CN115048377A - Time-space keyword query method under mixed storage block chain environment - Google Patents

Abstract

The invention provides a time-space keyword query method under a mixed storage block chain environment, which aims at the problem of time-space keyword query of a mixed storage block chain under a chain link, and relates to the technical field of computer block chain query; firstly, aiming at weak query semantics, constructing a block chain model CSBM which is classified according to attributes and endowed with semantics, dividing attribute types for transactions in blocks and adding semantics; secondly, for less efficient queries, the construction is based on B ² The chain two-level index structure of the F-BKM structure can index all blocks and transactions, and effectively improves the queryEfficiency; finally, aiming at the condition of high communication overhead, designing a method for searching uplink and downlink space-time keywords in a chain, and traversing B ² The F-BKM structure carries out on-chain condition query, and then carries out on-chain data connection query according to the connection attribute value, thereby reducing unnecessary communication overhead compared with the traditional query method; experiments prove that the index has good stability, the query efficiency is greatly improved, and the communication overhead is obviously reduced.

Description

A spatiotemporal keyword query method in a hybrid storage blockchain environment

technical field

The invention belongs to the technical field of computer block chain query, and in particular relates to a space-time keyword query method in a hybrid storage block chain environment.

Background technique

In recent years, with the increasing popularity of digital cryptocurrencies such as Bitcoin, its underlying technology, blockchain technology, has become the focus of current society. In essence, blockchain can be regarded as a distributed database, which has the characteristics of decentralization, non-tampering, traceability and data transparency, which can well solve the trust between multiple participants in supply chain and other scenarios. question. Spatiotemporal data is a common form of data representation in the supply chain, Internet of Things and other fields. With the application of blockchain in the above scenarios, a large amount of spatiotemporal data is stored on the blockchain. Time-space keyword query is to use time, space and keywords as query conditions to find all the information that meets the conditions, and is widely used in the above scenarios.

The current commonly used blockchain storage mode is the on-chain and off-chain hybrid storage mode, where operational transactions that can be shared by multiple parties are stored on the chain, while the private data of all parties is stored off-chain. When performing a spatiotemporal keyword query in this mode, it is necessary to return the result of the data connection on the chain and off the chain. Since most of the existing query methods use B+ tree indexes in blocks, the query efficiency is low for spatiotemporal keyword queries with multi-dimensional search space. In addition, the existing method needs to read all the data in the off-chain connection table for connection when querying, which leads to a lot of communication overhead and affects the user experience. Therefore, it is of great significance to study the spatiotemporal keyword query method in the hybrid storage blockchain environment.

Taking a real scenario as an example, in a supply chain scenario, multiple participants openly and transparently store their production information, logistics information, and sales information on the blockchain for data sharing and collaborative processing by multiple parties, while off-chain storage of each Participant's private data. When a user sends a query Q with the space-time keyword shown in the figure, it is necessary to connect the on-chain and off-chain transactions that satisfy the query conditions through the connection attribute and return it to the querying user. The existing query method needs to query according to the B+ tree index of each attribute column in the block, take the intersection of the query results to obtain all transactions that meet the query conditions, and then read all the data in the off-chain connection table to the blockchain for connection , and finally return the connected result to the query user. The shortcomings of this query method mainly lie in the following three aspects:

First, query semantics are weak. Although transactions on the blockchain contain various properties, existing blockchain platforms usually store transactions in the form of key-value or given semantics. The key-value mode does not support relational queries such as selection and connection, and cannot query transactions based on attribute names; the storage mode with semantics needs to build a separate index for each attribute column, and cannot merge indexes according to attribute types.

Second, the query efficiency is low. Transactions on the chain contain attributes such as time, space, and keywords, and are distributed in each block. Most of the existing blockchain query methods avoid traversing all transactions by using the B+ tree index in the block, but this structure has low query efficiency for spatiotemporal keyword queries with multi-dimensional search space.

Third, the communication overhead is large. Since both on-chain and off-chain data need to be obtained when querying, the existing method needs to read all the data in the off-chain connection table for connection. A lot of unnecessary communication overhead.

SUMMARY OF THE INVENTION

Aiming at the deficiencies of the prior art, the present invention designs a spatiotemporal keyword query method in a hybrid storage blockchain environment.

