A compression method based on UCIe protocol layer transmission and a storage medium

Abstract

The present application relates to chip design technical field, especially in the compression method and storage medium based on UCIe protocol layer transmission, it is through obtaining compression dictionary, compression dictionary includes multiple mapping relations, each mapping relation binds transmission mode and its index identifier, wherein, transmission mode is the data characteristic template that appears repeatedly in on-chip protocol channel;Get the main band channel of UCIe protocol layer, and the main band channel supports the normal transmission mode of uncompressed original data stream and the compression mode of compressed original data stream simultaneously;Get compression engine, and the compression engine is configured to be in the compression mode when the main band channel, the index identifier and the number of times are transmitted to the receiving end through the main band channel using the compression mode;Wherein, when the compression dictionary is updated, the updated compression dictionary is synchronized to the receiving end through the sideband channel of UCIe protocol layer.It greatly improves the bandwidth utilization.

Description

Technical Field

[0001] This invention relates to the field of chip design technology, and in particular to a compression method and storage medium based on UCIe protocol layer transmission. Background Technology

[0002] With the continuous advancement of chip design technology, chiplet technology has gradually become a focus of industry attention due to its ability to effectively solve problems related to size, cost, and yield in single-chip design. However, because the physical resources of chiplet interconnects are extremely expensive and limited, any underutilized bandwidth directly translates to a loss of system performance and a waste of cost. Therefore, how to improve the utilization of underlying links under the same physical connectivity is a core issue of chiplet interconnects. UCIe (Universal Chiplet Interconnect Express) is an open industry standard jointly promoted by global semiconductor giants such as Intel, AMD, Arm, TSMC, and Samsung. It aims to become the universal language and interconnect foundation of the chiplet era and is widely adopted by the industry. How to improve the bandwidth utilization of UCIe links is increasingly becoming a key issue for the industry.

[0003] Current solutions primarily employ the following approaches: First, simplifying packet headers to increase the effective payload ratio. Second, using credit-based flow control and adaptive routing to reduce idle waiting and balance network load. Third, network architecture innovation, such as adopting non-blocking topologies and integrating on-chip and inter-chip networks to optimize data flow paths. Fourth, hardware and software collaboration, such as enabling applications and compilers to understand interconnect topologies and optimize data layout and communication patterns to match the characteristics of the underlying links. However, all of these solutions have significant drawbacks: Regarding the first point, they only consider the traffic model of a single service, ignoring the temporal relationship of traffic models. Regarding the second point, they require complex logic interfaces and control methods, necessitating system-side cooperation. Regarding the third point, they require consideration at the chip architecture level, are strongly correlated with chip architecture, and are not easily generalized to all chips. Regarding the fourth point, they require the integration of hardware and software, imposing constraints on both.

[0004] Given the current situation, there is an urgent need for a protocol-layer compression method that completely breaks free from the constraints of the aforementioned solutions to address the bandwidth utilization problem. Summary of the Invention

[0005] To address the aforementioned technical problems, the present invention adopts the following technical solution: a compression method based on UCIe protocol layer transmission, applied at the sending end, the method comprising the following steps:

[0006] Obtain a compression dictionary, which includes multiple mapping relationships. Each mapping relationship is bound to a transmission mode and its index identifier. The transmission mode is a data feature template that appears repeatedly in the on-chip protocol channel.

[0007] Obtain the main channel of the UCIe protocol layer, which simultaneously supports both the normal transmission mode of uncompressed raw data stream and the compression mode of compressed raw data stream.

[0008] Obtain a compression engine, which is configured to perform the following steps using the compression mode when the main band channel is in the compression mode:

[0009] Obtain the raw data stream sent by the upper-layer protocol;

[0010] The compression engine identifies a recurring transmission pattern in the raw data stream, and the number of times the transmission pattern is repeated.

[0011] The compression engine matches the transmission mode with the transmission modes in the compression dictionary. When a match is successful, it does not send the original data stream, extracts the index identifier bound to the transmission mode from the compression dictionary, and transmits the index identifier and the number of times to the receiving end through the main band channel.

[0012] Furthermore, the present invention also provides a non-transitory computer-readable storage medium storing at least one instruction or at least one program segment, wherein the at least one instruction or the at least one program segment is loaded and executed by a processor to implement the above-described method.

