Differential time hopping estimation and multi-hop coherent combination method and device for satellite communication

Abstract

The invention discloses a differential time hopping estimation and multi-hop coherent combination method and device for satellite communication, and belongs to the field of digital signal processing. A sending end repeatedly sends an information symbol in a plurality of discontinuous time slots, and both communication parties appoint that the time interval of each time of sending changes randomly within a predetermined range. After a time hopping signal is transmitted through an AWGN channel, a receiving end performs joint estimation on a time interval of two adjacent hops and a multi-hop carrier phase by using a cross entropy iteration method. Establishing a mapping relationship between the hopping time interval and the carrier phase as well as the binary quantization bit, taking the combined signal-to-noise ratio loss as a target function, iteratively updating the binary quantization bit generation probability, regenerating the estimated values of the hopping time interval and the multi-hop carrier phase, and performing multi-hop coherent combination to obtain the binary quantization bit. And selecting the parameter estimation value with the minimum combined signal-to-noise ratio loss as a global optimal vector group. And realizing multi-hop coherent combination according to the global optimal vector group to obtain a signal-to-noise ratio gain. Compared with a traditional traversal method, the complexity can be reduced.

Description

Technical Field

[0001] The invention relates to a cross entropy iteration-assisted differential time hopping estimation and coherent combining method and device, belonging to the field of digital signal processing. Background Art

[0002] In the satellite communication scenario, time hopping technology is used to avoid interference from non-cooperative parties. The time hopping communication system selects the signal transmission time discontinuously based on the time hopping pattern jump, and can effectively avoid interference signals through time division. For the jammer, continuous transmission of interference signals can produce a strong interference effect on the system, increasing the interference cost. The transmitter disperses the signal in different time slots and transmits it repeatedly. In order to further improve the randomness of the signal in time, the time interval of each transmission varies randomly within a predetermined range. The receiver needs to estimate the transmission time interval to make full use of all signals scattered in each time slot to complete the merger. In addition, for the receiver, the carrier phase of each hop signal is random. In order to obtain a higher signal-to-noise ratio gain, the receiver performs coherent combination and demodulation of multi-hop signals in different time slots, which accurately estimates and compensates the carrier phase of each hop signal. Therefore, in order to improve the reliability of communication, the receiver needs to jointly estimate the time interval of each transmission and the carrier phase of each hop signal. The higher parameter estimation dimension leads to increased complexity.

[0003] As a global random optimization algorithm, cross entropy is widely used in multi-dimensional, large-scale, and nonlinear optimization problems. The optimization problem is converted into a probability estimation problem of a low-probability event, that is, the occurrence of the optimal solution is a low-probability event. An adaptive key sampling strategy is adopted to select samples with better effects according to the objective function to update the probability distribution. New samples are generated using the updated probability distribution, and the optimal solution is more likely to occur through iteration. Applying cross entropy to the joint estimation of multi-hop carrier phase and time-hopping interval can reduce the complexity of parameter estimation. Based on this, a cross-entropy iteration-assisted differential time-hopping estimation and multi-hop coherent merging method is proposed. Summary of the invention

[0004] In view of the problem of loss of combined signal-to-noise ratio caused by the deviation of time-hopping interval and carrier phase estimation in the time-hopping communication system, the purpose of the present invention is to provide a differential time-hopping estimation and multi-hop coherent combining method and device for satellite communication, which adaptively adjusts the probability distribution of the parameters to be estimated during the cross-entropy iteration process to obtain the time-hopping interval and multi-hop carrier phase estimation that make the combined signal-to-noise ratio loss smaller, thereby realizing the synchronization of the time-hopping interval and the multi-hop carrier phase; on this basis, the signal-to-noise ratio gain is obtained by coherent combining of multi-hop signals, and the complexity is reduced compared with the traditional ergodic method.

[0005] The objective of the present invention is achieved through the following technical solutions:

[0006] The invention discloses a differential time hopping estimation and multi-hop coherent merging method for satellite communication, in which the transmitting end repeatedly sends information symbols in multiple discontinuous time slots, and the communicating parties agree that the time interval of each transmission varies randomly within a predetermined range. After the time hopping signal is transmitted through the AWGN channel, the receiving end uses the cross entropy iteration method to jointly estimate the time interval between two adjacent hopping signals and the multi-hop carrier phase. A mapping relationship between the time hopping interval and the carrier phase and the binary quantization bit is established, and the combined signal-to-noise ratio loss is used as the objective function. The binary quantization bit generation probability is iteratively updated, and the estimated values ​​of the time hopping interval and the multi-hop carrier phase are regenerated, and multi-hop coherent merging is performed, and the parameter estimation value with the smallest combined signal-to-noise ratio loss is selected as the global optimal vector group. According to the global optimal vector group, multi-hop coherent merging is implemented to obtain a signal-to-noise ratio gain. Compared with the traditional traversal method, the present invention can reduce complexity.