A spatiotemporal keyword query method in a hybrid storage blockchain environment, the specific steps are as follows:

Step 1: Build a blockchain model classified by attributes and given semantics, namely CSBM;

Step 1.1: CSBM contains multiple blocks CS-B classified by attributes and given semantics, CSBM=CS-B ₁ +CS-B ₂ +...+CS-B _n , each CS-B contains block header B The head and the block body CS-B body classified by attributes and given semantics are two parts; where CS-B=B head+CS-B body; B head is basically the same as the block header structure of traditional blockchain, including connecting blocks The hash value PrevHash of the previous block in the chain, the block height BlockHeight that records the position of the current block on the chain, the timestamp Timestamp that records the block generation time, the mining difficulty Bits and the random number Nonce, while the original block header Merkle Root Merkle Root is replaced by BKM-tree Root, the root node of BKM-tree constructed in step 2, but is also generated based on the hash value of transactions in the block to ensure that transactions in the block cannot be tampered with; the CS-B body contains For all transactions in the block, classify attribute types for these transactions and add semantics;

Step 1.2: Define T _x as a transaction classified by attributes and given semantics in CSBM, T _x ={Key=v ₁ ,Time={time=v ₂ },Coordinate={longitude=v ₃ ,latitude=v ₄ } ,Keywords={kw ₁ ,kw ₂ ,...,kw _n },Others={oth ₁ ,oth ₂ ,...,oth _n }}, Key is the primary key of the transaction, by calculating the hash of the transaction The value is obtained, which is represented by a hexadecimal string with a length of 64; Time is the time attribute type, time is the timestamp of the transaction, Coordinate is the space attribute type, longitude is the longitude of the transaction, and latitude is the latitude of the transaction. v _i is the attribute value, Keywords is the keyword attribute type, kw _i is the attribute name of a certain keyword of the transaction; Others is the other attribute type, oth _i is the name of other attributes of the transaction; for different applications and transaction types , set different attribute names;

Step 2: Build a two-level index structure on the chain based on the B ² F-BKM structure; the structure consists of two parts, namely the inter-block B ² F-tree and the intra-block BKM-tree, which completes all blocks and transactions. index;

Step 2.1: Build a BKM tree in each block to replace the original Merkle tree; in the process of building a new block, extract all transaction space attribute type values and keyword attribute type values in the block; imitate the KD tree structure, according to the transaction The spatial attribute data is divided into regions until each region contains only one transaction, transactions are only stored in leaf nodes, and non-leaf nodes store information related to data division; then add bloom filter fields to the KD tree, according to the key of the transaction. The word attribute data constructs a Bloom filter, and the Bloom filter of a non-leaf node contains all the keywords of its child nodes; finally, according to the construction method of the Merkle tree, the hash value of each node is calculated from the bottom up to form a BKM-tree, And record the root node in the block header;

Step 2.2: Construct an inter-block B ² F-tree based on the block number, block creation time, and keywords in the block; obtain new block data, including the block number block-id of the new block, block creation time timestamp, The bloom filter bloomfilter-bf composed of keywords in the block, the address offset of the block in the blockchain file; construct a B ² F-tree node, and the format of the key in the leaf node is "block-id, timestamp, bf", where block-id comes from the block height BlockHeight located in the block header, timestamp comes from the timestamp Timestamp located in the block header, bf is the same as the Bloom filter in the BKM-tree root node in the block; the pointer record in the leaf node The address offset of the block in the blockchain file; the bloom filter of the non-leaf node contains all the keywords of its subtree, which is used to filter the pointer and determine whether the node pointed to by the pointer contains the corresponding keyword , update the B ² F-tree according to the block-id according to the B+ tree node insertion method;

Step 2.3: The construction of the two-level index structure B ² F-BKM structure on the chain is completed;

Step 3: Design an on-chain and off-chain spatiotemporal keyword query method; obtain query conditions, perform on-chain conditions query according to the two-level index structure B ² F-BKM structure on the chain, namely B ² F-tree and BKM-tree, and extract The connection attribute value of the queried transaction, according to which the on-chain and off-chain connection query is performed and the final result is returned;

