MMSE-FED and decision feedback interference removing method and device for molecular code division multiple access system

Abstract

The invention discloses an MMSE-FED (Molecular Code Division Multiple Access) and decision feedback interference removing method and device, and the method comprises the steps: carrying out the fast Fourier transform of an original observation value, obtaining a frequency domain observation matrix, carrying out the sorting according to the receiving power or distance of a nanometer transmitter at a fusion center, and obtaining a frequency domain observation matrix; the method comprises the following steps: taking a first nanometer transmitter as the current most reliable, taking an initial frequency domain observation matrix as a basis, obtaining a judgment variable through MMSE-FED, IFFT and de-spreading operation, then obtaining a hard judgment result through a judgment function as a bit sequence of a first nanometer transmitter, reconstructing signal contribution according to the bit sequence, obtaining an updated frequency domain receiving signal, and obtaining a second nanometer transmitter; for the kth nanometer transmitter after sorting, based on the updated frequency domain receiving signal, repeating the operation to detect the bit sequence, and reconstructing the signal again for interference elimination; and finally, detecting the bit sequence of the Kth transmitter based on the signal updated for the (K-1) th time through loop iteration.

Description

Technical Field

[0001] This invention relates to the field of molecular communication technology, and in particular to a molecular code division multiple access system (MMSE-FED) and a decision feedback method and apparatus for interference removal. Background Technology

[0002] With the development of nanotechnology and bioengineering, molecular communication is considered one of the key communication methods for future biological nanonetworks. Its basic principle is to use molecules as information carriers to transmit information at the microscale through diffusion or chemical reactions. Molecular communication has advantages such as low energy consumption, strong biocompatibility, and the ability to communicate in microenvironments that cannot be covered by traditional electromagnetic waves. Therefore, it has broad application prospects in fields such as in vivo medical monitoring, drug delivery, and intercellular signal transmission.

[0003] However, in diffusion-based molecular communication systems, information molecules are transmitted through Brownian motion in a fluid medium. Due to the randomness and nonlinearity of the molecular diffusion medium, there is significant inter-symbol interference and on-chip interference between the received signals. When multiple nanonodes simultaneously send molecular signals to the fusion center, the signals overlap, further generating multiple access interference, which significantly reduces the system's detection performance. Existing signal detection methods, such as matched filtering, maximum likelihood detection, and decision feedback equalization, suffer from high computational complexity or insufficient interference removal performance under conditions of non-Gaussianity of molecular channels, long channel memory, and multi-user interference. In particular, in molecular communication code division multiple access systems, traditional detection algorithms struggle to simultaneously address multi-user separation, inter-symbol interference suppression, and complexity control. Summary of the Invention

[0004] Based on this, it is necessary to address the above problems by proposing a molecular code division multiple access system (MMSE-FED) and a decision feedback method and apparatus for interference removal.

[0005] A molecular code division multiple access system (MMSE-FED) and a decision feedback method for interference removal, the method comprising: A molecular code division multiple access system is constructed, comprising K nano-emitters and a fusion center. The nano-emitters transmit information molecules through a diffusion medium, and the fusion center is used for signal detection. At the k-th nano-emitter, data is organized using a block transfer method, with each data block being M bits in length. Each nano-emitter is assigned a unique address code or extended sequence of length N. A single bit is expanded into N chip bits after address code expansion, and binary molecular shift keying is used to modulate and release two types of molecular pulses, A and B.

[0006] At the fusion center, the received molecular concentration signal is sampled to determine the observed concentration difference between type A and type B molecules; Based on the concentration difference observations, the expanded values ​​correspond to MN chip observations; and a cyclic prefix is ​​inserted between data blocks to eliminate inter-block interference. The fusion center preprocesses the discrete observations into time-domain observation vectors after removing the cyclic prefix. The time-domain observation signal is converted into a frequency-domain signal using a fast Fourier transform to obtain a frequency-domain observation vector; All nano-emitters are sorted according to their received power at the fusion center from high to low, or their distance from the fusion center from near to far. The first nano-emitter in the sorted list is selected as the priority nano-emitter for detection. Based on the frequency domain observation vector, decision variables are obtained through minimum mean square error frequency domain equalization, inverse fast Fourier transform, and despreading. The hard decision result derived from the decision variables is used as the bit sequence of the first nano-emitter through a decision function. Based on the detected bit sequence, the signal contribution of the first nano-emitter is reconstructed in the frequency domain and subtracted from the frequency domain observation vector to obtain... The updated frequency domain received signal; for the sorted k-th nano-emitter, where k ranges from 2 to K-1, iterative processing is performed: based on the current updated frequency domain received signal, the decision variable is obtained through minimum mean square error frequency domain equalization, inverse fast Fourier transform, and despreading operations, and the bit sequence of the k-th nano-emitter is detected; the signal contribution of the k-th nano-emitter is reconstructed and subtracted from the current frequency domain received signal to update the frequency domain received signal; based on the finally updated frequency domain received signal, the bit sequence of the k-th nano-emitter is detected through minimum mean square error frequency domain equalization, inverse fast Fourier transform, and despreading operations; For each bit of each nanoemitter, compare its decision variables. Zero threshold: If the decision variable is greater than 0, then the bit is judged to be 1; otherwise, it is judged to be -1, and the decoded bit sequence is output. As the final test result.

