CN109815225B - Parallel prefix data retrieval method and system based on prefix tree structure - Google Patents
Parallel prefix data retrieval method and system based on prefix tree structure Download PDFInfo
- Publication number
- CN109815225B CN109815225B CN201811511690.1A CN201811511690A CN109815225B CN 109815225 B CN109815225 B CN 109815225B CN 201811511690 A CN201811511690 A CN 201811511690A CN 109815225 B CN109815225 B CN 109815225B
- Authority
- CN
- China
- Prior art keywords
- prefix
- segmented
- tree
- network
- retrieval
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
Images
Abstract
The invention relates to a prefix data retrieval method and a system, comprising the following steps: a network prefix in a binary representation is obtained,constructing a prefix tree for the network prefix, and carrying out segmentation processing on the prefix tree to obtain a multi-segment segmented tree; obtaining binary expressed network prefix to be retrieved, carrying out segmentation processing on the network prefix to be retrieved to obtain a plurality of segmented prefixes, carrying out concurrent retrieval on the segmented trees corresponding to each segmented prefix to obtain retrieval result riAnd representing the retrieval result of the ith segment prefix in the ith segment tree, performing logic AND operation on all the retrieval results, and taking the operation result as the retrieval result of the network prefix to be retrieved. The invention can construct the information base of the prefix information by using the segmented prefix tree structure, can carry out parallelization on the specific prefix to obtain prefix matching search, and realizes the great improvement of the prefix matching efficiency.
Description
Technical Field
The invention relates to the field of information retrieval, in particular to a method and a system for retrieving parallelized prefix data based on a prefix tree structure.
Background
With the popularization of networks in the information era and the rapid development of emerging services such as big data and the like, network information shows an explosive growth trend. There is a huge amount of network prefix data in the global internet. In order to perform good management on the internet and perform timely detection and alarm on network abnormality, organization and construction of a network prefix information base become abnormally important.
A network prefix is a collection of a set of IP addresses, each prefix representing a small network. In order to perform good management on the internet and perform timely detection and alarm on network abnormality, organization and construction of a network prefix information base are indispensable.
There are two main ways for constructing the information base of the network prefix:
1. and constructing the information base in a hash table mode. In the mode, the prefix is used as a Key Value, the information to be stored is used as a Value, and the information is stored in a Key Value pair mode;
2. and constructing an information base based on the form of the binary prefix tree. The method expresses prefix information in a binary form and constructs a binary tree for the network prefix information. Each binary tree node represents a network prefix, and data information is stored in the node. The prefix query and matching process is carried out in the binary tree, namely, the branches of the binary tree are selected downwards layer by layer according to the bits of the binary representation of the prefix to be retrieved until the target prefix is found.
For the two modes, when the hash table mode faces massive network prefixes, the occupied storage space is large, byte-by-byte comparison is needed for processing the matching of the sub-prefixes or the longest prefix, and the real-time or quasi-real-time requirements under a big data scene are difficult to meet; the traditional prefix tree construction method solves the problems of the hash table method to a certain extent, but for an IPV4 network, the binary representation of the network prefix is 32 bits, and the binary representation always needs to be searched layer by layer from the root node in the prefix matching process, 8 layers need to be searched for the prefix with the length of 8, and 24 layers need to be searched for the prefix with the length of 24, so that the search time for the prefix with the longer length is longer.
Disclosure of Invention
The invention aims to solve the problems that in the storage and retrieval of massive network prefixes, the traditional prefix information base construction method (a hash table and a traditional prefix tree) has low efficiency in the execution of searching, inserting and deleting operations due to huge data volume, is difficult to maintain and the like, and provides a parallel prefix matching query technology based on a prefix tree structure. The technology of the invention is oriented to massive network prefixes, and the prefix tree structure is segmented to obtain the retrieval entries at each segment at the same time, so that the operations of searching, inserting and deleting are carried out concurrently, and the execution efficiency is greatly improved. The technology can remarkably improve the prefix information retrieval work under massive network prefixes in a parallelization mode, and greatly improves the retrieval efficiency.