Step 3.1: Define the space-time keyword query in the hybrid storage blockchain environment is composed of quaternions, Q=[[T _s ,T _e ],S,W,join-attr], where [T _s ,T _e ] is the time range of the user query, S is the spatial range of the user query, W is the keyword data set of the user query, including several keywords, join-attr is the connection condition in the form of an equation, and the left side of the equation is the link that needs to be connected on the chain The keyword attribute name, the right side of the equation is the connection keyword attribute name in the off-chain database table;

Step 3.2: Traverse the B ² F-tree according to the time and keyword conditions in the query Q; start from the root node based on the "timestamp, bf" part of the key in the B ² F-tree node, select the branch pointer according to the time condition, and then According to the Bloom filter, determine whether the node pointed to by the pointer contains the query keyword W, if it does, continue to traverse, otherwise terminate; according to this method until the leaf node, find out the query keyword W and the time is greater than or equal to 'T _s ', less than or equal to 'T s ' All keys equal to 'T _e ' and get the position of these blocks by pointer;

Step 3.3: After querying the specific block, according to the space and keyword conditions in the query Q, traverse the BKM-tree located in the block in turn; select the child nodes from the root node from the top to the bottom according to the KD-tree search method , and judge whether the Bloom filter of the child node contains all the query keywords, if it does, traverse the child node, if not, give up the branch; follow this method until the leaf node, that is, find the space and the key within the block that satisfy All transactions of the word condition; obtain the key data in the leaf node, that is, the hash value Key of the transaction, and then obtain the original data according to the key data, if the time condition is met, add the data to the onchain result set Onchain-RS;

Step 3.4: According to the connection condition join-attr in the query Q, extract the value of the connection attribute of the original data in Onchain-RS, query the data in the off-chain database table according to the value, and add the off-chain result set Offchain-RS; read After taking the data in Offchain-RS and Onchain-RS to connect on-chain and off-chain data through the connection attribute, the result is stored in the result set ResultSet;

Step 3.5: Return to ResultSet, terminate the current calculation and wait for the next call;

Beneficial technical effects of the present invention:

The present invention considers that when the blockchain based on the on-chain and off-chain hybrid storage mode is used to query the space-time keyword, the query efficiency of the existing method is low, and a lot of unnecessary communication overhead will be generated, which will affect the user experience; The purpose of the invention is to propose a spatiotemporal keyword query method in a hybrid storage blockchain environment, which improves query efficiency and reduces communication overhead compared with traditional query methods. The specific beneficial technical effects are:

1. The method provided by the present invention is based on the block chain model CSBM that is classified by attributes and given semantics to deal with the problem of space-time keyword query in the mixed storage blockchain environment, and can realize efficient space-time key in the mixed storage block chain environment. word query.

2. The spatiotemporal keyword query method of the present invention builds a two-level index structure on the chain based on the B ² F-BKM structure, and the structure consists of two parts, namely the inter-block B ² F-tree and the intra-block BKM-tree. For the index of all blocks and transactions, this structure can quickly locate the block position and find the query transaction, effectively improving the query efficiency.

3. The hybrid storage mode-oriented blockchain connection query method provided by the present invention performs on-chain conditional query by traversing the B ² F-BKM structure, and then performs on-chain and off-chain data connection query according to the connection attribute value. Compared with traditional query This method avoids unnecessary communication overhead caused by reading all the data in the off-chain connection table and provides users with a better experience.

Description of drawings

1 is a schematic diagram of a supply chain scenario in a spatiotemporal keyword query method in a hybrid storage blockchain environment according to an embodiment of the present invention; the left side is the blockchain model CSBM that is classified by attributes and given semantics on the chain, and the right side is the off-chain blockchain model CSBM database;

2 is a schematic structural diagram of a block CS-B classified by attributes and given semantics in a spatiotemporal keyword query method in a hybrid storage blockchain environment according to an embodiment of the present invention;

3 is a schematic diagram of the overall structure of the B2F-BKM structure of the ^two -level index structure on the chain in the spatiotemporal keyword query method in the hybrid storage blockchain environment according to the embodiment of the present invention;

4 is a flow chart of constructing a two-level index structure B ² F-BKM structure on the chain in the spatiotemporal keyword query method in the hybrid storage blockchain environment according to the embodiment of the present invention;

5 is a flowchart of a method for querying spatiotemporal keywords on and off the chain in a method for querying spatiotemporal keywords in a hybrid storage blockchain environment according to an embodiment of the present invention;

FIG. 6 is a query process diagram of an example of a spatiotemporal keyword query method on-chain and off-chain in a spatiotemporal keyword query method in a hybrid storage blockchain environment according to an embodiment of the present invention.

