A weighted cooperative frequency sensing method for NGSO-GSO downlink communication links

Abstract

The present invention discloses a weighted collaborative frequency sensing method for NGSO-GSO downlink communication links. The method comprises: using multiple ground stations as sensing nodes to collaboratively sense downlink transmission signals from authorized satellites, thereby constructing a ground station collaborative spectrum sensing downlink scenario; setting the downlink authorized satellite to be a GSO satellite, establishing a sensing signal model, and determining the sensing signal distribution characteristics; and establishing a time-invariant channel link optimization fusion decision strategy based on the spatial position of the authorized satellite, the sensing signal distribution characteristics, and the number of sensing ground stations. Using the objective function and constraints of the correct detection probability, the subgradient method of Lagrangian duality theory is used to jointly optimize and solve the decision weight allocation and detection threshold of each sensing ground station. Compared with traditional energy detection methods, the method of the present invention has a higher detection probability when the signal-to-noise ratio is lower than 15dB.

Description

Technical Field

[0001] The present invention belongs to the technical field of satellite communication spectrum sharing, and in particular relates to a weighted collaborative frequency sensing method for an NGSO-GSO downlink communication link. Background Art

[0002] In recent years, the field of satellite communications has been faced with the current situation of complex satellite orbit configurations, large numbers, and the use of multiple systems and multiple frequency bands. The international use of frequency resources is based on the principle of "first come, first served", and the use of frequencies is constrained by systems that started earlier. In order to alleviate the increasingly tense trend of spectrum resources, solve spectrum congestion, and improve the utilization efficiency of spectrum resources, it is urgent to explore effective spectrum sharing technologies. By utilizing spectrum sensing technology, the spectrum "hole" information of authorized satellite systems can be accurately detected, so that different satellite systems and different services can share the same frequency band, thereby realizing the effective utilization of space spectrum resources.

[0003] Spectrum sensing technology detects and uses idle spectrum within the authorized or primary node (PN) authorized frequency band through sensing or secondary node (SN), and quickly gives up the channel when the authorized node starts to use the frequency band to avoid co-channel interference to the authorized node.

[0004] A lot of research has been done on spectrum sensing technology at home and abroad. WUSW, ZHU JK, QIU L, et al. proposed a cooperative spectrum sensing technology based on optimal signal-to-noise ratio weighting in "SNR-based weighted cooperative spectrum sensing in cognitive radio networks". YE HY, JIANG J B. proposed a cooperative spectrum sensing technology based on optimal signal-to-noise ratio weighting in "Optimal linear weighted cooperative spectrum sensing for clustered-based cognitive radio networks". Liu Yulei, Liang Jun, Xiao Nan, et al. proposed a spectrum sensing method based on channel historical state information in "Spectrum sensing method based on channel historical state perception information" (Journal of Communications, 2017, 38(8):118-130). In addition, some scholars in this field have proposed a high-spectrum-efficiency joint optimization algorithm based on the establishment of a multi-node collaborative cognitive network model based on cooperative spectrum sensing. Some scholars have proposed a cooperative spectrum sensing method based on secondary user power control assistance to improve the performance of the communication system. In the field of satellite communication systems, existing technologies determine the optimal available spectrum by analyzing the effectiveness and availability of idle spectrum; some scholars have studied the spectrum sharing scenarios and cognitive networks of geostationary orbit and low-Earth orbit systems; some scholars have used the polarization state of the received signal to propose the optimal polarization joint technology for spectrum sensing to improve sensing efficiency; and some scholars have proposed a spectrum sensing strategy based on maximum a posteriori probability in the coexistence scenario of geostationary orbit and non-geostationary orbit satellites.

[0005] Currently, the application of spectrum sensing technology in satellite system spectrum sharing scenarios is mainly studied around geostationary satellite orbit (GSO) satellites, with ground stations as sensing nodes. In the simulation analysis of collaborative sensing scenarios, the impact of channel environment differences between different sensing ground stations on the perception effect is rarely considered. Summary of the Invention

[0006] The present invention aims to overcome the defects of the prior art and proposes a weighted cooperative frequency sensing method for NGSO-GSO downlink communication links. NGSO refers to Non-Geo-Stationary Orbit.

