Eye pattern generation method and system for die-to-die interface

Abstract

The invention provides an eye pattern generation method and system for a real-time full-duplex D2D interface, and the method comprises the steps: simulating all combinations of self-interference noise in a simulation model of a to-be-tested full-duplex D2D interface, and obtaining an edge response corresponding to each combination; establishing a probability density function of the self-interference noise based on the collected edge response; and combining the probability density function of the self-interference noise with the probability density functions of intersymbol interference, crosstalk and jitter to determine the bit error rate of the full-duplex D2D interface, thereby generating a bit error rate statistical eye pattern. In the scheme of the invention, the self-interference noise is introduced to reflect the interference between the sending signal and the receiving signal on the same side of the transmission channel of the full-duplex D2D interface, so that the accuracy of the eye pattern is improved, and the signal integrity analysis of the full-duplex D2D interface is improved.

Description

Technical Field

[0001] The present application relates to the field of chip design, and in particular to an eye diagram generation method and system for a die-to-die interface. Background Art

[0002] The D2D (die-to-die) interface refers to a functional module that provides a data interface between two chip dies in the same package, which can establish a reliable data link between the two dies. With the rapid development of high-performance computing and artificial intelligence applications, the amount of data that needs to be processed by chips has increased rapidly, which puts high demands on the transmission speed and bandwidth utilization of the D2D interface between dies. Using real-time full-duplex solutions to improve channel utilization is the future development trend of D2D interfaces.

[0003] In the design of D2D interfaces, it is necessary to consider various factors that affect the quality of signal transmission. Eye diagrams are currently widely used to help designers evaluate, correct and optimize signal integrity and signal quality. High-quality eye diagrams are essential to ensure the stability and reliability of the designed D2D interface. At present, there is an urgent need for an eye diagram generation solution for D2D interfaces that support real-time full-duplex.

[0004] It should be noted that the above content is only used to help understand the technical solution of the present application and should not be used as a basis for evaluating the prior art of the present application. Summary of the invention

[0005] This application provides a solution for quickly and accurately generating a statistical eye diagram for a real-time full-duplex D2D interface.

[0006] According to the first aspect of the embodiment of the present application, a method for generating a statistical eye diagram of a real-time full-duplex D2D interface is provided. The method mainly includes: simulating all code combinations of self-interference noise in a simulation model of the full-duplex D2D interface to be tested, and obtaining the edge response corresponding to each code combination; determining the probability density function of the self-interference noise based on the obtained edge response; and determining the probability density function of the transmission channel by convolving the probability density function of the self-interference noise with the probability density functions of inter-symbol interference, crosstalk, and jitter, thereby generating a statistical eye diagram. The self-interference noise refers to the interference between the transmitted signal and the received signal on the same side of the transmission channel of the full-duplex D2D interface.

[0007] In the above scheme, by modeling the interference between the transmitted signal and the received signal on the same side of the transmission channel of the full-duplex D2D interface (i.e., self-interference noise), the impact of this self-interference noise on the signal on the transmission channel is accurately obtained, and then such self-interference noise is combined with inter-symbol interference, crosstalk, and jitter to generate a statistical eye diagram, which is conducive to more efficient and accurate signal integrity analysis of the full-duplex D2D interface.

[0008] In some embodiments, the method also determines the bit error rate of the transmission channel by integrating the probability density function of the transmission channel based on a given decision threshold, thereby generating a bit error rate statistical eye diagram. Compared with the existing method of calculating the bit error rate based on inter-symbol interference, crosstalk and jitter, the solution of the present application also considers the influence of the interference between the transmitted signal and the received signal on the same side of the transmission channel on the signal quality when calculating the bit error rate, so it can more accurately reflect the signal quality of the full-duplex D2D interface.

[0009] In some embodiments, the number of code pattern combinations of self-interference noise is determined by the number and type of code elements sent by the transmitters on both sides of the transmission channel; wherein the first transmitter sends two consecutive code elements, and the second transmitter sends one code element at the same time, and the received signal is monitored at the first receiving end on the same side as the first transmitter to obtain an edge response. Wherein, each code element is "0" or "1", so the number of code pattern combinations used to simulate self-interference noise is 8, and the probability of each code pattern combination occurring is 0.25. During simulation, the signals corresponding to each code pattern combination can be loaded into the simulation model for simulation, and the signal waveform of the edge response corresponding to each code pattern combination can be obtained at the first receiving end. Through the small number of simulation code pattern combinations in this embodiment, the response behavior of the transmission channel to self-interference noise can be quickly and accurately obtained, which saves simulation time and is conducive to improving the efficiency of eye diagram generation.

[0010] In some embodiments, a leading sequence and an ending sequence of predetermined length may be set before and after the signal corresponding to each code type combination. The leading sequence and the ending sequence are sequences consisting of a plurality of consecutive identical code elements, thereby enabling the starting point of the edge response waveform to be detected more quickly and accurately.