[0007] The present invention discloses a method for estimating the time hopping interval and combining multiple hops coherently for satellite communication, comprising the following steps:

[0008] Step 1: The information symbol to be sent is modulated by BPSK and then repeatedly transmitted in multiple time slots that are discontinuous in time. Each time slot is called a hop. The time interval between two adjacent hop signals varies randomly within a predetermined range to generate a BPSK modulated time-hopping signal with a random hopping interval.

[0009] The BPSK modulated time-hopping signal with random time-hopping interval is generated as shown below:

[0010]

[0011] The signal is dispersed and sent repeatedly in M ​​hops, with K symbols sent in each hop. k represents the kth BPSK modulation symbol, b k ∈{-1,+1}, w(t) is a single pulse waveform, T s is the duration of a single symbol, L m is the number of time slots before the mth hop, T p is the duration of a single time slot, f0 is the carrier frequency of the transmitted signal, φ 0,m It is the initial phase of the carrier of the mth hop transmitted signal.

[0012] Let L δ,m represents the time hopping interval, that is, the number of time slots between the mth hop and the m-1th hop. Agreement between the two parties Varies randomly within a predetermined range.

[0013] Step 2: Send the BPSK modulated time-hopping signal obtained in step 1 into the AWGN channel, and receive the signal transmitted through the AWGN channel.

[0014] The received signal after passing through the AWGN channel is expressed as

[0015]

[0016] Where A is the amplitude of the transmitted signal when it reaches the receiver, τ is the delay of the receiving end relative to the transmitting end, and f d is the frequency deviation of the received signal relative to the transmitted signal, φ m is the initial phase of the carrier of the m-th hop received signal, n(t) represents additive complex Gaussian white noise, with a mean of 0 and a variance of Receiver delay τ and frequency offset f d Ideally synchronized.

[0017] Step 3: The search range of the initial phase of the carrier of each hop signal of the received signal is [0,2π). The two parties agree that the search range of the number of time slots between two adjacent hop signals is {1,2,…,L δ，max}, the mth hop carrier initial phase and time hopping interval search parameters are in A mapping relationship is established between the search value and the binary code. D1 and D2 are binary quantization bit numbers. Each carrier phase search value corresponds to a D1bit binary number, and each time hopping interval search value corresponds to a D2bit binary number, thereby realizing parameter quantization of the time hopping interval and the carrier phase of each hopping signal.

[0018] φ m,d represents the initial phase estimate of the dth carrier of the mth hop, d∈{1,2,...,N1}, and

[0019]

[0020] Represent d-1 as a binary number of D1 bits, and establish φ from the above formula m,d The mapping relationship with the D1bit binary number.

[0021] L δ,m,u represents the estimated value of the u-th hop time interval of the m-th hop, u∈{1, 2, ..., N2}, and N2=L δ,max ,have

[0022] L δ,m,u =u (4)

[0023] Represent u-1 as a D2-bit binary number, and establish L δ,m,u The mapping relationship with D2bit binary number.

[0024] Step 4: Generate N according to the preset initial probability of binary quantization bits c A group of binary codes, called candidate groups, consists of N cThe binary code mapping is used to obtain the M hop carrier phase and the N hop time interval estimation. c The initial solution of the group is generated, and the parameter solution is regenerated by updating the binary quantization bit generation probability during the cross entropy iteration process.

[0025] The probability of generating binary quantized bits for the i-th cross entropy iteration is By probability Randomly generate binary quantized bits Indicates generation The probability of taking the value 1. The set of carrier phase quantization bits for the mth hop is in represents the vbit of the carrier phase quantization bit of the mth hop in the i-th iteration, A group of hop interval quantization bits for the mth hop in represents the lth bit of the carrier phase quantization bit of the mth hop in the i-th iteration,

[0026] Depend on The carrier phase estimation obtained by mapping is

[0027]

[0028] Depend on The estimated time hopping interval obtained by mapping is

[0029]

[0030] According to formula (5) (6), Mapping obtains M hop carrier phase and time hopping interval estimation values

[0031] Step 5: Generate a local pulse template based on the estimated time hopping interval obtained in step 4, and perform correlation operation with the received signal, that is, match the local template with the time slot position of the received signal, and compensate the signal carrier phase of the selected time slot based on the carrier phase estimate, and coherently combine the compensated multi-hop signals.

[0032] The hop time interval of each hop is estimated by Get the number of time slots before the mth hop

[0033]

[0034] The receiving end Generate a local pulse template and perform correlation calculation and phase compensation on the kth data of all M jumps

[0035]

[0036] The signal after combining the k-th data of all M hops is further expressed as

[0037]

[0038] Wherein P≤M, P is the number of time slots where the useful signal in the M-hop data is located that matches the local pulse template.

[0039] Step 6: Estimate the signal-to-noise ratio of the combined signal obtained in step 5, calculate the difference between the estimated actual signal-to-noise ratio and the M-hop theoretical combined signal-to-noise ratio, and obtain the N generated in step 4. c The combined signal-to-noise ratio loss corresponding to the group carrier phase and time-hopping interval parameter solution is used as the objective function of the cross-entropy iteration.

