阅读说明：本技术 基于spad阵列非线性的sar检测方法 (SAR detection method based on SPAD array nonlinearity ) 是由 王昭诚 武燕敏 金爽 权进国 董宇涵 陈飞 吴海全 安俊 张颢 白勃 于 2019-10-31 设计创作，主要内容包括：本发明公开了一种基于SPAD阵列非线性的SAR检测方法，用于一光通信系统的接收端，包括门限计算步骤和符号判决步骤；所述门限计算步骤包括：根据平均光功率以及发送端星座点集合在接收端光子计数的期望集合，采用排序算法实时计算门限；所述符号判决步骤包括：根据所述门限，对接收的符号进行判决。本发明在SPAD阵列为弱光通信条件下的检测方法时，考虑由于死时间带来的非线性影响，采用排序的方法，使得提出的SAR检测方法的整体误码率性能好于一般的AR检测方法；同时，通过优化单调递减区内SLEA区域，使得检测方法的工作范围得以扩大。(The invention discloses an SAR detection method based on SPAD array nonlinearity, which is used for a receiving end of an optical communication system and comprises a threshold calculation step and a symbol judgment step; the threshold calculation step comprises: calculating a threshold in real time by adopting a sorting algorithm according to the average optical power and an expected set of photon counting of a constellation point set at a sending end at a receiving end; the symbol decision step includes: and judging the received symbol according to the threshold. When the SPAD array is a detection method under the condition of weak light communication, the invention considers the nonlinear influence caused by dead time and adopts a sequencing method, so that the overall error rate performance of the provided SAR detection method is better than that of a common AR detection method; meanwhile, the working range of the detection method is expanded by optimizing the SLEA area in the monotone decreasing area.)

1. A SAR detection method based on SPAD array nonlinearity is used for a receiving end of an optical communication system and is characterized by comprising a threshold calculation step and a symbol judgment step;

the threshold calculation step comprises: calculating a threshold in real time by adopting a sorting algorithm according to the average optical power and an expected set of photon counting of a constellation point set at a sending end at a receiving end;

the symbol decision step includes: and judging the received symbol according to the threshold.

2. The SAR detection method based on SPAD array nonlinearity of claim 1, wherein the step of calculating the threshold in real time with a sorting algorithm according to the average optical power and the expected set of the sending end constellation point set photon counts at the receiving end comprises:

s11, pair of expected setsPerforming AR transformation on each element to obtain an AR transformation set

s12, sorting the elements in the AR transformation set from small to large to obtain a sorted set

S13, sequentially averaging two adjacent elements in the sorted set

3. The SPAD array nonlinearity based SAR detection method of claim 2 wherein the set of constellation points

4. The method for SAR detection based on the non-linearity of the SPAD array of claim 2, wherein the step of deciding the received symbols based on the threshold comprises:

s21, carrying out AR transformation on the photon count z corresponding to the received symbol S to obtain z_AR；

S22, mixing z_ARAnd a threshold g_jAnd (3) comparison:

if z is_AR≤g₁The decision result of the symbol se₁I.e. as an index set

if g is_k＜z_AR≤g_k+1K 1, 2.. multidot.j-1, the result of the decision of the symbol s is obtained

If z is_AR＞g_M-1The decision result of the symbol s

Technical Field

The invention relates to the field of communication, in particular to an SAR detection method based on SPAD array nonlinearity.

Background

Visible light communication is a communication technology that can compensate for conventional wireless communication due to advantages such as safety to the human body, no electromagnetic interference, and a wide freely available spectrum. In visible light communication, an LED is generally used as an emitter, and a Photodiode (PD) is used as a detector. Due to the nature of light, visible light communications typically use light intensity modulation and direct detection (IM/DD) techniques, and therefore the transmitted signal is a non-negative real number. Generally, PD detectors can be divided into PIN PD and avalanche PD (apd). However, these conventional PD detectors do not have sufficient sensitivity to low light and do not receive signals well under low light conditions.

To solve such a problem, a Single-Photon avalanche diode (SPAD) detector is applied to this field. SPAD is an APD operating in geiger mode, which outputs a large current when a photon arrives, and can therefore be used as a photon counter. When a photon is received and counted, the SPAD will go to an inactive state and start charging to be able to detect the next photon. The time of charging is called dead time (dead time). The charging circuits are classified into an active quenching circuit (AQ) and a Passive quenching circuit (PQ). In PQ SPAD, the dead time is extended when a photon arrives within the dead time; whereas in AQSPAD the dead time is not changed. Dead time limits photon counting over time and thus causes non-linear distortion.