[0011] In some embodiments, determining the probability density function of self-interference noise based on the acquired edge response may include performing the following operations for the edge response of each code type combination: determining the starting point of the signal waveform of the edge response, dividing the signal waveform into multiple unit intervals UI, sampling multiple points in each UI at a set sampling rate, and each sampling point is represented by a sampling moment and its corresponding voltage amplitude; superimposing the sampling points corresponding to the same sampling moment in each UI to obtain the probability density distribution of the sampling moment; accumulating the probability density distribution of each sampling moment to obtain the probability density distribution corresponding to the edge response of the code type combination. Then, the probability density function of the self-interference noise is obtained based on the convolution of the probability of occurrence of each code type combination and the probability density distribution of the edge response of each code type combination. In the above manner, the impact of self-interference noise on the signal quality of the D2D interface transmission channel can be accurately evaluated.

[0012] In some embodiments, the method may further include: receiving parameters about the phase setting between the transmitting end and the receiving end on the same side; in response to the parameter indicating a random phase, before calculating the probability density function of the transmission channel, further uniformly distributing the probability density function of the self-interference noise; in response to the parameter indicating a fixed phase, before calculating the probability density function of the transmission channel, further convolving the probability density function of the self-interference noise with a dual Dirac function. In the above embodiment, when simulating the self-interference noise, the possible impact of the phase setting between the transmitting end and the receiving end on the same side on the signal is also taken into account, so that the impact of the self-interference noise on the signal quality of the D2D interface transmission channel can be more accurately evaluated, which is conducive to further improving the quality of the generated eye diagram.

[0013] According to the second aspect of the embodiment of the present application, a system for generating a statistical eye diagram of a real-time full-duplex D2D interface is provided, which includes a simulation data collection module and an eye diagram generation module. The simulation data collection module is used to simulate all code combinations of self-interference noise in the simulation model of the full-duplex D2D interface to be tested, and obtain the edge response corresponding to each code combination, wherein the self-interference noise refers to the interference between the transmitted signal and the received signal on the same side of the transmission channel of the full-duplex D2D interface. The eye diagram generation module is used to determine the probability density function of the self-interference noise based on the edge response obtained, and to combine the probability density function of the self-interference noise with the probability density function of inter-code interference, crosstalk, and jitter to determine the probability density function of the transmission channel, thereby generating a statistical eye diagram. In some embodiments, the eye diagram generation module is also configured to determine the bit error rate of the transmission channel by integrating the probability density function of the transmission channel based on a given decision threshold, thereby generating a bit error rate statistical eye diagram.

[0014] Compared with the prior art, the solution of the present application not only takes into account the impact of inter-symbol interference, crosstalk, and jitter on signal quality when generating a statistical eye diagram for a real-time full-duplex D2D interface, but also obtains the impact of such self-interference noise on the signal on the transmission channel through accurate simulation of the interference between the transmitted signal and the received signal on the same side of the transmission channel of the full-duplex D2D interface (i.e., self-interference noise), thereby improving the quality of the generated eye diagram and facilitating more efficient and accurate signal integrity analysis of the full-duplex D2D interface.

[0015] It should be understood that the foregoing general description and the following detailed description are exemplary and explanatory only and are not restrictive of the present application. BRIEF DESCRIPTION OF THE DRAWINGS

[0016] The drawings herein are incorporated into the specification and constitute a part of the specification, showing some embodiments consistent with the present application, and together with the specification, are used to explain the principles of the present application. Obviously, the drawings described below are only some embodiments of the present application, and for ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work. In the drawings:

[0017] Figure 1 Schematic diagram of self-interference noise in a full-duplex D2D interface;

[0018] Figure 2 A schematic diagram of a code pattern combination for simulating self-interference noise according to an embodiment of the present application;

[0019] Figure 3 The present invention is a flowchart of a method for generating an eye diagram of a full-duplex D2D interface according to an embodiment of the present application. DETAILED DESCRIPTION

[0020] In order to make the purpose, technical solutions and advantages of this application more clear, the present application is further described in detail by specific embodiments in conjunction with the accompanying drawings. It should be understood that the described embodiments are part of the embodiments of this application, rather than all of the embodiments. Based on the embodiments in this application, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of this application.

[0021] In addition, described feature, structure or characteristic can be combined in one or more embodiments in any suitable manner. In the following description, many specific details are provided to provide a full understanding of the embodiments of the present application. However, those skilled in the art will appreciate that the technical scheme of the present application can be put into practice without one or more of the specific details, or other methods, components, devices, steps, etc. can be adopted. In other cases, known methods, devices, realizations or operations are not shown or described in detail to avoid blurring the various aspects of the application.

[0022] The block diagrams shown in the accompanying drawings are merely functional entities and do not necessarily correspond to physically independent entities. That is, these functional entities may be implemented in software form, or in one or more hardware modules or integrated circuits, or in different networks and / or processor devices and / or microcontroller devices.

[0023] The flowcharts shown in the accompanying drawings are only exemplary and do not necessarily include all the contents and operations / steps, nor must they be executed in the order described. For example, some operations / steps can be decomposed, and some operations / steps can be combined or partially combined, so the actual execution order may change according to actual conditions.