In order to solve the above technical problem, an object of the present invention is to provide a prefix data retrieval method, including:
step 2, obtaining the binary expressed network prefix to be retrieved, carrying out segmentation processing on the network prefix to be retrieved to obtain a plurality of segmented prefixes, carrying out concurrent retrieval on the segmented trees corresponding to each segmented prefix to obtain a retrieval result riAnd representing the retrieval result of the ith segment prefix in the ith segment tree, performing logic AND operation on all the retrieval results, and taking the operation result as the retrieval result of the network prefix to be retrieved.
The prefix data retrieval method, wherein the segmentation processing in step 1 specifically includes: segmenting the prefix tree according to the preset segmentation quantity, wherein the obtained segmentation tree quantity is equal to the preset segmentation quantity;
the segmentation treatment in the step 2 specifically comprises the following steps: and segmenting the network prefix to be retrieved, wherein the number of the obtained approximate segmentation prefixes is equal to the preset segmentation number.
The prefix data retrieval method for performing segmentation processing on the prefix tree specifically comprises the following steps: the structure of the prefix tree is divided into one segment according to each 8 layers, wherein the ith segment is a segmented tree, and 256 segments existi-1And the index entry is used for being used as a matching entry of each parallelized segmented prefix during parallel retrieval, the matching entry of the 1 st segmented prefix is a root node of the prefix tree, the matching entry is 0, and the matching entry of the ith segmented prefix except the 1 st segmented prefix is an integer value of binary representation of the preceding (i-1) segment of the segmented prefix.
The prefix data retrieval method further comprises:
and 3, acquiring the network prefix to be inserted, carrying out segmentation processing on the network prefix to be inserted, and respectively carrying out parallel insertion operation on the segmented trees of the corresponding segments.
The prefix data retrieval method is characterized in that the network prefix is an IPV4 network prefix.
The invention also discloses a prefix data retrieval system, which comprises:
the prefix tree segmentation module is used for acquiring a network prefix represented by a binary system, constructing a prefix tree for the network prefix, and performing segmentation processing on the prefix tree to obtain a multi-segment segmented tree;
a retrieval module for obtaining binary expressed network prefix to be retrieved, carrying out segmentation processing on the network prefix to be retrieved to obtain a plurality of segmented prefixes, carrying out concurrent retrieval on the segmented tree corresponding to each segmented prefix to obtain retrieval result riAnd representing the retrieval result of the ith segment prefix in the ith segment tree, performing logic AND operation on all the retrieval results, and taking the operation result as the retrieval result of the network prefix to be retrieved.
In the prefix data retrieval system, the segmentation process in the prefix tree segmentation module specifically includes: segmenting the prefix tree according to the preset segmentation quantity, wherein the obtained segmentation tree quantity is equal to the preset segmentation quantity;
the segmentation processing in the retrieval module specifically comprises: and segmenting the network prefix to be retrieved, wherein the number of the obtained approximate segmentation prefixes is equal to the preset segmentation number.
The prefix data retrieval system, which performs segmentation processing on the prefix tree, specifically includes: the structure of the prefix tree is divided into one segment according to each 8 layers, wherein the ith segment is a segmented tree, and 256 segments existi-1And the index entry is used for being used as a matching entry of each parallelized segmented prefix during parallel retrieval, the matching entry of the 1 st segmented prefix is a root node of the prefix tree, the matching entry is 0, and the matching entry of the ith segmented prefix except the 1 st segmented prefix is an integer value of binary representation of the preceding (i-1) segment of the segmented prefix.
The prefix data retrieval system further includes:
and the inserting module is used for acquiring the network prefix to be inserted, carrying out segmentation processing on the network prefix to be inserted, and respectively carrying out parallel inserting operation on the segmented trees of the corresponding segments.
The prefix data retrieval system, wherein the network prefix is an IPV4 network prefix.
The invention thus has the technical advances comprising:
the network prefix and the segmentation representation method of the prefix tree structure can carry out segmentation representation on the traditional prefix tree structure, and a retrieval inlet of each segment can be determined; the parallelization prefix matching technology provided by the invention can perform prefix matching and searching in parallel on each section of prefix tree structure, thereby realizing the efficiency improvement of approximate linearity. The general technical effect is that the information base construction can be carried out on the prefix information by using a segmented prefix tree structure, the specific prefix can be parallelized to obtain prefix matching search, and the prefix matching efficiency is greatly improved.
Drawings
FIG. 1 is a schematic diagram of a segmented prefix tree;
FIG. 2 is a flow chart of the method of the present invention.
