Large dynamic underwater wireless optical signal detection method based on state counting

Abstract

The invention discloses a large dynamic underwater wireless optical signal detection method based on state counting, and belongs to the technical field of underwater wireless optical communication and signal detection. The method comprises the following steps of: firstly, performing M-time up-sampling on a received signal, and setting a judgment threshold to count photon state count in each symbol period; secondly, a truncated sub-Poisson distribution model is provided according to the truncation characteristic of the symbol 1 under the strong light condition, the probability quality function of the truncated sub-Poisson distribution model can accurately describe the state number distribution rule, and Poisson distribution modeling is adopted under the weak light condition; and finally, estimating distribution parameters of symbols' 0 'and' 1 'according to the pilot signal, and performing symbol judgment based on a maximum likelihood detection method. The method has the advantages of being simple in structure and low in calculation complexity, the detection performance and robustness of the system are remarkably improved, and the method is suitable for complex underwater wireless optical communication scenes.

Description

Technical Field

[0001] The present invention belongs to the technical field of underwater wireless optical communication and signal detection, and in particular relates to a large dynamic underwater wireless optical signal detection method based on state counting. Background Art

[0002] With the rapid development of marine resource development, environmental monitoring, and military applications, the importance of underwater wireless communication technology has become increasingly prominent. Currently, underwater wireless communication relies primarily on three methods: underwater acoustic communication, underwater radio frequency communication, and underwater wireless optical communication (UWOC). While underwater acoustic communication offers long transmission distances, it suffers from limited bandwidth, high latency, and susceptibility to multipath interference. Underwater radio frequency communication, however, suffers from the severe attenuation of electromagnetic waves in water, making it suitable only for very short-distance transmission. In contrast, underwater wireless optical communication leverages the low attenuation of the blue-green light band in water to achieve high-speed, low-latency communication, making it an ideal choice for short- to medium-range underwater communications.

[0003] However, underwater wireless optical communication systems still face many challenges in practical applications, particularly robust signal detection at the receiving end. Due to the complexity of the underwater environment, optical signals are affected by factors such as water absorption, scattering, turbulence, and dynamic occlusion, resulting in wide fluctuations in the received optical power. Photomultiplier tubes (PMTs), commonly used photon-counting detectors, have output characteristics that vary significantly with light intensity. These characteristics are typically divided into three operating ranges: the pulse region (outputting sparse pulses under weak light conditions), the transition region (pulses gradually become denser under medium light intensities), and the waveform region (outputting a continuous waveform under strong light conditions). The statistical characteristics of the signals corresponding to these different ranges vary significantly, making it difficult for traditional detection methods to adapt to these dynamic changes.

[0004] In low-light conditions, the photon count per symbol period typically follows a Poisson distribution, making the traditional Poisson Maximum Likelihood Detection (PMLD) method more applicable. In strong-light conditions, the signal amplitude tends to be continuous, so the Gaussian Maximum Likelihood Detection (GMLD) method is typically used. However, when light intensity changes dynamically during communication, a single detection model cannot accurately match the statistical characteristics of the signal, resulting in degraded detection performance. For example, in transition and waveform regions, the Poisson distribution will produce large fitting errors due to signal saturation effects, while the Gaussian distribution has difficulty accurately describing the discrete characteristics under low-light conditions. Furthermore, existing methods lack an adaptive switching mechanism for varying light intensity conditions, limiting the system's reliability in complex underwater environments.

[0005] Therefore, there is an urgent need for a unified signal detection method that can adapt to large dynamic light intensity changes to improve the robustness and detection accuracy of underwater wireless optical communication systems in complex environments. Summary of the Invention

[0006] To solve the above technical problems, the present invention provides a large-dynamic underwater wireless optical signal detection method based on state counting. It characterizes the signal characteristics under strong light conditions by truncating the sub-Poisson distribution, and combines it with the Poisson distribution to achieve accurate detection under weak light conditions, thereby maintaining excellent performance when the light intensity fluctuates over a wide range.