[0024] The die-to-die (D2D) interface is a functional block that provides a data interface between two chip dies in the same package, which can establish a reliable data link between the two dies. With the rapid development of high-performance computing and artificial intelligence applications, the amount of data that needs to be processed by chips has increased rapidly, which has put forward high requirements on the transmission speed and bandwidth utilization of the D2D interface between dies. The D2D interface has become one of the important factors that restrict and affect chip performance.

[0025] Using single-pin full-duplex technology to improve channel utilization is the development trend of D2D interfaces in the future. Single-pin full-duplex aims to allow simultaneous transmission (TX) and reception (RX) operations on one physical pin, thereby significantly improving the efficiency of data transmission and bandwidth utilization under a limited number of physical pins, while reducing the number of physical pins required for the chip to a certain extent, which is of great significance for reducing chip size and cost. In order to achieve real-time two-way communication on a single pin, some form of signal modulation technology is usually used; for example, different frequencies or phases can be used to distinguish between transmitted and received signals. But this also brings complexity to the design of D2D interfaces. In the design of D2D interfaces, it is necessary to consider a variety of factors that affect the quality of signal transmission, such as inter-symbol interference, crosstalk, jitter, etc.

[0026] Eye diagrams are commonly used in D2D interface design as a tool for evaluating and analyzing signal quality, so as to timely correct and optimize the D2D interface design. Eye diagrams are "eye"-shaped graphics formed by accumulating and superimposing signal waveforms over a long period of time, which can intuitively reflect the various changes in signal edges and voltage amplitudes. High-quality eye diagrams are essential to ensure the stability and reliability of the designed D2D interface. In the design process of D2D interfaces, the rapid generation of high-quality eye diagrams can help designers quickly evaluate signal quality, discover potential problems in the design in a timely manner, and optimize the design accordingly, thereby reducing rework and testing costs caused by design problems in the later stage. In short, the rapid generation of high-quality eye diagrams is conducive to accelerating the efficiency of D2D interface design verification, improving design quality, reducing costs, and accelerating product time to market.

[0027] At present, eye diagrams are mainly generated by time domain simulation and statistical domain simulation. When performing time domain eye diagram simulation, in order to obtain highly accurate results, it is necessary to consider various code stream combinations and use a very long bit stream as input information for simulation, which takes up a lot of simulation time. However, the D2D interface is different from the traditional communication system. It has ultra-high transmission speed and ultra-low bit error rate. Taking the commonly used UCIE (Universal Chiplet Interconnect Express) industry standard as an example, it requires the bit error rate (BER) to be set to 1e-15 and below, which means that only a maximum of 1 error bit is allowed in every 1 trillion bits. If the time domain simulation method is used to form the eye diagram of the D2D interface, it takes too long to accumulate time, and the total amount of data processed and stored is too large, resulting in very low simulation test efficiency. Therefore, for the ultra-high-speed and low bit error rate signal transmission channel provided by the D2D interface, the statistical domain simulation method is usually used to quickly generate a statistical eye diagram. The statistical eye diagram is usually a three-dimensional eye diagram, which can not only show the voltage amplitude that may appear at each sampling moment, but also show the probability of the voltage amplitude corresponding to the sampling moment. The statistical eye diagram is obtained by obtaining the response distribution of the link through statistical calculation and analysis based on the response of some specific code patterns of the known link.

[0028] For ease of understanding, before introducing the specific embodiments of the present application, some relevant knowledge about statistical domain simulation of eye diagrams is briefly introduced.

[0029] The theoretical basis of the statistical domain simulation method is the linear time-invariant (LTI) characteristic of the transmission channel. The response of the LTI system to the input signal is proportional, satisfying the superposition principle, so the output signal of the system can be predicted by simple algebraic operations. In addition, the statistical characteristics of the output signal of the LTI system do not change over time. The current output depends only on the current and past input values, not on future input values. All the characteristics of the LTI system can be fully described by its unit impulse response δ(t) (also known as the Dirac delta function), that is, for any input signal x(t), the system's output signal y(t) is the convolution of the input signal x(t) and the system's impulse response δ(t).

[0030] The ideal situation of channel transmission usually refers to the signal quality and integrity being fully maintained without any interference and loss, without any bit error or distortion. However, due to the interference of various channel noises, the signal waveform will be wrong or distorted during the transmission process. The noise that interferes with the signal usually includes inter-symbol interference (ISI), total jitter (TJ), crosstalk interference (CCI), etc. The eye diagram can intuitively reflect the impact of ISI, TJ, CCI and other noises on the signal quality in a graphical way. The statistical domain simulation method usually needs to model the impact of these noises on the signal to obtain the probability density function (PDF) of the corresponding noise, so as to generate a statistical eye diagram based on these PDFs. In addition, based on the PDFs corresponding to these noises, the bit error rate of the system can be quickly calculated to obtain the bit error rate statistical eye.