Detailed Description
When the inventor conducts prefix tree algorithm optimization research, the matching search algorithm of the traditional prefix tree is found to be highly efficient for processing prefixes with short length, and the processing time is gradually increased along with the increase of the prefix length. Now, with the improvement of CPU performance and the rapid development of parallel computing, the computing power of a computer is no longer the bottleneck of improving system efficiency, and parallel programming becomes a fundamental and important part of system design. The inventors consider applying parallel programming to algorithmic optimization of prefix trees. For all prefixes, each plurality of bits is divided into a plurality of sections, and index entries in each section can be found through the prefixes during prefix matching, so that parallel searching, inserting and deleting operations can be performed on each section, and the retrieval time of the prefixes with the length of 8 can be approximated even if the longest prefixes are matched, and the retrieval efficiency is improved.
In practical application, the invention can process matching query of massive network prefixes. The technology provided by the invention firstly segments the network prefix information so as to complete a prefix information base and store the prefix information in a segmented prefix tree; when searching and inserting a new prefix, the parallelization prefix matching technology is used for operating in the prefix tree to complete the parallelization matching query of the prefix.
As shown in fig. 2, the method of the present invention comprises: step 1, segmented representation of network prefixes and prefix tree structures. And acquiring a network prefix represented by a binary system, constructing a prefix tree for the network prefix, and performing segmentation processing on the prefix tree to obtain a multi-segment segmented tree.
And performing binary representation on the network prefix, and dividing each plurality of bits into a plurality of sections. Taking IPV4 network prefix as an example, each 8 bits is a segment, and its binary representation and corresponding prefix tree structure have 32 levels, it is divided into 4 segments, i.e. the prefix tree structure is also divided into 4 segments each 8 levels, where the 32 levels is determined by the network prefix itself of IPV 4. The network binary of IPV4 requires 32 bits, namely 32 layers; if 8 bits are taken as 1 segment, 4 segments are needed; for segments where i is a positive integer, there are 256i-1An index entry, and for a particular prefix its index entry is specified by an integer number of binary representations of the first (i-1) segment prefix. In fig. 1, a prefix tree structure divided into 4 segments is shown, with dashed lines indicating the index of each segment and numbers above the nodes indicating the cable entries. 256 are leaf nodes of a full binary tree with a depth of 8. For a prefix tree with 32 layers of depth, every 8 layers are one segment, and the number of leaf nodes is 256, and the total number is 4 segments. Segment 1 has a root node as an index entry, 2561-11 inlet; paragraph 2 depends on the leaf node index of the first paragraph for a total of 2562-1256 inlets; similarly, the 3 rd section and the 4 th section respectively have 2563-1、2564-1An inlet. The index entry serves as a matching entry for each part (i.e., segment) of the parallelization when performing parallelization prefix matching.
The specific prefix refers to a target prefix given to be retrieved by a query. Taking 124.26.3.0/24, a 24-bit decimal network prefix as an example, because the length is 24, the invention only needs to consider the first 3 segments for each 8 bits: the 1 st segment index entry is the root node, and the index position (matching entry) is 0; the 2 nd segment index entry depends on the integer value of the 1 st segment, the index entry position is 124-1-123, the 3 rd segment index entry depends on the previous two segments, and the index entry position is 124-26-1-3223. Note: the index position value starts from 0.
And (3) representing the network prefix by segments, correspondingly segmenting the prefix tree, finding a retrieval inlet of each segment, and laying a foundation for parallel searching and matching. The tree structure layering is the operation corresponding to the prefix segmentation, and aims to perform parallelization retrieval on each layer after layering; it is the network prefix that is segmented that naturally leads to tree structure layering. The tree structure and the network prefix are in a corresponding relationship, and the 32-bit prefix corresponds to 32 layers of the tree structure. Every 8 bits of the prefix is a segment, and every 8 layers of the corresponding tree structure are segments.
The prefix tree is used as a storage structure for storing network prefixes, and the structure supports insertion and retrieval of prefixes. The prefix tree may be different in content in different scenarios. For example, a system needs a large amount of network prefix information to construct a prefix tree, and then prefixes used for constructing the prefix tree in the early stage are few (in an extreme case, the prefix tree is empty initially); the prefixes stored in the tree then grow progressively over time. At any point in time, a search can be made in the prefix tree.