[0007] To achieve the above objectives, the present invention proposes a weighted collaborative frequency sensing method for NGSO-GSO downlink communication links, the method comprising:

[0008] Step 1) Multiple ground stations are used as sensing nodes to collaboratively sense downlink transmission signals from authorized satellites, thereby building a ground station collaborative spectrum sensing downlink scenario.

[0009] Step 2) Setting the downlink authorized satellite to a GSO satellite, establishing a perception signal model, and determining the perception signal distribution characteristics;

[0010] Step 3) Based on the spatial position of the authorized satellite, the distribution characteristics of the sensing signal, and the number of sensing ground stations, a time-invariant channel link optimization fusion decision strategy is established. Through the objective function and constraints of the correct detection probability, the subgradient method of Lagrangian duality theory is used to jointly optimize the decision weight distribution and detection threshold of each sensing ground station.

[0011] As an improvement to the above method, step 1) specifically includes:

[0012] The signal power C received by the i-th sensing node from the downlink signal transmitted by the authorized satellite di and signal-to-noise ratio γ di They are:

[0013]

[0014] Where p pn is the transmission power of the authorized node, in W, G pnt (θ pnt,i ) is the transmitting antenna gain of the authorized node, θ pnt,i G is the off-axis angle of the authorized node's transmitting antenna on the sensing link corresponding to the i-th sensing node; snr (θ snr,i ) is the receiving antenna gain of the sensing node, θ snr,i is the receiving antenna off-axis angle of the i-th sensing node; k B is the Boltzmann constant, which is 1.38×10 -23 J / K; T is the equivalent noise temperature of the receiving end of the sensing node, in K; W is the bandwidth of the sensing communication link, in Hz; l i is the satellite-to-ground link loss of the i-th sensing node, satisfying the following formula:

[0015]

[0016] Where, f is the communication frequency, in Hz; d sn,i→pn is the distance of the sensing link corresponding to the i-th sensing node, in meters; A r,i is the attenuation value caused by rainfall in the satellite-to-ground link, in dB; A c,i is the attenuation value caused by clouds and fog in the satellite-to-ground link, in dB.

[0017] As an improvement to the above method, the perception signal model of step 2) is:

[0018] For downlink authorized satellites that are GSO satellites, the established sensing link is a time-invariant channel link. The k-th time slot signal x received by the authorized node at the i-th sensing node is i,k The model is:

[0019]

[0020] Where, The idle state of the spectrum of the authorized node received by the i-th sensing node; The spectrum occupancy status of the authorized node received by the i-th sensing node; is the sampling signal of the kth time slot transmitted by the authorized node, which obeys Gaussian distribution; i is the channel phase; n i,k is the k-th time slot additive Gaussian white noise sampling signal received by the i-th sensing node, that is, Taking into account With {n i,k} are independent of each other, then Under the condition x i,k It also obeys the Gaussian distribution.

[0021] As an improvement to the above method, the step 2) of determining the distribution characteristics of the perceived signal specifically includes:

[0022] Detection statistic R using energy detection method i Can be approximated as a Gaussian distribution;

[0023] The false detection probability P of the i-th sensor node e,i and the correct detection probability P c,i They are:

[0024]

[0025]

[0026] Where, P d,i and P f,i are the conditional detection probability and false alarm probability respectively; and are the probabilities of spectrum idle and occupied states, respectively.

[0027] As an improvement to the above method, the time-invariant channel link optimization fusion decision strategy in step 3 is jointly calculated based on the optimal weight and threshold, with the global error detection probability P g,e As the objective function to be optimized:

[0028]

[0029] The false alarm probability P when ensuring cooperative perception g,f and spectrum conflict probability P g,sc Under the condition that no limit is exceeded, the nonlinear constraint conditions are constructed as follows:

[0030]

[0031] Where, P f,th , P sc,th are the false alarm probability limit and spectrum conflict probability limit respectively, w i is the optimal weight of the i-th sensor node, and M is the total number of sensor nodes.