[0007] To achieve the above object, the present invention adopts the following technical solutions:

[0008] A large dynamic underwater wireless optical signal detection method based on state counting, the method comprising:

[0009] Step 1: upsample the received signal by a factor of M, and perform state determination on each sampling point based on a set determination threshold, counting the number of photon states in each symbol period;

[0010] Step 2: The number of photon states corresponding to the symbol "1" under strong light conditions is modeled using a truncated sub-Poisson distribution, the number of photon states corresponding to the symbol "1" under weak light conditions is modeled using a Poisson distribution, and the number of photon states corresponding to the symbol "0" is uniformly modeled using a Poisson distribution to construct a binary hypothesis testing model;

[0011] Step 3: Use the pilot signal to estimate the state number distribution parameters of symbol "0" and symbol "1" respectively, adaptively select the truncated sub-Poisson distribution or Poisson distribution model according to the light intensity conditions, and use the maximum likelihood detection method to complete symbol decision.

[0012] The beneficial effects of the present invention are:

[0013] Unified signal modeling framework: By statistically counting photon states within a symbol period and combining truncated sub-Poisson and Poisson distributions, accurate modeling of signal characteristics under varying light intensities is achieved. This approach avoids the limitations of traditional single-distribution models and is adaptive to low-light, transitional, and high-light environments, significantly improving detection accuracy.

[0014] Enhanced adaptability to strong light: To address the truncation effect caused by signal saturation under strong light conditions, an innovative truncated sub-Poisson distribution model is proposed. Its probability mass function can accurately characterize the distribution law of the state number, solving the problem of large fitting error of the traditional Poisson distribution at high light intensity, enabling the system to maintain stable performance over a wide dynamic range.

[0015] Low complexity and high robustness: The symbol decision method based on maximum likelihood detection is computationally simple and easy to implement. It also uses pilot signals to estimate distribution parameters in real time, ensuring the algorithm's adaptability to dynamic light intensity changes. Experiments have shown that this method can reduce the optical power requirement by approximately 3 dB compared to traditional methods while maintaining the same bit error rate, significantly improving system energy efficiency and reliability.

[0016] Wide applicability: Suitable for complex underwater environments, such as those with turbulence, obstruction, or dynamic light intensity fluctuations, it provides a high-performance, low-complexity signal detection solution for underwater wireless optical communication systems. BRIEF DESCRIPTION OF THE DRAWINGS

[0017] Figure 1 This is a flow chart of the large dynamic underwater wireless optical signal detection method based on state counting of the present invention;

[0018] Figure 2 This is a block diagram of the underwater wireless optical communication system of the present invention;

[0019] Figure 3 The distribution histogram and fitting comparison diagram of the state number under different received optical power conditions;

[0020] Figure 4 This is the curve showing the system bit error rate changes with the received optical power. DETAILED DESCRIPTION

[0021] The present invention will be further described below with reference to the accompanying drawings and examples.

[0022] like Figure 1 As shown, the main purpose of the present invention is to provide a large dynamic underwater wireless optical signal detection method based on state counting. The detection method is simple in form, has low complexity, and improves the robustness of the underwater wireless optical communication system. The method includes the following steps:

[0023] Step 1: upsample the received signal by a factor of M, and perform state determination on each sampling point based on a set determination threshold, counting the number of photon states in each symbol period;

[0024] Step 2: The number of photon states corresponding to the symbol "1" under strong light conditions is modeled using a truncated sub-Poisson distribution, the number of photon states corresponding to the symbol "1" under weak light conditions is modeled using a Poisson distribution, and the number of photon states corresponding to the symbol "0" is uniformly modeled using a Poisson distribution, and a binary hypothesis testing model is constructed to achieve symbol judgment;

[0025] Step 3: Use the pilot signal to estimate the state number distribution parameters of symbol "0" and symbol "1" respectively, adaptively select the truncated sub-Poisson distribution or Poisson distribution model according to the light intensity conditions, and use the maximum likelihood detection method to complete symbol decision.

[0026] Furthermore, the step 1 includes:

[0027] Step 1.1: Upsample the analog signal at the receiving end by M times to obtain a sampling point sequence. ,in, Indicates the symbol time Amplitude of sampling points;

[0028] Step 1.2: Set the threshold for determining the “photon entering” state , will symbol time The state of each sampling point is expressed as:

[0029] ;

[0030] Step 1.3, define The number of photon states within the symbol duration is .

[0031] Furthermore, the truncated sub-Poisson distribution and its characteristics in step 2 include:

[0032] The state number corresponding to the symbol "1" obeys the parameter The probability mass function of the truncated sub-Poisson distribution is:

[0033] ,

[0034] in, is the normalization coefficient, is the number of photon states within the symbol period.

