CN107341193A - Mobile object querying method in road network - Google Patents
Mobile object querying method in road network Download PDFInfo
- Publication number
- CN107341193A CN107341193A CN201710446636.2A CN201710446636A CN107341193A CN 107341193 A CN107341193 A CN 107341193A CN 201710446636 A CN201710446636 A CN 201710446636A CN 107341193 A CN107341193 A CN 107341193A
- Authority
- CN
- China
- Prior art keywords
- mobile object
- data
- road network
- inquiry
- querying method
- 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.)
- Granted
Links
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
- G06F16/90—Details of database functions independent of the retrieved data types
- G06F16/95—Retrieval from the web
- G06F16/953—Querying, e.g. by the use of web search engines
- G06F16/9537—Spatial or temporal dependent retrieval, e.g. spatiotemporal queries
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
- G06F16/20—Information retrieval; Database structures therefor; File system structures therefor of structured data, e.g. relational data
- G06F16/29—Geographical information databases
Landscapes
- Engineering & Computer Science (AREA)
- Databases & Information Systems (AREA)
- Theoretical Computer Science (AREA)
- Data Mining & Analysis (AREA)
- Physics & Mathematics (AREA)
- General Engineering & Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Remote Sensing (AREA)
- Mobile Radio Communication Systems (AREA)
Abstract
The embodiment of the present invention provides mobile object querying method in a kind of road network, belongs to high-performance computing sector and database field.Mobile object querying method in the road network, for server end, and including：Obtain mobile object and update the data the inquiry data inputted with user；Using multi-core CPU, the node based on mobile object place in road network updates the data to acquired mobile object carries out aggregat ion pheromones and based on the side at inquiry data place in road network come to acquired inquiry data progress aggregat ion pheromones；And the data after aggregat ion pheromones are put into GPU and carry out the calculating based on arest neighbors KNN algorithms, to obtain Query Result.The embodiment of the present invention proposes mobile object querying method in a kind of high-throughput road network based under new hardware environment, it has given full play to the characteristics of big internal memory, multi-core CPU, GPU, so as to improve the query processing efficiency of mobile object, it can more meet user's query demand based on location-based service under big data.
Description
Technical field
The present invention relates to high-performance computing sector and database field, is inquired about more particularly to mobile object in a kind of road network
Method.
Background technology
With the continuous development of the widely available and location-based service of mobile device, the query processing for mobile object turns into
The focus of research.Typical inquiry, such as：Search the supermarket or gas station nearest apart from user；The taxi nearest apart from user
Car etc..In daily life, all objects are all based on road network and moved.For any object X in road network and
Y, the road network distance d (x, y) between object is than Euclidean distance | | x-y | | it can preferably reflect the actual range between object.For example,
In two-way lane, vehicle q to positioned at opposite side gas station p road network distance d (q, p) be far longer than between the two it is European away from
From | | q-p | |.Therefore, the road network distance d (q, p) between two objects is than Euclidean distance | | q-p | | it is more meaningful.
But present inventor has found during the present invention is realized, mobile object query operator in existing road network
Method has the disadvantage that：With being continuously increased for number of users, application scenarios are enriched constantly, and conventional KNN algorithms can not be very
The good emerging application of reply.
For example, for typical application of calling a taxi, vehicle can produce a vehicle location update data stream in the process of running.
Meanwhile the request of calling a taxi that a large number of users is initiated can form an inquiry data flow.System first has to carry out vehicle position information
Real-time update, to ensure the validity of Query Result, while the inquiry request of all users of real-time response is needed again, to ensure to take
Business quality.For this kind of application scenarios, when KNN algorithms are when by the way of one per treatment inquiry (one by one), greatly
Measuring user needs to be lined up to wait inquiry response, and the queue waiting time of user will have a strong impact on service quality.
Therefore, mobile object search algorithm can not tackle emerging demand well, it is necessary to find in existing road network
New mobile object query scheme, to meet the query demand of location-based service under big data.