[0031] The bit error rate is a key indicator for evaluating the quality of signal transmission. It indicates the ratio of the number of error bits received to the total number of bits during data transmission. The relationship between noise and bit error rate is related to the decision level V ref When the transmitter sends the symbol "1", any voltage that appears in the area below the decision level at the receiver will be misjudged as 0, resulting in a bit error. The sum of the probabilities of these bit errors corresponds to a bit error rate of 1. Similarly, when the transmitter sends the symbol "0", any voltage that appears at the receiver that is higher than the decision level due to noise will be misjudged as 1, resulting in a bit error. The sum of the probabilities of these bit errors corresponds to a bit error rate of 0. In this way, the system's bit error rate is the sum of the bit error rate of 1 and the bit error rate of 0.

[0032] For example, assuming the sampling time is t = t0 and the decision level is V ref, the probability density function of a certain noise when sending codeword "1" is recorded as pdf1(Δv), and the probability density function when sending codeword "0" is recorded as pdf0(Δv), P1 is the probability of sending codeword "1", P0 is the probability of sending codeword "0", then the bit error rate F1 of bit "1" under the influence of this noise is:

[0033]

[0034] Similarly, the bit error rate F0 of bit "0" under the influence of this noise is:

[0035]

[0036] Then, the sum of the bit error rates of bit "1" and bit "0" is the system bit error rate F at sampling time t1. 10 :

[0037]

[0038] In the statistical domain simulation method, obtaining the probability density function PDF that can reflect the interference of noise on the signal through statistical analysis is a very critical step. The process of determining the probability density function mainly includes the following steps: (a) constructing a code pattern that simulates a certain noise, loading the signal sequence corresponding to the code pattern into the simulation model for simulation and obtaining the edge response; (b) performing statistical analysis on the edge response to determine the probability density function. For example, first determine the UI where the main cursor is located from the edge response data. The UI here refers to the duration of a bit (its time length can be recorded as T), which is determined by the data transmission rate of the system. The UI where the main cursor is located can be determined by finding the maximum voltage in the edge response as the main cursor, and taking two points with the closest voltage values ​​and a time interval of T on the left and right of the main cursor. For another example, the location where the edge jumps can also be used as the main cursor area. After determining the main cursor area, the edge response waveform is divided into multiple UI intervals in units of UI length, and each UI interval is sampled according to the set sampling frequency to obtain a set of sampling data. This set of sampling data can be understood as a set of voltage values ​​corresponding to each sampling moment, that is, a row of voltage values ​​spaced at a fixed time (i.e., sampling interval), which can also be regarded as a vector. The sampling frequency can be set according to the specific analysis accuracy requirements. According to the linear time-invariant characteristics, the response waveform output at a certain sampling moment (expressed as the probability of the voltage value) is a linear fit or superposition of the response waveforms at all sampling moments before the sampling moment. Therefore, the sampling data of each UI can be superimposed together in the form of convolution, traversing each sampling moment, and calculating the probability distribution of each cursor (i.e., each voltage value of the edge response collected) at each sampling moment, thereby obtaining the probability density function PDF of the noise (which can be understood as a function of time and voltage), so that a statistical domain eye diagram representing the probability distribution of the voltage at each sampling moment can be drawn.

[0039] It should be understood that the introduction of the above-mentioned related knowledge is only a schematic introduction to the common process of statistical domain simulation, which is only used to help understand the technical solution of the present application, and does not impose any limitation on the embodiments of the present application.

[0040] The existing noise probability density function for the D2D interface is obtained by estimating the probability density functions of inter-symbol interference, crosstalk, and jitter respectively and convolving each probability density function. However, the inventors found that such a probability density function cannot accurately reflect the signal transmission quality of the real-time full-duplex D2D interface.

[0041] like Figure 1As shown in the figure, in a full-duplex D2D interface, each side of the signal transmission channel can be used as a transmitter (TX) and a receiver (RX) at the same time. At each TX end, the main driver (Main-Drv) converts the digital signal into a suitable electrical signal and drives it to the transmission channel, and the auxiliary driver (Rep-drv) is used to eliminate the transmission signal, thereby obtaining a relatively clean receiving signal under the same pin pad. Figure 1 The full-duplex communication channel shown in comprises a first transmitter TX0 and a first receiver RX0 as well as a second transmitter TX1 and a second receiver RX1. The signal sent by the first transmitter TX0 is received by the first receiver RX0, and the signal sent by the second transmitter TX1 is received by the second receiver RX1. The first transmitter TX0 and the second receiver RX1 are located on one side of the channel, while the second transmitter TX1 and the first receiver RX0 are located on the other side of the channel; that is, TX0 is adjacent to RX1, and TX1 is adjacent to RX0.

[0042] The inventors found that the signal quality received at each RX end is not only affected by the inter-code interference, crosstalk, and jitter during the transmission of the signal from the transmission to the reception channel, but also by the interference of the adjacent TX end signal. In this application, the interference between the TX signal and the RX signal on the same side of the transmission channel is referred to as self-interference noise. In the embodiments of the present application, the impact of self-interference noise on the signal is accurately estimated, and a probability density function for obtaining self-interference noise and a solution for generating a statistical eye diagram of a full-duplex D2D interface based on a probability density function that takes into account inter-code interference, crosstalk, jitter, and self-interference noise are provided.