[0040] If the time-hopping interval is estimated accurately, but Thus, P=M in formula (9). If the carrier phase estimation is accurate, then Then the signal after combining the kth data of all M hops is

[0041]

[0042] Let γ1 represent the single-hop signal-to-noise ratio. From formula (10), we can get the theoretical signal-to-noise ratio of the combined signal:

[0043] γ comb =γ1+10log 10 (M) (11)

[0044] The signal-to-noise ratio of the actual combined signal is estimated using the method shown in formula (12):

[0045]

[0046] Among them, mean() means taking the mean, var() means taking the variance, then the combined signal-to-noise ratio loss is

[0047]

[0048] According to formula (11)(12)(13), N c Group parameters to get N c The combined signal-to-noise ratio loss estimation result is used as the objective function of the cross entropy iteration.

[0049] Step 7: With the goal of minimizing the combined signal-to-noise ratio loss, select samples with better objective function effects from the candidate group to update the binary quantization bit generation probability. cSort the combined signal-to-noise ratio losses from small to large, and select the top N with the smallest combined signal-to-noise ratio loss e The time hopping interval and carrier phase quantization parameter corresponding to the group are taken as the preferred group, and the probability of each bit in the binary quantization bit of the preferred group taking 0 or 1 is calculated. According to the probability obtained by statistics of the preferred group, the binary code generation probability of the next cross entropy iteration is obtained.

[0050] The multidimensional parameter estimation problem of carrier phase and time hopping interval is transformed into selecting the optimal carrier phase and time hopping interval quantization bit b to minimize the combined signal-to-noise ratio loss, that is,

[0051]

[0052] in, is a value set of M hop carrier phase and time hopping interval quantization parameters.

[0053] In order to obtain the carrier phase and time hopping interval quantization parameters that minimize the combined signal-to-noise ratio loss, the binary quantization bit generation probability is iterated so that the probability distribution of the estimated parameters is close to the optimal probability distribution. The probability iteration rule is:

[0054]

[0055] Among them, I{γ loss ≤T H} is a binary indicator function, T H Threshold

[0056]

[0057] According to the probability iteration rule shown in formula (15), the binary quantization bit generation probability is iteratively updated. c Among the binary quantized bits, select N bits that minimize the combined signal-to-noise ratio loss. e The threshold T in formula (15) is H With N c and N e The combined signal-to-noise ratio loss ranking result obtained by the i-th cross entropy iteration is recorded as The corresponding time hopping interval and carrier phase binary quantization bits are recorded as Depend on The estimated value obtained by mapping is recorded as l∈{1, 2, ..., N c}, merge the top N with the smallest signal-to-noise ratio loss e Group binary quantization bits into preferred groups The corresponding time hopping interval and carrier phase estimation value are Calculate the probability p of each quantized bit in the preferred group being 1 i+1

[0058]

[0059] P i+1 Perform smoothing and update the generation probability of the next iteration

[0060]

[0061] Among them, α is the smoothing coefficient, α∈(0,1], the larger the α, the faster the convergence speed.

[0062] Step 8: By setting the global optimal vector group to avoid falling into the local optimal solution, the time hopping interval and carrier phase estimation value with the smallest signal-to-noise ratio loss in the cross entropy iteration process are selected as the global optimal vector group. When each element in becomes 0 or 1, or the number of iterations reaches the set maximum number of iterations, the cross entropy iteration stops and the global optimal vector group is output. The time slot selection and carrier phase compensation of the received signal are performed according to the global optimal vector group obtained by random search, and the signal-to-noise ratio is improved through multi-hop coherent combining.

[0063] set up Used to record the minimum combined signal-to-noise ratio loss from the 1st iteration to the i-th iteration, set As a record The corresponding time hopping interval and carrier phase estimation value, when the N obtained in the i-th iteration in step 5 c The minimum value of the combined signal-to-noise ratio loss satisfies When Update the value to and Update the value to otherwise and No change. Calculation The number of elements in D1 that are 0 or 1 is N. When N<M(D1+D2) and the current number of iterations is less than the set maximum number of iterations, As the input condition for the next cross entropy iteration, repeat steps 4 to 8. When N = M (D1 + D2) or the current number of iterations is equal to the set maximum number of iterations, the cross entropy iteration stops and the global optimal vector group is output

[0064] The invention also discloses a differential time hopping estimation and multi-hop coherent merging device for satellite communication, which is used to implement a differential time hopping estimation and multi-hop coherent merging method for satellite communication. The time hopping interval estimation and multi-hop coherent merging device for satellite communication includes a parameter quantization unit, a parameter compensation and multi-hop merging unit, a signal-to-noise ratio estimation unit, and a cross entropy iteration unit.

[0065] The parameter quantization unit is used to establish a mapping relationship between the time hopping interval and the carrier phase search value and the binary quantization bit.