The content of the invention
The purpose of the embodiment of the present invention is to provide mobile object querying method in a kind of road network, the mobile object querying method
Meet the mobile object query scheme of the query demand of location-based service under big data for realizing.
To achieve these goals, the embodiment of the present invention provides mobile object querying method in a kind of road network, for servicing
Device end, and mobile object querying method includes in the road network：Obtain mobile object and update the data the inquiry number inputted with user
According to；Using multi-core CPU, based on mobile object in road network where node acquired mobile object updated the data gathered
Collection indexes and carries out aggregat ion pheromones to acquired inquiry data based on the side at inquiry data place in road network；And will
Data after aggregat ion pheromones, which are put into GPU, carries out the calculating based on arest neighbors KNN algorithms, to obtain Query Result.
Alternatively, the acquisition mobile object is updated the data and included with the inquiry data of user's input：Periodically collection moves
Dynamic object is updated the data, and the mobile object is updated the data including mobile object identifier and mobile object coordinate；Real-time reception is used
The inquiry data of family input；And mobile object described in use buffer updates the data and the inquiry data, and according to
The Thread Count that needs to use divides the data in the buffer, wherein one piece of mobile object of each thread process update the data or
Inquire about data.
Alternatively, the mobile object is updated the data and delayed described in the inquiry data Cun Chudao by the way of snapshot
In storage.
Alternatively, based on mobile object in road network where node acquired mobile object updated the data gathered
Collection index includes：Calculate mobile object with its node at two end points in the paths distance；By with the section at end points
The mobile object for the mobile object that point distance is no more than its path length half, which updates the data, is gathered in the node；And will aggregation
The mobile object of completion, which updates the data, to be put into an Object table structure.
Alternatively, aggregat ion pheromones bag is carried out to acquired inquiry data based on the side at inquiry data place in road network
Include：All inquiry data are assembled according to place path, and the inquiry data on same paths are put into an inquiry
The neighboring storage locations of table structure.
Alternatively, the data after aggregat ion pheromones are put into GPU and carry out the calculating based on KNN algorithms and include：Based on aggregation
Data after index, the KNN result sets of two end points in path where calculating any one node in path；Based on aggregation rope
Data after drawing, calculate the mobile object set in path；And from the KNN result sets and the conjunction of the mobile object set
Integrated query goes out the KNN result sets of selected node.
Alternatively, the aggregat ion pheromones carried out using multi-core CPU are Grid Indexes, and the Grid Index includes：By each shifting
Dynamic object, which updates the data, is indexed to the corresponding unit of the two-dimentional theorem in Euclid space coordinate updated the data in grid with the mobile object
In lattice；And the list where each inquiry data directory is updated the data into grid with the mobile object of the inquiry data match
In first lattice.
Alternatively, the Grid Index also includes：When the amount for the data assembled in the cell in grid is beyond setting
During threshold value, the cell is divided at least two subelement lattice at GPU ends, and each data respective stored is single to corresponding son
First lattice.
Alternatively, mobile object querying method also includes in the road network：The Query Result is distributed to user, and deleted
Caused intermediate result during except being calculated.
On the other hand, the present invention provides a kind of calculating readable storage medium storing program for executing, is stored with the computer-readable recording medium
Computer instruction, the computer instruction are used to cause the computer to perform the above-mentioned method of the application.
Pass through above-mentioned technical proposal, beneficial effect possessed by the embodiment of the present invention are：The embodiment of the present invention proposes one
For kind based on mobile object querying method in the high-throughput road network under new hardware environment, it has given full play to big internal memory, multinuclear
The characteristics of CPU, GPU, so as to improve the query processing efficiency of mobile object, it can more meet under big data based on location-based service
User's query demand.
The further feature and advantage of the embodiment of the present invention will be described in detail in subsequent specific embodiment part.