Detailed ways

The present invention will be further described below in conjunction with the accompanying drawings and embodiments;

The present invention proposes a spatiotemporal keyword query method in a hybrid storage blockchain environment, aiming at the problem of space-time keyword query in the mixed storage blockchain on the chain and off the chain. First of all, for the weak query semantics, a Categorical and Semantic Blockchain Model (CSBM) that is classified by attributes and given semantics is constructed, which divides the attribute types and adds semantics for the transactions in the block. Secondly, in view of the low query efficiency, a two-level index structure on the chain based on the B ² F-BKM structure is constructed, which can index all blocks and transactions, effectively improving the query efficiency. Finally, in view of the large communication overhead, a space-time keyword query method on the chain and off the chain is designed. The conditional query on the chain is performed by traversing the B ² F-BKM structure, and then the data connection query on the chain and off the chain is performed according to the connection attribute value. Compared with the traditional The query method reduces unnecessary communication overhead.

A spatiotemporal keyword query method in a hybrid storage blockchain environment, the specific steps are as follows:

Step 1: Build a Categorical and Semantic Blockchain Model (CSBM) classified by attributes and given semantics. The structure is shown in the left part of Figure 1.

Step 1.1: CSBM contains multiple blocks (CS-B) classified by attributes and given semantics, CSBM=CS-B ₁ +CS-B ₂ +...+CS-B _n , each CS-B contains a block The block header (B head) and the block body (CS-Bbody), which are classified by attributes and given semantics, are structured as shown in Figure 2. CS-B=B head+CS-B body. The structure of B head is basically the same as that of the traditional blockchain, including the hash value PrevHash of the previous block that connects the block into the chain, the Block Height that records the position of the current block on the chain, and the block height that records the block generation time. Timestamp, mining difficulty Bits and random number Nonce, and the Merkle Root (Merkle Root) in the original block header is replaced by the BKM-tree Root (BKM-tree Root) constructed in step 2 of the present invention, but also based on block The hash value generation of the inner transaction is used to ensure that the transaction within the block cannot be tampered with. The CS-B body contains all transactions in the block, divides attribute types for these transactions and adds semantics;

Step 1.2: Define T _x as a transaction classified by attributes and given semantics in CSBM, T _x ={Key=v ₁ ,Time={time=v ₂ },Coordinate={longitude=v ₃ ,latitude=v ₄ } ,Keywords={kw ₁ ,kw ₂ ,...,kw _n },Others={oth ₁ ,oth ₂ ,...,oth _n }}, Key is the primary key of the transaction, by calculating the hash of the transaction The value is obtained, which is represented by a hexadecimal string with a length of 64; Time is the time attribute type, time is the timestamp of the transaction, Coordinate is the space attribute type, longitude is the longitude of the transaction, and latitude is the latitude of the transaction. v _i is the attribute value, Keywords is the keyword attribute type, kw _i is the attribute name of a certain keyword of the transaction; Others is the other attribute type, and oth _i is the name of the other attribute of the transaction. Set different attribute names for different applications and transaction types;

In this example, the transaction Keywords={operate,goods,factory} for recording production information, and the transaction Keywords={operate,goods,factory,driver} for recording logistics information. Taking transaction T _j+1 in Figure 2 as an example, the primary key of the transaction is '8C...9B', and a piece of logistics information is recorded: 'At time t ₂ , the driver Tom transports goods ₂ produced by factory ₂ to (134.3 °E, 42.5°N) position';

Step 2: Build an on-chain two-level index structure based on the B ² F-BKM structure, as shown in Figure 3. The structure consists of two parts, namely the inter-block B ² F-tree and the intra-block BKM-tree, which completes the indexing of all blocks and transactions. The construction process is shown in Figure 4. The specific process is as follows:

