Channel compression coding method for multi-channel detector signal array

Abstract

The invention discloses a channel compression coding method for a multi-channel detector signal array, which comprises the following steps of: 1) numbering channels of a target detector system 1-N in sequence, and numbering channels of a target electronics system 1-M in sequence; 2) recording each channel from the target detector system as an input channel; marking each channel of the target electronics system as an output channel; distributing two output channels with different codes for each input channel as a two-bit code combination of the corresponding input channel; generating a coding plate according to the two-bit coding combination of each input channel; 3) selecting a number m, so that no repeated two-bit coding combination output signal exists in m signals which are randomly and continuously read out on all channels of the target detector system; (4) repeating the steps (2)-(3), and finding the coding plate corresponding to the maximum value of m meeting the condition; and 5) establishing a mapping relation between input and output channels according to the final coding plate, and completing channel compression coding.

Description

Technical Field

[0001] The present invention belongs to the field of high-energy charged particle detection, and relates to a channel compression coding method for a multi-channel detector signal array, mainly for channel multiplexing of large multi-channel detectors. Background Art

[0002] In fields such as medical imaging and particle detection, it is often necessary to use large-area detectors to achieve high-precision detection of particles. The media of these detectors are generally gas, scintillator or crystal. Taking scintillator as an example, when using granularly arranged scintillators or a planar array composed of a large number of strip scintillators, linear photoelectric devices are generally required for readout. As the system scale expands or the accuracy requirement increases, the number of photoelectric devices is often large, which brings certain challenges to the subsequent signal processing. For example, the positron emission tomography (PET) technology used in medical imaging needs to achieve high-precision γ photon detection, and there are often tens of thousands of silicon photomultipliers (SiPMs) forming a readout array inside; another example is that when detecting cosmic rays, a large-area imaging plane is required to detect the incident position of cosmic rays, so that the number of electronic channels reaches several hundred to several thousand. If a corresponding electronic readout system is configured for each channel, it is difficult to implement in engineering and the cost will become very high. Therefore, researching a technology that compresses the signals of a large number of channels into a smaller number of channels is a common problem in large-size detector technology.

[0003] In order to achieve signal channel compression, relevant professionals have been exploring various solutions. Currently, there are mainly three methods to achieve the compression of the number of analog signal channels: resistor network, delay line network, and channel coding-decoding.

[0004] When using a resistor network for channel compression, low resistors with specific resistance values are arranged between different signals. Therefore, the signals will flow in different directions according to the different resistors, and then by reading the charge values of multiple readout channels, the location of the original signal is determined by the distribution of charges between different readout channels; however, for different signal access points, the equivalent impedance of the resistor network is also different, resulting in a large difference in the output amplitude of different detectors.

[0005] When using the delay line method for channel compression, delay devices are inserted between different signals, and the corresponding time difference is read when reading the signals, so as to determine the channel value where the signal is located; this method has two challenges. One is that it is difficult to perform impedance matching, and the other is that this method makes the channels of the entire detector linear, greatly increasing the recovery time (or dead time). Once the signal frequency is high, signal pile-up will be caused, which is not conducive to detecting high-flux particle beams or particle beams in a large area.

[0006] The channel coding method is to establish a mapping set, which can inversely deduce the location of the original signal with fewer readout channels. Generally, an original signal needs to be mapped to multiple readout channels, and by using software and hardware methods, it is allocated to several preset readout channels when the signal arrives. The coding readout method has high flexibility and a strong data compression ratio, but there are difficulties in signal strength distribution and impedance matching, and challenges such as signal crosstalk also need to be faced in the specific implementation. In short, although there are various methods for analog signal channel compression, there are still difficulties and challenges in engineering.

[0007] Currently, multi-channel large detectors are still constantly evolving, gradually shifting from scientific research to applied sciences, such as medical and imaging fields. Due to the above problems, it is necessary to propose a method that makes certain improvements based on the existing channel multiplexing method and combines it with the basic properties of the detector to achieve strengths and avoid weaknesses, and to achieve a high-efficiency and high-performance channel compression technology. Summary of the Invention

[0008] Aiming at the technical problems existing in the prior art, the purpose of the present invention is to provide a channel compression coding method for a multi-channel detector signal array. Based on the basic idea of coding channel multiplexing, the present invention realizes a type of coding method with specific functions. This coding method can achieve no duplicate codes within adjacent several channels, so that the channel number can be greatly saved without sacrificing too much counting rate ability; on the basis of coding, a matching resistor network is used to specifically implement the coding, and it is ensured that the load of the SiPM at each point of the resistor network is basically the same, so as to ensure that there are no obvious inconsistencies in the output amplitude, frequency range, etc. of the signal, so as to ensure the consistency during readout.