Brief description of the drawings
Accompanying drawing is that the embodiment of the present invention is further understood for providing, and a part for constitution instruction, with
The embodiment in face is used to explain the embodiment of the present invention together, but does not form the limitation to the embodiment of the present invention.Attached
In figure：
Fig. 1 is the schematic flow sheet of mobile object querying method in road network according to embodiments of the present invention；
Fig. 2 is the schematic flow sheet that acquisition mobile object according to embodiments of the present invention updated the data and inquired about data；
Fig. 3 is the schematic flow sheet according to embodiments of the present invention for being updated the data to mobile object and carrying out aggregat ion pheromones；
Fig. 4 is the schematic diagram for the example that aggregat ion pheromones are carried out to mobile object；
Fig. 5 is the schematic flow sheet calculated by GPU the data after aggregat ion pheromones；
Fig. 6 is the configuration diagram of GPGPU model according to embodiments of the present invention；And
Fig. 7 is the signal according to embodiments of the present invention that object KNN inquiries are moved based on big internal memory and GPGPU model
Figure.
Embodiment
The embodiment of the embodiment of the present invention is described in detail below in conjunction with accompanying drawing.It should be appreciated that this
The embodiment of place description is merely to illustrate and explain the present invention embodiment, is not intended to limit the invention embodiment.
Present inventor during realizing the object of the invention find for major applications, single query
As long as the response time reaches second level can meet demand, it is not necessary to deliberately pursue the response time of single query, such as
Typical application of calling a taxi, it can just meet to require as long as most of user can meet with a response in seconds, for inquiry faster
Response time, the service experience of user do not have significant change.Therefore, in order to ensure service quality, it is necessary within the unit interval to the greatest extent
The inquiry of response user more than possible, the i.e. handling capacity of system become the key for meeting service-seeking demand in position under big data
Factor.
Based on this thinking, the embodiment of the present invention proposes mobile object querying method in a kind of road network.Specifically introducing
Before this method, first road network involved in the embodiment of the present invention is illustrated.
In the embodiment of the present invention, road network is represented with one without phase weighted graph G=(V, E), and wherein V is the set of node,
The junction in road network is represented, E is the set without phase side, represents the line segment between two nodes in road network, simultaneouslyFor any road network G (E, V), each edge can be expressed as e (v1, v2), and wherein v1, v2 are Liang Ge UNICOMs
Node, v1 are start node, and v2 is finish node, and the weight of each edge is nonnegative value.It should be noted that above-mentioned parameter is used for body
The concept of existing road network, and in practice, can be as needed, the implication of parameter is represented using different characters, such as can also
Node is represented with n1, n2.
Fig. 1 shows the schematic flow sheet of mobile object querying method in the road network of the embodiment of the present invention, the mobile object
Querying method is applied in server end, may comprise steps of：
Step S100, obtain mobile object and update the data the inquiry data inputted with user.
Wherein, mobile object is updated the data including the information such as mobile object identifier (ID) and mobile object coordinate, such as
Embody the information of the change of the real time position for the vehicle being related in typical application of calling a taxi；Inquiry data are, for example, that typical case calls a taxi application
In the inquiry of the user that is related to the taxi in neighbouring two kilometers.
Preferably, as shown in Fig. 2 step S100 may comprise steps of：
Step S101, periodically gather mobile object and update the data.
Step S102, the inquiry data of real-time reception user input.
Step S103, the mobile object is cached using buffer (hereinafter referred to as Buffer) and updated the data and the inquiry
Data, and the data in the Thread Count division Buffer used as required.
Wherein, settable each one piece of mobile object of thread process is updated the data or one piece is inquired about data.
It is further preferable that the embodiment of the present invention is updated the data the mobile object and the inquiry by the way of snapshot
In data Cun Chudao Buffer, follow-up calculating can also be based on snapshot and carry out, such as the inquiry base that user submits at the T1 moment
Calculated in the position snapshot formed at the T0 moment, wherein T0≤T1, T1-T0 ＜ Δ T, Δ T are between a regular time
Every Query Result is effective in T0+ Δ T time sections.
Wherein, the principle and specific method on snapshot, refers to existing pertinent literature, and the embodiment of the present invention is herein no longer
Repeat.