[0007] Preferably, the step of inserting a cyclic prefix between data blocks to eliminate inter-block interference, and the preprocessing of discrete observations into time-domain observation vectors by the fusion center after removing the cyclic prefix, specifically includes: through... Determine the time-domain observation vector, where, , That is, a single molecular pulse in continuous Concentration values ​​at each sampling time on the chip. It is a size of The matrix, by Constructed, that is , yes The noise vector.

[0008] Preferably, the step of converting the time-domain observation signal into a frequency-domain signal using a Fast Fourier Transform (FFT) to obtain a frequency-domain observation vector specifically includes: transforming the matrix using an FFT... Time-domain observation vector Convert to frequency domain observation vector in, It is a circular matrix The diagonalized matrix can be represented as: .

[0009] Preferably, the step of using the hard decision result derived from the decision variable through the decision function as the bit sequence of the first nano-emitter specifically includes: through... Determine the bit sequence of the first nanoemitter, where It is an equilibrium matrix generated using the MMSE-FED principle.

[0010] Preferably, the method further includes: obtaining an estimate of the signal received by the first nanotransmitter on the p-th subcarrier based on the bit sequence of the first nanotransmitter. ;in, It is a length of The column vector is represented as The elements within it This indicates that the first nano-emitter was in the first... Estimates of the received signal on each subcarrier; The estimated value From length Subtract from the frequency domain observation vector to obtain the updated frequency domain observation vector. .

[0011] Preferably, the step of detecting the bit sequence of the k-th nano-emitter after obtaining the decision variable through minimum mean square error frequency domain equalization, inverse fast Fourier transform, and despreading operation based on the currently updated frequency domain received signal, and then reconstructing the signal contribution of the k-th nano-emitter and subtracting it from the current frequency domain received signal to update the frequency domain received signal, specifically includes: through... Determine the signal sent by the k-th nano-emitter A bit sequence, an estimated value, and an updated frequency domain observation vector, wherein... This represents the decision function.

[0012] Preferably, the method further includes: employing binary molecular shift keying modulation, wherein each nano-emitter spreads the data using a spreading sequence with values ​​of {+1, -1}, and releases molecular concentration signals representing bit information by releasing type A and type B molecules; Preferably, the binary molecular shift keying modulation is employed, and each nano-emitter uses a spreading sequence with values ​​{+1, -1} to spread the data, specifically including: in molecular code division multiple access, the first... The information to be transmitted by each nano-emitter is represented as follows: ,in Each nano-emitter sends the first Each bit is passed through a length of The spread spectrum sequence is extended, and this sequence can be represented as... ,in ,and When using binary molecular shift keying modulation, each pulse is emitted... The first molecule, type A molecule represents '1', type B molecule represents '-1', and the second molecule... The nano-emitter transmits the first... When the data is bit-sized, its molecular emission signal can be expressed as: in It is a nano-emitter At the moment of molecular pulse emission, the pulses of type A and type B molecules are respectively generated by... and It is represented and defined as: ( ), and when hour, ( ).

[0013] Preferably, the method further includes: performing equalization processing on the frequency domain signal by applying a minimum mean square error frequency domain equalization algorithm, specifically including: through... Obtain the equalized signal ,in, It is a diagonal balanced matrix, which can be represented as , Corresponding to the The equalization coefficient of each subcarrier.

[0014] A molecular code division multiple access system (MMSE-FED) and a decision feedback de-interference device are disclosed. The device includes: K nano-emitters, each configured to: transmit information molecules through a diffusion medium and employ binary molecular shift keying modulation; spread the data using a spreading sequence with values ​​of {+1, -1}, and release molecular concentration signals representing bits "1" (A-type molecules) and bits "-1" (B-type molecules).