SPAD arrays can improve the ability to count photons compared to a single SPAD. SPAD arrays have some fundamental parametric features, the ratio C of the active area to the total area_FFCan be used to represent the probability that a beam of light, when arriving, is directed directly at the activation region; when a photon reaches the activation region, it has a certain probability of not being able to excite an avalanche and thus being able to be detected, this probability being denoted by C_PDPAnd (4) showing. Dark count rate N_DCRShowing an avalanche caused by thermal motion inside the device when no photons arrive per unit time, it can be seen as a noise. P_APIndicating the probability that additional avalanches are triggered after receiving an avalanche caused by a photon or thermal motion. Based on the above characteristics, the count of SPADs can be approximately represented by a poisson random variable. For slave constellation point sets when the effect of dead time is not considered

In which an arbitrarily selected transmitted symbol s is assumed to have a duration T_SThen the duration in the symbol period is T_ST(T_S≥T_ST) The expectation of photon counts of (a) can be expressed as:

wherein x is the average optical power, N_SPADIs the number of SPADs in the array,

is the energy of a single photon, h is the Planckian constant, c_LIs the speed of light, λ_LIs the wavelength of the light. The expectation of photon counting, taking into account the effect of dead time, is:

wherein, tau_dIs the dead time.

According to the above, when the average optical power x is smaller, the expectation of different constellation points monotonically increases with the constellation point, and the region is referred to as a monotonically increasing region; when the average optical power x increases to a certain extent, the expected curves of the different constellation points will intersect and the expectation becomes no longer monotonically increasing, this region being called the intersection region, approximately represented by the region between the maximum values expected for the first and last symbols. After the intersection region, the situation that the expectation of different constellation points monotonically decreases with the constellation points, that is, the monotonically decreasing region, may occur. Finally, the SPAD will quickly reach saturation due to the large optical power, and thus almost overlap will eventually occur.

In addition, the error performance of the detection method in the intersection region can also be analyzed. As the optical power approaches the optical power at the intersection point, the bit error rate increases as the spacing between the expected curves decreases; when the cross point is reached, the error rate reaches a maximum value; as the optical power moves away from the crossover point, the bit error rate decreases as the spacing between the desired curves becomes larger. Therefore, the error performance of the intersection region fluctuates. In the monotone decreasing area, because no cross condition exists, a better error rate can be achieved by designing a detection method. In the last almost overlapping region, the error rate increases with decreasing pitch.

In conjunction with the above equation, where the transmitted symbol is s, the conditional probability density of the photon count z is:

where n is a natural number representing the value of photon count z.

The AR (anscombe root) transformation can convert variables subject to the poisson process into variables subject to variance 1 and gaussian distribution, and the transformed mean value can also be approximated to AR transformation of the mean value, so that the detection problem under the gaussian channel can be converted, and the complexity is greatly reduced by adopting a hard decision method. However, the conventional AR detection method does not consider the non-linear factor caused by the dead time, and thus cannot be applied to all cases when the SPAD array is used as the detection method.

The above background disclosure is only for the purpose of assisting understanding of the inventive concept and technical solutions of the present invention, and does not necessarily belong to the prior art of the present patent application, and should not be used for evaluating the novelty and inventive step of the present application in the case that there is no clear evidence that the above content is disclosed before the filing date of the present patent application.

Disclosure of Invention

The invention mainly aims to overcome the defects of the prior art and provides an SAR detection method based on SPAD array nonlinearity.

A SAR detection method based on SPAD array nonlinearity is used for a receiving end of an optical communication system and comprises a threshold calculation step and a symbol judgment step; the threshold calculation step comprises: calculating a threshold in real time by adopting a sorting algorithm according to the average optical power and an expected set of photon counting of a constellation point set at a sending end at a receiving end; the symbol decision step includes: and judging the received symbol according to the threshold.

Further:

the step of calculating the threshold in real time by adopting a sorting algorithm according to the average optical power and the expected set of the photon counting of the constellation point set at the receiving end at the transmitting end comprises the following steps:

s11, pair of expected sets

Performing AR transformation on each element to obtain an AR transformation set

Wherein the content of the first and second substances,

is a set of constellation points

A desired set at an average optical power; i 1, 2., M represents the number of constellation points; lambda [ alpha ]_iWhen the transmitted symbol is a constellation point s_iPhoton count z of SPAD array_iThe expected value of (d);

denotes λ_iAR transformation of (1);

s12, sorting the elements in the AR transformation set from small to large to obtain a sorted set

And collecting each element

Corresponding original index to obtain index set

S13, sequentially averaging two adjacent elements in the sorted set

Obtaining a set of thresholds

Where j is 1, 2.