Step S200, using multi-core CPU, the node based on mobile object place in road network is to acquired mobile object
Update the data and carry out aggregat ion pheromones and based on the side at inquiry data place in road network come to acquired inquiry data progress
Aggregat ion pheromones.
Preferably, as shown in figure 3, the node based on mobile object place in road network updates to acquired mobile object
Data carry out aggregat ion pheromones and may comprise steps of：
Step S201, calculate mobile object with its node at two end points in the paths distance.
Step S202, the mobile object of the mobile object of its path length half will be no more than with the nodal distance at end points
Update the data and be gathered in the node.
Step S203, the mobile object for assembling completion is updated the data and is put into an Object table structure.
For step S201- step S203, illustrate below by example.Fig. 4 is shown to be assembled to mobile object
The example of index, wherein, n1-n8 represent node, p1-p5 represent mobile object, as we know from the figure p2 and p3 to node n4 away from
From the half of the length no more than corresponding path n3n4 and n5n4, so as to which p2 and p3 to be gathered in the adjacent bit of Object table structure
Put.Similarly, positions of the p4 and p5 in Object table structure is also to be determined using similar approach.
Being preferably based on that side of the inquiry data where in road network to carry out aggregat ion pheromones to acquired inquiry data can
With including：All inquiry data are assembled according to place path, and the inquiry data on same paths are put into one
The neighboring storage locations of inquiry table structure.In this way, understand, all inquiry numbers different from the mode that mobile object updates the data processing
According to being to be assembled by path, storage location of the inquiry data in inquiry table structure on same paths is close.
By operation above, mobile object is updated the data and inquired about data and is carried out according to the side at place node and place
Aggregat ion pheromones, it is respectively stored in Object table and inquiry table, so as to using more nuclear properties of multi-core CPU, pass through batch processing
Mode, single treatment multiple queries.
In addition, further relating to the utilization to Large Copacity memory techniques using multi-core CPU, i.e., by workload, (i.e. Buffer delays
The inquiry data and mobile object deposited update the data) all it is placed in multi-core CPU and is handled, so as to take full advantage of data
Spatial locality, the hit rate of cache is added, so as to be advantageous to the follow-up execution efficiency for improving related algorithm.
Further, the aggregat ion pheromones that the embodiment of the present invention is carried out using multi-core CPU are grid (Grid) indexes, specifically
Step is：Each mobile object is updated the data to the two-dimentional theorem in Euclid space coordinate for being indexed to and being updated the data in grid with the mobile object
In corresponding cell；And mobile object of each inquiry data directory into grid with the inquiry data match is updated
In cell where data.
In addition, in step 200, multi-core CPU conveys the data for building the Grid indexes completed in units of cell
To GPU ends, to perform step S300.
Step S300, the data after aggregat ion pheromones are put into GPU and carry out the calculating based on KNN algorithms, to be inquired about
As a result.
Here, it should be noted that, in step s 200, mobile object is updated the data and inquired about data and is gathered in cell
When middle, it is possible that data twisted phenomena, i.e., excessive data, which are concentrated in one or several cells, causes data torsion
It is bent.On the other hand, can set when the amount for the data assembled in the cell in Grid exceeds given threshold, at GPU ends by the list
First lattice are divided at least two subelement lattice, and by each data respective stored to corresponding subelement lattice.In this way, cell is entered
The secondary division of row, has obtained the subelement lattice of smaller particle size, wherein secondary division is preferably what can be divided first according to cell
Same way is carried out, for example, forming 3*3 cell after dividing for the first time, the data volume in two of which cell, which exceedes, to be set
Determine threshold value, then individually continue the two cells to be divided into 3*3 cell.
In addition, the core concept of KNN algorithms is if big in the k in feature space most adjacent samples of a sample
Majority belongs to some classification, then the sample falls within this classification, and with the characteristic of sample in this classification.The present invention is implemented
The method of example substantially gives a kind of modified KNN algorithms using multi-core CPU, GPU and feature, the innovatory algorithm principle
For：For in pathIn KNN inquiry q, its Query ResultIts
Middle Rq is inquiry q KNN results, and O (ni, nj) be path ni, the mobile object in nj, and Rni is node ni KNN results, Rnj
For node nj KNN results, the KNN result sets of some node two end points KNN knots where the node in the path so as to knowable to
Fruit collects the subset with object set in path.