[0015] A fusion center, connected to the nano-emitter via a diffusion medium, is configured to: sample the received molecular concentration signal and determine the observed concentration difference between type A and type B molecules; Based on the concentration difference observations, MN chip observations are generated after expansion; a cyclic prefix is ​​inserted between data blocks to eliminate inter-block interference. The fusion center preprocesses the discrete observations after removing the cyclic prefix into a time-domain observation vector; the time-domain observation signal is converted to a frequency-domain signal using a fast Fourier transform to obtain a frequency-domain observation vector; all nano-emitters are sorted according to their received power at the fusion center from high to low or their distance from the fusion center from near to far, and the first nano-emitter in the sorted order is selected as the priority detection nano-emitter. Based on the frequency-domain observation vector, a decision variable is obtained through minimum mean square error frequency domain equalization, inverse fast Fourier transform, and despreading; the hard decision result derived from the decision variable is used as the bit sequence of the first nano-emitter using a decision function. Based on the detected bit sequence, in... The signal contribution of the first nanotransmitter is reconstructed in the frequency domain and subtracted from the frequency domain observation vector to obtain the updated frequency domain received signal. For the k-th nanotransmitter in the sorted sequence, where k is from 2 to K-1, iterative processing is performed: based on the current updated frequency domain received signal, the decision variables are obtained through minimum mean square error frequency domain equalization, inverse fast Fourier transform, and despreading operations, and the bit sequence of the k-th nanotransmitter is detected. The signal contribution of the k-th nanotransmitter is reconstructed and subtracted from the current frequency domain received signal to update the frequency domain received signal. After completing the interference cancellation of the first K-1 nanotransmitters, the bit sequence of the k-th nanotransmitter is detected based on the final updated frequency domain received signal through minimum mean square error frequency domain equalization, inverse fast Fourier transform, and despreading operations. For each bit of each nanotransmitter, its decision variables are compared. Zero threshold: If the decision variable is greater than 0, then the bit is judged to be 1; otherwise, it is judged to be -1, and the decoded bit sequence is output. As the final test result.

[0016] The embodiments of the present invention have the following beneficial effects: This invention achieves a balance between low complexity and high reliability in molecular code division multiple access (MCD) systems by combining minimum mean square error frequency domain equalization with single-stage decision feedback interference elimination. It utilizes a cyclic prefix block transmission structure to convert signal processing to the frequency domain, significantly reducing the processing complexity of inter-symbol and inter-chip interference. Through a single-stage decision feedback mechanism sorted by received power, it gradually eliminates the signal contribution of strong-interference nanotransmitters, effectively suppressing multiple access interference. Ultimately, while maintaining the computational efficiency of nanodevices, it significantly improves the accuracy of multi-user detection and the system's anti-interference capability. Attached Figure Description

[0017] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0018] in: Figure 1 This is a flowchart illustrating an embodiment of the present invention using a molecular code division multiple access system (MMSE-FED) and a decision feedback de-interference method. Figure 2 This is a schematic diagram of a diffusion-type molecular communication system model using a molecular code division multiple access (MMSE-FED) system and a decision feedback de-interference method, as described in an embodiment of the present invention. Detailed Implementation

[0019] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0020] This invention provides an embodiment of a molecular code division multiple access system (MMSE-FED) and a decision feedback de-interference method, such as... Figure 1 , 2 As shown, the method includes: Step 1: Construct a molecular code division multiple access system, including K nano-emitters and a fusion center. The nano-emitters transmit information molecules through a diffusion medium, and the fusion center is used for signal detection. Specifically, the molecular code division multiple access system consists of The system consists of multiple nano-emitters and a spherical transparent receiver located at the fusion center. During information transmission, the nano-emitters release two types of information molecules (Type A and Type B) into a three-dimensional fluid medium. This medium is infinitely spatial, homogeneous, stable, and without flow boundaries. The fusion center can then... Molecular concentration signals are sampled and detected within a spherical detection space.

[0021] In molecular code division multiple access, nano-emitters release... Each molecular pulse is used to transmit one bit of information, and each pulse contains Each molecule. The time interval between two adjacent pulses is called a chip, and its duration is... Second. Definition If the extension factor is used, then the symbol duration is... When binary molecular shift keying is used, This also corresponds to one bit period. The duration of the bit.

[0022] In a code-division multiple access system, the first... The information to be transmitted by each nano-emitter is represented as follows: ,in , representing the j-th bit information transmitted by the k-th nanotransmitter. To achieve multiple access, each nanotransmitter transmits the j-th bit information. Each bit is passed through a length of The spreading sequence is extended, and the spreading sequence can be represented as follows: ,in ,and , representing a pseudo-random sequence. Since the transmission distances between different nano-emitters and the fusion center vary, a constant can be defined for the entire molecular code division multiple access system to achieve synchronous transmission between different nano-emitters. .

[0023] Furthermore, regardless of the distance between the nano-emitter and the fusion center, it is assumed that the fusion center samples at the desired peak of the molecular concentration pulse. To achieve this, a time delay is set. To adjust the initial emission time of different nano-emitters, so that the molecular pulses emitted by different nano-emitters are aligned at the concentration peak time of the fusion center.

[0024] Step 2: At the k-th nano-emitter, organize the data using a block transfer method, with each data block being M bits in length. Each nano-emitter is assigned a unique address code or extended sequence of length N. A single bit is expanded into N chip bits after address code expansion, and binary molecular shift keying is used to modulate and release two types of molecular pulses, A and B. Specifically, in a code-division multiple access system, the first... The information to be transmitted by each nano-emitter is represented as follows: ,in To achieve multiple access, each nano-emitter sends the first... Each bit is passed through a length of The extended sequence is expanded, and this sequence can be represented as... ,in ,and Since the transmission distances between different nano-emitters and the fusion center vary, a constant can be defined for the entire molecular code division multiple access system to achieve synchronous transmission between different nano-emitters. .