[0009] A relatively traditional coding method is relatively similar to the carry when using numbers. For example, in order to code A×B channels, first divide every B channels into a group, and finally a total of A groups can be obtained. Therefore, each channel can obtain a separate A and B as the unique location identifier. On the electronic board, channels with the same A are connected to the same output channel, and channels with the same B are also connected to the same output channel. Finally, the total number of output channels required is A + B. This channel compression method can compress A×B channels into A + B channels, that is, compress N channels to the minimum extent of channels. Although this readout method can relatively well achieve channel compression, when multiple signals are triggered simultaneously, the decoding result is not unique. For example, if signals appear simultaneously on channels 1 and 2 in the compressed packet A, and also on channels 1 and 2 in packet B, it is impossible to determine whether the signals come from (1, 2) and (2, 1), or from (1, 1) and (2, 2). Therefore, this decoding method cannot effectively decode the situation where signals may stack and arrive simultaneously.

[0010] The above is only a discussion of a class of problems that often exist in encoding and decoding. In the fields that require a large amount of signal compression, due to the differences in each signal, there are some difficulties and challenges in each field, and there are already many different solutions. Due to the complexity of these problems and limiting conditions, and the difficulties faced in implementing on a specific circuit board, no encoding scheme can generally solve the problems. The present invention also faces a class of complex problems with specific usage scenarios, and also needs to cooperate with the basic properties and usage methods of signal generating devices and signal amplifiers.

[0011] For a large-area detector, if the detected particles have a certain randomness, and often particles will simultaneously excite signals in one or several adjacent channels, and the position of the particles is restored through several signals. Therefore, if a number n is defined, when signals occur in every adjacent n channels, the signals read out after encoding can accurately restore the positions and signal magnitudes of the consecutive n signals, then it can be successfully applied to a device for precise position measurement.

[0012] The technical solution of the present invention is as follows:

[0013] A channel compression encoding method for a multi-channel detector signal array, the steps of which include:

[0014] 1) According to the number of channels N of the target detector system and the number of channels M of the target electronics system, encode the two respectively. Number the channels of the target detector system sequentially from 1 to N, and number the channels of the target electronics system sequentially from 1 to M;

[0015] 2) Denote each channel from the target detector system as an input channel; denote each channel of the target electronics system as an output channel; allocate two output channels with different encodings for each input channel as the two-bit encoding combination corresponding to the input channel; generate an encoding board according to the two-bit encoding combinations of each input channel;

[0016] 3) Select a number m such that among any continuously read m signals on all channels of the target detector system, there are no signals with repeated two-bit encoding combinations output;

[0017] 4) Repeat steps 2) - 3) to find the encoding board corresponding to the maximum value of m that meets the conditions;

[0018] 5) Establish the mapping relationship between the input channels and the output channels according to the encoding board determined in step 4) to complete the channel compression encoding.

[0019] Furthermore, M is less than N, and M is a power of 2.

[0020] Furthermore, the channel arrangement mode of the target detector system is a linear arrangement in a column.

[0021] Furthermore, the shape of the target detector system after channel arrangement is an arc, a circular ring, an S-shaped curve, or an irregular curve.

[0022] Furthermore, the channel arrangement mode of the target detector system is a linear arrangement in two columns; along the same direction, the channel number of the i-th channel arranged in sequence in the first column is 2i - 1, and the channel number of the i-th channel arranged in sequence in the second column is 2i.

[0023] The advantages of the present invention are as follows:

[0024] The present invention is mainly used for linearly arranging and combining the original signals of the detector when a multi-channel signal detector is required for signal readout, and signals may exist in several consecutive channels. The present invention is a unique encoding and decoding method, which ensures that even if signals exist in adjacent channels or multiple signals simultaneously, a unique solution can be guaranteed during signal decoding, so as to accurately restore the position of the original signal. Brief Description of the Drawings

[0025] Figure 1 It is the readout channels with a one-column linear or two-column linear relationship applicable to the present invention.

[0026] Figure 2 It is the calculation and solution method of the encoding table (encoding board) of the present invention. Detailed Embodiment

[0027] The present invention discloses a method for multiplexing the readout channels of a detector by using an encoding method. The purpose of this encoding method is to read a larger number of original signals of the detector with a smaller number of electronic channels, thereby greatly reducing the demand for the backend electronic channels. The following further describes the invention in detail. Some specific numbers are only used to describe the used method or explain the basic idea of the present invention. The present invention is an encoding and implementation method, and does not limit the specific number of channels, the specific encoding table, or the specific circuit wiring during engineering implementation.