Accordingly, as shown in figure 5, the step of data after aggregat ion pheromones are put into the calculating carried out in GPU based on KNN algorithms
Can be as follows：
Step S301, based on the data after aggregat ion pheromones, two of path where calculating any one node in path
KNN ((k-nearest neighbor, k nearest neighbor) result sets of end points.
Wherein, the calculating on KNN result sets refers to existing pertinent literature progress, is then repeated no more at this.
Step S302, based on the data after aggregat ion pheromones, calculate the mobile object set in path.
Step S303, selected node is inquired from the KNN result sets and the intersection of the mobile object set
KNN result sets.
By aforesaid operations, Grid corresponding unit lattice are performed by GPU, so that mobile object associates with inquiry, are obtained
Corresponding Query Result.In addition, in a preferred embodiment, it is also necessary to Query Result is distributed to user, the Query Result
It is GPU result of calculation, so as to be deleted while Query Result is distributed caused by during being calculated
Intermediate result, to ensure GPU execution efficiency.
In this way, step S200 and step S300 are combined, realize and object KNN is moved based on multi-core CPU and GPU looked into
The scheme of inquiry, and multi-core CPU and GPU characteristic have fully been used, looked into for the mobile object renewal KNN to be arrived in Buffer
Ask, by way of rebuilding index, given full play to the characteristic of multi-core CPU；For the index structure built, utilize
GPU is skillful in the feature of efficient data computing, improves the performance of algorithm queries, and by multiple queries and performs, and makes full use of and looks into
It is parallel between inquiry in asking, improve the efficiency of algorithm.
Therefore, it is known that be configured with the mobile object KNN query schemes of the embodiment of the present invention based on Large Copacity internal memory, multinuclear
The inquiry framework of CPU and GPU new hardware environment, the concrete configuration of the inquiry framework can be such as：Configuring every can provide
Some 8 road servers of 12TB internal memories, maximum reachable 18 server of configuration CPU core number, and be configured to do greatly
The GPU of the concurrent operation of scale.Wherein, it is data cached by Buffer, it is responsible for performing complex logic processing and transaction management by CPU
Etc. the calculating of unsuitable data parallel, it is responsible for the Large-scale parallel computing of computation-intensive by GPU.
Here, multi-core CPU and GPU cooperation constitute GPGPU (General Purpose GPU, general purpose GPU) mould
Type.As shown in fig. 6, in GPGPU model, CPU is as main frame (Host), and GPU is as coprocessor or equipment (Device).
There may be a main frame and several equipment in a system, you can be interpreted as：By a multi-core CPU to the movement
Object updates the data rebuilds index with the inquiry data, and the index structure built is entered by several GPU
Row calculates.
In addition, multi-core CPU is that the data for building the Grid indexes completed are delivered into GPU ends in units of cell, such as
Shown in Fig. 6, GPU ends remain the data storage scheme of the cell based on Grid indexes at multi-core CPU end, and GPU is to identical list
Mobile object in first lattice updates the data and inquired about data and calculated, and, can be parallel. between unit lattice independently of each other
Perform.Wherein, the cell (2,0) such as in Fig. 6 refers to the cell that two-dimentional theorem in Euclid space coordinate is (2,0), and GPU is to the list
Data in first lattice, which carry out calculating, can obtain relevant query result.
In this GPGPU model, CPU and GPU cooperates, and Each performs its own functions, and CPU is responsible for carrying out the strong things of logicality
Processing and serial computing, GPU, which is then absorbed in, performs highly threading parallel processing task.Therefore once it is determined that program and
Row part, it is possible to consider this part evaluation work to give GPU.
It should be noted that CPU, GPU each possess separate memory address space, i.e.,：The internal memory of host side and
The video memory of equipment end.
