A digital logistics platform historical data classification transmission storage processing method and system
Patent Information
- Application Number
- CN202610959550.9
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2026-06-30
- Publication Date
- 2026-10-13
AI Technical Summary
[0003]1.数据无差异化统一传输，所有物流业务历史数据采用相同传输协议与带宽分配策略，高频轨迹实时数据与低频结算归档数据混合同步，带宽资源浪费，实时数据传输延迟高；
[0029]本发明实施例提供的技术方案，通过按数据的分类标签划分传输通道，热数据优先占用低延迟带宽，冷数据错峰批量传输，从而避免带宽资源浪费，实时数据传输延迟高的问题；
Smart Images
Figure CN122903448A_ABST
Abstract
Description
Technical Field
[0001] This invention relates to the field of digital logistics data processing technology, specifically to a method and system for classifying, transmitting, storing, and processing historical data of a digital logistics platform. Background Technology
[0002] Existing digital logistics platforms continuously generate massive amounts of historical business data, such as orders, transportation, warehousing, settlement, and tracking information. This data is then transmitted and stored. However, current general data transmission and storage architectures have the following drawbacks:
[0003] 1. Data is transmitted uniformly without differentiation. All historical data of logistics business adopts the same transmission protocol and bandwidth allocation strategy. High-frequency trajectory real-time data and low-frequency settlement archive data are mixed and synchronized, resulting in wasted bandwidth resources and high latency in real-time data transmission.
[0004] 2. Lack of automatic classification mechanism; all original business data is uniformly stored in the database without being classified according to data business type, access frequency, and confidentiality level, resulting in long data retrieval and query time.
[0005] 3. There is no hierarchical storage strategy for cold and hot data. Long-term archived cold data and frequently accessed hot data share high-performance storage media, resulting in high hardware storage costs and low storage resource utilization.
[0006] Therefore, it can be seen that the existing technology has not formed an integrated solution for pre-classification, differentiated transmission and hierarchical storage. Summary of the Invention
[0007] In view of the technical defects mentioned in the background art, the purpose of this invention is to provide a method and system for classifying, transmitting, storing and processing historical data of a digital logistics platform, aiming to at least solve one of the technical problems in the related art to a certain extent.
[0008] To achieve the above objectives, in a first aspect, embodiments of the present invention provide a method for classifying, transmitting, storing, and processing historical data of a digital logistics platform, the method comprising:
[0009] Collect historical raw data from the digital logistics platform's business, extract features, and then combine the extracted features with preset rules to divide the data into real-time hot data, regular warm data, and archived cold data, and bind corresponding classification tags;
[0010] Based on the classification labels, the corresponding transmission channels are matched for transmission, and differentiated transmission scheduling is performed.
[0011] Based on the classification labels, a tiered storage strategy is executed, and the data is stored in the corresponding storage pool;
[0012] The system periodically counts the access frequency of various types of data, dynamically updates the classification labels, and adaptively adjusts the bandwidth of each transmission channel and the resource allocation of each storage pool.
[0013] As a preferred implementation of this application, the method further includes:
[0014] During storage, a unified index table is synchronously established across storage pools; when a data query request is received, the corresponding storage pool is located based on the unified index table.
[0015] As a preferred embodiment of this application, the differentiated transmission scheduling specifically comprises:
[0016] Real-time thermal data is transmitted using a high-bandwidth, low-latency real-time channel; regular temperature data is transmitted using a balanced bandwidth general channel; and archived cold data is transmitted in batches using off-peak compression.
[0017] The classification tags are sent to the storage terminal along with the data during the transmission process.
[0018] As one specific implementation of this application, the hierarchical storage strategy is as follows:
[0019] Real-time thermal data is stored in a high-speed memory cache and a high-performance solid-state storage pool; regular temperature data is stored in a regular mechanical hard disk storage pool; archived cold data is compressed and encrypted and then stored in a low-cost offline archive storage pool.