And 2, matching parallelized prefixes. Based on the constructed segmented prefix tree structure, when a given prefix is subjected to matching query, concurrent search can be performed on each segment of the prefix tree structure, and after the search of each segment is completed, the search result of each segment is operated to obtain a final result. Obtaining binary expressed network prefix to be retrieved, carrying out segmentation processing on the network prefix to be retrieved to obtain a plurality of segmented prefixes, carrying out concurrent retrieval on the segmented trees corresponding to each segmented prefix to obtain retrieval result riAnd representing the retrieval result of the ith segment prefix in the ith segment tree, performing logic AND operation on all the retrieval results, and taking the operation result as the retrieval result of the network prefix to be retrieved.
The step 2 specifically comprises the following steps: and step 21, parallelizing prefix search. Segmenting prefixes to be searched, respectively searching on prefix trees of corresponding segments, and respectively obtaining results ri(i is 1,2,3 … N, N is the number of segments of the prefix division, and if 8 bits are one segment, N is 4), the search is performed in parallel, and the segments are performed simultaneously. Still taking the prefix 124.26.3.0/24, 8 bits as a segment as an example, the specific search process:segment 1 search 124, binary represented as 01111100, sequentially finds whether nodes exist in the segment 1 prefix tree, if so, the result r1Returning true, otherwise false; the 2 nd section and the 3 rd section are simultaneously carried out in the same way to obtain r2And r3. "corresponding" refers to a segment corresponding to a prefix segment to be searched. r isiRepresenting the search result in the ith segment, true represents that the search is successful, false represents that the search is failed, and the final search result is:
r＝r1&&r2&&r3…&&rN
wherein "& &" represents a logical AND.
And 3, parallelizing prefix insertion. And segmenting the prefix to be inserted, and respectively performing insertion operation on the prefix trees of the corresponding segments. Taking the prefix 124.26.3.0/24 to be inserted, 8 bits as a segment, the specific insertion process is as follows: inserting 124 the 1 st segment, inserting 01111100 binary representation bits into the 1 st segment prefix tree according to 0/1, if the node exists, then continuing downward without repeated insertion; similarly, the 2 nd and 3 rd sections are inserted with 26(00011010) and 3(00000011) at the same time. Thus, the parallelized insert operation is complete. The method and the device realize the matching query of the prefixes by segmentation and concurrence, realize the parallelization search and insertion operation and improve the operation efficiency.
The following are system examples corresponding to the above method examples, and this embodiment can be implemented in cooperation with the above embodiments. The related technical details mentioned in the above embodiments are still valid in this embodiment, and are not described herein again in order to reduce repetition. Accordingly, the related-art details mentioned in the present embodiment can also be applied to the above-described embodiments.
The invention also discloses a prefix data retrieval system, which comprises:
the prefix tree segmentation module is used for acquiring a network prefix represented by a binary system, constructing a prefix tree for the network prefix, and performing segmentation processing on the prefix tree to obtain a multi-segment segmented tree;
retrieval module for obtaining a binary representation to be examinedNetwork prefix is required, the network prefix to be retrieved is segmented to obtain a plurality of segmented prefixes, concurrent retrieval is carried out on the segmented tree corresponding to each segmented prefix to obtain a retrieval result riAnd representing the retrieval result of the ith segment prefix in the ith segment tree, performing logic AND operation on all the retrieval results, and taking the operation result as the retrieval result of the network prefix to be retrieved.
In the prefix data retrieval system, the segmentation process in the prefix tree segmentation module specifically includes: segmenting the prefix tree according to the preset segmentation quantity, wherein the obtained segmentation tree quantity is equal to the preset segmentation quantity;
the segmentation processing in the retrieval module specifically comprises: and segmenting the network prefix to be retrieved, wherein the number of the obtained approximate segmentation prefixes is equal to the preset segmentation number.
The prefix data retrieval system, which performs segmentation processing on the prefix tree specifically includes that the structure of the prefix tree is divided into one segment according to each 8 layers, wherein the ith segment of the segmented tree has 256 segmentsi-1And the index entry is used for being used as a matching entry of each parallelized segmented prefix during parallel retrieval, the matching entry of the 1 st segmented prefix is a root node of the prefix tree, the matching entry is 0, and the matching entry of the ith segmented prefix except the 1 st segmented prefix is an integer value of binary representation of the preceding (i-1) segment of the segmented prefix.