In addition, above-mentioned steps S100- steps S300 performs by server, i.e., it can pass through only one in the embodiment of the present invention
Server manages mobile object, so as to being advantageous to simplify system.
Carry out specific introduce in the present embodiment below by example and object KNN is moved based on big internal memory and GPGPU model
The implementation detail of inquiry.
As shown in fig. 7, when two threads perform, the processing path of thread 1 firstIn inquiry q1,
Q2, the processing path of thread 2On inquiry q3.During execution, query node is needed for thread 1
N1, n2 KNN result sets, thread 2 need the KNN result sets on query path n3, n4.When inquiry continues executing with, at thread 1
ReasonInquiry q5 on path, thread 2 are handledInquiry q6 on path.For thread 1,
Because node n3 KNN result sets have obtained, index only needs query node n5 result set.Need to inquire about for thread 2
Node n5 and n6 KNN result sets, shared for node n5 in two inquiries, so only needing a thread pool can
With.In query node n5 KNN result sets, avoid the need for searching for node n2 and n4 again after node n3 is searched.Because
N3 KNN result sets have obtained, so reducing the scope of search.It is can see by Fig. 2 example in query execution mistake
Cheng Zhong, it is only necessary to seldom node is searched for regard to KNN inquiry can be completed, while in multithreading implementation procedure, many nodes
KNN result sets have obtained, it is not necessary to the calculating repeated, improve the search efficiency of algorithm.
In summary, the embodiment of the present invention proposes in a kind of high-throughput road network based under new hardware environment mobile pair
As querying method, the characteristics of it has given full play to big internal memory, multi-core CPU, GPU, so as to improve the query processing of mobile object
Efficiency, it can more meet user's query demand based on location-based service under big data.
The optional embodiment of example of the present invention, still, the embodiment of the present invention and unlimited are described in detail above in association with accompanying drawing
Detail in above-mentioned embodiment, can be to the embodiment of the present invention in the range of the technology design of the embodiment of the present invention
Technical scheme carry out a variety of simple variants, these simple variants belong to the protection domain of the embodiment of the present invention.
It is further to note that each particular technique feature described in above-mentioned embodiment, in not lance
In the case of shield, it can be combined by any suitable means.In order to avoid unnecessary repetition, the embodiment of the present invention pair
Various combinations of possible ways no longer separately illustrate.
It will be appreciated by those skilled in the art that realize that all or part of step in above-described embodiment method is to pass through
Program instructs the hardware of correlation to complete, and the program storage is in the storage medium, including some instructions are causing one
Individual (can be single-chip microcomputer, chip etc.) or processor (processor) perform the whole of each embodiment methods described of the application
Or part steps.And foregoing storage medium includes：USB flash disk, mobile hard disk, read-only storage (ROM, Read-Only
Memory), random access memory (RAM, Random Access Memory), magnetic disc or CD etc. are various can store journey
The medium of sequence code.
In addition, it can also be combined between a variety of embodiments of the embodiment of the present invention, as long as it is not
The thought of the embodiment of the present invention is run counter to, it should equally be considered as disclosure of that of the embodiment of the present invention.
Claims (9)
1. mobile object querying method in a kind of road network, for server end, it is characterised in that mobile object is looked into the road network
Inquiry method includes：
Obtain mobile object and update the data the inquiry data inputted with user；
Using multi-core CPU, based on mobile object, the node at place updates the data progress to acquired mobile object in road network
Aggregat ion pheromones and based on inquiry data in road network where side aggregat ion pheromones are carried out to acquired inquiry data；And
Data after aggregat ion pheromones are put into GPU and carry out the calculating based on arest neighbors KNN algorithms, to obtain Query Result.
2. mobile object querying method in road network according to claim 1, it is characterised in that the acquisition mobile object is more
New data and the inquiry data of user's input include：
Periodically collection mobile object is updated the data, and the mobile object is updated the data including mobile object identifier and mobile object
Coordinate；
The inquiry data of real-time reception user input；And
Updated the data using mobile object described in buffer and the inquiry data, and the Thread Count used as required is drawn
The data divided in the buffer, wherein one piece of mobile object of each thread process updates the data or inquired about data.