[0043] For the convenience of discussion, the following settings are used as examples in this application, but are not intended to impose any restrictions: a code element carries one bit of information, such as "0" or "1"; the duration of a code element (or a bit) is taken as the time period of the eye diagram, which is the unit interval (UI); the ideal sampling point (main cursor) is the center of the code element, that is, half of the unit interval, to ensure that the bit information carried by the code element can be accurately read; the sampling is triggered at the rising and falling edges of the code element (also known as the double-edge response) to facilitate synchronization or adjustment of the sampling time; the code element is sent using the non-return to zero coding NRZ (Non-Return to Zero) mechanism; the code type indicates a sequence consisting of multiple code elements. It should be understood that the above settings can be adaptively modified according to different communication standards of the D2D interface, but will not affect the application and implementation of the solution provided in this application.

[0044] The inventors found during the research that the self-interference noise generated by the different states of the TX signal and the common mode point of the RX signal on the same side of the transmission channel is different, and the self-interference noise under different signal states needs to be considered. Therefore, if the probability density function PDF of the self-interference noise is to be accurately estimated, the influence of the self-interference noise on the signal under different signal states needs to be considered.

[0045] Figure 2 An example of the classification of signal states according to an embodiment of the present application is given. When a codeword is sent in the transmission channel of the D2D interface, the codeword corresponds to a pulse excitation with a length of one UI. This pulse excitation will be interfered by various factors on the channel. Therefore, the pulse response received at the receiving end may be deformed. In one embodiment of the present application, the second receiving end RX1 is used as an observation reference point to observe how the signal from TX1 received at RX1 is affected by the signal transmitted from TX0 adjacent to RX1. It has been found that the signal waveform of each codeword received at RX1 is affected by at least the current codeword and the previous codeword transmitted by its adjacent TX0. In an embodiment of the present application, the signal states that may affect the received signal at RX1 due to self-interference noise are divided into the following types:

[0046] 000 state: TX1 transmits bit "0", and the adjacent TX0 signal has no jump, and the output is 00;

[0047] 110 state: TX1 transmits bit "0", and the adjacent TX0 signal has no jump, and the output is 11;

[0048] F0 state: TX1 transmits bit "0", and the adjacent TX0 signal has a jump, and the output is 01;

[0049] R0 state: TX1 transmits bit "0", and the adjacent TX0 signal jumps, and the output is 10;

[0050] 001 state: TX1 transmits bit "1", and the adjacent TX0 signal has no jump, and the output is 00;

[0051] 111 state: TX1 transmits bit "1", and the adjacent TX0 signal has no transition, so the output is 11;

[0052] F1 state: TX1 transmits bit "1", and the adjacent TX0 signal changes, and the output is 01;

[0053] R1 state: TX1 transmits bit "1", and the adjacent TX0 signal changes, and the output is 10;

[0054] like Figure 2As shown, in the absence of coding, the probability that the code element received from TX1 at RX1 is "0" or "1" is 0.5, and there are four combinations between two adjacent code elements sent at TX0, "00", "11", "01 (R)", "10 (F)", and the probability of each combination is 0.25. Therefore, the probability that the above states may appear is 1 / 8 (i.e. 0.5*0.25). In the subsequent eye diagram generation process, it is necessary to perform system response simulation on the above states respectively, collect relevant signal data at RX1, and obtain the probability density function of self-interference noise in each state based on the analysis of the collected signal. It should be understood that if RX0 is used as the observation reference point, it is only necessary to exchange the content transmitted by TX1 and TX0 in the above states.

[0055] Figure 3 The process of a method for generating an eye diagram of a full-duplex D2D interface according to an embodiment of the present application is given. The method mainly includes the following steps: S1) simulating all code combinations of self-interference noise in a simulation model of the full-duplex D2D interface to be tested, and obtaining the edge response corresponding to each code combination; S2) determining the probability density function of the self-interference noise based on the obtained edge response; S3) combining the probability density function of the self-interference noise with the probability density functions of inter-symbol interference, crosstalk, and jitter to determine the probability density function of the full-duplex D2D interface, and then generating a statistical eye diagram.

[0056] More specifically, in step S1), according to Figure 2 The signal state shown constructs all code combinations for simulating self-interference noise, that is, eight code combinations need to be simulated. The correspondence between each signal state and code combination is as follows:

[0057] 000 state: TX1 code type 0, TX0 code type 00;

[0058] 110 state: TX1 code type 0, TX0 code type 11;

[0059] F0 state: TX1 code type 0, TX0 code type 01;

[0060] R0 state: TX1 code type 0, TX0 code type 10;

[0061] 001 state: TX1 code type 1, TX0 code type 00;

[0062] 111 state: TX1 code type 1, TX0 code type 11;

[0063] F1 state: TX1 code type 1, TX0 code type 01;

[0064] R1 status: TX1 code type 1, TX0 code type 10;

[0065] In another embodiment, in order to obtain a complete edge response more accurately, a leading sequence and an ending sequence may be added before and after each code pattern. The leading sequence and the ending sequence may be a sequence consisting of a preset number of consecutive code elements "0" or "1". The code pattern 1 to be simulated may be "00000000 1 00000000", code type 01 can be "00000000 01 00000000".