[0066] The parameter compensation and multi-hop merging unit performs parameter compensation and multi-hop merging based on the estimation result obtained by the cross entropy iteration unit, generates a local pulse template according to the time hopping interval, performs correlation operation with the received signal, selects the time slot where the signal is located, compensates the carrier phase of the selected time slot signal according to the carrier phase estimation value, and coherently combines the compensated multi-hop signals.

[0067] The signal-to-noise ratio estimation unit estimates the signal-to-noise ratio of the combined signal obtained by the parameter compensation and combining unit, calculates the difference between the estimated signal-to-noise ratio and the theoretical signal-to-noise ratio, and obtains the combined signal-to-noise ratio loss.

[0068] The cross entropy iteration unit generates a binary code from a binary quantization bit generation probability, generates a time hopping interval and a carrier phase estimation value according to the binary code and a mapping relationship obtained by the parameter quantization unit, outputs the estimation value to the parameter compensation and multi-hop merging unit, takes the combined signal-to-noise ratio loss calculated by the signal-to-noise ratio estimation unit as an objective function, selects samples with a smaller combined signal-to-noise ratio loss as a preferred group, regenerates the binary quantization bit generation probability for the next iteration according to the preferred group, records the time hopping interval and the carrier phase estimation value corresponding to the minimum combined signal-to-noise ratio loss as a global optimal vector group during the iteration process, and outputs the global optimal vector group after the iteration stops.

[0069] Beneficial effects:

[0070] 1. The cross-entropy iteration-assisted differential time hopping estimation and multi-hop coherent merging method and device disclosed in the present invention, in the process of random search, adaptively iteratively updates the probability distribution so that the solution with less loss of the combined signal-to-noise ratio is more likely to occur, thereby realizing the joint estimation of the time interval between two adjacent hop signals and the multi-hop carrier phase. Compared with the traditional ergodic method, which requires searching all possible parameter values ​​one by one and has a higher complexity, the use of the cross-entropy iteration method to assist the differential time hopping estimation and multi-hop coherent merging can significantly reduce the number of searches and reduce the complexity of parameter estimation.

[0071] 2. The cross-entropy iteration-assisted differential time hopping estimation and multi-hop coherent combining method and device disclosed in the present invention coherently combines the signals of multiple time slots on the basis of joint estimation of the time hopping interval and the carrier phase, and demodulates the combined signals. Under the condition of low single-hop signal-to-noise ratio, the demodulation signal-to-noise ratio is improved through multi-hop coherent combining.

[0072] 3. In the satellite communication scenario, the wireless communication link is susceptible to interference from non-cooperative parties. The cross-entropy iteration-assisted differential time hopping estimation and multi-hop coherent merging method and device disclosed in the present invention jumps and discontinuously transmits signals in the time domain at random time intervals. The receiving end uses the cross-entropy iteration method to jointly estimate the time hopping interval and the carrier phase, thereby realizing multi-hop coherent merging and demodulation. For non-cooperative parties, the jammer needs to send interference continuously to cause a strong interference effect, which increases the difficulty of interference for non-cooperative parties and improves the anti-interference ability of communication.

[0073] 4. The cross-entropy iteration-assisted differential time hopping estimation and multi-hop coherent merging method and device disclosed in the present invention randomly generate estimated values ​​of the time hopping interval and the carrier phase during the cross-entropy iteration process. The randomness of parameter generation may lead to a better solution than the parameter solution obtained in the last iteration in the middle of the iteration. By setting a global optimal vector group to select a solution that achieves the best effect for the objective function during the iteration process, the cross-entropy iteration can be prevented from falling into a local optimal solution, and the parameter estimation error of the cross-entropy iteration can be reduced. BRIEF DESCRIPTION OF THE DRAWINGS

[0074] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the following briefly introduces the drawings required for use in the embodiments or the prior art descriptions. Obviously, the drawings described below are some embodiments of the present invention, and for ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work.

[0075] Figure 1 Flowchart of the joint estimation of time hopping interval and carrier phase assisted by cross entropy iteration.

[0076] Figure 2 Block diagram of cross-entropy iteration-assisted differential hop time estimation and multi-hop coherent merging.

[0077] Figure 3 Schematic diagram of differential time hopping interval.

[0078] Figure 4 Iteration curve for merging signal-to-noise ratio loss.

[0079] Figure 5 is the combined bit error rate assisted by cross entropy iteration. DETAILED DESCRIPTION

[0080] The present invention is further described and described in detail below with reference to the accompanying drawings and embodiments.

[0081] like Figure 1 As shown, the cross entropy iteration-assisted differential time hopping estimation and multi-hop coherent merging method disclosed in this embodiment has the following specific implementation steps:

[0082] Step 1: The time slot structure of the time-hopping signal based on the differential time-hopping interval is as follows: Figure 2 As shown, in this embodiment, M=8, L δ,max =4, K = 1000, generate BPSK modulated time hopping signal as shown in formula (1), repeatedly transmit 1000 BPSK modulation symbols to be sent in 8 discontinuous hops, and the communication parties agree that the number of time slots between two adjacent hop signals is in the range of L δ,m ∈{1,2,3,4}.

