination process based on the quadratic method. In a phase recovery coherent receiver, the frequency drift of the local light on the transmitting side is estimated and compensated for as part of the channel response _hij . On the other hand, the residual phase noise component is reconstructed by phase recovery as shown in the inset of FIG. 5, and can be removed by post-processing as in the conventional coherent receiver. This suggests the possibility of realizing a phase recovery coherent receiver without using an extremely narrow linewidth laser as in this experiment. Finally, in the case of two-channel spatial multiplexing transmission (FIG. 6), it can be seen that the BER characteristic is degraded due to the correlation between channels. Mode coupling in the MMF is weak and cannot be said to be sufficiently random interference. Also, oversampling itself causes channel correlation. However, 20%-FEC error-free operation could be achieved by using 24 channels (ie, 12 PDs) with N>400.

The system of the present invention can be used, for example, as a low-cost coherent receiver in short-distance, ultra-large-capacity optical space multiplexing transmission in data centers and access systems, and in optical wireless communication and optical sensing. Thus, the present invention can be used, for example, in the information and communications industry.

REFERENCE SIGNS LIST 1 system 3 interference element 5 detector 7 detector array 9 processing circuit

Claims (5)

A system (1) for receiving an optical signal and detecting information including optical phase, comprising:
an interfering element (3) for scattering the optical signal to obtain scattered light and causing random interference between the optical signal and the scattered light ;
a detector array (7) comprising a plurality of detectors (5) for measuring the optical intensity of said optical signals interfered by said interfering elements (3);
and processing circuitry (9) for recovering phase information of said optical signal using information about the light intensity measured by said detector array (7).
The system of claim 1, wherein
A system, wherein Nr is the number of the plurality of detectors (5) and Nt is the number of channels of the optical signal, Nr being four or more times Nt. The system of claim 1, wherein
A system according to claim 1, wherein said processing circuit (9) recovers phase information of said optical signal using a factor relating to the discreteness of said optical signal. The system of claim 1, wherein
The processing circuit (9) uses the known signal to estimate the randomness factor of the interference factor (3), and uses the estimated randomness factor of the interference factor (3) to estimate the randomness of the optical signal. A system for recovering phase information. A method for receiving an optical signal and detecting information including optical phase, comprising:
an interfering element (3) scattering the optical signal to obtain scattered light and causing random interference between the optical signal and the scattered light ;
a detector array (7) comprising a plurality of detectors (5) measuring the optical intensity of said optical signals interfered by said interfering elements (3);
a processing circuit (9) using information about the light intensity measured by the detector array (7) to recover phase information of the optical signal.

Images