[0066] The code patterns corresponding to each transmitter in the above code pattern combination are loaded into the two transmitters (TX0 and TX1) of the simulation model for simulation, and the signal is collected at one of the receivers (RX1 or RX2) to obtain the edge response corresponding to each code pattern combination. During the design and testing process, the D2D interface simulation model can be generated by the EDA (Electronic Design Automation) tool, which includes the settings of relevant parameters such as the transmitter, receiver, and channel characteristics of each channel of the D2D interface. For example, the data transmission rate and sampling rate of the D2D transmission channel can be set according to requirements. In the simulation environment, the edge response data corresponding to the input signal can be directly exported from the simulation software.

[0067] In step S2), a probability density function of the self-interference noise is determined based on the acquired edge response.

[0068] Through the simulation of step S1, the edge response waveform corresponding to each code combination can be obtained at RX1. The following processing can be performed on the edge response of each code combination: determine the main cursor area of ​​the edge response waveform, and divide the edge response waveform into multiple UI intervals according to the UI length; sample multiple points in each UI interval according to the set sampling rate, and each sampling point is represented by the sampling moment and its corresponding voltage amplitude; superimpose the sampling points at the same sampling moment in each UI to obtain the probability density distribution of the sampling moment; accumulate the probability density distribution of each sampling moment to obtain the probability density distribution corresponding to the edge response of the code combination. Finally, the probability density function of the self-interference noise is obtained based on the convolution of the probability of occurrence of each code combination and the probability density distribution of the edge response of each code combination.

[0069] For example, dividing the edge response waveform according to the UI length to obtain multiple UI intervals can be understood as dividing the waveform into multiple voltage segments. The voltage probability distribution of the kth segment under the above eight signal states can be recorded as:

[0070]

[0071]

[0072]

[0073]

[0074] in, represents the voltage probability distribution of the kth segment in the "000" state. It is a matrix with a set of voltage values ​​in the first row (the number of which is determined by the sampling rate) and probability values ​​in the second row, where the voltage values ​​correspond to the probability values ​​one by one. 000 represents the probability of the "000" state occurring, t represents the time of each sampling point (its variation range is the kth time interval of the edge response waveform on the time axis), T is the width of a unit time (i.e., the duration of UI), t-kT is used to indicate a sampling moment of a UI interval, V 000 (t-kT) represents the voltage that appears in the kth segment at the sampling time t-kT in the "000" state.

[0075] By combining the probability distribution of each voltage segment under different states through convolution, we can get the probability density function PDF of self-interference noise. Since it is a double-edge trigger mechanism, that is, the state of the last bit after each jump is either "0" or "1", two probability density functions are obtained according to the state of the last bit after each jump. One of them is the cumulative probability density function ending with the "0" state. And the other one ends with the cumulative probability density function of "1" In one embodiment, the two cumulative probability density functions can be calculated in the following ways:

[0076]

[0077]

[0078] It can be seen that in this embodiment, the cumulative probability density function of the current kth segment ending in the "0" state is It is obtained by convolving the voltage probability distribution of the four signal states ending in the "0" state in the kth segment with the eight cumulative probability density functions of the previous segment (i.e., the k-1th segment). The current cumulative probability density function of the kth segment ending in the "1" state is It is obtained by convolving the voltage probability distribution under the four signal states ending with the "1" state in the kth segment with the eight cumulative probability density functions of the previous segment (i.e., the k-1th segment).

[0079] In step S3, the probability density function of the self-interference noise is compared with the probability density function of the inter-symbol interference ISI. Probability density function of crosstalk CCI Probability density function of jitter TJ Convolution is performed to obtain the probability density function of the real-time full-duplex interface, as shown in the following formula:

[0080]

[0081]

[0082] In this embodiment, is the cumulative probability density function ending in the "0" state obtained in step S2 is the cumulative probability density function ending in the "1" state obtained in step S2 Each of the probability density functions of ISI, CCI and TJ corresponds to a cumulative probability density function ending in a "0" state and ending in a "1" state, which can be obtained using the method provided by the prior art and will not be described in detail here.

[0083] In step S4): a statistical eye diagram is generated based on the probability density function of the full-duplex interface obtained through the above steps. There are many ways to generate the corresponding eye diagram based on PDF, which will not be described in detail here.