The prefix data retrieval system further includes:
and the inserting module is used for acquiring the network prefix to be inserted, carrying out segmentation processing on the network prefix to be inserted, and respectively carrying out parallel inserting operation on the segmented trees of the corresponding segments.
Claims (10)
1. A prefix data retrieval method, comprising:
step 1, obtaining a network prefix represented by binary, constructing a prefix tree for the network prefix, and performing segmentation processing on the prefix tree to obtain a multi-segment segmented tree;
step 2, obtaining binary expressed network prefix to be retrieved and using the prefixThe network prefix to be retrieved is segmented to obtain a plurality of segmented prefixes, concurrent retrieval is carried out on the segmented tree corresponding to each segmented prefix to obtain a retrieval result riAnd representing the retrieval result of the ith segment prefix in the ith segment tree, performing logic AND operation on all the retrieval results, and taking the operation result as the retrieval result of the network prefix to be retrieved.
2. The prefix data retrieval method of claim 1, wherein the segmentation processing in step 1 specifically includes: segmenting the prefix tree according to the preset segmentation quantity, wherein the obtained segmentation tree quantity is equal to the preset segmentation quantity;
the segmentation treatment in the step 2 specifically comprises the following steps: and segmenting the network prefix to be retrieved, wherein the obtained number of the segmented prefixes is equal to the preset number of the segments.
3. The prefix data retrieval method of claim 2, wherein the step of segmenting the prefix tree specifically comprises: the structure of the prefix tree is divided into one segment according to each 8 layers, wherein the ith segment is a segmented tree, and 256 segments existi-1And the index entry is used for being used as a matching entry of each parallelized segmented prefix during parallel retrieval, the matching entry of the 1 st segmented prefix is a root node of the prefix tree, the matching entry is 0, and the matching entry of the ith segmented prefix except the 1 st segmented prefix is an integer value of binary representation of the preceding (i-1) segment of the segmented prefix.
4. The prefix data retrieval method of claim 1 or 3, further comprising:
and 3, acquiring the network prefix to be inserted, carrying out segmentation processing on the network prefix to be inserted, and respectively carrying out parallel insertion operation on the segmented trees of the corresponding segments.
5. The prefix data retrieval method of claim 1, wherein the network prefix is an IPV4 network prefix.
6. A prefix data retrieval system, comprising:
the prefix tree segmentation module is used for acquiring a network prefix represented by a binary system, constructing a prefix tree for the network prefix, and performing segmentation processing on the prefix tree to obtain a multi-segment segmented tree;
a retrieval module for obtaining binary expressed network prefix to be retrieved, carrying out segmentation processing on the network prefix to be retrieved to obtain a plurality of segmented prefixes, carrying out concurrent retrieval on the segmented tree corresponding to each segmented prefix to obtain retrieval result riAnd representing the retrieval result of the ith segment prefix in the ith segment tree, performing logic AND operation on all the retrieval results, and taking the operation result as the retrieval result of the network prefix to be retrieved.
7. The prefix data retrieval system of claim 6, wherein the segmentation process in the prefix tree segmentation module specifically includes: segmenting the prefix tree according to the preset segmentation quantity, wherein the obtained segmentation tree quantity is equal to the preset segmentation quantity;
the segmentation processing in the retrieval module specifically comprises: and segmenting the network prefix to be retrieved, wherein the obtained number of the segmented prefixes is equal to the preset number of the segments.
8. The prefix data retrieval system of claim 7, wherein segmenting the prefix tree specifically includes: the structure of the prefix tree is divided into one segment according to each 8 layers, wherein the ith segment is a segmented tree, and 256 segments existi-1And the index entry is used for being used as a matching entry of each parallelized segmented prefix during parallel retrieval, the matching entry of the 1 st segmented prefix is a root node of the prefix tree, the matching entry is 0, and the matching entry of the ith segmented prefix except the 1 st segmented prefix is an integer value of binary representation of the preceding (i-1) segment of the segmented prefix.
9. The prefix data retrieval system of claim 6 or 8, further comprising:
and the inserting module is used for acquiring the network prefix to be inserted, carrying out segmentation processing on the network prefix to be inserted, and respectively carrying out parallel inserting operation on the segmented trees of the corresponding segments.