[0083] Step 2: Send the BPSK modulated time-hopping signal obtained in step 1 into the AWGN channel, and receive the signal transmitted through the AWGN channel. The received signal after passing through the AWGN channel is shown in formula (2). The receiving end delay τ and frequency deviation f d Ideally synchronized.

[0084] Step 3: The initial phase search range of the received signal carrier for each hop is [0,2π). The two parties agree on L δ,m The search range is {1, 2, 3, 4}. In this embodiment, D1=6, D2=2, N1=64, N2=4, and the search parameters of the initial phase and time hopping interval of the received signal carrier of the mth hop are respectively A mapping relationship is established between the search value and the binary code. Each carrier phase search value corresponds to a 6-bit binary number, and each time hopping interval search value corresponds to a 2-bit binary number, thereby realizing parameter quantization of the time hopping interval and the carrier phase of each hopping signal.

[0085] The initial phase estimate of the dth carrier of the mth hop is shown in formula (3). d-1 is represented as a 6-bit binary number. For example, d-1=30 is represented as the binary number 011110, thereby establishing φ m,d The estimated value of the u-th hop interval of the m-th hop is shown in formula (4). u-1 is represented as a 2-bit binary number, for example, u-1=2 is represented as the binary number 10, thus establishing Mapping relationship with 2-bit binary numbers.

[0086] Step 4: Number of candidate groups N in this embodiment c = 1300, according to the preset initial probability of binary quantization bits, 1300 groups of binary codes are generated, called candidate groups, and N c The binary code mapping is used to obtain 1300 sets of initial solutions for M-hop carrier phase and time-hopping interval estimation. The parameter solutions are regenerated by updating the binary quantization bit generation probability in the cross entropy iteration process.

[0087] In this embodiment, the carrier phase and time hopping interval estimation value of 8 hops are binary quantized into 64 bits in total. Randomly generate binary quantized bits The set of carrier phase quantization bits for the mth hop is According to formula (5), The mapping results in a set of carrier phase estimation values ​​for the mth hop, and a set of hop interval quantization bits for the mth hop is: According to formula (6), The mapping results in a set of estimated time intervals for the mth hop. Mapping obtains M hop carrier phase and time hopping interval estimation values

[0088] Step 5: Generate a local pulse template based on the estimated time hopping interval obtained in step 4, and perform correlation operation with the received signal, that is, match the local template with the time slot position of the received signal, and compensate the signal carrier phase of the selected time slot based on the carrier phase estimate, and coherently combine the compensated multi-hop signals.

[0089] Step 6: Estimate the signal-to-noise ratio of the combined signal obtained in step 5, calculate the difference between the estimated actual signal-to-noise ratio and the 8-hop theoretical combined signal-to-noise ratio, and obtain the combined signal-to-noise ratio loss corresponding to the 1300 sets of carrier phase and time-hopping interval parameter solutions generated in step 4, and use the combined signal-to-noise ratio loss as the objective function of the cross-entropy iteration. According to formula (11), the theoretical signal-to-noise ratio of the combined signal is γ1+10log 10 8≈γ1+9, the signal-to-noise ratio of the actual merged signal is estimated according to formula (12), and the merged signal-to-noise ratio loss is calculated according to formula (13), and the estimated results of the merged signal-to-noise ratio loss of 1300 candidate groups are obtained.

[0090] Step 7: With the goal of minimizing the combined signal-to-noise ratio loss, select samples with better objective function effects from the candidate group to update the binary quantization bit generation probability. Sort the 1300 combined signal-to-noise ratio losses obtained in step 6 from small to large. In this embodiment, N e =130, select the first N with the smallest combined signal-to-noise ratio loss e The time hopping interval and carrier phase quantization parameter corresponding to the group are taken as the preferred group, and the probability of each bit in the binary quantization bit of the preferred group taking 0 or 1 is calculated. According to the probability obtained by statistics of the preferred group, the binary code generation probability of the next cross entropy iteration is obtained.

[0091] Sort the 1300 combined signal-to-noise ratio loss estimation results obtained from the i-th cross entropy iteration from small to large, and we get The corresponding time hopping interval and carrier phase binary quantization bits are recorded as Depend on The estimated value obtained by mapping is recorded as l∈{1, 2, ..., Nc}, merge the top N with the smallest signal-to-noise ratio loss e Group binary quantization bits into preferred groups The corresponding time hopping interval and carrier phase estimation value are According to formula (17), the probability p of each quantization bit in the preferred group being 1 is calculated: i+1 , update the generation probability of the next iteration according to formula (18) In this embodiment, α=1, that is,

[0092] Step 8: By setting the global optimal vector group to avoid falling into the local optimal solution, the time hopping interval and carrier phase estimation value with the smallest signal-to-noise ratio loss in the cross entropy iteration process are selected as the global optimal vector group. When each element in becomes 0 or 1, or the number of iterations reaches the set maximum number of iterations, the cross entropy iteration stops and the global optimal vector group is output. The time slot selection and carrier phase compensation of the received signal are performed according to the global optimal vector group obtained by random search, and the signal-to-noise ratio is improved through multi-hop coherent combining.