[0020] As a preferred implementation of this application, the method further includes:
[0021] After the initial classification and labeling of the data is completed, a preset improved popularity prediction model is used to intelligently correct the initial classification results. The classification labels are dynamically updated by combining the logistics time-series decay weight and access trend characteristics to achieve adaptive grading of data popularity. The improved popularity prediction model is based on a time-series LSTM network as the training algorithm and constructs a multi-level improved architecture that includes a logistics business feature embedding layer, a two-layer heterogeneous stacked LSTM layer, a popularity score correction output layer, and a fine-tuning adaptation layer.
[0022] As a specific implementation of this application, the dual-layer heterogeneous LSTM is divided into a shallow LSTM that captures short-term access peaks and a deep LSTM that fits long-term heat decay patterns. A feature fusion gating unit is set between the two layers to automatically allocate long and short time-series feature weights.
[0023] Secondly, embodiments of the present invention also provide a system for classifying, transmitting, storing, and processing historical data of a digital logistics platform, the system comprising:
[0024] The data acquisition module is used to collect historical raw data from the digital logistics platform's operations.
[0025] The data classification module is used to extract features from the historical raw data, combine the extracted features with preset rules to divide the data into real-time hot data, regular temperature data and archived cold data, and bind the corresponding classification labels.
[0026] The transmission scheduling module is used to match the corresponding transmission channel based on the classification label and perform differentiated transmission scheduling.
[0027] A tiered storage module is used to execute a tiered storage strategy based on the classification tags and store the data in the corresponding storage pool.
[0028] The dynamic control module is used to periodically count the access frequency of various types of data, dynamically update the classification labels, and adaptively adjust the bandwidth of each transmission channel and the resource allocation of each storage pool.
[0029] The technical solution provided by this invention divides the transmission channels according to the data classification labels, with hot data preferentially occupying low-latency bandwidth and cold data being transmitted in batches during off-peak hours, thereby avoiding the problems of wasted bandwidth resources and high latency in real-time data transmission.
[0030] To address the issues of unclassified mixed storage and low retrieval efficiency, based on the classification tags and a unified cross-storage index, query requests can be directly directed to the corresponding storage pool, avoiding the time-consuming query caused by full database scanning.
[0031] To address the issues of mixed storage of hot and cold data and high hardware costs, cold data is archived offline at low cost, while only hot data occupies high-performance storage hardware, significantly reducing investment in solid-state storage devices, lowering overall storage hardware operation and maintenance costs, and significantly improving storage resource utilization.
[0032] This forms an integrated solution that combines pre-classification, differentiated transmission, and hierarchical storage. Attached Figure Description
[0033] To more clearly illustrate the specific embodiments of the present invention or the technical solutions in the prior art, the accompanying drawings used in the description of the specific embodiments or the prior art will be briefly introduced below.
[0034] Figure 1 This is a flowchart of a method for classifying, transmitting, storing, and processing historical data of a digital logistics platform, provided in an embodiment of the present invention.
[0035] Figure 2 This is a schematic diagram of a historical data classification, transmission, storage and processing system for a digital logistics platform provided in an embodiment of the present invention. Detailed Implementation
[0036] 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, not all, of the embodiments of the present invention. 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.
[0037] It should be understood that, when used in this specification and the appended claims, the terms "comprising" and "including" indicate the presence of the described features, integrals, steps, operations, elements and / or components, but do not exclude the presence or addition of one or more other features, integrals, steps, operations, elements, components and / or collections thereof.
[0038] Throughout this specification, references to "an embodiment," "an example," or "an example" mean that a particular feature, structure, or characteristic described in connection with that embodiment or example is included in at least one embodiment of the invention. Therefore, the phrases "in an embodiment," "in an embodiment," "an example," or "an example" appearing in various places throughout the specification do not necessarily refer to the same embodiment or example. Furthermore, specific features, structures, or characteristics can be combined in one or more embodiments or examples in any suitable combination and / or sub-combination.
[0039] It should be noted that, unless otherwise stated, the technical terms used in this embodiment have the common meaning as understood in the relevant technical field.
[0040] Please refer to Figure 1 This invention provides a method for classifying, transmitting, storing, and processing historical data of a digital logistics platform, the method comprising:
[0041] S101 collects historical raw data of digital logistics platform business, extracts features, and then combines the extracted features with preset rules to divide the data into real-time hot data, regular warm data and archived cold data, and binds the corresponding classification labels.