10. The prefix data retrieval system of claim 6, wherein the network prefix is an IPV4 network prefix.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201811511690.1A CN109815225B (en) | 2018-12-11 | 2018-12-11 | Parallel prefix data retrieval method and system based on prefix tree structure |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201811511690.1A CN109815225B (en) | 2018-12-11 | 2018-12-11 | Parallel prefix data retrieval method and system based on prefix tree structure |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN109815225A CN109815225A (en) | 2019-05-28 |
| CN109815225B true CN109815225B (en) | 2021-07-09 |
Family
ID=66602856
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201811511690.1A Active CN109815225B (en) | 2018-12-11 | 2018-12-11 | Parallel prefix data retrieval method and system based on prefix tree structure |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN109815225B (en) |
Citations (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103064841A (en) * | 2011-10-20 | 2013-04-24 | 北京中搜网络技术股份有限公司 | Retrieval device and retrieval method |
Family Cites Families (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US7440461B2 (en) * | 2003-12-23 | 2008-10-21 | Intel Corporation | Methods and apparatus for detecting patterns in a data stream |
| CN101430755A (en) * | 2008-11-13 | 2009-05-13 | 王新锋 | Label recognition anti-collision processing method in radio frequency recognition technology |
| US8160069B2 (en) * | 2009-01-30 | 2012-04-17 | Palo Alto Research Center Incorporated | System for forwarding a packet with a hierarchically structured variable-length identifier |
| US8880507B2 (en) * | 2010-07-22 | 2014-11-04 | Brocade Communications Systems, Inc. | Longest prefix match using binary search tree |
| CN105005621A (en) * | 2015-07-23 | 2015-10-28 | 张真 | Method for constructing distributed storage and parallel indexing system for big data |
| CN107967219B (en) * | 2017-11-27 | 2021-08-06 | 北京理工大学 | TCAM-based large-scale character string high-speed searching method |
-
2018
- 2018-12-11 CN CN201811511690.1A patent/CN109815225B/en active Active
Patent Citations (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103064841A (en) * | 2011-10-20 | 2013-04-24 | 北京中搜网络技术股份有限公司 | Retrieval device and retrieval method |
Also Published As
| Publication number | Publication date |
|---|---|
| CN109815225A (en) | 2019-05-28 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| JP6028567B2 (en) | Data storage program, data search program, data storage device, data search device, data storage method, and data search method | |
| US7499912B2 (en) | Search method using coded keys | |
| US10671586B2 (en) | Optimal sort key compression and index rebuilding | |
| US11630864B2 (en) | Vectorized queues for shortest-path graph searches | |
| US20180089243A1 (en) | Bloom filter index for device discovery | |
| US10241979B2 (en) | Accelerated detection of matching patterns | |
| CN111801665A (en) | Hierarchical Locality Sensitive Hash (LSH) partition indexing for big data applications | |
| CN102867049A (en) | Chinese PINYIN quick word segmentation method based on word search tree | |
| CN114780502B (en) | Database method, system, device and medium based on compressed data direct computation | |
| CN111813744A (en) | File searching method, device, equipment and storage medium | |
| CN109815225B (en) | Parallel prefix data retrieval method and system based on prefix tree structure | |
| CN110941730B (en) | Retrieval method and device based on human face feature data migration | |
| CN1235169C (en) | Data storage and searching method of embedded system | |
| CN111382170B (en) | Sentence automatic conversion method and device | |
| CN110807092A (en) | Data processing method and device | |
| CN113076330A (en) | Query processing method and device, database system, electronic equipment and storage medium | |
| CN113946580B (en) | Massive heterogeneous log data retrieval middleware | |
| KR102388458B1 (en) | Method and apparatus for conversing data key value | |
| CN113726342B (en) | Segmented difference compression and inert decompression method for large-scale graph iterative computation | |
| US20220345315A1 (en) | Associative hash tree | |
| Luo | Learning Augmented Binary Search Trees | |
| CN113946580A (en) | Mass heterogeneous log data retrieval middleware | |
| KR20080050296A (en) | Look up apparatus and method of forwarding table | |
| Pinkas et al. | A simple recursive tree oblivious ram | |
| CN116303414A (en) | Branching of tree structures in database systems |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |