CN109769136A - Online Video buffer memory management method and system based on instantaneous access rate value - Google Patents
Online Video buffer memory management method and system based on instantaneous access rate value Download PDFInfo
- Publication number
- CN109769136A CN109769136A CN201811491430.2A CN201811491430A CN109769136A CN 109769136 A CN109769136 A CN 109769136A CN 201811491430 A CN201811491430 A CN 201811491430A CN 109769136 A CN109769136 A CN 109769136A
- Authority
- CN
- China
- Prior art keywords
- video block
- video
- block
- user
- cache
- 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
Abstract
The present invention provides a kind of Online Video buffer memory management method and system based on instantaneous access rate value.This method comprises: video is cut into sequentially in time, multiple byte numbers are identical and continuous video block, after video cache device receives the request for some video block that user sends, from locally or when necessary some video block is obtained from remote video server, some video block is sent to user;Video cache device is according to the information recorded in the data fifo structure of each video block and in user behavior file, the maximum instantaneous access rate value for calculating each video block and some video block in its spatial cache, the cache way of some video block is determined according to the maximum instantaneous access rate value of each video block.Present invention utilizes user's watching behavior characteristics of video-on-demand service system, the foundation of instantaneous access rate is calculated from online user's future watching behavior, so that the importance value of video block and the prediction accuracy of video-see are very high, so as to promote the efficiency of video cache.
Description
Technical field
The present invention relates to video storage tube reason technical field more particularly to a kind of online views based on instantaneous access rate value
Frequency buffer memory management method and system.
Background technique
Increasingly with the number of users of online viewing video more and more and available for download number of videos and size
Greatly, video server is faced with huge file storage pressure and bandwidth cost.File video cache device is to alleviate this pressure
The common method of power.
In video cache field, existing buffer replacing method include LRU (Least Recently Used, earliest recently
Access) tactful and LFU (Least Frequently Used, least referenced frequency) strategy.Their common ways all assume that
The behavior in user future is similar to behavior over, and then predicts the visit of file to the history access record of file according to user
Ask mode;When caching is full, these cache policies preferentially reject most unlikely accessed file from caching.Wherein, LRU plan
It is slightly based on history access record, the earliest file of last access time is preferentially rejected from caching;Its essence is file-based
The access module of short-term Popularity prediction this document, that is, if file is accessed recently, the following accessed probability
It is higher.It is minimum that LFU strategy rejects access frequency according to the requested frequency of history access record statistics file, preferentially from caching
File;Its essence is the access module of this document of the long-term Popularity prediction based on long-term documents, it may be assumed that if file was in the past
It is accessed many times, then the following accessed frequency is higher.
The shortcomings that buffer replacing method of above-mentioned video in the prior art are as follows: LRU policy and LFU strategy are all with file
History be accessed information replace file future be accessed mode, often will appear misjudgment, accuracy is limited.
Summary of the invention
The embodiment of the invention provides a kind of Online Video buffer memory management method and system based on instantaneous access rate value,
To improve the efficiency of video cache.
To achieve the goals above, this invention takes following technical solutions.
According to an aspect of the invention, there is provided a kind of Online Video cache management side based on instantaneous access rate value
Video, is cut into that multiple byte numbers are identical and continuous video block by method sequentially in time, which comprises
After video cache device receives the request for some video block that user sends, obtained from remote video server
Some described video block is sent to the user by some described video block, while recording the history access of user and video block
Information;
Video cache device calculates each of its video piece of collection according to the history access information of user and video block
The maximum instantaneous access rate MIRR of video block and some video block, according to the MIRR of each video block determine it is described some
The cache way of video block.
It is further, described that video is cut into multiple byte numbers are identical and continuous video block sequentially in time,
Include:
Video is divided into multiple byte numbers are identical and continuous video block sequentially in time, video cache device is in it
Storage and management are carried out to each video in a manner of video block in the spatial cache in portion, remote video server is with video block
All videos of mode storage and management, remote video server carry out data by communication network and each video cache device and lead to
Letter;
The historical data information of video cache device storage and management video block, the historical data information includes: user
Schedule file and/or video block access count file are watched, the user watches schedule file for recording each online user
The video block of the last time viewing and viewing moment, the video block access count file is for recording each video block by every
The access times added up when secondary accessed.