[0025] Furthermore, regardless of the distance between the nano-emitter and the fusion center, the fusion center is configured to sample at the desired peak of the molecular concentration pulse. To achieve this, a time delay is set. This allows for adjustment of the initial emission times of different nano-emitters, aligning the molecular pulses emitted by each emitter with the concentration peak at the fusion center. When using binary molecular shift-bonding modulation, each pulse emission... The number of molecules is as follows: type A molecules represent '1', and type B molecules represent '-1'. Therefore, the number of molecules is... The nano-emitter transmits the first... When the data is bit-sized, its molecular emission signal can be expressed as:

[0026] (1)

[0027] in, It is a nano-emitter The moment of molecular pulse emission, u is the bit index, one bit is divided into N chips, q is the chip index, and t is the time scale. The pulses of type A and type B molecules are respectively generated by... and They are defined as follows: ( ), and when hour, ( ), where x is the independent variable of the impulse function.

[0028] Molecules diffuse freely in a stagnant, boundaryless fluid medium, and this medium is homogeneous and stable; nano-emitters At any moment Launch a When a pulse composed of molecules is applied, the fusion center is at... The molecular concentration sampled at time t obeys Fick's second law, given as: (2)

[0029] in, Where is the diffusion coefficient. For the first Distance between each emitter and the fusion center, molecular concentration It is a time-domain impulse function representing channel state information for a given distance. Molecular concentration at The maximum value is reached at time , and its maximum value is , where e is the natural base.

[0030] To ensure that the maximum desired concentrations from different nanoemitters arrive at the fusion center simultaneously, a uniform sampling time is defined. To satisfy The premise is launch delay Given .

[0031] Step 3: At the fusion center, sample the received molecular concentration signal to determine the observed concentration difference between type A and type B molecules; Specifically, the fusion center can distinguish between type A and type B molecules; when there are When the nano-emitters send signals to the fusion center according to equation (1), at the first... The first bit At each chip sampling time (where The observed concentration difference between type A and type B molecules can be derived and expressed as: (3)

[0032] in, This indicates a round-down operation. This represents the modulo operation; while and They represent the first The nano-emitter sends the first... Particle counting noise caused by different types of molecules during a molecular pulse.

[0033] When the number of molecules in each chip Or the total number of molecules in each symbol When large enough, and Both can be approximated as Gaussian noise with a mean of zero, and their variances are respectively as well as ,in, To accommodate the volume of the fused central spherical detector, the detection radius is... Therefore, these counting noises follow the following normal distribution: , .

[0034] According to equation (2) Characteristics, when a certain nano-emitter is at time After emitting a molecular pulse, the desired molecular concentration at the fusion center will be It reaches its maximum value at that time. Therefore, in order to detect from The first nano-emitter Each bit, the fusion center at time Molecular concentrations are sampled at various points, and correspondingly, the observed concentration difference between type A and type B molecules at these sampling times can be expressed as:

[0035] (4)

[0036] According to the pulse function As can be seen from the characteristics, inter-symbol interference (ISI) and on-chip interference decay rapidly over time. Therefore, we assume that the maximum duration of ISI / on-chip interference is... In this case, the chip in equation (4) It can be rewritten in the following form:

[0037] (5)

[0038] In equation (5), and Having the same statistical properties, similarly, and It also has the same statistical properties.

[0039] Therefore, equation (5) can be expressed in an equivalent form as: (6)

[0040] make Then equation (6) can be rewritten in a more manageable form: (7)

[0041] in, This term can be well approximated as noise following a Gaussian distribution, with the distribution being: ,in .

[0042] Step 4: Based on the concentration difference observations, expand them to correspond to MN chip observations; and insert cyclic prefixes between data blocks to eliminate inter-block interference. The fusion center preprocesses the discrete observations after removing the cyclic prefixes into time-domain observation vectors. Specifically, after the received signal is sampled and represented, the fusion center begins to process the signal based on a length of... Bit-by-bit detection is performed on the observed vector. For ease of detection, it is assumed that the data is transmitted in blocks, each block being of length [missing information]. . No. The data block transmitted by a nano-emitter can be represented as: In this case, for a given transmitted data block, its spread spectrum signal will be distributed in... On each observation, the range is from arrive .

[0043] (8)

[0044] When considering inter-block interference The front of the middle Each observation should be added from the previous data block. The interference consists of bits. This interference can be represented as... This proves that the current data block contains... All observations corresponding to each sampling time point are represented as follows:

[0045] (9)

[0046] in, It is a dimension The matrix is ​​of the form: (10)

[0047] The left block matrix (10) is a block of size 10. The matrix, whose beginning is filled with... One zero. In equation (9), ,here This represents the Kronecker product. Therefore, It is a size of The matrix. Includes all influence observation vectors bits, i.e. Furthermore, due to this Each observation has experienced a complete length of [length missing]. On-chip interference, observation noise vector It can be approximated as having a mean of zero and a covariance matrix of... The Gaussian distribution.