[0032] As an improvement to the above method, step 3) uses the subgradient method of Lagrangian duality theory to jointly optimize the decision weight distribution and detection threshold of each sensing ground station; specifically, it includes:

[0033] Construct the Lagrangian function J1(w,ξ g )for:

[0034]

[0035] Where, κ d,i =(κ d,1 ,...,κ d,M )≥0 is the Lagrangian factor of J1(·), w=[w1,w2,...,w i ,...,w M ] T represents the optimal weight matrix, the superscript T represents the transpose, κ a ,κ b ,κ c is the Lagrangian factor of J1(·); ξ g Indicates the threshold value;

[0036] Let J1(·) be i and ξ g The partial derivative function value is 0, and the optimal weight value is obtained by Newton iteration method. and the optimal threshold value

[0037] Compared with the prior art, the advantages of the present invention are:

[0038] Aiming at the downlink sensing scenario of satellite systems, the present invention proposes a weighted collaborative spectrum sensing method for ground stations based on an established sensing signal model. This method takes into account the influence of factors such as the ground station's sensing channel model and radio wave propagation model, as well as the differences in the sensing channel environment between ground stations. By establishing an objective function and constraints for the correct detection probability, parameters such as the decision weight allocation and detection threshold of each sensing ground station are jointly optimized, while ensuring that the spectrum conflict of the authorized system does not exceed the limit. In a simulation analysis of a downlink spectrum sensing scenario constructed using actual satellite data registered by the International Telecommunication Union (ITU), the influence of spectrum occupancy status and the number of sensing signal samples on the sensing effect is studied. Under the same signal-to-noise ratio (SNR) condition, the correct perception results of different sensing methods are compared and analyzed. Compared with the traditional energy detection (ED) method, the proposed method has a higher detection probability when the SNR is lower than -15dB. BRIEF DESCRIPTION OF THE DRAWINGS

[0039] Figure 1 It is a downlink multi-ground station collaborative perception scenario

[0040] Figure 2 It is a schematic diagram of spectrum conflicts between M sensing nodes and authorized nodes;

[0041] Figure 3(a) is the characteristic curve of the conditional detection probability of different sampling numbers as the signal-to-noise ratio changes (P g,f ≤0.1);

[0042] Figure 3(b) is the characteristic curve of the spectrum conflict probability with different sampling numbers and the change of signal-to-noise ratio (P g,f ≤0.1);

[0043] Figure 3(c) shows the characteristic curve of the global correct detection probability with different sampling numbers as the signal-to-noise ratio changes (P g,f ≤0.1). DETAILED DESCRIPTION

[0044] The technical solution of the present invention is described in detail below with reference to the accompanying drawings and embodiments.

[0045] Example

[0046] To solve the spectrum sharing problem between satellite systems, an embodiment of the present invention proposes a weighted collaborative frequency sensing method for NGSO-GSO downlink communication links, the method comprising:

[0047] Step 1) Multiple ground stations are used as sensing nodes to collaboratively sense downlink transmission signals from authorized satellites, thereby building a ground station collaborative spectrum sensing downlink scenario.

[0048] Step 2) Setting the downlink authorized satellite to a GSO satellite, establishing a perception signal model, and determining the perception signal distribution characteristics;

[0049] Step 3) Based on the spatial position of the authorized satellite, the distribution characteristics of the sensing signal, and the number of sensing ground stations, a time-invariant channel link optimization fusion decision strategy is established. Through the objective function and constraints of the correct detection probability, the subgradient method of Lagrangian duality theory is used to jointly optimize the decision weight distribution and detection threshold of each sensing ground station.

[0050] The following is a detailed analysis:

[0051] 1. Construction of ground station collaborative spectrum sensing downlink scenario

[0052] Considering the positional relationship between the sensing ground station and the space orbit of the authorized satellite system, the downlink ground station collaborative sensing scenario is constructed as follows: Figure 1 As shown in FIG, the sensing scenario is to use multiple ground stations as sensing nodes to collaboratively sense the downlink transmission signals from the authorized satellite.