Set of constellation points

As a PAM constellation point set, or as an expressionA set of constellation points; and the parameter d is obtained by maximizing the SLEA area by adopting a traversal method according to different data rates and different modulation orders. In the intersection region, constellation points can be designedThe expected monotone decreasing area is enlarged, and the purpose of enlarging the working area is further achieved.

The step of deciding the received symbol according to the threshold comprises:

s21, carrying out AR transformation on the photon count z corresponding to the received symbol S to obtain z_AR；

S22, mixing z_ARAnd a threshold g_jAnd (3) comparison:

if z is_AR≤g₁The decision result of the symbol s

e₁I.e. as an index set1 st element in (1);

if g is_k＜z_AR≤g_k+1K 1, 2.. multidot.j-1, the result of the decision of the symbol s is obtained

If z is_AR＞g_M-1The decision result of the symbol s

When the SPAD array is a detection method under the condition of weak light communication, the invention considers the nonlinear influence caused by dead time and adopts a sequencing method, so that the overall error rate performance of the provided SAR detection method is better than that of a common AR detection method; meanwhile, the working range of the detection method is expanded by optimizing the SLEA area in the monotone decreasing area.

Drawings

FIG. 1 shows that conventional 8PAM is at T_STExpected curves for different transmitted symbols at 1 μ s;

FIG. 2 is a block diagram of a conventional PAM combined with a weak optical communication system of the SAR detection method of the present invention;

FIG. 3 is a block diagram of a weak optical communication system in which Geo-PAM is combined with the SAR detection method of the present invention;

FIG. 4 shows that the SAR detection method and the conventional AR detection method of the present invention are performed at T when conventional 8PAM is used_STComparing the error code performance at 1 mu s;

FIG. 5 shows the modulation order at T of 3_STWhen 1 mu s is adopted, Geo-PAM and the traditional PAM are respectively adopted, the error code performance is compared when the SAR detection method is adopted;

FIG. 6 shows the modulation order at T of 3_STWhen 1ms is adopted, Geo-PAM and the traditional PAM are respectively adopted, the error code performance is compared when the SAR detection method is adopted;

FIG. 7 is a graph at T_STWhen the modulation order is 1 mu s, the modulation order is 3, and when the modulation order is 4, the error code performance of the SAR detection method and the SAR detection method is compared with that of a system adopting the traditional PAM and AR detection methods;

FIG. 8 is a graph at T_STWhen the modulation order is 1ms, the modulation order is 3, and when the modulation order is 4, the error code performance of the system is compared by adopting a Geo-PAM and SAR detection method and a traditional PAM and AR detection method.

Detailed Description

The invention is further described with reference to the following figures and detailed description of embodiments.

The specific implementation mode of the invention provides an SAR detection method based on SPAD array nonlinearity aiming at an optical communication system, in particular to a weak optical communication system, and symbol detection is carried out at a receiving end. The method is realized based on a traditional AR (Antscombe root) detection method, and is called an SAR (SAR) (sortingAntscombe root) detection method because a Sorting algorithm and AR transformation are adopted when a threshold is calculated. The SAR detection method of the invention comprises a threshold calculation step and a symbol decision step, wherein the threshold calculation step comprises the following steps: calculating a threshold in real time by adopting a sorting algorithm according to the average optical power and an expected set of photon counting of a constellation point set at a sending end at a receiving end; the symbol decision step includes: and judging the received symbol according to the threshold.

The step of calculating the threshold in real time by adopting a sorting algorithm according to the average optical power and the expected set of the photon counting of the constellation point set at the receiving end at the transmitting end comprises the following steps of S11-S13:

s11, pair of expected sets

Performing AR transformation on each element to obtain an AR transformation setWherein the content of the first and second substances,

is a set of constellation points

A desired set at an average optical power; i 1, 2., M represents the number of constellation points; lambda [ alpha ]_iWhen the transmitted symbol is a constellation point s_iPhoton count z of SPAD array_iThe expected value of (d);

denotes λ_iAR transform of (3).

Wherein, the constellation point set

May be a conventional PAM constellation point set (e.g. 8 PAM).