[0093] set up Used to record the minimum combined signal-to-noise ratio loss from the 1st iteration to the i-th iteration, set As a record The corresponding time hopping interval and carrier phase estimation value, when the N obtained in the i-th iteration in step 5 c The minimum value of the combined signal-to-noise ratio loss satisfies When Update the value to and Update the value to otherwise and No change. Calculation The number of elements in the number of 0 or 1 is M. When M < 64 and the current number of iterations is less than the set maximum number of iterations, As the input condition for the next cross entropy iteration, repeat steps 4 to 6. When M = 64 or the current number of iterations is equal to the set maximum number of iterations, the cross entropy iteration stops and the global optimal vector group is output.

[0094] like Figure 4 As shown, N c =1300, change the ratio of the number of preferred groups to the number of candidate groups N e / N c , simulate the iterative convergence curve, and obtain the relationship between the average combined signal-to-noise ratio loss and the number of iterations. e / N c= 0.01, 10 iterations are sufficient to converge, and the combined signal-to-noise ratio loss is about 1dB. e / N c = 0.15, the combined signal-to-noise ratio loss is reduced to about 0.4 dB, and convergence is achieved after about 25 iterations. e / N c When N is smaller, the convergence speed is faster, but it is easy to converge to the local optimal solution. e / N c As increases, it is easy to obtain the global optimal solution, but the number of iterations increases and the complexity increases.

[0095] Figure 5 The cross-entropy iteration-assisted method is compared with the ergodic method and the theoretical combination of bit error rates. The ergodic method performs point-by-point searches on a two-dimensional grid of time-hopping intervals and carrier phases. Ideally, this method can accurately estimate discrete time-hopping intervals, and the performance loss comes from the grid search step of the carrier phase. The cross-entropy iteration-assisted method uses an adaptive key sampling strategy to randomly search on a two-dimensional grid of time-hopping intervals and carrier phases to reduce the number of grid searches. In this embodiment, the number of grids in the time-hopping interval dimension of the ergodic method is 2 16 , the number of grids in the carrier phase dimension is 2 24 , then the traversal method searches for 2 16 ×2 24 =2 40 In the cross entropy iterative auxiliary method, N c =1300 and N e / N c = 0.1, it converges after about 20 iterations, and the number of cross entropy iteration searches is about 1300 × 20 = 26000, which is significantly reduced compared to the ergodic method. The received signal carrier phase is set in the middle of the grid. b / N0 range is -6dB~1dB, under the condition of simulating demodulation bit error rate, such as Figure 5 As shown in the figure, the bit error rate of the cross entropy iteration-assisted method is basically the same as that of the ergodic method (@BER=1×10 -4 ), the performance loss compared to the theoretical combination is about 0.5dB (@BER=1×10 -4 ). Compared with the ergodic method, the parameter estimation complexity of the cross entropy iterative assisted method is reduced.

[0096] The specific description above further illustrates the purpose, technical solutions and beneficial effects of the invention in detail. It should be understood that the above is only a specific embodiment of the present invention and is not intended to limit the scope of protection of 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 method for estimating time hopping intervals and combining multiple hops coherently for satellite communications, characterized in that: The following steps are included: Step 1: The information symbol to be sent is modulated by BPSK and then repeatedly transmitted in multiple time slots that are not continuous in time. Each time slot is called a hop. The time interval between two adjacent hop signals varies randomly within a predetermined range to generate a BPSK modulated time-hopping signal with a random hop interval. Step 2: Send the BPSK modulated time-hopping signal obtained in step 1 into the AWGN channel, and receive the signal transmitted through the AWGN channel; Step 3: The search range of the initial phase of the carrier of each hop signal of the received signal is [0,2π). The two parties agree that the search range of the number of time slots between two adjacent hop signals is {1,2,...,L δ，max }, the mth hop carrier initial phase and time hopping interval search parameters are in A mapping relationship is established between the search value and the binary code. D1 and D2 are binary quantization bit numbers. Each carrier phase search value corresponds to a D1bit binary number, and each time hopping interval search value corresponds to a D2bit binary number, that is, the parameter quantization of the time hopping interval and the carrier phase of each hopping signal is realized; Step 4: Generate N according to the preset initial probability of binary quantization bits c A group of binary codes, called candidate groups, consists of N c The binary code mapping is used to obtain the M hop carrier phase and the N hop time interval estimation. c The initial solution of the group is generated, and the parameter solution is regenerated by updating the binary quantization bit generation probability during the cross entropy iteration process; Step 5: Generate a local pulse template according to the estimated time hopping interval obtained in step 4, and perform correlation operation with the received signal, that is, match the local template with the time slot position of the received signal, and compensate the signal carrier phase of the selected time slot according to the carrier phase estimation value, and coherently combine the compensated multi-hop signals; Step 6: Estimate the signal-to-noise ratio of the combined signal obtained in step 5, calculate the difference between the estimated actual signal-to-noise ratio and the M-hop theoretical combined signal-to-noise ratio, and obtain the N generated in step 4. c The combined signal-to-noise ratio loss corresponding to the group carrier phase and time-hopping interval parameter solution is used as the objective function of the cross entropy iteration; Step 7: With the goal of minimizing the combined signal-to-noise ratio loss, select samples with better objective function effects from the candidate group to update the binary quantization bit generation probability; The N obtained in step 6 c Sort the combined signal-to-noise ratio losses from small to large, and select the top N with the smallest combined signal-to-noise ratio loss e The time hopping interval and carrier phase quantization parameter corresponding to the group are taken as the preferred group, and the probability of each bit in the binary quantization bit of the preferred group taking 0 or 1 is calculated. According to the probability obtained by statistics of the preferred group, the binary code generation probability of the next cross entropy iteration is obtained; Step 8: By setting the global optimal vector group to avoid falling into the local optimal solution, the time hopping interval and carrier phase estimation value with the smallest signal-to-noise ratio loss in the cross entropy iteration process are selected as the global optimal vector group. When each element in becomes 0 or 1, or the number of iterations reaches the set maximum number of iterations, the cross entropy iteration stops and the global optimal vector group is output; the time slot selection and carrier phase compensation of the received signal are performed according to the global optimal vector group obtained by random search, and the signal-to-noise ratio is improved by multi-hop coherent combining.