[0048] To eliminate inter-block interference and introduce cyclic characteristics into the channel state information matrix for frequency domain detection, a cyclic prefix is ​​inserted between two adjacent data blocks as a guard interval. The length of the cyclic prefix should be greater than the length of the on-chip interference. To completely cancel inter-block interference, its length can be set to the last block length after spread spectrum. Values.

[0049] After the spread spectrum operation is completed The end Each chip is copied and placed at the beginning of the data block. Subsequently, the nano-emitter sends a sequence containing... cyclic prefix chips and A complete data block of a normal data chip. After receiving it, the fusion center will... One chip observation is discarded, and the rest are retained. Several observations are used for signal processing. Due to the insertion of the aforementioned cyclic prefix, the channel state information matrix... Converted to a size of Circular matrix Its diagonal element is , denoted as:

[0050] (11)

[0051] Therefore, the observation equation (9) can now be written in the following form: (12)

[0052] in, It is a size of The matrix, by Constructed, that is .

[0053] Step 5: Convert the time-domain observation signal into a frequency-domain signal using a fast Fourier transform to obtain the frequency-domain observation vector; Specifically, to achieve frequency domain signal detection, the received signal needs to be converted to the frequency domain for processing. At the fusion center, the received time-domain discrete signal in equation (12) is... Through a The point-wise Fast Fourier Transform (FFT) converts a signal into the frequency domain. The FFT transform matrix is ​​defined as follows:

[0054] (13)

[0055] Among them, matrix The elements ( ) is defined as ,matrix It is a unitary matrix that satisfies ,in This represents the conjugate transpose operation, while It is called the inverse fast Fourier transform matrix.

[0056] The time-domain signal is transformed by the FFT transform matrix. Converting to the frequency domain signal yields its frequency domain representation. : (14)

[0057] matrix It is a circular matrix The diagonalized matrix can be represented as: (15)

[0058] Furthermore, the spread spectrum data block in the frequency domain is represented as ,in (16)

[0059] Indicates by the first Spread vectors transmitted by nano-emitters The One chip. The noise vector in the frequency domain is expressed as: , of which The component represents the first... The noise on each subcarrier is defined as:

[0060] (17)

[0061] Since the FFT / IFFT transformation is a unitary transformation, after the FFT transformation, It has zero mean and variance. .

[0062] Step 6: Apply the Minimum Mean Square Error Frequency Domain Equalization (MMSE-FED) algorithm to equalize the frequency domain signal; Specifically, after obtaining the frequency domain representation of the received signal, the fusion center performs frequency domain equalization before signal detection to mitigate the effects caused by inter-symbol interference and on-chip interference.

[0063] Based on the fundamental principle of linear frequency domain equalization, the equalized signal It can be represented as: (18)

[0064] in, It is a diagonal balanced matrix, which can be represented as ,in Corresponding to the The equalization coefficient of each subcarrier.

[0065] The diagonal equalization matrix is ​​determined using the principle of minimum mean square error (MSE) frequency domain equalization. To minimize the MSE, the equalization matrix... This can be derived by solving the following optimization problem:

[0066] (19)

[0067] The solution is: (20)

[0068] Therefore, it can be written as: (twenty one)

[0069] Here, diag represents a diagonal matrix.

[0070] When the channel remains constant, the matrix is ​​also fixed, so the computational complexity of minimum mean square error frequency domain equalization is also low.

[0071] Step 7: Using the balance matrix from Step 6 Multiply Obtain the equalized signal After performing the equalization operation, the IFFT matrix is... Multiply by the equilibrium result The signal is converted from the frequency domain back to the time domain; then, it is multiplied by... Complete the despreading to obtain the result used for detecting the first... Emits from a nano-emitter A decision variable of 1 bit; Specifically, we can obtain: (twenty two)

[0072] in

[0073] When using minimum mean square error frequency domain equalization, the equalization matrix of minimum mean square error frequency domain equalization is... Substituting into equation (22), at this point, after inverse fast Fourier transform and despreading, the th Nano-emitters The decision variable block can be represented as: (twenty three)

[0074] in, Represents a circular matrix .

[0075] During the judgment process, the first The first nano-emitter The decision variable corresponding to each bit is denoted as . It is dimensional vector The first in Each element. Based on this decision variable. It can detect the first The first nano-emitter 1 bit, as shown below:

[0076] (twenty four)

[0077] in ,

[0078] Step 8: Sort all nano-transmitters according to their received power at the fusion center from high to low or their distance from the fusion center from near to far. Select the first nano-transmitter in the sorted order as the priority nano-transmitter for detection. Based on the frequency domain observation vector, obtain the decision variable through minimum mean square error frequency domain equalization, inverse fast Fourier transform, and despreading. Use the hard decision result derived from the decision variable through the decision function as the bit sequence of the first nano-transmitter. Based on the detected bit sequence, reconstruct the signal contribution of the first nano-transmitter in the frequency domain and subtract it from the frequency domain observation vector to obtain the updated frequency domain reception. For the sorted k-th nano-emitter, where k is from 2 to K-1, perform iterative processing: based on the currently updated frequency domain received signal, obtain the decision variable through minimum mean square error frequency domain equalization, inverse fast Fourier transform, and despreading operations, and then detect the bit sequence of the k-th nano-emitter; reconstruct the signal contribution of the k-th nano-emitter and subtract it from the current frequency domain received signal to update the frequency domain received signal; after completing the interference cancellation of the first K-1 nano-emitters, detect the bit sequence of the k-th nano-emitter based on the finally updated frequency domain received signal through minimum mean square error frequency domain equalization, inverse fast Fourier transform, and despreading operations.