[0035] When the parameters meet Conditions, where the truncation condition , we can get the following two approximate conclusions: 1) The normalization coefficient satisfies ; 2) The expectation and variance of the distribution are approximately: Under this condition, , indicating that the truncated distribution has sub-Poisson distribution characteristics.

[0036] Furthermore, the binary hypothesis testing model in step 2 includes:

[0037] Construct a binary hypothesis testing model to realize symbol judgment, assuming that the number of states corresponding to the symbol "0" obeys the parameter Poisson distribution; the symbol "1" corresponds to the number of states according to the parameter Whether it meets the preset range Modeling: When When , it is considered to be under strong light conditions, and the model is parameter The truncated sub-Poisson distribution is otherwise considered to be in low light conditions and is modeled as parameter Poisson distribution. Based on this, the following hypothesis testing model is established:

[0038] ,

[0039] in, Indicates the Send symbols, and Respectively represent the assumption that the current symbol is symbol "1" and symbol "0", Indicates known sending symbols Conditions, No. The number of observed states within a transmission symbol period is probability.

[0040] Furthermore, the step 3 includes:

[0041] Step 3.1: According to the known positions of the symbol "0" and the symbol "1" in the pilot sequence, the corresponding state number sets are counted, and the corresponding estimated distribution parameters are , , for:

[0042] , ,

[0043] in, and Respectively represent the position index sets of symbol "1" and symbol "0" in the pilot sequence.

[0044] Step 3.2: When the estimated parameter of symbol “1” satisfies , using the maximum likelihood test based on truncated sub-Poisson distribution, construct the The expression of the log-likelihood ratio (LLR) of a symbol is:

[0045] ,

[0046] And according to Make symbol decisions.

[0047] when , using the conventional Poisson distribution for maximum likelihood detection, and according to Perform symbol decision, where The LLR expression of the symbol is:

[0048] .

[0049] Step 3.3: Complete the final symbol determination based on the calculated log-likelihood ratio result and output the detection result.

[0050] like Figure 2The communication block diagram shown in the figure shows a transceiver placed in a 30m underwater environment. The transmitter uses a blue laser, and the receiver uses a PMT. An attenuator is placed before the transmitter to adjust the received signal optical power, and a filter is placed before the receiver to filter out noise. To simulate the dynamic underwater environment, a bubble generator is introduced into the underwater channel to increase channel dynamics and evaluate the robustness of the algorithm. System performance is tested in this environment.

[0051] Figure 3 The statistical histogram of the number of states within a symbol period at different received signal optical powers is presented and compared with the fitted curves of the Poisson distribution and the truncated sub-Poisson distribution. Intervals I, II, and III represent the PMT operating in the pulse, transition, and waveform regions, respectively. As can be seen from the figure, when the PMT operates in the pulse or transition regions, the number of states can be fitted by the Poisson distribution. However, when the PMT transitions to the waveform region, the truncation effect occurs, and the Poisson distribution fitting error increases significantly. The truncated sub-Poisson distribution proposed in this invention provides a more accurate fit.

[0052] Figure 4 The system bit error rates of the detection method of the present invention (the proposed method) and traditional detection methods (Poisson and Gaussian maximum likelihood detection) at different received optical powers are demonstrated. The results show that the bit error rates of both the proposed method and Gaussian maximum likelihood detection show a downward trend with increasing optical power; however, Poisson maximum likelihood detection experiences an increase in bit error rate at high optical power, demonstrating a certain degree of inadaptability. Furthermore, across the entire optical power range, the proposed method outperforms traditional methods in bit error rate performance, achieving at least a 3 dB optical power gain. Specifically, when the BER is 10⁻³, the proposed method and Gaussian and Poisson maximum likelihood detection require optical powers of -46dBm, -43dBm, and -38dBm, respectively. In summary, the detection method proposed in this invention improves the performance and robustness of the system under complex underwater optical channel conditions while balancing bit error rate performance and implementation complexity, making it suitable for underwater wireless optical communication systems in dynamic environments.

[0053] The specific embodiments described above further illustrate the objectives, technical solutions and beneficial effects of the present invention in detail. It should be understood that the above are only specific embodiments of the present invention and are not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the present invention should be included in the scope of protection of the present invention.