[0028] The following discusses the characteristics and implementation process of the encoding method involved in this patent:

[0029] In specific terms, we specify that the number of input channels is the channels before compression, and denote its quantity as N; we specify that the channels after data compression are the output channels, and denote its quantity as M; we define the process of compressing N input channels into M output channels as "channel compression"; we define the table that realizes the specific corresponding relationship between the N input channels and the M output channels for channel compression as the "coding table" or "coding board", and will express it according to convenience in subsequent specific descriptions.

[0030] First, for the specified number of input channels N, according to the specific arrangement method of each channel, number them from 1 to N in sequence. In the situation targeted by the present invention, the detectors are arranged linearly or nearly linearly. In each group of channel multiplexing, only consider the geometric arrangement methods as Figure 1 shown: The first is a single-column linear; the second is a two-column staggered linear, where one line is all odd-numbered and the other line is all even-numbered, and adjacent odd and even numbers are always adjacent on the detector. Note that the linear relationship is not necessarily a straight line, and in some cases it is bent or looped, such as forming a circular shape, an S shape or an irregular curve. If the logical relationship of the signal is still linear, it also applies to the method proposed by the present invention. If it is a readout of more than three rows, or even a larger array, it is not within the scope covered by the present invention. Starting from a starting point of the linearity as number 1, number it up to N in a specific order. The numbers cannot jump between each other to reflect the real geometric relationship, and the coding methods are as Figure 1 shown in several cases.

[0031] The second step is to determine the number of channels M that need to be compressed. M is a number less than N, but M usually depends on the total number of digital electronics readout channels that can be provided. For the original number of channels N, if each channel is encoded with two readout channels and the method of the present invention is adopted, for the known number of output channels M (because in most cases, the output channels need to correspond to a specific electronics readout system, and the number of channels of this readout system is usually fixed and unchangeable. For example, if we use an 8-channel electronics board, regardless of the number of channels of the detector itself, the number of channels that the electronics can read is 8, so it is 8 channels at this time), encoding each output channel from 1 to M, the minimum number of multiplexed channels is N = C 2 MFor example, if it is known that there are 8 output channels (the maximum number of output channels M = 8), theoretically N = 2^8 channels can be encoded; if it is known that there are 16 digital electronics channels, theoretically 120 channels can be encoded. When the number of electronics channels increases, the theoretical compression ratio will become higher and higher. In short, for a determined number of channels N to be compressed, the value of M needs to be determined according to the situation of the backend electronics, ensuring that the value of M is theoretically feasible and there is also a certain degree of redundancy. Generally speaking, M is a power of 2.

[0032] In the third step, an encoding board needs to be generated. At this stage, for each of the N input channels, two encodings from 1 to M are selected, ensuring that for each of the N channels, there is a unique two-digit encoding combination. This encoding combination comes from two numbers between 1 and M, and the two numbers cannot be repeated. For the case where N is relatively small, this code board will be relatively simple and can be obtained by calculation. When N is relatively large, a computer needs to be used to traverse or randomly generate the codes. And in this step, it is necessary to ensure that the number of input channels corresponding to all M channels is basically the same, and the difference is preferably not greater than 1. This step is used to ensure that the signal does not undergo too much distortion and that each signal channel has a certain degree of consistency.

[0033] In the fourth step, code constraints need to be imposed. According to specific requirements, a number m needs to be determined such that there are no repeated two-digit encoding combinations among m consecutive signals in the N input channels. For example, in the case of m = 5, if the encoding of the 10th channel of the detector is (16, 21), then the encodings of the 6th to 9th channels and the 11th to 14th channels must not have any of the numbers 16 or 21. There are various factors to consider in determining m. The factor that limits the smallest value of m is the signal characteristic of the detector itself. For example, in many cases, the charge centroid method needs to be used for readout, and each individual signal will have multiple adjacent input channels generating signals. Then m must be greater than this number. On this premise, the larger m is, the less likely it is to have interference, and the better. The factor that limits the maximum value of m needs to be inferred in the fifth step below because when m is greater than a certain value, no code board can be solved. In this case, it means that m has reached the upper limit and cannot be larger.

[0034] After this limitation, for any result obtained on the output channel, there can be a corresponding unique solution; if no limitation is imposed, it will be impossible to distinguish between (1, 2) + (2, 1) and (1, 1) + (2, 2). Further, for every consecutive two - to - m signal two - bit coding combinations of N input channels, there are no duplicate two - bit coding combinations. If the code is obtained by traversal in the third step and duplicates are found, it proves that this code plate does not meet the requirements and it is necessary to go back to the third step to generate the next set of codes and perform the constraint check in the fourth step again. If the code plate is obtained by random generation in the third step and it is found that the constraint conditions of this step cannot be met, then go back to the third step and randomly generate a code plate again.