[0084] The inventors also found during the research that the different phases of the transmitted TX signal and the received RX signal may also cause the generated self-interference noise to be different. Therefore, the self-interference noise at different phases also needs to be considered. In another embodiment, in order to more accurately estimate the probability density function of the self-interference noise, the probability density function obtained in step S2 is further processed as follows:

[0085] (1) In the case where there is no synchronous training mechanism (also called training mechanism) for the phases of adjacent TX and RX ends in the D2D interface simulation model, considering that the phases of adjacent TX and RX are random, the probability density function of the self-interference noise obtained in step S2 can be calculated as follows: and Further seek uniform distribution to obtain the final self-interference noise probability density function and The calculation process is as follows:

[0086]

[0087]

[0088] Where n is an integer, which is a distribution coefficient preset according to actual needs, and T is the time length of 1 UI. Since the phases of adjacent TX and RX are random, interference will be superimposed at any time within 1 UI, so 1UI is further divided into n parts, and the probability of each phase difference is

[0089] (2) When a training mechanism for the phases of adjacent TX and RX is set in the D2D interface simulation model, considering that the phases of adjacent TX and RX are fixed, only the worst jitter caused by the quantization error is required. Therefore, the probability density function of the self-interference noise obtained in step S2 can be calculated as and Further convolve the double Dirac function to obtain the final self-interference noise probability density function and The calculation process is as follows:

[0090]

[0091]

[0092] Where δ() is the Dirac function (also called impulse function), D n is the jitter in the negative direction, D p is the jitter in the positive direction, P p is the probability of a negative Dirac function, P n The probability of being a positive Dirac function. In signal processing and system analysis, the double Dirac delta function usually refers to the superposition or interaction of Dirac delta functions at two different positions, which is used to analyze the response of the system to two instantaneous impulses. It is determined by the characteristics of the system itself and will not be elaborated here.

[0093] As mentioned above, inter-symbol interference, crosstalk, jitter and the self-interference noise mentioned in this article may cause the bit error rate of the D2D interface transmission channel. Therefore, in another embodiment, the method may also include: calculating the total bit error rate based on the probability density function of the D2D interface obtained in step S3, thereby generating a bit error rate statistical eye. For the preset decision level v ref , when the signal sent by the transmitter is 1, and due to the influence of the above factors, its voltage is lower than the decision level, the system will mistakenly judge it as 0, and a bit error will occur at this time. The sum of the probabilities of these bit errors corresponds to a bit error rate of 1. Similarly, when the signal sent by the transmitter is 0, and due to the influence of the above factors, its voltage is lower than the decision level, the system will mistakenly judge it as 1, and a bit error will occur at this time. The sum of the probabilities of these bit errors corresponds to a bit error rate of 0. The bit error rate BER is the sum of the bit error rate of sending code type 0 and the bit error rate of sending code type 1, which can be calculated by the following formula:

[0094]

[0095] where v ref is the decision threshold, P0 is the probability of sending code type 0, P1 is the probability of sending code type 1, and dx represents the integration of the voltage variable x.

[0096] In another embodiment of the present application, a system for generating a statistical eye diagram of a full-duplex D2D interface is also provided. The system includes a simulation data collection module and an eye diagram generation module. The simulation data collection module is used to simulate all code combinations of self-interference noise in the simulation model of the full-duplex D2D interface to be tested, and obtain the edge response corresponding to each code combination, wherein the self-interference noise refers to the interference between the transmitted signal and the received signal on the same side of the transmission channel of the full-duplex D2D interface. The eye diagram generation module is used to determine the probability density function of the self-interference noise based on the edge response obtained, and to combine the probability density function of the self-interference noise with the probability density function of inter-code interference, crosstalk, and jitter to determine the probability density function of the transmission channel of the full-duplex D2D interface, thereby generating a statistical eye diagram. In some embodiments, the eye diagram generation module is also configured to determine the bit error rate of the transmission channel based on a given decision threshold by further integrating the probability density function of the transmission channel, thereby generating a bit error rate statistical eye diagram.

[0097] In another embodiment, the system may further include a parameter configuration module for receiving parameters regarding phase settings between a transmitting end and a receiving end on the same side. The eye diagram generation module may also be configured to further uniformly distribute the probability density function of the self-interference noise before calculating the bit error rate in response to the parameter indicating a random phase, or to convolve the probability density function of the self-interference noise with a dual Dirac function before calculating the bit error rate in response to the parameter indicating a fixed phase.

[0098] It should be understood that, for the steps, processes or methods mentioned in this article, in addition to being implemented in a purely computer-readable program code, it is entirely possible to implement the same functions by logically programming the corresponding functional modules, processes or steps so that these modules are in the form of logic gates, switches, application-specific integrated circuits, programmable logic controllers and embedded microcontrollers. Therefore, the module or device implemented in this way can be considered as a hardware component, and the submodules included therein for implementing various functions can also be regarded as the internal structure of the hardware component. Alternatively, the device for implementing various functions can be regarded as both a software module for implementing the relevant process or method steps and a structure within the hardware component.