Claims

1. A large dynamic underwater wireless optical signal detection method based on state counting, characterized in that: The method comprises: Step 1: upsample the received signal by a factor of M, and perform state determination on each sampling point based on a set determination threshold, counting the number of photon states in each symbol period; Step 2: The number of photon states corresponding to the symbol "1" under strong light conditions is modeled using a truncated sub-Poisson distribution, the number of photon states corresponding to the symbol "1" under weak light conditions is modeled using a Poisson distribution, and the number of photon states corresponding to the symbol "0" is uniformly modeled using a Poisson distribution to construct a binary hypothesis testing model; Step 3: Use the pilot signal to estimate the state number distribution parameters of symbol "0" and symbol "1" respectively, adaptively select the truncated sub-Poisson distribution or Poisson distribution model according to the light intensity conditions, and use the maximum likelihood detection method to complete symbol decision.

2. The large dynamic underwater wireless optical signal detection method based on state counting according to claim 1 is characterized in that: The step 1 comprises: Step 1.1: Upsample the analog signal at the receiving end by M times to obtain a sampling point sequence. ,in Indicates the symbol time Amplitude of sampling points; Step 1.2: Set the threshold for the "photon entering" state , will symbol time The state of each sampling point is expressed as: ； Step 1.3, define The number of photon states within the symbol duration is .

3. The large dynamic underwater wireless optical signal detection method based on state counting according to claim 2 is characterized in that: The truncated sub-Poisson distribution and its characteristics in step 2 include: The number of states corresponding to the symbol "1" obeys the parameter The probability mass function of the truncated sub-Poisson distribution is: ， in, is the normalization coefficient, Count the number of photon states in a symbol period; Among them, when the parameters meet , where the cutoff condition , we get: The normalization coefficient satisfies ; The expectation and variance are approximately: , under this condition, satisfying , indicating that the truncated distribution has sub-Poisson distribution characteristics.

4. The large dynamic underwater wireless optical signal detection method based on state counting according to claim 3 is characterized in that: The step 2 includes: Assume that the symbol "0" corresponds to the state number that follows the parameter Poisson distribution; the symbol "1" corresponds to the number of states according to the parameter Whether it meets the preset range Modeling: When When , it is considered to be under strong light conditions, and the model is parameter The truncated sub-Poisson distribution is otherwise considered to be in low light conditions and is modeled as parameter Poisson distribution.

5. The large dynamic underwater wireless optical signal detection method based on state counting according to claim 4 is characterized in that: The binary hypothesis testing model in step 2 is specifically: ， in, Indicates the Send symbols, and Respectively represent the assumption that the current symbol is symbol "1" and symbol "0", Indicates known sending symbols Conditions, No. The number of observed states within a transmission symbol period is probability.

6. The large dynamic underwater wireless optical signal detection method based on state counting according to claim 1 is characterized in that: The step 3 comprises: Step 3.1: Using the known positions of the symbol "1" and symbol "0" in the pilot sequence, count the corresponding state number sets and estimate their respective distribution parameters; Step 3.2: When the estimated parameters of the symbol "1" meet the truncation condition, the maximum likelihood detection method based on the truncated sub-Poisson distribution is used to calculate the log-likelihood ratio and make a symbol decision. If the truncation condition is not met, the maximum likelihood detection method based on the conventional Poisson distribution is used for decision making. Step 3.3: Complete the final symbol determination based on the calculated log-likelihood ratio result and output the detection result.

7. The large dynamic underwater wireless optical signal detection method based on state counting according to claim 6 is characterized in that: The step 3.1 includes: according to the known positions of the symbol "0" and the symbol "1" in the pilot sequence, respectively counting the corresponding state number sets, which correspond to the estimated distribution parameters , , for: ， ， in, and Respectively represent the position index sets of symbol "1" and symbol "0" in the pilot sequence.

8. The large dynamic underwater wireless optical signal detection method based on state counting according to claim 6 is characterized in that: The step 3.2 includes: when the estimated parameter of the symbol "1" satisfies , using the maximum likelihood test based on truncated sub-Poisson distribution, construct the The log-likelihood ratio of the symbols is expressed as: ， And according to Make symbol decisions; when , using the conventional Poisson distribution for maximum likelihood detection, and according to Perform symbol decision, where The LLR expression of the symbol is: 。

Images