The expression can also be adopted as

The constellation point set of (a) can be called as a geometric pulse amplitude modulation (Geo-PAM) constellation point because the form is similar to the geometric series. The region for which the constellation points are primarily directed is the desired non-monotonic increasing region caused by SPAD array non-linearity. And the parameter d is obtained by maximizing the SLEA area by adopting a traversal method according to different data rates and different modulation orders. When working area optimization is carried out by using Geo-PAM constellation points, the error rate in a monotone decreasing area is less than 10^-3Range of (1)As an optimization target, this region is called a stable Low Error rate Area (SLEA), that is, the SLEA region is expanded as much as possible, so that the region of Error performance fluctuation is reduced, the relatively stable region is increased, and the working range is expanded. The optimization method adopts a traversal method, and the upper bound of the optimization method can be calculated by adopting the traditional M-PAM method, and is as follows:

optimized optimal d^★Expressed as:

s12, sorting the elements in the AR transformation set from small to large to obtain a sorted set

e_i∈[1,M]. It is understood that in the collection

In (1), in addition, each element was collected

Corresponding original index to obtain index set

S13, sequentially averaging two adjacent elements in the sorted set

Obtaining a set of thresholds

Where j is 1, 2.

The decision of the received symbol according to the calculated threshold specifically includes the following steps S21 to S22:

s21, carrying out AR transformation on the photon count z corresponding to the received symbol S to obtain z_AR；

S22, mixing z_ARAnd a threshold g_jAnd (3) comparison:

if z is_AR≤g₁The decision result of the symbol s

e₁I.e. as an index set

1 st element in (1);

if g is_k＜z_AR≤g_k+1K 1, 2.. multidot.j-1, the result of the decision of the symbol s is obtained

If z is_AR＞g_M-1The decision result of the symbol s

An exemplary application of the SAR detection method provided by the present invention is shown in fig. 2, which is a block diagram of a weak optical communication system combining the conventional PAM with the SAR detection method of the present invention. For conventional PAM, the constellation point set is set when the modulation order is 3The resulting expected curve is plotted against the light irradiance as shown in fig. 1, and it can be seen that the curve for each symbol tends to rise first and then fall, with a maximum. The maximum value of the light irradiance is p_iThen can obtain：

Said monotonically increasing region is less than p_MThe region of intersection is [ p ]_M,p₁]Length D ═ p₁-p_M(ii) a The non-monotonic increasing region is larger than p₁The area of (a).

As can also be seen from fig. 1, in the illustrated monotone increasing region, it is desirable to increase as the value of the symbol increases when the light irradiance is fixed, and in the non-monotone increasing region, there are an intersection region and a monotone decreasing region.

The performance of the system shown in fig. 2 when transmitting conventional 8PAM is shown in fig. 4. From the simulation result, the performance of the receiver of the SAR detection method is the same as that of the AR detection method in a monotone increasing area, and the integral error rate of the SAR detection method is not more than that of the AR detection method in a non-monotone increasing area, so that the integral performance is better. It can also be seen, however, that a portion of the SAR performance curve fluctuates due to the aforementioned intersection region.

Fig. 3 is a block diagram of a weak optical communication system using Geo-PAM using SPAD array, using SAR detection method. For Geo-PAM, the constellation points are

At an accuracy of 0.01, respectively at T_ST1 μ s and T_STThe results of the optimization parameters for modulation orders 3 and 4 at 1ms are given in the following table:

FIGS. 5 and 6 are comparisons of conventional PAM and Geo-PAM. Since the AR detection method hardly works normally in a non-monotonic region, the SAR detection method is adopted. Taking the modulation order as 3 as an example, it can be seen from the figure that at T_ST1 μ s and T_STGeo-PAM provides additional operating regions of about 2.5dB and 5.5dB, respectively, at 1ms。

Fig. 7 and 8 are final performance representations of two systems of the conventional PAM-bound AR detection method and the Geo-PAM-bound SAR detection method. As can be seen from fig. 7 and 8, the system performance of the Geo-PAM combined SAR detection method has a significant advantage in a non-monotonic region. At T_STFor modulation orders of 3 and 4, providing additional operating regions of about 6.4dB and 5.2dB, respectively, at 1 mus; at T_STFor modulation orders of 3 and 4 at 1ms, additional operating regions of about 11.5dB and 10.5dB are provided, respectively.

It can be seen that the SAR detection method provided by the specific embodiment of the present invention can provide the following advantages for the weak optical communication system receiver using the SPAD array: 1. aiming at the nonlinear influence caused by the SPAD dead time, the threshold is calculated by adopting a sequencing method, so that the overall error rate performance of the method is better than that of a common AR detection method; 2. the working range of the detection method is expanded by the optimization of the constellation design aiming at the non-monotone increasing region.

The foregoing is a more detailed description of the invention in connection with specific preferred embodiments and it is not intended that the invention be limited to these specific details. For those skilled in the art to which the invention pertains, several equivalent substitutions or obvious modifications can be made without departing from the spirit of the invention, and all the properties or uses are considered to be within the scope of the invention.