[0053] Figure 1 In the example, the signal power C of the downlink signal received by the i-th sensing node from the authorized satellite is di and signal-to-noise ratio γ di They are

[0054]

[0055] Where p pn is the transmission power of the authorized node, W; G pnt (θ pnt,i ) is the transmitting antenna gain of the authorized node, θ pnt,i G is the off-axis angle of the authorized node's transmitting antenna on the i-th sensing link; snr (θ snr,i ) is the receiving antenna gain of the sensing node, θ snr,i is the receiving antenna off-axis angle of the i-th sensing node; k B is the Boltzmann constant, which is 1.38×10 -23 J / K; T is the equivalent noise temperature of the receiving end of the sensing node, K; W is the bandwidth of the sensing communication link, Hz; l i is the satellite-to-ground link loss of the i-th sensing node, Figure 1 The specific form is

[0056]

[0057] Ignoring line loss and antenna pointing error loss, l i In addition to spatial loss, the attenuation caused by rainfall and fog must also be considered. In formula (2), f is the communication frequency, Hz; d sn,i→pn is the distance of the i-th sensing link, m; A r,i is the attenuation value caused by rainfall in the satellite-to-ground link, dB; A c,i is the attenuation value caused by clouds and fog in the satellite-to-ground link, dB.

[0058] 2 Perception Signal Model and Distribution Characteristics

[0059] 2.1 Perception Signal Model

[0060] like Figure 1 If the downlink authorized satellite in is a GSO satellite, then the established sensing link is a time-invariant channel link. Under this condition, the k-th time slot signal x received by the i-th sensing node is the k-th time slot signal of the authorized node. i,k The model is

[0061]

[0062] Among them, K i =2U i ,k=1,2,...,K i .

[0063] Substitute into formula (1,2), that is

[0064]

[0065] Where, The idle state of the spectrum of the authorized node; is the spectrum occupancy status of the authorized node; K i is the number of samples in the i-th sensing cycle, U i is the delay-bandwidth product; is the sampling signal of the kth time slot transmitted by the authorized node, which obeys Gaussian distribution; C i is the power of the signal received by the i-th sensing node; φ i is the channel phase; n i,k is the k-th time slot additive Gaussian white noise sampling signal received by the i-th sensing node, that is, Taking into account With {n i,k} are independent of each other, then Under the condition x i,k It also obeys the Gaussian distribution.

[0066] 2.2 Perception Signal Distribution Characteristics

[0067] Detection statistic R using energy detection methodi (x i,k )for

[0068]

[0069] Where R i Follows chi-square distribution Its distribution characteristics are

[0070]

[0071] Considering that sufficient sampling number is required to ensure the accuracy of perception results in low signal-to-noise ratio conditions, i >>1 Under the premise, using the central limit theorem (CLT), R i It can be approximated as a Gaussian distribution, that is

[0072]

[0073] According to formula (6,7), the i-th sensor node in K i >>1 Conditional detection probability of the ED algorithm using CLT approximation and false alarm probability for

[0074]

[0075]

[0076] Where Q(·) is the complementary cumulative distribution function.

[0077] The false detection probability P of the i-th sensor node e,i and the correct detection probability P c,i They are

[0078]

[0079]

[0080] Where, P d,i and P f,i are the conditional detection probabilities ( Conditions) and false alarm probability ( condition); and are the probabilities of spectrum idle and occupied states, respectively.

[0081] 3 Basic principles and solution methods of ground station collaborative perception fusion decision strategy

[0082] Based on the research on the perception signal model in Section 2, the weighted threshold decision algorithm corresponding to different perception nodes is designed according to the spatial position of the authorized satellite, the perception channel characteristics and the number of perception ground stations to perform fusion decision of each perception node.

[0083] 3.1 Basic principles of OFD-TIC strategy

[0084] Taking the ground station cooperative sensing downlink scenario in Section 2.1 as an example, the detection statistic R of the signals received by different ground stations as sensing nodes is i (x i,k ), which is affected by factors such as its geographical location, the elevation angle of the constructed sensing link and the link loss. Therefore, the statistical weights calculated by different sensing nodes are considered in the fusion decision.