Step 2.1: Build a BKM-tree within each block to replace the original Merkle tree. In the process of constructing a new block, all transaction space attribute type values and keyword attribute type values in the block are extracted. Imitating the KD-tree structure, divide the area according to the spatial attribute data of the transaction, until each area contains only one transaction, the transaction is only stored in the leaf nodes, and the non-leaf nodes store the information related to the data division; then add a layout to the KD-tree In the Bloom filter field, a Bloom filter is constructed according to the keyword attribute data of the transaction. The Bloom filter of a non-leaf node contains all the keywords of its child nodes; finally, each node is calculated from the bottom up according to the construction method of the Merkle tree. The hash value of the BKM-tree is formed, and the root node is recorded in the block header. The structure is shown in Figure 3;

In this example, as shown on the right side of Figure 3, an intra-block BKM-tree is constructed for all transactions in block n. For example, the root node contains the hash value of the node, space division related information and Bloom filter, Bloom filter BF ₇ contains keyword data for all transactions in block n. Leaf node P _j+1 records the space information (134.3°E, 42.5°N), keyword information (transport, factory ₂ , goods ₂ , Tom) and Key value (8C...9B) of transaction T _j+1 , The hash value of this node is calculated by concatenating the hashes of the above information.

Step 2.2: Construct an inter-block B ² F-tree based on block number, block creation time, and intra-block key. Obtain the new block data, including the block ID of the new block (block-id), the block creation time (timestamp), the bloom filter (bloomfilter-bf) composed of keywords in the block, and the block located in the blockchain file The address offset in . Construct a B ² F-tree node. The format of the key in the leaf node is "block-id, timestamp, bf", where block-id comes from the block height (Block Height) located in the block header, and timestamp comes from the timestamp located in the block header. (Timestamp), bf is the same as the Bloom filter in the intra-block BKM-tree root node. The pointer in the leaf node records the address offset of the block in the blockchain file. The Bloom filter of a non-leaf node contains all the keywords of its subtree, which is used to filter the pointer, determine whether the node pointed to by the pointer contains the corresponding keyword, and update B ² F according to the block-id according to the B+ tree node insertion method - tree, the structure is shown in Figure 3;

In this example, Fig. 3 shows the B ² F-tree constructed for all blocks. For example, the Bloom filter BF ₅ of the root node includes all of its subtree Bloom filters BF ₁ and BF ₂ Keyword, leaf node ((1,ts ₁ ,bf ₁ ),p ₁ ) indicates that the first block is created at ts ₁ time, the key data of transactions in the block constitutes Bloom filter bf ₁ , p ₁ points to this block location.

Step 2.3: The construction of the two-level index structure B ² F-BKM structure on the chain is completed;

Step 3: Design on-chain and off-chain spatiotemporal keyword query methods. Obtain query conditions, perform on-chain conditional query according to the two-level index structure B ² F-BKM structure on the chain, namely B ² F-tree and BKM-tree, and extract the connection attribute value of the queried transaction, and perform chaining according to this value. Connect the query on and off the chain and return the final result. The process is shown in Figure 5, and the instance query process is shown in Figure 6. The specific process is as follows:

Step 3.1: Define the space-time keyword query in the hybrid storage blockchain environment is composed of quaternions, Q=[[T _s ,T _e ],S,W,join-attr], where [T _s ,T _e ] is the time range of the user query, S is the spatial range of the user query, W is the keyword data set of the user query, including several keywords, join-attr is the connection condition in the form of an equation, and the left side of the equation is the link that needs to be connected on the chain The keyword attribute name, the right side of the equation is the connection keyword attribute name in the off-chain database table;

In this example, query Q=[[t ₇ , t ₉ ], [S], [goods=goods ₂ , operate=transport], driver=Driver.dname].

Step 3.2: Traverse the B ² F-tree according to the time and keyword conditions in query Q. Based on the "timestamp, bf" part of the key in the B ² F-tree node, start from the root node, select the branch pointer according to the time condition, and then judge whether the node pointed to by the pointer contains the query keyword W according to the Bloom filter, and if it does, continue Traverse, otherwise terminate; according to this method until the leaf node, find out all the keys that contain the query keyword W and the time is greater than or equal to 'T _s ', less than or equal to 'T _e ', and obtain the location of these blocks through the pointer;

In this example, by traversing the B ² F-tree, it is obtained that only the block with block number 9 meets the query time condition and contains all query keywords, and the location information of the block is obtained.