[0013] The present invention has at least the following beneficial effects:

[0014] The compression method provided by this invention transmits compressed data through the main channel and updates the compression dictionary through the side channel, both in complete parallel operation. During dictionary updates, the main channel does not need to pause to wait for dictionary writing, eliminating pipeline interruptions caused by dictionary switching in traditional solutions. Since dictionary updates do not consume the main channel's bus bandwidth, the main channel's bandwidth resources are primarily used to transmit the compressed payload, resulting in higher bus utilization. Because data and the dictionary are physically separated, the decompression engine does not require additional parsing logic to distinguish between dictionary and data input, simplifying the hardware design logic. Attached Figure Description

[0015] To more clearly illustrate the technical solutions in the embodiments of the present invention, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the accompanying 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.

[0016] Figure 1 A flowchart of a compression method based on UCIe protocol layer transmission is provided for an embodiment of the present invention. Detailed Implementation

[0017] 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.

[0018] Unless otherwise defined, all technical and scientific terms used in the embodiments of this invention have the same meaning as commonly understood by those skilled in the art.

[0019] Please see Figure 1 It illustrates a UCIe protocol layer-based transmission compression method applied to the sending end, the method comprising the following steps:

[0020] S100, Obtain a compression dictionary, which includes multiple mapping relationships. Each mapping relationship is bound to a transmission mode and its index identifier. The transmission mode is a data feature template that appears repeatedly in the on-chip protocol channel.

[0021] The transmission modes include a complete transaction mode and an independent channel mode. The complete transaction mode corresponds to a data feature template of all behavioral patterns of a complete transmission transaction, and the independent channel mode corresponds to a data feature template of all behavioral patterns of a data load carried by an independent channel.

[0022] The on-chip protocols include AXI, ACE, CHI, etc., and other types of on-chip protocols also fall within the protection scope of this invention.

[0023] In one implementation, the transmission mode is a data feature template representing the entire behavioral pattern of a complete transmission transaction. This is the coarsest-grained compression unit, and its compression object is a complete bus transaction request, all its associated data, and the final response. In other words, a complete transmission transaction is identified and compressed as an indivisible whole. Taking the AXI protocol as an example, its compression object is a complete AXI write transaction, including the write address, write data, and write response. Taking the CHI protocol as an example, its compression object is a complete CHI read transaction, including the request, listen / data response, and read / read response.

[0024] In one implementation, the transmission mode is a data feature template representing the entire behavioral pattern of the data load carried by an independent channel. This is a finer-grained compression unit, and its compression object is the data stream within a single independent channel. For example, in channel-specific protocols such as AXI, ACE, and CHI, channels for reading address, writing address, reading data, writing data, and writing response are independent of each other and may be transmitted in parallel. Taking the AXI / ACE protocol as an example, the compression engine compresses the data stream of the AXI / ACE write data channel. Taking the CHI protocol as an example, the compression engine compresses the data stream corresponding to the Data field of the CHI request / response channel.

[0025] Each index identifier is an index number for a transmission mode in the compression dictionary.

[0026] This method binds the transmission mode to an index identifier. During subsequent transmissions, only the corresponding index identifier needs to be transmitted; the data length of the index identifier is shorter than the transmitted data stream. By using extremely short codes, such as a few bits for the index, to replace long data blocks, near-zero overhead compression is achieved.

[0027] Before data compression begins, the transmission mode in the compression dictionary needs to be configured. The system provides three options: it can include only the preset static compression dictionary, it can include only the dynamic compression dictionary, or it can include both types of dictionaries.

[0028] S200, Obtain the main channel of the UCIe protocol layer, wherein the main channel simultaneously supports both the normal transmission mode without compressing the raw data stream and the compression mode that compresses the raw data stream.

[0029] It should be noted that the UCIe protocol layer includes a main channel and a sideband channel. In this embodiment of the invention, the data stream is transmitted through the main channel, and the compression dictionary is updated through the sideband channel. The transmission speed of the sideband channel is slower than that of the main channel, but its reliability is higher.

[0030] S300, Obtain the compression engine, the compression engine being configured to perform the following steps using the compression mode when the main band channel is in the compression mode:

[0031] S310: Obtain the raw data stream sent by the upper-layer protocol.

[0032] S320, the compression engine identifies that the raw data stream includes a recurring transmission pattern and the number of times the transmission pattern is repeated.

[0033] S330, the compression engine matches the transmission mode with the transmission mode in the compression dictionary. When the match is successful, the original data stream is not sent. Instead, the index identifier bound to the transmission mode in the compression dictionary is extracted, and the index identifier and the number of times are transmitted to the receiving end through the main channel.

[0034] It should be noted that the compressed data units, after being processed by the link adapter and physical layer, are transmitted to the receiving end through physical channels between the chips, such as microbumps on the silicon interposer. Because the data has been compressed, the actual channel time and energy consumption occupied by this transmission are significantly reduced, improving bandwidth utilization.

[0035] In one embodiment, the method includes a preset compression mode, wherein the compression dictionary includes a preset static compression dictionary, which includes predefined fixed transmission patterns that are known to repeat.

[0036] In one implementation, the system designer or software predefines and loads a set of fixed transmission patterns known to occur frequently during link initialization. These fixed transmission patterns may include all zeros, specific command headers, or buffer line padding values. These transmission patterns are permanently stored in the compression dictionaries of both the sender and receiver, serving as a compressed cipher.

[0037] The preset compression mode is suitable for scenarios where communication patterns are highly predictable and repetitive. In one implementation, the preset compression mode is applicable to scenarios including specific commands in cache coherence protocols, memory initialization data, or control packets with fixed formats. Other highly predictable and repetitive scenarios also fall within the scope of protection of this invention.

[0038] In one implementation, the following steps are performed when the preset compression mode is used:

[0039] S301, during system startup or link training phase, loads the preset static compression dictionary into the compression and decompression engines of the sending and receiving ends.

[0040] S302, when the upper layer protocol needs to send a data stream, the compression engine identifies the transmission mode in the data stream and compares it with the preset static compression dictionary line by line; if the match is successful, the original data stream is not sent, but the index identifier bound in the static compression dictionary is extracted and encapsulated into the transmission unit; if the match is unsuccessful, the original data stream is not compressed and is sent as the original data stream.

[0041] Compression methods using a preset compression mode only require transmitting a short index code, greatly saving bandwidth. When decompressing, the receiving end directly queries its local preset static compression dictionary (which is identical to the sending end's) based on the received index code, restoring the index code to the complete data mode. This compression method is simple to implement, has extremely low compression and decompression latency, no dynamic negotiation overhead, and can achieve maximum compression ratio for fixed modes.

[0042] In one embodiment, the method includes a dynamic compression mode, wherein the compression dictionary is a dynamic compression dictionary, and the dynamic compression dictionary includes transmission patterns that are dynamically identified during system operation.

[0043] The dynamic compression mode is an adaptive, learning-based compression mode. During operation, the system analyzes the flowing data stream in real time, dynamically identifies, learns, and negotiates new transmission patterns that occur more frequently than a threshold, adding them to a dynamically updated dynamic compression dictionary. This allows the compression algorithm to adapt to the actual workload, continuously learning to find high-frequency transmission patterns in the data stream, thus achieving efficient compression.

[0044] The dynamic compression mode is suitable for applications with unknown or variable transmission modes, or those containing a large number of repetitive but non-preset data blocks. In one implementation, the dynamic compression mode is applicable to scenarios including specific gradient tensor patterns in AI training, and duplicate key-value pairs in database queries. Other scenarios with unknown or variable transmission modes, or those containing a large number of repetitive but non-preset data blocks, also fall within the scope of this invention.

[0045] In one implementation, the following steps are performed when the dynamic compression mode is used:

[0046] S3001 initializes the dynamic compression dictionary to an empty or a transmission mode containing a small amount of basic data.

[0047] S3002, for the sending end, the compression engine identifies the transmission pattern. When the number of times a data block is repeated reaches a set minimum threshold, the data block is marked as candidate data. The specific minimum threshold can be flexibly configured according to the actual scenario;

[0048] S3003, the sending end pauses and starts the update negotiation process. The update negotiation process sends a new request to the receiving end through the sideband channel. The new request is used to request that the candidate data be added to the dynamic dictionary of both parties and a new index identifier be assigned. After the negotiation is confirmed, when the same data block appears again, it is compressed with the new index identifier.

[0049] S3004, the compression engine continuously monitors the dynamic compression dictionary. If a transmission mode has not been used for a long time, an update negotiation process is initiated. This process sends a removal request to the receiving end via a sideband channel, which is used to remove the mode from the dynamic compression dictionary. In other words, updating the dynamic compression dictionary, including adding or removing transmission modes, requires going through the update negotiation process via the sideband channel. The receiving end synchronously executes the sending end's dynamic compression dictionary update operation, and during decompression, it looks up and restores the corresponding transmission mode based on the index identifier. This compression method is highly flexible, dynamically updating the dictionary as the workload changes, and maintains high long-term compression efficiency.

[0050] In one embodiment, the method includes a hybrid compression mode, wherein the compression dictionary includes two types of dictionaries: a preset static compression dictionary, which includes predefined fixed transmission patterns that are known to recur; and a dynamic compression dictionary, which includes transmission patterns that are dynamically identified during system operation.

[0051] This hybrid compression mode combines the advantages of preset and dynamic compression modes through an intelligent strategy. It intelligently determines which compression mode to use for a given data block, or switches dictionaries at runtime, by using preset selection strategies such as cost models or hit rate statistics. The preset compression mode ensures zero-overhead compression for basic and high-frequency transmission modes, while the dynamic compression mode captures and optimizes repetitive patterns specific to the workload, achieving the best balance between coverage and operational efficiency.

[0052] This hybrid compression mode is suitable for the vast majority of practical systems. It needs to handle both known standard protocol traffic and application-specific data efficiently.

[0053] In one implementation, the following steps are performed when the hybrid compression mode is used:

[0054] S31, the system loads the preset static compression dictionary and initializes an empty dynamic compression dictionary;

[0055] S32, when the data stream arrives, the compression engine identifies the transmission pattern in the data stream and queries the preset static compression dictionary according to the transmission pattern. If a match is found, the index identifier is extracted; if no match is found, the dynamic dictionary is queried. If a match is found, the index identifier is extracted; if no match is found, the compression engine starts the decision logic: send the original data stream while tracking the frequency of occurrence of the transmission pattern. If the number of occurrences is greater than the minimum number threshold, it is added to the dynamic compression dictionary.

[0056] At the receiving end, the system determines the dictionary from which the data originates by using the index range of the index identifier, and then decompresses and recovers the original data stream from the corresponding dictionary. This hybrid compression mode combines high performance with high adaptability, enabling it to handle complex and varied real-world workloads and balancing efficiency and flexibility.

[0057] S400, when the compressed dictionary is updated, the updated compressed dictionary is synchronized to the receiving end through the sideband channel of the UCIe protocol layer.

[0058] It should be noted that the compression method provided by this invention transmits compressed data through the main channel and updates the compression dictionary through the side channel, both in complete parallel operation. During dictionary updates, the main channel does not need to pause to wait for dictionary writing, eliminating pipeline interruptions caused by dictionary switching in traditional solutions. Since dictionary updates do not consume the main channel's bus bandwidth, the main channel's bandwidth resources are primarily used for transmitting the compressed payload, resulting in higher bus utilization. Because data and the dictionary are physically separated, the decompression engine does not require additional parsing logic to distinguish between dictionary and data input, simplifying the hardware design logic.

[0059] In one implementation, when the compressed dictionary is updated, the main channel switches to normal transmission mode to transmit the original data stream; when the compressed dictionary is updated or not updated, the main channel switches to compressed mode to transmit the corresponding index identifier.

[0060] It should be noted that the main channel does not need to stop transmitting data due to dictionary updates, but directly switches to normal transmission mode, thus improving bandwidth utilization.

[0061] In one implementation, when the compression dictionary is a dynamic compression dictionary, the dynamic compression dictionary is updated when transmission patterns that occur more than a preset update threshold are updated in the compression dictionary.

[0062] In one implementation, when the compression dictionary is a dynamic compression dictionary, the dynamic compression dictionary is updated within a preset time window by updating the transmission mode with the highest recurrence frequency to the compression dictionary; if the compression dictionary is full, a stored transmission mode is deleted according to a preset elimination policy, and the transmission mode with the highest recurrence frequency is written into the compression dictionary.

[0063] In one implementation, when the compression dictionary is a dynamic compression dictionary, the dynamic compression dictionary is updated within a preset time window. If there are multiple transmission modes that occur most frequently, the latest transmission mode is updated into the compression dictionary. If the compression dictionary is full, a stored transmission mode is deleted according to a preset elimination policy, and the latest transmission mode is written into the compression dictionary.

[0064] In one implementation, when the access addresses of data in a recurring transmission pattern are different, the main band channel transmission also includes the base address corresponding to the transmission pattern and the rules followed by its address sequence.

[0065] As an example, taking AXI write transactions using a preset compression mode, the compression object is a complete write address, write data, and write response process. The data payload of the AXI write transaction is a 128-byte all-zero data block, and the static compression dictionary includes a mapping relationship of "preset mode #3, 128-byte all-zero data block". When a GPU core frequently writes a 128-byte all-zero initialization block to the on-chip cache, the compression engine recognizes that the data payload of the current AXI write transaction matches the preset mode #3 in the preset compression dictionary, and its address sequence follows a continuous increment, and the response is always "OKAY". The compression engine does not send the AXI write transaction (detailed address, data, and response signals), but instead generates a special, short transaction token. This token encodes: execute preset mode #3 once, with the target address being the base address + offset N. After receiving the transaction token, the receiving end recovers the entire write transaction locally according to preset rules, completes the data write, and generates a response. This preset compression mode compresses hundreds of bits of raw transmission into a token of tens of bits, greatly saving bandwidth.

[0066] As an example, taking CHI read transactions with dynamic compression mode, the compression object is a complete request, listener / data response. Its application scenario is as follows: In a consistency system, multiple cores poll the state of the same shared lock, i.e., reading from a specific address. In the early stages of system operation, the compression engine dynamically identifies multiple read requests to that specific address and the same data they return, such as the lock's unoccupied state, and adds them to the dynamic compression dictionary and assigns index identifiers. Subsequent read requests to the same address no longer send the complete request data packet (including source ID, target ID, address, etc.), but instead send its index identifier. Upon receiving the index identifier, the receiving end directly recovers the complete data packet with the data response, significantly reducing network congestion caused by frequent state queries within the consistency domain.

[0067] As an example, taking the AXI write data channel in hybrid compression mode, the compression target is the data stream of the AXI write data channel. The application scenario is: the video processing unit writes a continuous background area with the same color value from a frame of image into memory. In this scenario, the compression engine only focuses on the data stream on the WDATA channel, ignoring the associated address and response channels. The engine identifies a large number of continuous, repeating 32-bit pixel values ​​(e.g., 0xFF0000FF representing red) on the WDATA channel. This may be preset or dynamically learned. The sender no longer sends 32 / 64 / 128 bits of data per cycle, but a short header including an index identifier and the repetition count. The address channel normally transmits the start address and burst length information. This allows for extremely high compression ratios on the WDATA channel, which has the largest data volume and consumes the most bandwidth.

[0068] As an example, taking the Data field of the CHI request channel, the compression object is the data stream corresponding to the Data field of the CHI request channel. Its application scenario is: transmitting large amounts of small data blocks, such as 4KB data blocks read from an NVMe disk, on CHI-based transaction layers like CCIX or CXL.io. If the data block content has high entropy, lossless compression algorithms are used for dynamic compression. Other key fields in the same data packet, such as TxnID, SrcID, Opcode, and RespErr, are used as control information and are usually not compressed or compressed using a preset mode to ensure reliability and low latency. This effectively reduces the pressure on inter-core link bandwidth caused by the transmission of large, regularly occurring data blocks.

[0069] In one implementation, each type of compressed object corresponds to a compression dictionary.

[0070] Embodiments of the present invention also provide a non-transitory computer-readable storage medium that can be disposed in an electronic device to store at least one instruction or at least one program related to implementing a method in the method embodiments, wherein the at least one instruction or the at least one program is loaded and executed by the processor to implement the method provided in the above embodiments.

[0071] Embodiments of the present invention also provide an electronic device, including a processor and the aforementioned non-transitory computer-readable storage medium.

[0072] Embodiments of the present invention also provide a computer program product including program code, which, when the program product is run on an electronic device, causes the electronic device to perform the steps of the methods described above in various exemplary embodiments of the present invention.

[0073] Those skilled in the art will clearly understand that, for the sake of convenience and brevity, the above-described division of functional units and modules is used as an example. In practical applications, the above functions can be assigned to different functional units and modules as needed, that is, the internal structure of the device can be divided into different functional units or modules to complete all or part of the functions described above.

[0074] While specific embodiments of the invention have been described in detail by way of example, those skilled in the art should understand that the above examples are for illustrative purposes only and are not intended to limit the scope of the invention. Those skilled in the art should also understand that various modifications can be made to the embodiments without departing from the scope and spirit of the invention. The scope of this invention is defined by the appended claims.

Claims

1. A compression method based on UCIe protocol layer transmission, characterized in that, Applied to the sending end, the method includes the following steps: Obtain a compression dictionary, which includes multiple mapping relationships. Each mapping relationship is bound to a transmission mode and its index identifier. The transmission mode is a data feature template that appears repeatedly in the on-chip protocol channel. Obtain the main channel of the UCIe protocol layer, which simultaneously supports both the normal transmission mode of uncompressed raw data stream and the compression mode of compressed raw data stream. Obtain a compression engine, which is configured to perform the following steps using the compression mode when the main band channel is in the compression mode: Obtain the raw data stream sent by the upper-layer protocol; The compression engine identifies a recurring transmission pattern in the raw data stream, and the number of times the transmission pattern is repeated. The compression engine matches the transmission mode with the transmission mode in the compression dictionary. When the match is successful, it does not send the original data stream, extracts the index identifier bound to the transmission mode in the compression dictionary, and transmits the index identifier and the number of times to the receiving end through the main channel. Specifically, when the compressed dictionary is updated, the updated compressed dictionary is synchronized to the receiving end through the sideband channel of the UCIe protocol layer.

2. The method according to claim 1, characterized in that, The method also includes a step of updating the compressed dictionary: When the compression dictionary is updated, the main channel switches to normal transmission mode to transmit the original data stream; When the compressed dictionary is updated or not updated, the main channel switches to compressed mode to transmit the corresponding index identifier.

3. The method according to claim 1, characterized in that, The method includes a preset compression mode, in which the compression dictionary includes a preset static compression dictionary, which includes a predefined fixed transmission pattern that is known to repeat.

4. The method according to claim 1, characterized in that, The method includes a dynamic compression mode, wherein the compression dictionary is a dynamic compression dictionary, and the dynamic compression dictionary includes transmission patterns that are dynamically identified during system operation.

5. The method according to claim 1, characterized in that, The method includes a hybrid compression mode, in which the compression dictionary includes two types of dictionaries: A preset static compression dictionary, which includes predefined fixed transmission patterns that are known to repeat. A dynamic compression dictionary, which includes transmission patterns that are dynamically identified during system operation.

6. The method according to claim 4 or 5, characterized in that, When the compressed dictionary is a dynamic compressed dictionary, the dynamic compressed dictionary is updated when transmission patterns that occur more than a preset update threshold are updated in the compressed dictionary.

7. The method according to claim 4 or 5, characterized in that, When the compression dictionary is a dynamic compression dictionary, the dynamic compression dictionary is updated within a preset time window by updating the compression dictionary with the transmission pattern that appears most frequently. If the compression dictionary is full, a stored transmission mode is deleted according to a preset elimination policy, and the transmission mode with the highest recurrence frequency is written into the compression dictionary.

8. The method according to claim 4 or 5, characterized in that, When the compression dictionary is a dynamic compression dictionary, the dynamic compression dictionary is updated within a preset time window. If there are multiple transmission modes that occur most frequently, the latest transmission mode is updated into the compression dictionary. If the compression dictionary is full, a stored transmission mode is deleted according to a preset elimination policy, and the latest transmission mode is written into the compression dictionary.

9. The method according to claim 1, characterized in that, When the access addresses of data in a recurring transmission pattern are different, the main band channel transmission also includes the base address corresponding to the transmission pattern and the rules followed by its address sequence.

10. A non-transitory computer-readable storage medium storing at least one instruction or at least one program segment, characterized in that, The at least one instruction or the at least one program segment is loaded and executed by the processor to implement the method as described in any one of claims 1-9.

Images