2. The method for estimating the time hopping interval and combining multiple hops coherently for satellite communication according to claim 1, characterized in that: In step one, The BPSK modulated time-hopping signal with random time-hopping interval is generated as shown in the following formula: The signal is dispersed and sent repeatedly in M ​​hops, with K symbols sent in each hop. k represents the kth BPSK modulation symbol, b k ∈{-1,+1}, w(t) is a single pulse waveform, T s is the duration of a single symbol, L m is the number of time slots before the mth hop, T p is the duration of a single time slot, f0 is the carrier frequency of the transmitted signal, φ 0,m It is the initial phase of the carrier of the mth hop transmitted signal.

3. The method for estimating the time hopping interval and combining multiple hops coherently for satellite communication according to claim 2, characterized in that: In step 2, The received signal after passing through the AWGN channel is expressed as Where A is the amplitude of the transmitted signal when it reaches the receiver, τ is the delay of the receiving end relative to the transmitting end, and f d is the frequency deviation of the received signal relative to the transmitted signal, φ m is the initial phase of the carrier of the m-th hop received signal, n(t) represents additive complex Gaussian white noise, with a mean of 0 and a variance of 4. The method for estimating the time hopping interval and combining multiple hops coherently for satellite communication according to claim 3, characterized in that: In step three, φ m,d represents the initial phase estimate of the dth carrier of the mth hop, d∈{1,2,...,N1}, and Represent d-1 as a binary number of D1 bits, and establish φ from the above formula m,d The mapping relationship with the D1bit binary number; L δ,m,u represents the estimated value of the u-th hop interval of the m-th hop, u∈{1, 2, ..., N2}, and N2=L δ,max ,have L δ,m,u =and (4) Represent u-1 as a D2-bit binary number, and establish L from the above formula δ,m,u The mapping relationship with D2bit binary number.

5. The method for estimating the time hopping interval and combining multiple hops coherently for satellite communication according to claim 4, characterized in that: In step four, The probability of generating binary quantized bits for the i-th cross entropy iteration is By probability Randomly generate binary quantized bits Indicates generation The probability of taking the value 1; the set of carrier phase quantization bits for the mth hop is in represents the vbit of the carrier phase quantization bit of the mth hop in the i-th iteration, A group of hop interval quantization bits for the mth hop in represents the lth bit of the carrier phase quantization bit of the mth hop in the i-th iteration, Depend on The carrier phase estimation obtained by mapping is Depend on The estimated time hopping interval obtained by mapping is According to formula (5) (6), Mapping obtains M hop carrier phase and time hopping interval estimation values 6. The method for estimating the time hopping interval and combining multiple hops coherently for satellite communication according to claim 5, characterized in that: The implementation method of step five is: The hop time interval of each hop is estimated by Get the number of time slots before the mth hop The receiving end Generate a local pulse template and perform correlation calculation and phase compensation on the kth data of all M jumps The signal after combining the k-th data of all M hops is further expressed as Wherein P≤M, P is the number of time slots where the useful signal in the M-hop data is located that matches the local pulse template.