[0079] Specifically, the original observation values ​​are used as the frequency domain observation vector of the first nano-emitter; pass Determine the bit sequence of the first nanoemitter. The corresponding decision function operation is equation (24) in step 7.

[0080] The estimated value of the signal received by the first nanotransmitter on the p-th subcarrier is obtained based on the bit sequence of the first nanotransmitter. ;in, It is a length of The column vector is represented as The elements within it This indicates that the first nano-emitter was in the first... Estimates of the received signal on each subcarrier; The estimated value From length Subtract from the frequency domain observation vector to obtain the updated frequency domain observation vector. .

[0081] pass Determine the signal sent by the k-th nano-emitter A bit sequence, an estimated value, and an updated frequency domain observation vector, wherein... This represents the decision function.

[0082] For example, the signal emitted by the second nanoemitter can be detected using the same minimum mean square error frequency domain equalization, despreading, and decision operations. The expression for bits is: in, This represents the decision function defined in equation (24) in step 7.

[0083] Similarly, based on the updated observation vector It can detect the signal emitted by the third nano-emitter. bits, of which .

[0084] This invention proposes a molecular code division multiple access system based on binary molecular shift keying modulation (BSK) for establishing uplinks from multiple nanotransmitters to a single fusion center in diffuse molecular communication. The system assigns values ​​to all nanotransmitters. A unique extended sequence is used to enable parallel multiple access communication between them and the fusion center. By using a pair of isomers (type A and type B) with the same diffusion coefficient as information carriers, representing "+1" and "-1" respectively, this design can convert the counting noise at the receiver into Gaussian additive noise with constant variance under time-invariant channel conditions, thereby simplifying detection modeling and analysis.

[0085] The molecular code division multiple access system proposed in this invention transmits data in blocks and avoids inter-block interference by using a cyclic prefix. It shifts the signal detection process from traditional time-domain processing to frequency-domain processing, realizing matrix batch processing of signals, reducing the computational complexity of the receiver, and improving detection efficiency and anti-inter-code crosstalk capability. Thus, it achieves a balance between low complexity and high reliability in molecular communication systems.

[0086] To address the prevalent inter-symbol interference (ISI) and inter-chip interference (ICI) in molecular code division multiple access (MCD) signals, and considering the increasing counting noise with the number of nano-emitters, this invention proposes a low-complexity frequency domain equalization algorithm—Minimum Mean Square Error Frequency Domain Equalization. This method effectively suppresses ISI and ICI in molecular CCD systems with relatively low computational complexity.

[0087] Furthermore, the single-stage decision feedback algorithm proposed in this invention can be used in conjunction with the above equalizer to further combat multiple access interference in molecular code division multiple access received signals, thereby achieving efficient and reliable multi-user signal detection in a nano-communication network environment.

[0088] This invention also provides a molecular code division multiple access system (MMSE-FED) and a decision feedback de-interference device, the device comprising: K nano-emitters, each configured to use binary molecular shift-bonding modulation; The fusion center, connected to the nano-emitter via a diffusion medium, is configured as follows: Based on the concentration difference observations, MN chip observations are generated after expansion; a cyclic prefix is ​​inserted between data blocks to eliminate inter-block interference. The fusion center preprocesses the discrete observations after removing the cyclic prefix into a time-domain observation vector; the time-domain observation signal is converted to a frequency-domain signal using a fast Fourier transform to obtain a frequency-domain observation vector; all nano-emitters are sorted according to their received power at the fusion center from high to low or their distance from the fusion center from near to far, and the first nano-emitter in the sorted order is selected as the priority detection nano-emitter. Based on the frequency-domain observation vector, a decision variable is obtained through minimum mean square error frequency domain equalization, inverse fast Fourier transform, and despreading; the hard decision result derived from the decision variable is used as the bit sequence of the first nano-emitter using a decision function. Based on the detected bit sequence, in... The signal contribution of the first nanotransmitter is reconstructed in the frequency domain and subtracted from the frequency domain observation vector to obtain the updated frequency domain received signal. For the k-th nanotransmitter in the sorted sequence, where k is from 2 to K-1, iterative processing is performed: based on the current updated frequency domain received signal, the decision variables are obtained through minimum mean square error frequency domain equalization, inverse fast Fourier transform, and despreading operations, and the bit sequence of the k-th nanotransmitter is detected. The signal contribution of the k-th nanotransmitter is reconstructed and subtracted from the current frequency domain received signal to update the frequency domain received signal. After completing the interference cancellation of the first K-1 nanotransmitters, the bit sequence of the k-th nanotransmitter is detected based on the final updated frequency domain received signal through minimum mean square error frequency domain equalization, inverse fast Fourier transform, and despreading operations. For each bit of each nanotransmitter, its decision variables are compared. Zero threshold: If the decision variable is greater than 0, then the bit is judged to be 1; otherwise, it is judged to be -1, and the decoded bit sequence is output. As the final test result.