[0085] Considering that the authorized node is a GSO satellite, assuming that the number of samples of each sensing node is K i Same, that is, K i = K. Considering the relative stability of the link established by the GSO satellite, the weighted detection statistic R of the M ground station nodes for cooperative sensing is g Defined as

[0086]

[0087] Where w is the weight vector, w i ≥0. w and R are

[0088]

[0089] From formula (5), we can get R g The distribution characteristics of

[0090]

[0091] Where μ and υ are

[0092]

[0093] Using CLT to obtain the above conditional global detection probability P g,d and false alarm probability P g,f for

[0094]

[0095]

[0096] The threshold ξ of collaborative sensing can be obtained from formula (16,17): g and judgment criteria,

[0097]

[0098] For the i-th node, the link loss l i The smaller it is, the greater the SNR is, and the detection probability P d,i Using the model of formula (12), in the ground station cooperative sensing downlink scenario, when selecting the ground station, it is necessary to consider whether the sensing link it builds meets the minimum elevation angle requirement.

[19] , considering the receiving antenna beam characteristics of the sensing node, link propagation loss and other factors, in order to improve the accuracy of the collaborative sensing results, an optimal weight w=[w1,w2,...,w M ] T Assignment and threshold ξ g The joint calculation method is called the optimal fusion decision for time-invariant channel link (OFD-TIC) strategy. This strategy uses the global error detection probability P g,e As the objective function to be optimized, its form is

[0099]

[0100] When using multiple ground stations in the constellation system to collaboratively sense GSO satellites, taking into account the round-trip delay of electromagnetic wave propagation in the satellite-to-ground link and the "protection" status of GSO satellites by the International Telecommunication Union (ITU), when further studying the constraints corresponding to the objective function of Equation (19), the proposed OFD-TIC strategy needs to consider not only the false alarm probability P of the collaborative sensing node, but also the false alarm probability P of the cooperative sensing node. g,f In addition to the limit conditions met, the corresponding limit conditions of spectrum conflict probability also need to be studied.

[0101] Figure 2 It reflects the spectrum conflict between M sensing nodes and authorized nodes. is the duration of the spectrum idle state of the authorized node; t d,i is the duration of data transmission of the i-th sensor node; t h,i is the spectrum idle time caused by the i-th sensing node not being used; t delay,i is the delay difference between detection and spectrum utilization of the i-th sensing node.

[0102] When studying the spectrum conflict model during spectrum sensing, the idle state duration X before the authorized node spectrum state transition is idle Obeying the exponential distribution with parameter λ, the maximum likelihood estimate of λ for in, For X idleThe sample mean of .

[0103] This section assumes that when the spectrum state of the authorized node changes, the sensing node will immediately make a detection response, that is, t delay,i =t h,i For the i-th sensing node, the effective probability P of idle spectrum is eff,i and the availability probability P ava,i Defined as [9]

[0104]

[0105]

[0106] The spectrum conflict probability P between the i-th sensing node and the authorized node at a certain moment is sc,i for

[0107]

[0108] Substituting equation (20,21) into (22), we can get

[0109]

[0110] Spectrum conflict probability P based on OFD-TIC strategy g,sc Can be defined as

[0111]

[0112] Where, and are the average effective probability and average available probability of idle spectrum respectively.

[0113] In summary, this section ensures P g,f and P g,sc Under the condition that both are not exceeded, the nonlinear constraint condition corresponding to formula (19) is constructed as follows:

[0114]

[0115] 3.2 OFD-TIC Strategy Solution Method

[0116] Equations (19, 25) can be solved by the subgradient method using Lagrange duality theory. The specific solution process is described as follows.

[0117] Construct the Lagrangian function J1(w,ξ) of Eq. (19) g )for

[0118]

[0119] Where, κ d =(κd,1 ,...,κ d,M )≥0,w=[w1,w2,...,w M ] T κ a ,κ b ,κ c is the Lagrange factor of J1(·). Let J1(·) be the Lagrange factor of w i and ξ g The partial derivative function value is 0, and the optimal value can be obtained by Newton iteration method and The solution algorithm is shown in Table 1.