[0099] References in this specification to "various embodiments," "some embodiments," "one embodiment," or "an embodiment," etc. refer to a particular feature, structure, or property described in conjunction with the above-described embodiment being included in at least one embodiment. Thus, the appearance of the phrases "in various embodiments," "in some embodiments," "in one embodiment," or "in an embodiment," etc., in various places throughout the specification do not necessarily refer to the same embodiment. Furthermore, particular features, structures, or properties may be combined in any suitable manner in one or more embodiments. Thus, particular features, structures, or properties shown or described in conjunction with one embodiment may be combined in whole or in part with features, structures, or properties of one or more other embodiments without restriction, as long as the combination is not illogical or inoperable.

[0100] The expressions of "including", "having" and similar meanings in this specification are intended to cover non-exclusive inclusions. For example, a process, method, system, product or device that includes a series of steps or units is not limited to the listed steps or units, but may optionally include steps or units that are not listed, or may optionally include other steps or units inherent to these processes, methods, products or devices. "One" or "an" does not exclude multiple situations. In addition, the elements in the drawings of this application are for illustrative purposes only and are not drawn to scale.

[0101] Although the present application has been described through the above-mentioned embodiments, the present application is not limited to the embodiments described herein, and also includes various changes and modifications made without departing from the scope of the present application.

Claims

1. A method for generating a statistical eye diagram for a full-duplex D2D interface, comprising: Simulate all code pattern combinations of self-interference noise in the simulation model of the full-duplex D2D interface to be tested, and obtain the edge response corresponding to each code pattern combination, wherein the self-interference noise refers to the interference between the transmission signal and the reception signal on the same side of the transmission channel of the full-duplex D2D interface; determining a probability density function of the self-interference noise based on the acquired edge response; The probability density function of the transmission channel is determined by convolving the probability density function of the self-interference noise with the probability density functions of the inter-symbol interference, the crosstalk and the jitter, thereby generating a statistical eye diagram.

2. The method according to claim 1 further includes determining the bit error rate of the transmission channel by integrating the probability density function of the transmission channel based on a given decision threshold, and then generating a bit error rate statistical eye diagram.

3. The method according to claim 1, wherein the number of code pattern combinations of self-interference noise is determined by the number and type of code elements sent by the transmitters on both sides of the transmission channel; wherein the first transmitter sends two consecutive code elements, and the second transmitter sends one code element at the same time, and the received signal is monitored at the first receiving end on the same side as the first transmitter to obtain an edge response.

4. The method according to claim 3, wherein simulating all code pattern combinations of self-interference noise and obtaining edge responses corresponding to each code pattern combination comprises: The signal corresponding to each code pattern combination is loaded into the simulation model for simulation, and a signal waveform of the edge response corresponding to each code pattern combination is obtained at the first receiving end.

5. The method according to claim 4, wherein: A leading sequence and an ending sequence of a predetermined length are arranged before and after the signal corresponding to each code pattern combination. The leading sequence and the ending sequence are sequences composed of a plurality of consecutive identical code elements.

6. The method according to claim 1, wherein determining the probability density function of the self-interference noise based on the acquired edge response comprises: For each pattern combination, the edge response is: Determine the starting point of the signal waveform of the edge response, divide the signal waveform into multiple unit intervals UI, sample multiple points in each UI according to the set sampling rate, and each sampling point is represented by the sampling time and its corresponding voltage amplitude; The sampling points corresponding to the same sampling time in each UI are superimposed to obtain the probability density distribution of the sampling time; Accumulating the probability density distribution at each sampling moment to obtain the probability density distribution corresponding to the edge response of the code pattern combination; The probability density function of the self-interference noise is obtained based on the convolution of the probability of occurrence of each code pattern combination and the probability density distribution of the edge response of each code pattern combination.

7. The method according to claim 6, wherein: The number of code pattern combinations used to simulate self-interference noise is 8, and the probability of each code pattern combination appearing is 0.

25.

8. The method according to claim 1, further comprising: receiving parameters regarding a phase setting between a transmitter and a receiver on the same side; In response to the parameter indicating a random phase, before calculating the probability density function of the transmission channel, further uniformly distributing the probability density function of the self-interference noise; In response to the parameter indicating a fixed phase, the probability density function of the self-interference noise is further convolved with a dual Dirac function before calculating the probability density function of the transmission channel.

9. A system for generating a statistical eye diagram for a full-duplex D2D interface, comprising: A simulation data collection module, used to simulate all code combinations of self-interference noise in the simulation model of the full-duplex D2D interface to be tested, and obtain the edge response corresponding to each code combination, wherein the self-interference noise refers to the interference between the transmission signal and the reception signal on the same side of the transmission channel of the full-duplex D2D interface; The eye diagram generation module is used to determine the probability density function of the self-interference noise based on the acquired edge response, and to combine the probability density function of the self-interference noise with the probability density functions of inter-symbol interference, crosstalk, and jitter to determine the probability density function of the transmission channel, thereby generating a statistical eye diagram.

10. The system according to claim 9, wherein the eye diagram generation module is further configured to determine the bit error rate of the transmission channel by integrating the probability density function of the transmission channel based on a given decision threshold, and then generate a bit error rate statistical eye diagram.

Images