[0089] The technical features of the above embodiments can be combined in any way. For the sake of brevity, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0090] The embodiments described above are merely illustrative of several implementation methods of this application, and while the descriptions are specific and detailed, they should not be construed as limiting the scope of this patent application. It should be noted that those skilled in the art can make various modifications and improvements without departing from the concept of this application, and these all fall within the protection scope of this application. Therefore, the protection scope of this patent application should be determined by the appended claims.

Claims

1. A molecular code division multiple access system (MMSE-FED) and a decision feedback method for interference removal, characterized in that, The method includes: A molecular code division multiple access system is constructed, comprising K nano-emitters and a fusion center. The nano-emitters transmit information molecules through a diffusion medium, and the fusion center is used for signal detection. At the k-th nano-emitter, data is organized using a block transfer method, with each data block being M bits in length. Each nano-emitter is assigned a unique address code or extended sequence of length N. A single bit is expanded into N chip bits after address code expansion, and binary molecular shift keying is used to modulate and release two types of molecular pulses, A and B. At the fusion center, the received molecular concentration signal is sampled to determine the observed concentration difference between type A and type B molecules; Based on the concentration difference observations, the expanded values ​​correspond to MN chip observations; and a cyclic prefix is ​​inserted between data blocks to eliminate inter-block interference. The fusion center preprocesses the discrete observations into time-domain observation vectors after removing the cyclic prefix. The time-domain observation signal is converted into a frequency-domain signal using a fast Fourier transform to obtain a frequency-domain observation vector; All nanotransmitters are sorted according to their received power at the fusion center from high to low or their distance from the fusion center from near to far. The first nanotransmitter in the sorted order is selected as the priority nanotransmitter for detection. Based on the frequency domain observation vector, decision variables are obtained through minimum mean square error frequency domain equalization, inverse fast Fourier transform, and despreading. The hard decision result derived from the decision variables is used as the bit sequence of the first nanotransmitter through the decision function. Based on the detected bit sequence, the signal contribution of the first nanotransmitter is reconstructed in the frequency domain and subtracted from the frequency domain observation vector to obtain the updated frequency domain received signal. For the sorted k-th nano-emitter, where k is from 2 to K-1, perform iterative processing: based on the currently updated frequency domain received signal, obtain the decision variable through minimum mean square error frequency domain equalization, inverse fast Fourier transform, and despreading operations, and then detect the bit sequence of the k-th nano-emitter; reconstruct the signal contribution of the k-th nano-emitter and subtract it from the current frequency domain received signal to update the frequency domain received signal; after completing the interference cancellation of the first K-1 nano-emitters, detect the bit sequence of the k-th nano-emitter based on the finally updated frequency domain received signal through minimum mean square error frequency domain equalization, inverse fast Fourier transform, and despreading operations; For each bit of each nanoemitter, compare its decision variables. Zero threshold: If the decision variable is greater than 0, then the bit is judged to be 1; otherwise, it is judged to be -1, and the decoded bit sequence is output. As the final test result.

2. The molecular code division multiple access system MMSE-FED and decision feedback de-interference method according to claim 1, characterized in that, The process of inserting cyclic prefixes between data blocks to eliminate inter-block interference, and the preprocessing of discrete observations into time-domain observation vectors by the fusion center after removing the cyclic prefixes, specifically includes: through... Determine the time-domain observation vector, where, , That is, a single molecular pulse in continuous Concentration values ​​at each sampling time on the chip. It is a size of The matrix, by Constructed, that is , yes The noise vector.

3. The molecular code division multiple access system MMSE-FED and decision feedback de-interference method according to claim 1 or 2, characterized in that, The step of converting the time-domain observation signal into a frequency-domain signal using a Fast Fourier Transform (FFT) to obtain a frequency-domain observation vector specifically includes: transforming the time-domain observation vector using an FFT transform matrix. Convert to frequency domain observation vector in, The Fourier transform matrix is... It is a circular matrix The diagonalized matrix can be represented as: .

4. The molecular code division multiple access system MMSE-FED and decision feedback interference removal method according to claim 3, characterized in that, The step of using the hard decision result derived from the decision variable through the decision function as the bit sequence of the first nano-emitter specifically includes: through... Determine the bit sequence of the first nanoemitter, where It is an equilibrium matrix generated using the MMSE-FED principle.