[0035] Step 5, after the loop of the third and fourth steps, for specific N, M, and m, if there exists a code plate that meets the requirements, then after a sufficient amount of time, this code plate can finally be obtained. If there is no corresponding solution due to N being too small, M being too small, or m being too large, then after the traversal ends, or when all possibilities have been randomly traversed, the coding is terminated. Generally speaking, N and M are difficult to change. At this time, according to specific needs, change the value of m, relax the constraints, and it is possible to obtain a solution.

[0036] This coding method has the following obvious characteristics: First, the weights or statuses of each channel in this coding method are the same. For most coding methods, different channels often have a priority order or weight order. In this coding method, there is no grouping or grading of signals, and all channels only exist as coding digits. Second, this coding method does not have obvious digital rules, texture characteristics, repetitive rules, or unit - type structures. From the coding results, there are no obvious rules. Whether the scale of the coding digits is large or small, it cannot be translated or reused.

[0037] In specific implementation, it is only necessary to split each read - out signal into two, and according to the two digits corresponding to each N given by the code plate, map each of the N input channels to two channels among the specific M output channels, and take measures to prevent crosstalk of the coding signals. After such connection, each of the N input channels is connected to 2 output channels; each of the M read - out channels will be connected to several input channels (taking Table 1 as an example, each output channel is connected to more than 10 input channels).

[0038] This utility model patent provides a specific requirement and a corresponding coding table as an example to further illustrate this coding method. The content protected by this patent is not limited to the following example, or the specific code plate, specific numbers, or specific wiring methods in the following example, but covers all similar specific methods for processing coding.

[0039] Specific examples are as follows:

[0040] Taking the coding board in Table 1 as an example, in this specific example, we have 85 signal channels to be compressed (N = 85), and these signals are arranged in a linear manner of 1 column or 2 columns. It is necessary to compress these 85 signal channels onto 16 readout channels (M = 16). First, according to the formula, it can be known that 16 channels can encode at most 120 signals. Therefore, it is first determined that it is possible to adopt the coding method in the present invention. Then, according to specific requirements, it is determined that there should be no code overlap for every consecutive 5 channels. Using the generation method of the present invention, a set of code tables that meet the requirements is obtained, as Figure 2 shown. For each of the 16 readout channels of this set of code tables, the corresponding number of input channels is 10 or 11, and there is no code overlap for every adjacent 5 channels.

[0041] Table 1 shows an exemplary coding table given by using the method of the present invention for a specific coding condition.

[0042]

[0043] Table 1 represents a coding board under a specific compression requirement and with a solution. Different coding tables obtained by using the method of the present invention in different situations or even in the same situation, as long as the same solution and compilation method are used, are also within the scope protected by this patent.

[0044] Although specific embodiments of the present invention are disclosed for illustrative purposes, the purpose is to help understand the content of the present invention and implement it accordingly. Those skilled in the art can understand that: without departing from the spirit and scope of the present invention and the appended claims, various substitutions, changes, and modifications are possible. Therefore, the present invention should not be limited to the content disclosed in the best embodiments, and the scope of protection claimed by the present invention shall be subject to the scope defined by the claims.

Claims

1. A channel compression coding method for a multi-channel detector signal array, the steps comprising: 1) According to the number of channels N of the target detector system and the number of channels M of the target electronics system, the two are encoded respectively, and the channels of the target detector system are numbered 1 to N in sequence, and the channels of the target electronics system are numbered 1 to M in sequence; 2) Record each channel from the target detector system as an input channel; Record each channel of the target electronics system as an output channel; Allocate two output channels with different codes to each input channel as the two-bit code combination of the corresponding input channel; generate a code board according to the two-bit code combination of each input channel; 3) Select a number m so that there is no repeated signal output by the two-bit code combination among any m signals read out continuously on all channels of the target detector system; 4) Repeat steps 2) to 3) to find the corresponding coding plate when the maximum value of m meets the conditions; 5) According to the coding board determined in step 4), a mapping relationship between the input channel and the output channel is established to complete the channel compression coding.

2. The method according to claim 1, characterized in that M is less than N, and M is a multiple power of 2.

3. The method according to claim 1 or 2, characterized in that: The channels of the target detector system are arranged in a linear row.

4. The method according to claim 3, characterized in that The channels of the target detector system are arranged in a shape of an arc, a ring, an S-shaped curve or an irregular curve.

5. The method according to claim 1 or 2, characterized in that: The channels of the target detector system are arranged in two linear columns; along the same direction, the i-th channel arranged in sequence in the first column is numbered 2i-1, and the i-th channel arranged in sequence in the second column is numbered 2i.

Images