7. The method for estimating the time hopping interval and combining multiple hops coherently for satellite communication according to claim 6, characterized in that: The implementation method of step six is: If the time-hopping interval is estimated accurately, but Thus, P=M in formula (9). If the carrier phase estimation is accurate, then Then the signal after combining the kth data of all M hops is Let γ1 represent the single-hop signal-to-noise ratio. From formula (10), we can get the theoretical signal-to-noise ratio of the combined signal: c comb =γ1+10log 10 (M) (11) The signal-to-noise ratio of the actual combined signal is estimated using the method shown in formula (12): Among them, mean() means taking the mean, var() means taking the variance, then the combined signal-to-noise ratio loss is According to formula (11)(12)(13), N c Group parameters to get N c The combined signal-to-noise ratio loss estimation result is used as the objective function of the cross entropy iteration.

8. The method for estimating the time hopping interval and combining multiple hops coherently for satellite communication according to claim 7, characterized in that: The implementation method of step seven is: The multidimensional parameter estimation problem of carrier phase and time hopping interval is transformed into selecting the optimal carrier phase and time hopping interval quantization bit b to minimize the combined signal-to-noise ratio loss, that is, in, is a value set of M hopping carrier phase and time hopping interval quantization parameters; In order to obtain the carrier phase and time hopping interval quantization parameters that minimize the combined signal-to-noise ratio loss, the binary quantization bit generation probability is iterated so that the probability distribution of the estimated parameters is close to the optimal probability distribution. The probability iteration rule is: Among them, I{γ loss ≤T H } is a binary indicator function, T H Threshold According to the probability iteration rule shown in formula (15), the binary quantization bit generation probability is iteratively updated. c Among the groups of binary quantized bits, select N bits that minimize the combined signal-to-noise ratio loss. e The threshold T in formula (15) is H With N c and N e The combined signal-to-noise ratio loss ranking result obtained by the i-th cross entropy iteration is recorded as The corresponding time hopping interval and carrier phase binary quantization bits are recorded as Depend on The estimated value obtained by mapping is recorded as l∈{1,2,…,N c }, merge the top N with the smallest signal-to-noise ratio loss e Group binary quantization bits into preferred groups The corresponding time hopping interval and carrier phase estimation value are Calculate the probability p of each quantized bit in the preferred group being 1 i+1 P i+1 Perform smoothing and update the generation probability of the next iteration Among them, α is the smoothing coefficient, α∈(0,1], the larger the α, the faster the convergence speed.

9. The method for estimating the time hopping interval and combining multiple hops coherently for satellite communication according to claim 8, characterized in that: The implementation method of step eight is: set up Used to record the minimum combined signal-to-noise ratio loss from the 1st iteration to the i-th iteration, set As a record The corresponding time hopping interval and carrier phase estimation value, when the N obtained in the i-th iteration in step 5 c The minimum value of the combined signal-to-noise ratio loss satisfies When Update the value to and Update the value to otherwise and unchanged; calculated The number of elements in D1 that are 0 or 1 is N. When N<M(D1+D2) and the current number of iterations is less than the set maximum number of iterations, As the input condition for the next cross entropy iteration, repeat steps 4 to 8; when N = M (D1 + D2) or the current number of iterations is equal to the set maximum number of iterations, the cross entropy iteration stops and the global optimal vector group is output 10. A satellite communication-oriented differential time hopping estimation and multi-hop coherent combining device, used to implement the satellite communication-oriented time hopping interval estimation and multi-hop coherent combining method as claimed in claim 1, 2, 3, 4, 5, 6 or 7, characterized in that: It includes parameter quantization unit, parameter compensation and multi-hop merging unit, signal-to-noise ratio estimation unit, and cross entropy iteration unit; The parameter quantization unit is used to establish a mapping relationship between the time hopping interval and the carrier phase search value and the binary quantization bit; The parameter compensation and multi-hop merging unit performs parameter compensation and multi-hop merging based on the estimation result obtained by the cross entropy iteration unit, generates a local pulse template according to the time hopping interval, performs correlation operation with the received signal, selects the time slot where the signal is located, compensates the carrier phase of the selected time slot signal according to the carrier phase estimation value, and coherently combines the compensated multi-hop signals; The signal-to-noise ratio estimation unit estimates the signal-to-noise ratio of the combined signal obtained by the parameter compensation and combining unit, calculates the difference between the estimated signal-to-noise ratio and the theoretical signal-to-noise ratio, and obtains the combined signal-to-noise ratio loss; The cross entropy iteration unit generates a binary code from a binary quantization bit generation probability, generates a time hopping interval and a carrier phase estimation value according to the binary code and a mapping relationship obtained by the parameter quantization unit, outputs the estimation value to the parameter compensation and multi-hop merging unit, takes the combined signal-to-noise ratio loss calculated by the signal-to-noise ratio estimation unit as an objective function, selects samples with a smaller combined signal-to-noise ratio loss as a preferred group, regenerates the binary quantization bit generation probability for the next iteration according to the preferred group, records the time hopping interval and the carrier phase estimation value corresponding to the minimum combined signal-to-noise ratio loss as a global optimal vector group during the iteration process, and outputs the global optimal vector group after the iteration stops.

Images