5. The molecular code division multiple access system MMSE-FED and decision feedback de-interference method according to claim 4, characterized in that, The method further includes: obtaining an estimate of the signal received by the first nanotransmitter on the p-th subcarrier based on the bit sequence of the first nanotransmitter. ;in, It is a length of The column vector is represented as The elements within it This indicates that the first nano-emitter was in the first... Estimates of the received signal on each subcarrier; The estimated value From length Subtract from the frequency domain observation vector to obtain the updated frequency domain observation vector. .

6. The molecular code division multiple access system MMSE-FED and decision feedback de-interference method according to claim 5, characterized in that, The process involves obtaining the decision variable based on the currently updated frequency domain received signal through minimum mean square error frequency domain equalization, inverse fast Fourier transform, and despreading operations, then detecting the bit sequence of the k-th nano-emitter; reconstructing the signal contribution of the k-th nano-emitter and subtracting it from the current frequency domain received signal to update the frequency domain received signal, specifically including: through... Determine the signal sent by the k-th nano-emitter A bit sequence, an estimated value, and an updated frequency domain observation vector, wherein... This represents the decision function.

7. The molecular code division multiple access system MMSE-FED and decision feedback de-interference method according to claim 6, characterized in that, The method further includes: employing binary molecular shift keying modulation, wherein each nano-emitter spreads the data using a spreading sequence with values ​​of {+1, -1}, and releases molecular concentration signals representing bit information by releasing type A and type B molecules; According to claim 7, the molecular code division multiple access system MMSE-FED and decision feedback de-interference method are characterized in that, the binary molecular shift keying modulation is adopted, and each of the nano-emitters uses a spreading sequence with values ​​{+1, -1} to spread the data, specifically including: in molecular code division multiple access, the first... The information to be transmitted by each nano-emitter is represented as follows: ,in Each nano-emitter sends the first Each bit is passed through a length of The spread spectrum sequence is extended, and this sequence can be represented as... ,in ,and When using binary molecular shift keying modulation, each pulse is emitted... The first molecule, type A molecule represents '1', type B molecule represents '-1', and the second molecule... The nano-emitter transmits the first... When the data is bit-sized, its molecular emission signal can be expressed as: in It is a nano-emitter At the moment of molecular pulse emission, the pulses of type A and type B molecules are respectively generated by... and It is represented and defined as: ( ), and when hour, ( ).

8. The molecular code division multiple access system MMSE-FED and decision feedback de-interference method according to claim 8, characterized in that, The method further includes: performing equalization processing on the frequency domain signal by applying a minimum mean square error frequency domain equalization algorithm, specifically including: through... Obtain the equalized signal ,in, It is a diagonal balanced matrix, which can be represented as , Corresponding to the The equalization coefficient of each subcarrier.

9. A molecular code division multiple access system (MMSE-FED) and a decision feedback interference removal device, characterized in that, The device includes: K nano-emitters, each configured to: transmit information molecules through a diffusion medium and employ binary molecular shift keying modulation; spread the data using a spreading sequence with values ​​{+1, -1}, and release molecular concentration signals representing bit "1" by type A molecules and bit "-1" by type B molecules; A fusion center, connected to the nano-emitter via a diffusion medium, is configured to: sample the received molecular concentration signal and determine the observed concentration difference between type A and type B molecules; Based on the concentration difference observations, MN chip observations are generated after expansion; a cyclic prefix is ​​inserted between data blocks to eliminate inter-block interference. The fusion center preprocesses the discrete observations after removing the cyclic prefix into a time-domain observation vector; the time-domain observation signal is converted to a frequency-domain signal using a fast Fourier transform to obtain a frequency-domain observation vector; all nano-emitters are sorted according to their received power at the fusion center from high to low or their distance from the fusion center from near to far, and the first nano-emitter in the sorted order is selected as the priority detection nano-emitter. Based on the frequency-domain observation vector, a decision variable is obtained through minimum mean square error frequency domain equalization, inverse fast Fourier transform, and despreading; the hard decision result derived from the decision variable is used as the bit sequence of the first nano-emitter using a decision function. Based on the detected bit sequence, in... The signal contribution of the first nanotransmitter is reconstructed in the frequency domain and subtracted from the frequency domain observation vector to obtain the updated frequency domain received signal. For the k-th nanotransmitter in the sorted sequence, where k is from 2 to K-1, iterative processing is performed: based on the current updated frequency domain received signal, the decision variables are obtained through minimum mean square error frequency domain equalization, inverse fast Fourier transform, and despreading operations, and the bit sequence of the k-th nanotransmitter is detected. The signal contribution of the k-th nanotransmitter is reconstructed and subtracted from the current frequency domain received signal to update the frequency domain received signal. After completing the interference cancellation of the first K-1 nanotransmitters, the bit sequence of the k-th nanotransmitter is detected based on the final updated frequency domain received signal through minimum mean square error frequency domain equalization, inverse fast Fourier transform, and despreading operations. For each bit of each nanotransmitter, its decision variables are compared. Zero threshold: If the decision variable is greater than 0, then the bit is judged to be 1; otherwise, it is judged to be -1, and the decoded bit sequence is output. As the final test result.

Images