[0042] S102, based on the classification label, match the corresponding transmission channel for transmission and perform differentiated transmission scheduling;
[0043] S103, Based on the classification label, execute the hierarchical storage strategy and store it in the corresponding storage pool;
[0044] S104, periodically count the access frequency of various types of data, dynamically update the classification labels, and adaptively adjust the bandwidth of each transmission channel and the resource allocation of each storage pool.
[0045] In this embodiment, the historical raw data includes all historical raw data of business operations, such as order data, transportation trajectory data, warehouse entry and exit data, and settlement ledgers. The metadata information of each data item is read, and the core features are extracted.
[0046] When applying the data, four core features are extracted: data service type, access frequency in the last 30 days, data confidentiality level, and data generation timestamp.
[0047] Construct a pre-defined hierarchical rule base: Pre-configure the judgment criteria for three types of data: hot, warm, and cold. The specific judgment logic is as follows:
[0048] 1) Real-time hot data judgment rules: Data generated within 7 days, accessed ≥ 20 times / day in the past 30 days, and belongs to high-time-sensitive business data such as vehicle trajectory, real-time orders, and warehouse dynamics, with no confidentiality limit; 2) Regular warm data judgment rules: Data generated between 7 and 90 days, accessed 1 to 20 times / day in the past 30 days, and belongs to regular business data such as daily inbound and outbound documents, regular logistics reconciliation, and transit transportation records; 3) Archived cold data judgment rules: Data generated more than 90 days ago, accessed < 1 time / day in the past 30 days, and belongs to low-frequency archived data such as annual settlement ledgers, historical archived orders, and expired transportation records;
[0049] The multidimensional features of each extracted data point are compared one by one with the preset hierarchical rule base. When a single data point meets all the threshold conditions of the corresponding category, the data point is automatically classified into the corresponding data type.
[0050] For the real-time hot data, regular temperature data and archived cold data that have been divided, preset hot data tags, temperature data tags and cold data tags are bound to them respectively. The tags carry data type, generation time, confidentiality level and access frequency attributes, and each data tag is unique and flows with the data throughout the transmission and storage process.
[0051] Furthermore, based on the above technical solution, to reduce retrieval time, the method further includes:
[0052] During storage, a unified index table is synchronously established across storage pools; when a data query request is received, the corresponding storage pool is located based on the unified index table.
[0053] In this embodiment, the differentiated transmission scheduling specifically refers to:
[0054] Real-time thermal data is transmitted using a high-bandwidth, low-latency real-time channel; regular temperature data is transmitted using a balanced bandwidth general channel; and archived cold data is transmitted in batches using off-peak compression.
[0055] The classification tags are sent to the storage terminal along with the data during the transmission process.
[0056] The tiered storage strategy is specifically as follows:
[0057] Real-time thermal data is stored in a high-speed memory cache and a high-performance solid-state storage pool; regular temperature data is stored in a regular mechanical hard disk storage pool; archived cold data is compressed and encrypted and then stored in a low-cost offline archive storage pool.
[0058] By periodically counting the number of queries for various types of data, if the access frequency of a certain type of cold data increases, the tag is automatically updated to warm data, and the transmission channel and storage medium are adjusted to achieve adaptive optimization.
[0059] Furthermore, to address the shortcomings of traditional fixed rules, such as being outdated, inaccurate in classification, and unable to adapt to fluctuations in logistics operations, the method also includes:
[0060] After the initial classification and labeling of the data is completed, a preset improved popularity prediction model is used to intelligently correct the initial classification results. The classification labels are dynamically updated by combining the logistics time-series decay weight and access trend characteristics to achieve adaptive grading of data popularity. The improved popularity prediction model is based on a time-series LSTM network as the training algorithm and constructs a multi-level improved architecture that includes a logistics business feature embedding layer, a two-layer heterogeneous stacked LSTM layer, a popularity score correction output layer, and a fine-tuning adaptation layer.
[0061] Specifically, the dual-layer heterogeneous LSTM is divided into a shallow LSTM that captures short-term access peaks and a deep LSTM that fits long-term heat decay patterns. A feature fusion gating unit is set between the two layers to automatically allocate long and short time-series feature weights.