Step 3.3: After querying a specific block, according to the space and keyword conditions in the query Q, traverse the BKM-tree located in the block in turn. According to the KD-tree search method, the child nodes are selected from the root node from the top to the bottom, and it is judged whether the Bloom filter of the child node contains all the query keywords. this branch. Follow this method until the leaf node, i.e. find all transactions within the block that satisfy the space and key conditions. Obtain the key data in the leaf node, that is, the hash value (Key) of the transaction, and then obtain the original data according to the key data. If the time condition is met, add the data to the on-chain result set (Onchain-RS);

In this example, by traversing the BKM-tree in block 9, it is obtained that nodes P _j and P _j+1 meet the space and keyword query conditions, and then the hash value Key stored in nodes P _j and P _j+1 is read. _j and Key _j+1 , find the original data T _j and T _j+1 and the original data meets the time condition, and add the original data of transaction T _j and T _j+1 to Onchain-RS.

Step 3.4: According to the connection condition join-attr in the query Q, extract the value of the connection attribute of the original data in Onchain-RS, query the data in the off-chain database table according to the value, and add the off-chain result set (Offchain-RS) . After reading the data in Offchain-RS and Onchain-RS and connecting the data on the chain and off-chain through the connection attribute, the result is stored in the result set ResultSet; in this example, the values of the connection attribute driver Tom and James are obtained according to the original data, and the values of the connection attribute driver are obtained from the original data. The data about the drivers Tom and James is found in the off-chain Driver table, and then the on-chain data is connected with the off-chain data through the connection attribute value, and the results R1 and R2 are obtained, which are stored in the result set ResultSet.

Step 3.5: Return to ResultSet, terminate the current computation and wait for the next call.

Claims (7)

1. A time-space keyword query method under a mixed storage block chain environment is characterized by comprising the following steps:

step 1: constructing a block chain model (CSBM) which is classified according to attributes and is endowed with semantics;

step 2: construction based on B ² An on-chain two-level index structure of the F-BKM structure; theThe structure is composed of two parts, namely a block B ² F-tree and intra-block BKM-tree, completing indexing of all blocks and transactions;

and step 3: designing a chain uplink and downlink time-space keyword query method; obtaining query conditions according to two-stage index structure B on the chain ² F-BKM structure, i.e. B ² And performing on-chain condition query on the F-tree and the BKM-tree, extracting a connection attribute value of the queried transaction, performing on-chain and off-chain connection query according to the value, and returning a final result.

2. The method for querying the space-time keyword under the mixed storage block chain environment according to claim 1, wherein the step 1 specifically comprises:

step 1.1: the CSBM comprises multiple blocks CS-B classified by attribute and assigned with semantics ₁ +CS-B ₂ +···+CS-B _n Each CS-B comprises a block head B head and a block body CS-B body which is classified according to attributes and is endowed with semantics; wherein CS-B ═ B head + CS-B body; the B head has basically the same structure as the block head of the traditional block chain; the Merkle Root in the original block head is replaced by the Root node BKM-tree Root of the BKM-tree constructed in the step 2, but is generated based on the hash value of the transaction in the block and is used for ensuring that the transaction in the block cannot be tampered; CS-B body contains all the transactions in the block, and attribute types are divided for the transactions and semantics are added;

step 1.2: definition of T _x For a transaction, T, classified by attribute and assigned semantics in a CSBM _x ＝{Key＝v ₁ ,Time＝{time＝v ₂ },Coordinate＝{longitude＝v ₃ ,latitude＝v ₄ },Keywords＝{kw ₁ ,kw ₂ ,···,kw _n },Others＝{oth ₁ ,oth ₂ ,···,oth _n Key is a primary Key of the transaction, is obtained by calculating a hash value of the transaction, and is represented by a 16-system character string with the length of 64; time is a Time attribute type, Time is a timestamp of the transaction, Coordinate is a space attribute type, longitude is longitude of the transaction, latitude is latitude of the transaction, v _i Is attribute value, and Keywords is offType of key attribute, kw _i An attribute name of a key word for the transaction; other attribute types, oth _i Name of other attributes for the transaction; different attribute names are set for different application occasions and transaction types.

3. The method according to claim 2, wherein the B-head comprises a previous Block hash value PrevHash for connecting blocks into a chain, a Block Height for recording the position of a current Block on the chain, a Timestamp for recording the Block generation time, and mining difficulty Bits and Nonce.