3. the mobile object querying method in road network according to claim 2, it is characterised in that will by the way of snapshot
The mobile object updates the data inquires about in buffer described in data Cun Chudao with described.
4. mobile object querying method in road network according to claim 1, it is characterised in that based on mobile object in road network
The node at middle place, which updates the data progress aggregat ion pheromones to acquired mobile object, to be included：
Calculate mobile object with its node at two end points in the paths distance；
The mobile object for the mobile object for being no more than its path length half with the nodal distance at end points is updated the data into aggregation
In the node；And
The mobile object for assembling completion is updated the data and is put into an Object table structure.
5. mobile object querying method in road network according to claim 1, it is characterised in that based on inquiry data in road network
The side at middle place includes to carry out aggregat ion pheromones to acquired inquiry data：
All inquiry data are assembled according to place path, and the inquiry data on same paths are put into an inquiry
The neighboring storage locations of table structure.
6. mobile object querying method in road network according to claim 1, it is characterised in that by the data after aggregat ion pheromones
It is put into GPU and carries out the calculating based on KNN algorithms and include：
Based on the data after aggregat ion pheromones, the KNN results of two end points in path where calculating any one node in path
Collection；
Based on the data after aggregat ion pheromones, the mobile object set in path is calculated；And
The KNN result sets of selected node are inquired from the KNN result sets and the intersection of the mobile object set.
7. mobile object querying method in road network according to claim 1, it is characterised in that carried out using multi-core CPU
Aggregat ion pheromones are Grid Indexes, and the Grid Index includes：
Each mobile object is updated the data to the two-dimentional theorem in Euclid space coordinate phase for being indexed to and being updated the data in grid with the mobile object
In corresponding cell；And
Cell where each inquiry data directory is updated the data into grid with the mobile object of the inquiry data match
In.
8. mobile object querying method in road network according to claim 7, it is characterised in that the Grid Index also wraps
Include：
When the amount for the data assembled in the cell in grid exceeds given threshold, the cell is divided at GPU ends
At least two subelement lattice, and by each data respective stored to corresponding subelement lattice.
9. mobile object querying method in road network as claimed in any of claims 1 to 8, it is characterised in that described
Mobile object querying method also includes in road network：
The Query Result is distributed to user, and delete calculated during caused intermediate result.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201710446636.2A CN107341193B (en) | 2017-06-14 | 2017-06-14 | Method for inquiring mobile object in road network |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201710446636.2A CN107341193B (en) | 2017-06-14 | 2017-06-14 | Method for inquiring mobile object in road network |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN107341193A true CN107341193A (en) | 2017-11-10 |
| CN107341193B CN107341193B (en) | 2021-02-02 |
Family
ID=60221583
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201710446636.2A Active CN107341193B (en) | 2017-06-14 | 2017-06-14 | Method for inquiring mobile object in road network |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN107341193B (en) |
Cited By (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN107917716A (en) * | 2018-01-02 | 2018-04-17 | 广东工业大学 | Fixed circuit air navigation aid, device, terminal and computer-readable recording medium |
| CN109408738A (en) * | 2018-09-10 | 2019-03-01 | 中南民族大学 | The querying method and system of spatial entities in a kind of transportation network |
| CN110515972A (en) * | 2019-08-27 | 2019-11-29 | 陈东升 | Database quick reference system and data query method |
Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103473248A (en) * | 2012-12-17 | 2013-12-25 | 东北大学 | Continuously-moving K adjacent query method in road network environment |
| CN104182210A (en) * | 2014-09-02 | 2014-12-03 | 重庆大学 | Multiple GPU (graphics processing unit) data parallel comparison method based on CUDA (compute unified device architecture) |
-
2017
- 2017-06-14 CN CN201710446636.2A patent/CN107341193B/en active Active
Patent Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103473248A (en) * | 2012-12-17 | 2013-12-25 | 东北大学 | Continuously-moving K adjacent query method in road network environment |
| CN104182210A (en) * | 2014-09-02 | 2014-12-03 | 重庆大学 | Multiple GPU (graphics processing unit) data parallel comparison method based on CUDA (compute unified device architecture) |
Non-Patent Citations (5)
| Title |
|---|
| C.BALKESEN: "Main-Memory Hash Joins on Modern Processor Architectures", 《IEEE TRANSACTIONS ON KNOWLEDGE AND DATA ENGINEERING》 * |
| 沈钧毅: "《数据库系统原理》", 28 February 2014, 西安交通大学出版社 * |
| 薛忠斌等: "双流模式下高吞吐量移动对象范围查询算法", 《软件学报》 * |
| 薛忠斌等: "路网中高吞吐量移动对象实时查询算法", 《计算机科学》 * |
| 韦春丹: "一种基于GPU的移动对象并行处理框架", 《计算机应用与软件》 * |
Cited By (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN107917716A (en) * | 2018-01-02 | 2018-04-17 | 广东工业大学 | Fixed circuit air navigation aid, device, terminal and computer-readable recording medium |
| CN107917716B (en) * | 2018-01-02 | 2021-07-06 | 广东工业大学 | Fixed line navigation method, device, terminal and computer readable storage medium |
| CN109408738A (en) * | 2018-09-10 | 2019-03-01 | 中南民族大学 | The querying method and system of spatial entities in a kind of transportation network |
| CN110515972A (en) * | 2019-08-27 | 2019-11-29 | 陈东升 | Database quick reference system and data query method |
Also Published As
| Publication number | Publication date |
|---|---|
| CN107341193B (en) | 2021-02-02 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN102521406B (en) | Distributed query method and system for complex task of querying massive structured data | |
| CN106055563B (en) | A kind of parallel spatial querying method and its system based on grid dividing | |
| CN104361113B (en) | A kind of OLAP query optimization method under internal memory flash memory mixing memory module | |
| CN102521405B (en) | Massive structured data storage and query methods and systems supporting high-speed loading | |
| CN107341193A (en) | Mobile object querying method in road network | |
| CN105468439A (en) | Adaptive parallel algorithm for traversing neighbors in fixed radius under CPU-GPU (Central Processing Unit-Graphic Processing Unit) heterogeneous framework | |
| US7921103B2 (en) | Adaptive materialized view selection for databases | |
| CN110852046B (en) | Block induction sequencing method and system for text suffix index | |
| Deng et al. | Parallel processing of dynamic continuous queries over streaming data flows | |
| CN113032427B (en) | Vectorization query processing method for CPU and GPU platform | |
| CN109063194A (en) | Data retrieval method and device based on space encoding | |
| CN105359142B (en) | Hash connecting method and device | |
| CN109739646A (en) | A kind of data processing method and device | |
| CN105393210A (en) | Memory unit for emulated shared memory architectures | |
| Gowanlock et al. | Distance threshold similarity searches: Efficient trajectory indexing on the GPU | |
| Nouri et al. | GPU-based parallel indexing for concurrent spatial query processing | |
| Malik et al. | Task scheduling for GPU accelerated hybrid OLAP systems with multi-core support and text-to-integer translation | |
| Huang et al. | A grid and density based fast spatial clustering algorithm | |
| CN107341126A (en) | Mobile object inquiry unit | |
| CN110618963A (en) | Heterogeneous computing hardware acceleration system and method integrating computing and storage | |
| CN105573834B (en) | A kind of higher-dimension vocabulary tree constructing method based on heterogeneous platform | |
| CN112100446B (en) | Search method, readable storage medium, and electronic device | |
| Yelmewad et al. | Near optimal solution for traveling salesman problem using GPU | |
| CN106354433A (en) | Hot spot data mining method and device of distributed memory storage system | |
| Heiler et al. | Comparing implementation variants of distributed spatial join on spark |
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 |