[0062] The embedding layer enables the vectorized fusion of discrete logistics business features, and the output layer introduces heat compensation and attenuation weights to correct the prediction results. At the same time, an incremental local fine-tuning training method is adopted to complete lightweight iterations based on new business logs every day, without the need for overall retraining, thus balancing the accuracy of time series prediction with the efficiency of online inference and training.
[0063] Introducing a logistics time-series decay weight factor, we compensate for the heat of newly generated logistics business data with a short-term surge in access (such as peak promotion capacity scheduling and temporary trunk line transportation data), so as to avoid being misjudged as warm data by fixed rules.
[0064] By introducing access trend slope features, historical hot data with continuously decreasing access frequency is corrected for heat decay and downgraded to warm data in advance to avoid occupying high-heat storage resources for a long time.
[0065] By improving the heat prediction model to achieve dynamic intelligent classification, the shortcomings of traditional fixed rules that cannot adapt to the tidal fluctuations of logistics business are overcome, which greatly improves the accuracy of data hot and cold classification, further reduces the waste of transmission resources, and improves the utilization rate of storage resources and data retrieval efficiency.
[0066] The above solution divides the transmission channels according to the data classification labels, with hot data taking priority to occupy low-latency bandwidth and cold data being transmitted in batches during off-peak hours, thereby avoiding the waste of bandwidth resources and the problem of high latency in real-time data transmission.
[0067] To address the issues of unclassified mixed storage and low retrieval efficiency, based on the classification tags and a unified cross-storage index, query requests can be directly directed to the corresponding storage pool, avoiding the time-consuming query caused by full database scanning.
[0068] To address the issues of mixed storage of hot and cold data and high hardware costs, cold data is archived offline at low cost, while only hot data occupies high-performance storage hardware, significantly reducing investment in solid-state storage devices, lowering overall storage hardware operation and maintenance costs, and significantly improving storage resource utilization.
[0069] This forms an integrated solution that combines pre-classification, differentiated transmission, and hierarchical storage.
[0070] Reference Figure 2 Based on the same inventive concept, embodiments of the present invention also provide a system for classifying, transmitting, storing, and processing historical data of a digital logistics platform, the system comprising:
[0071] The data acquisition module is used to collect historical raw data from the digital logistics platform's operations.
[0072] The data classification module is used to extract features from the historical raw data, combine the extracted features with preset rules to divide the data into real-time hot data, regular temperature data and archived cold data, and bind the corresponding classification labels.
[0073] The transmission scheduling module is used to match the corresponding transmission channel based on the classification label and perform differentiated transmission scheduling.
[0074] A tiered storage module is used to execute a tiered storage strategy based on the classification tags and store the data in the corresponding storage pool.
[0075] The dynamic control module is used to periodically count the access frequency of various types of data, dynamically update the classification labels, and adaptively adjust the bandwidth of each transmission channel and the resource allocation of each storage pool.
[0076] During storage, a unified index table is synchronously established across storage pools; when a data query request is received, the corresponding storage pool is located based on the unified index table.
[0077] Furthermore, the differentiated transmission scheduling specifically includes:
[0078] Real-time thermal data is transmitted using a high-bandwidth, low-latency real-time channel; regular temperature data is transmitted using a balanced bandwidth general channel; and archived cold data is transmitted in batches using off-peak compression.
[0079] The classification tags are sent to the storage terminal along with the data during the transmission process.
[0080] In another embodiment, the data classification module is further configured to:
[0081] After the initial classification and labeling of the data is completed, a preset improved popularity prediction model is used to intelligently correct the initial classification results. The classification labels are dynamically updated by combining the logistics time-series decay weight and access trend characteristics to achieve adaptive grading of data popularity. The improved popularity prediction model is based on a time-series LSTM network as the training algorithm and constructs a multi-level improved architecture that includes a logistics business feature embedding layer, a two-layer heterogeneous stacked LSTM layer, a popularity score correction output layer, and a fine-tuning adaptation layer.
[0082] It should be noted that for a more detailed description of the workflow and implementation of the system embodiments, please refer to the aforementioned method embodiments section, which will not be repeated here.
[0083] The entire system divides transmission channels according to data classification labels, prioritizing low-latency bandwidth for hot data and staggering the transmission of cold data in batches, thereby avoiding bandwidth waste and high latency in real-time data transmission.
[0084] To address the issues of unclassified mixed storage and low retrieval efficiency, based on the classification tags and a unified cross-storage index, query requests can be directly directed to the corresponding storage pool, avoiding the time-consuming query caused by full database scanning.
[0085] To address the issues of mixed cold and hot data storage and high hardware costs, cold data is archived offline at low cost, while only hot data occupies high-performance storage hardware, significantly reducing investment in solid-state storage devices. This reduces overall storage hardware operation and maintenance costs and significantly improves storage resource utilization, thus forming an integrated solution that combines front-end classification, differentiated transmission, and tiered storage.
[0086] In this embodiment, a computer-readable storage medium is also provided, which stores a computer program. When the computer program is executed by a processor, the processor performs the steps of a method for classifying, transmitting, storing, and processing historical data of a digital logistics platform as described in the method embodiment.
[0087] Specifically, the processor can be a Central Processing Unit (CPU), but it can also be other general-purpose processors, digital signal processors (DSPs), application-specific integrated circuits (ASICs), field-programmable gate arrays (FPGAs), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. A general-purpose processor can be a microprocessor or any conventional processor.
[0088] The code of the computer program can be in the form of source code, object code, executable file, or some intermediate form.
[0089] Computer-readable storage media may include cache, high-speed random access memory (RAM), such as the common double data rate synchronous dynamic random access memory (DDR SDRAM), and may also include non-volatile memory (NVRAM), such as one or more read-only memory (ROM), disk storage devices, flash memory devices, or other non-volatile solid-state storage devices such as optical discs (CD-ROM, DVD-ROM), floppy disks, or data tapes.
[0090] In the embodiments provided in this application, it should be understood that the disclosed methods and apparatus can also be implemented in other ways. The method embodiments described above are merely illustrative. It should also be noted that in some alternative implementations, the functions marked in the blocks may occur in a different order than that shown in the figures. For example, two consecutive blocks may actually be executed substantially in parallel, or they may sometimes be executed in reverse order, depending on the functions involved. It should also be noted that each block in the block diagram and / or flowchart, and combinations of blocks in the block diagram and / or flowchart, can be implemented using dedicated hardware-based apparatus that performs the specified functions or actions, or can be implemented using a combination of dedicated hardware and computer instructions.
[0091] Furthermore, the functional modules in the various embodiments of this invention can be integrated together to form an independent part, or each module can exist independently, or two or more modules can be integrated to form an independent part. The integrated modules can be implemented in hardware or as software functional modules. When in use, each module will only collect and store user information with the user's full authorization and in compliance with relevant laws and regulations, protecting the security and privacy of user data and strictly prohibiting unauthorized access; data processing will be conducted within the scope stipulated by law and will not exceed the purpose and scope authorized by the user; at the same time, users have the rights to access, correct, delete, restrict, and refuse processing of their personal data; and must strictly comply with applicable laws and regulations and conduct compliance reviews.
[0092] If the aforementioned functions are implemented as software functional modules and sold or used as independent products, they can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention, in essence, or the part that contributes to the prior art, or a part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of the present invention. The above descriptions are merely preferred embodiments of the present invention and are not intended to limit the present invention. For those skilled in the art, the present invention can have various modifications and variations. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the scope of protection of the present invention.
Claims
1. A method for classifying, transmitting, storing, and processing historical data on a digital logistics platform, characterized in that: The method includes: Collect historical raw data from the digital logistics platform's business, extract features, and then combine the extracted features with preset rules to divide the data into real-time hot data, regular warm data, and archived cold data, and bind corresponding classification tags; Based on the classification labels, the corresponding transmission channels are matched for transmission, and differentiated transmission scheduling is performed. Based on the classification labels, a tiered storage strategy is executed, and the data is stored in the corresponding storage pool; The system periodically counts the access frequency of various types of data, dynamically updates the classification labels, and adaptively adjusts the bandwidth of each transmission channel and the resource allocation of each storage pool.
2. The method for classifying, transmitting, storing, and processing historical data of a digital logistics platform as described in claim 1, characterized in that, The method further includes: During storage, a unified index table is synchronously established across storage pools; when a data query request is received, the corresponding storage pool is located based on the unified index table.
3. The method for classifying, transmitting, storing, and processing historical data of a digital logistics platform as described in claim 1, characterized in that, The differentiated transmission scheduling specifically refers to: Real-time thermal data is transmitted using a high-bandwidth, low-latency real-time channel; regular temperature data is transmitted using a balanced bandwidth general channel; and archived cold data is transmitted in batches using off-peak compression. The classification tags are sent to the storage terminal along with the data during the transmission process.
4. The method for classifying, transmitting, storing, and processing historical data of a digital logistics platform as described in claim 1, characterized in that, The tiered storage strategy is specifically as follows: Real-time thermal data is stored in a high-speed memory cache and a high-performance solid-state storage pool; regular temperature data is stored in a regular mechanical hard disk storage pool; archived cold data is compressed and encrypted and then stored in a low-cost offline archive storage pool.
5. A method for classifying, transmitting, storing, and processing historical data of a digital logistics platform as described in any one of claims 1 to 4, characterized in that, The method further includes: After the initial classification and labeling of the data is completed, a preset improved popularity prediction model is used to intelligently correct the initial classification results. The classification labels are dynamically updated by combining the logistics time-series decay weight and access trend characteristics to achieve adaptive grading of data popularity. The improved popularity prediction model is based on a time-series LSTM network as the training algorithm and constructs a multi-level improved architecture that includes a logistics business feature embedding layer, a two-layer heterogeneous stacked LSTM layer, a popularity score correction output layer, and a fine-tuning adaptation layer.
6. The method for classifying, transmitting, storing, and processing historical data of a digital logistics platform as described in claim 5, characterized in that, The dual-layer heterogeneous LSTM is divided into a shallow LSTM that captures short-term access peaks and a deep LSTM that fits long-term heat decay patterns. A feature fusion gating unit is set between the two layers to automatically allocate the weights of long and short time-series features.
7. A system for classifying, transmitting, storing, and processing historical data of a digital logistics platform, characterized in that: The system includes: The data acquisition module is used to collect historical raw data from the digital logistics platform's operations. The data classification module is used to extract features from the historical raw data, combine the extracted features with preset rules to divide the data into real-time hot data, regular temperature data and archived cold data, and bind the corresponding classification labels. The transmission scheduling module is used to match the corresponding transmission channel based on the classification label and perform differentiated transmission scheduling. A tiered storage module is used to execute a tiered storage strategy based on the classification tags and store the data in the corresponding storage pool. The dynamic control module is used to periodically count the access frequency of various types of data, dynamically update the classification labels, and adaptively adjust the bandwidth of each transmission channel and the resource allocation of each storage pool.
8. The historical data classification, transmission, storage, and processing system for a digital logistics platform as described in claim 7, characterized in that, During storage, a unified index table is synchronously established across storage pools; when a data query request is received, the corresponding storage pool is located based on the unified index table.
9. A historical data classification, transmission, storage, and processing system for a digital logistics platform as described in claim 7, characterized in that, The differentiated transmission scheduling specifically refers to: Real-time thermal data is transmitted using a high-bandwidth, low-latency real-time channel; regular temperature data is transmitted using a balanced bandwidth general channel; and archived cold data is transmitted in batches using off-peak compression. The classification tags are sent to the storage terminal along with the data during the transmission process.
10. The historical data classification, transmission, storage, and processing system for a digital logistics platform as described in claim 7, characterized in that, The data classification module is also used for: After the initial classification and labeling of the data is completed, a preset improved popularity prediction model is used to intelligently correct the initial classification results. The classification labels are dynamically updated by combining the logistics time-series decay weight and access trend characteristics to achieve adaptive grading of data popularity. The improved popularity prediction model is based on a time-series LSTM network as the training algorithm and constructs a multi-level improved architecture that includes a logistics business feature embedding layer, a two-layer heterogeneous stacked LSTM layer, a popularity score correction output layer, and a fine-tuning adaptation layer.