4. The method for querying a space-time keyword under a mixed storage block chain environment according to claim 1, wherein the step 2 specifically comprises:

step 2.1: constructing a BKM-tree in each block to replace the original Merkle tree; in the process of constructing a new block, extracting all transaction space attribute type values and keyword attribute type values in the block; simulating a KD-tree structure, and dividing regions according to spatial attribute data of the transactions until each region only contains one transaction, the transactions are only stored in leaf nodes, and non-leaf nodes store information related to data division; then adding a bloom filter field for the KD-tree, constructing a bloom filter according to the key word attribute data of the transaction, wherein the bloom filter of the non-leaf node comprises all the key words of the child nodes; finally, calculating the hash value of each node from bottom to top according to the construction mode of the Merkle tree to form a BKM-tree, and recording the root node in the block head;

step 2.2: constructing an inter-block B based on block number, block creation time, and intra-block key ² F-tree; obtaining new block data, constructing B ² F-tree nodes, wherein the format of keys in leaf nodes is 'Block-id, Timestamp and bf', the Block-id is from the Block Height at the Block head, the Timestamp is from the Timestamp at the Block head, and the bf is the same as a bloom filter in a BKM-tree root node in a Block; leaf of Chinese characterThe pointer in the node records the address offset of the block in the blockchain file; the bloom filter of the non-leaf node comprises all keywords of the subtree thereof, and is used for filtering the pointer, judging whether the node pointed by the pointer contains the corresponding keyword, and updating the B according to the block-id and in a B + tree node insertion mode ² F-tree;

step 2.3: two-stage index structure B on chain ² And completing the construction of the F-BKM structure.

5. The method according to claim 4, wherein the new block data obtained comprises a block number block-id of the new block, a block creation time timestamp, a bloom filter block-bf composed of intra-block keywords, and an address offset of the block in the block chain file.

6. The method for querying a space-time keyword under a mixed storage block chain environment according to claim 1, wherein step 3 specifically comprises:

step 3.1: defining a space-time key query in a mixed-memory blockchain environment consists of a quadruplet, Q [ [ T [ ] _s ,T _e ],S,W,join-attr]；

Step 3.2: traversing B according to time and keyword conditions in query Q ² F-tree; based on B ² Starting from a root node, selecting a branch pointer according to a time condition, judging whether a node pointed by the pointer contains a query keyword W according to a bloom filter, continuously traversing if the node contains the query keyword W, and stopping if the node does not contain the query keyword W; by the method, the leaf nodes are searched for a search keyword W with time greater than or equal to' T _s 'less than or equal to' T _e All keys of' and get the location of these blocks through pointers;

step 3.3: after a specific block is queried, sequentially traversing the BKM-trees in the block according to the space and keyword conditions in the query Q; selecting child nodes from top to bottom from a root node according to a searching mode of a KD-tree, judging whether bloom filters of the child nodes contain all query keywords, traversing the child nodes if the bloom filters of the child nodes contain all the query keywords, and abandoning the branch if the bloom filters of the child nodes do not contain all the query keywords; according to the method, all the affairs which meet the conditions of space and key words in the block are found until the leaf nodes; obtaining Key data in leaf nodes, namely a hash value Key of a transaction, obtaining original data according to the Key data, and adding the data into a result set Onchain-RS on a chain if a time condition is met;

step 3.4: extracting the value of the connection attribute of the original data in the Onchain-RS according to the connection condition join-attr in the query Q, querying the data in a database table under the chain according to the value, and adding a result set Offcain-RS under the chain; reading data in the Offchain-RS and the Onchhain-RS, performing chain uplink and downlink data connection through connection attributes, and storing results into a result set;

step 3.5: returning to ResultSet, terminating the current calculation and waiting for the next call.

7. The method according to claim 6, wherein [ T ] in step 3.1 _s ,T _e ]The method comprises the steps of querying a time range for a user, querying a space range for the user, querying a keyword data set for the user, wherein the keyword data set comprises a plurality of keywords, join-attr is a connection condition in an equation form, the left side of the equation is a keyword attribute name needing to be connected on a chain, and the right side of the equation is a connection keyword attribute name in a database table under the chain.

Images