[0120] Table 1 OFD-TIC strategy solution algorithm

[0121]

[0122] 4 Simulation Implementation and Verification

[0123] 4.1 OFD-TIC Strategy Simulation Experiment

[0124] The orbital parameters and beam air interface parameters of the authorized satellite refer to the downlink KADWB beam and air interface parameters of the CHNNEWSAT-G1-118E satellite registered with the ITU. The antenna beam is pointed to its sub-satellite point as shown in Table 2. The beam and air interface parameters of the sensing ground station are shown in Table 3.

[0125] Table 2 Authorized system satellite downlink orbit, beam and air interface parameters

[0126]

[0127] Table 3 Perception ground station downlink beam and air interface parameters

[0128]

[0129] The sensing ground stations were deployed in the range of 25°N to 35°N latitude and 110°E to 120°E longitude according to the Monte Carlo method, and the minimum elevation angle requirement was met. The random deployment was then used for experimental verification and analysis. Figure 3(a) shows the characteristic curve of the conditional detection probability (P) with different sampling numbers as a function of the signal-to-noise ratio. g,f ≤0.1); Figure 3(b) is the characteristic curve of the spectrum conflict probability with different sampling numbers as the signal-to-noise ratio changes (P g,f ≤0.1); Figure 3(c) shows the characteristic curve of the global correct detection probability with different sampling numbers as the signal-to-noise ratio changes (P g,f ≤0.1).

[0130] Figure 3(c) shows how the global correct detection probability varies with the average signal-to-noise ratio when the OFD-TIC strategy is adopted, where K = 5000 and the spectrum conflict probability limit P sc,th The average effective probability limit and the average available probability limit of idle spectrum are 0.8. Considering that the IEEE 802.22 protocol requires the false alarm probability P of spectrum sensing terminal to be g,f ≤0.1, so the false alarm probability limit P f,th Take 0.1.

[0131] In Figure 3(c), under the condition of K = 5000 samples, the OFD-TIC strategy is adopted. When the average signal-to-noise ratio (SNR) received by the sensing ground station is ≥ -14dB, the correct detection probability gradually approaches 1. At the same time, this strategy can also alleviate the impact of differences in the perceived channel environment on the correct detection probability.

[0132] 5 Conclusion

[0133] For downlink ground station perception scenarios, the present invention:

[0134] 1) A multi-ground station downlink weighted cooperative spectrum sensing method is proposed. This method takes into account the downlink sensing channel environment, the influence of the radio wave propagation model, and the spectrum conflict with the authorized satellite system.

[0135] 2) Based on the signal model of a time-invariant channel, we designed the corresponding objective function and constraints, and proposed the OFD-TIC strategy for collaborative ground station sensing. We jointly optimized the parameters such as the weights and detection thresholds assigned by different sensing ground stations, and solved the algorithm.

[0136] 3) In the simulation analysis, the correct perception effect of OFD-TIC strategy was quantitatively analyzed and verified. The simulation results show that when the number of samples of the perception signal K ≥ 1000 and SNR ≥ -12dB, the correct detection probability P of OFD-TIC strategy is g,c and conditional detection probability P g,d gradually tend to 1.

[0137] Finally, it should be noted that the above embodiments are intended only to illustrate the technical solutions of the present invention and are not intended to limit the present invention. Although the present invention has been described in detail with reference to the embodiments, it should be understood by those skilled in the art that modifications or equivalent substitutions to the technical solutions of the present invention do not depart from the spirit and scope of the technical solutions of the present invention and are intended to be encompassed by the claims of the present invention.

Claims

1. A method for weighted collaborative frequency sensing of an NGSO-GSO downlink communication link, the method comprising: Step 1) Multiple ground stations are used as sensing nodes to collaboratively sense downlink transmission signals from authorized satellites, thereby building a ground station collaborative spectrum sensing downlink scenario. Step 2) Setting the downlink authorized satellite to a GSO satellite, establishing a perception signal model, and determining the perception signal distribution characteristics; Step 3) Based on the spatial position of the authorized satellite, the distribution characteristics of the sensing signal, and the number of sensing ground stations, a time-invariant channel link optimization fusion decision strategy is established. Based on the objective function and constraints of the correct detection probability, the subgradient method of Lagrangian duality theory is used to jointly optimize the decision weight allocation and detection threshold of each sensing ground station. The step 3) time-invariant channel link optimization fusion decision strategy is based on the joint calculation of the optimal weight and threshold, with the global error detection probability P g,e As the objective function to be optimized: Among them, P g,d is the global detection probability, P g,f is the false alarm probability, and are the probabilities of spectrum idle and occupied states respectively; Q(·) is the complementary cumulative distribution function; w=[w1,w2,...,w i ,...,w M ] T represents the optimal weight matrix, the superscript T represents the transpose, ξ g represents the threshold; μ and υ are: The false alarm probability P when ensuring cooperative perception g,f and spectrum conflict probability P g,sc Under the condition that no limit is exceeded, the nonlinear constraint conditions are constructed as follows: Where, P f,th , P sc,th are the false alarm probability limit and spectrum conflict probability limit respectively, w i is the optimal weight of the i-th sensor node, M is the total number of sensor nodes; The step 3) uses the subgradient method of Lagrange duality theory to jointly optimize the decision weight distribution and detection threshold of each sensing ground station; specifically includes: Construct the Lagrangian function J1(w,ξ g )for: Where, κ d,i =(κ d,1 ,...,κ d,M )≥0 is the Lagrangian factor of J1(·), κ a ,κ b ,κ c is the Lagrangian factor of J1(·); Let J1(·) be i and ξ g The partial derivative function value is 0, and the optimal weight value is obtained by Newton iteration method. and the optimal threshold value .

2. The NGSO-GSO downlink communication link weighted collaborative frequency sensing method according to claim 1, characterized in that: The step 1) specifically includes: The signal power C received by the i-th sensing node from the downlink signal transmitted by the authorized satellite di and signal-to-noise ratio γ di They are: Where p pn is the transmission power of the authorized node, in W, G pnt (θ pnt,i ) is the transmitting antenna gain of the authorized node, θ pnt,i G is the off-axis angle of the authorized node's transmitting antenna on the sensing link corresponding to the i-th sensing node; snr (θ snr,i ) is the receiving antenna gain of the sensing node, θ snr,i is the receiving antenna off-axis angle of the i-th sensing node; k B is the Boltzmann constant, which is 1.38×10 -23 J / K; T is the equivalent noise temperature of the receiving end of the sensing node, in K; W is the bandwidth of the sensing communication link, in Hz; l i is the satellite-to-ground link loss of the i-th sensing node, satisfying the following formula: Where, f is the communication frequency, in Hz; d sn,i→pn is the distance of the sensing link corresponding to the i-th sensing node, in meters; A r,i is the attenuation value caused by rainfall in the satellite-to-ground link, in dB; A c,i is the attenuation value caused by clouds and fog in the satellite-to-ground link, in dB.

3. The NGSO-GSO downlink communication link weighted collaborative frequency sensing method according to claim 2, characterized in that: The perception signal model of step 2) is: For downlink authorized satellites that are GSO satellites, the established sensing link is a time-invariant channel link. The k-th time slot signal x received by the authorized node at the i-th sensing node is i,k The model is: Where, The idle state of the spectrum of the authorized node received by the i-th sensing node; The spectrum occupancy status of the authorized node received by the i-th sensing node; is the sampling signal of the kth time slot transmitted by the authorized node, which obeys Gaussian distribution; i is the channel phase; n i,k is the k-th time slot additive Gaussian white noise sampling signal received by the i-th sensing node, that is, Taking into account With {n i,k } are independent of each other, then Under the condition x i,k It also obeys the Gaussian distribution.

4. The NGSO-GSO downlink communication link weighted collaborative frequency sensing method according to claim 3, characterized in that: The step 2) of determining the distribution characteristics of the perceived signal specifically includes: Detection statistic R using energy detection method i Can be approximated as a Gaussian distribution; The false detection probability P of the i-th sensor node e,i and the correct detection probability P c,i They are: Where, P d,i and P f,i are the conditional detection probability and false alarm probability respectively; and are the probabilities of spectrum idle and occupied states, respectively.

Images