Further, after the video cache device receives the request of some video block that user sends, from long-range
Some described video block is obtained in video server, some described video block is sent to the user, comprising:
After video cache device receives the request for some video block that user sends, searches and compare video cache device
With the presence or absence of some described video block in the video piece of collection of storage, if there is some described video block, then will it is described some
Video block returns to user, while updating the historical data information of some video block, this process terminates;
If there is no some described video block, then video cache device to remote video server obtains some described view
Frequency block, and some described video block is returned into the user, then sequence executes some video block described in step S2, update
Historical data information, this process terminate.
Further, the cache way that some video block is determined according to the MIRR of each video block, comprising:
The memory for some video block that the video cache device caches its idle spatial cache with needs itself is big
It is small to be compared, when idle spatial cache is more than or equal to the memory of some video block itself, then the video
The buffer storage spatial cache that the deposit of some described video block is idle;
When idle spatial cache is less than the memory of some video block itself, the video cache device is according to it
The historical data information of each video block of storage calculates each video block and some described video in its video piece of collection
The MIRR value of block;When the smallest video block of MIRR value is not some described video block, then in spatial cache described in deletion
The smallest video block of MIRR value, and some described video block is stored on corresponding spatial cache;When the smallest video block of MIRR value
When being some described video block, then some described video block is stored not in local cache.
Further, the video cache device calculates its view according to the history access information of user and video block
The maximum instantaneous access rate MIRR of each video block and some video block in frequency set of blocks, comprising:
The collection of video block and some video block that setting video buffer storage has cached is combined into Γ, to each of Γ
Video block executes following operation:
According to the viewing demand of online user, maximum online user's request rate of video block is calculated;
According to the viewing demand of offline user, the offline user request rate of video block is calculated;
According to the position of video block in video, the maximum online user with certain strategy in conjunction with the video block requests speed
The offline user request rate of rate and the video block, forms the MIRR of the video block.
Further, the viewing demand according to online user, the maximum online user for calculating video block request speed
Rate, comprising:
For any one video block (v ', i ') in set Γ, wherein v ', i ' are that the corresponding video of video block is compiled respectively
Number and video block number, giving a maximum observation window length is [0, i'-1], watched according to online user described in
Video block (v', j in maximum observation window length [0, i'-1] rangek) position, jk≤ i', and it is based on video block access
The video block (v', i') and video block (v', j that accounting file obtainsk) continuously watched total degree, obtain the online user
To the viewing probability and request time of video block (v', i'), using the request time as the maximum observation window length [0,
I'-1] observation window length in range;
In each observation window length, by all users to the sum of viewing probability of video block (v', i') as at this
The total degree that the video block (v', i') measured in observation window length will be watched, by the total degree divided by corresponding observation window
Mouth length, obtains the online user access rate IRR of the corresponding video block of the observation window length (v', i'), by all sights
It surveys the corresponding IRR of length of window to be compared, the maximum value of IRR is taken to request as the maximum online user of video block (v', i')
Rate.
Further, the viewing demand according to offline user calculates the offline user request rate of video block, packet
It includes:
For any one video block (v ', i ') in set Γ, according to video block (v ', i ') in video block access count
Earliest primary requested moment, the last requested moment and the accumulative access total degree recorded in file, calculates
The offline user request rate of current time video block (v ', i ').
Further, the position according to video block in video, with certain strategy in conjunction with the video block most
The offline user request rate of big online user's request rate and the video block, forms the MIRR of the video block, comprising:
The number threshold θ for setting video block is greater than the video block of θ for numbering, directly with its maximum online user's request
MIRR of the rate as the video block;It is less than or equal to the video block of θ for numbering, with the maximum online user of the video block
MIRR of the maximum value of request rate and offline user request rate as the video block;
Alternatively,
The number threshold θ for setting video block is greater than the video block of θ for numbering, directly with its maximum online user's request
MIRR of the rate as the video block;It is less than or equal to the video block of θ for numbering, is used with the ARR of the video block online
The viewing demand at family compensates, and calculates new balanced user's request rate IRR+ (v,i), with the maximum of the video block
Online user's request rate and IRR+ (v,i)MIRR of the maximum value as the video block.
According to an aspect of the invention, there is provided a kind of Online Video cache management system based on instantaneous access rate value
System, comprising: video cache device and remote video server;
The remote video server is identical and continuous for video to be cut into multiple byte numbers sequentially in time
Video block, all videos of storage and management in a manner of video block are counted by communication network and each video cache device
According to communication;
The video cache device, after the request for receiving some video block that user sends, from long-distance video
Some described video block is obtained in server, some described video block is sent to the user, while recording user and video
The history access information of block;
According to the history access information of the user and video block, calculate each video block in its video piece of collection and
The maximum instantaneous access rate MIRR of some video block determines some video block according to the MIRR of each video block
Cache way.
Further, comprising: the video cache device includes video cache module, user's viewing video historical data letter
Cease management module, video block transmission module, video block access rate computing module and video block buffer process module;
The video cache module is identical and continuous for video to be cut into multiple byte numbers sequentially in time
Video block passes through the video block of each video of spatial cache storage and management;
The user watches video historical data information management module, the historical data for storage and management video block
Information, the historical data information include: that user watches schedule file and/or video block access count file, and the user sees
See schedule file for recording the video block and viewing moment that each online user's the last time is watched, the video block access meter
Number file is used to record each video block access times accumulative when each be accessed;
The video block transmission module after the request for receiving some video block that user sends, is searched and is compared
To some video block with the presence or absence of described in the video piece of collection of video cache device storage, if there is some described video
Some described video block is then returned to user by block, while updating the historical data information of some video block, this process
Terminate;If there is no some described video block, then video cache device to remote video server obtains some described video
Block, and some described video block is returned into the user;
The video block access rate computing module is calculated for the history access information according to user and video block
The maximum instantaneous access rate MIRR of each video block in its video piece of collection and some video block out;
The collection of video block and some video block that setting video buffer storage has cached is combined into Γ, to each of Γ
Video block executes following operation: according to the viewing demand of online user, calculating maximum online user's request rate of video block;
According to the viewing demand of offline user, the offline user request rate of video block is calculated;According to the position of video block in video,
With certain strategy in conjunction with maximum online user's request rate of the video block and the offline user request rate of the video block,
Form the MIRR of the video block;
The video block buffer process module, some video block sheet for caching idle spatial cache and needs
The memory size of body is compared, when idle spatial cache is more than or equal to the memory of some video block itself,
The then spatial cache that the deposit of some described video block is idle;When idle spatial cache is less than some video block itself
When memory, obtain each video block in the calculated video piece of collection of video block access rate computing module and it is described certain
The MIRR value of a video block is then deleted in spatial cache when the smallest video block of MIRR value is not some described video block
The smallest video block of MIRR value, and some described video block is stored on corresponding spatial cache;When the smallest view of MIRR value
When frequency block is some described video block, then some described video block is stored not in local cache.
As can be seen from the technical scheme provided by the above-mentioned embodiment of the present invention, view is utilized in the method for the embodiment of the present invention
User's watching behavior characteristic of demand service system, the foundation for calculating instantaneous access rate will be watched from online user's future
Behavior, so that the importance value of video block and the prediction accuracy of video-see are very high, so as to promote the effect of video cache
Rate improves QoS of customer.
The additional aspect of the present invention and advantage will be set forth in part in the description, these will become from the following description
Obviously, or practice through the invention is recognized.
Detailed description of the invention
In order to illustrate the technical solution of the embodiments of the present invention more clearly, required use in being described below to embodiment
Attached drawing be briefly described, it should be apparent that, drawings in the following description are only some embodiments of the invention, for this
For the those of ordinary skill of field, without creative efforts, it can also be obtained according to these attached drawings others
Attached drawing.
Fig. 1 is a kind of Online Video buffer memory management method based on instantaneous access rate value provided in an embodiment of the present invention
Process flow diagram;
Fig. 2 is a kind of exemplary diagram of the observation window of video block provided in an embodiment of the present invention.
Fig. 3 is that a kind of Online Video of instantaneous access rate value based on video block provided in an embodiment of the present invention caches system
The structural schematic diagram of system.
Specific embodiment
Embodiments of the present invention are described below in detail, the example of the embodiment is shown in the accompanying drawings, wherein from beginning
Same or similar element or element with the same or similar functions are indicated to same or similar label eventually.Below by ginseng
The embodiment for examining attached drawing description is exemplary, and for explaining only the invention, and is not construed as limiting the claims.
Those skilled in the art of the present technique are appreciated that unless expressly stated, singular " one " used herein, " one
It is a ", " described " and "the" may also comprise plural form.It is to be further understood that being arranged used in specification of the invention
Diction " comprising " refer to that there are the feature, integer, step, operation, element and/or component, but it is not excluded that in the presence of or addition
Other one or more features, integer, step, operation, element, component and/or their group.It should be understood that when we claim member
Part is " connected " or when " coupled " to another element, it can be directly connected or coupled to other elements, or there may also be
Intermediary element.In addition, " connection " used herein or " coupling " may include being wirelessly connected or coupling.Wording used herein
"and/or" includes one or more associated any cells for listing item and all combinations.
Those skilled in the art of the present technique are appreciated that unless otherwise defined, all terms used herein (including technology art
Language and scientific term) there is meaning identical with the general understanding of those of ordinary skill in fields of the present invention.Should also
Understand, those terms such as defined in the general dictionary, which should be understood that, to be had and the meaning in the context of the prior art
The consistent meaning of justice, and unless defined as here, it will not be explained in an idealized or overly formal meaning.
In order to facilitate understanding of embodiments of the present invention, it is done by taking several specific embodiments as an example below in conjunction with attached drawing further
Explanation, and each embodiment does not constitute the restriction to the embodiment of the present invention.
Embodiment one
The embodiment of the present invention according to online and offline user to the importance of the total demand predicted video block of video block, and
The cache decision of video block is carried out according to the importance of video block.
In actual line stream media system, user will not usually watch entire video from the beginning to the end, so that same video
Different fragments have different popularities.Each video is averagely divided into N according to the natural order of video by the embodiment of the present invention
A continuous videos block, for video v, the play time length of each video block after cutting is Tv, number is 0 to sequence respectively,
1,2 ..., N-1 is respectively 0, T at the beginning of corresponding to video block in videov, 2 × Tv..., (N-1) × Tv.It therefore, can be with
The popularity of entire video file is refine to the popularity of each video block in video file.In turn, in the present invention, caching system
System is designed to preferentially store most popular video block, rather than the entire video file of most popular video, empty to improve caching
Between utilization rate, and then promoted caching system overall performance.
MIRR (Maximum Instant Request Rate, the maximum instantaneous access of video block of the embodiment of the present invention
Rate) Lai Hengliang video block importance.It would generally be watched according to the natural order of video after system and can since user enters
It can be left halfway (according to certain probability), therefore, can predict that the viewing of subsequent each video block is general according to its current viewing position
Rate and start viewing at the time of.In view of a certain given current time, there may be multiple users just to watch same video at the same time
Different location.Therefore, for any given video block, possible multiple users are subsequent to will do it viewing.A given observation
Time window, the user for being in different viewing progresses can watch the video block in different moments with different probability;These viewings
The sum of probability is exactly the total degree observation time window in will online watched of the video block in statistical significance;Into one
Step, the total degree watched show that online user of the video block on the observation window asks divided by observation time window length
Rate is sought, which is referred to as IRR.The present invention calculates in different sights each video block
Corresponding IRR value in window size is surveyed, and maximum online user's request rate value is denoted as MIRRonline.Video block
MIRRonlineThe bigger expression of value has that more online users will watch this video block or/and user gets within the observation period
Urgent needs to watch the video block, therefore the importance of the video block is higher.Therefore, the present invention utilizes each video block
MIRRonlineValue assesses corresponding video block for the importance of online user.
On the other hand, each video block has the offline user request rate based on user's history viewing record, this
Invention is referred to as ARR.In the present invention, a video block is given, ARR value is equal to the video within past a period of time
The total degree that block has been watched divided by this period duration.For this purpose, to each video block, one regular length of the present invention
FIFO (First Input First Output, first in, first out data structure) records the access information of the video block.Specifically
For, when the video block is requested, FIFO will record its requested moment and be accessed by the end of current time total secondary
Several and video is by the total degree under continuous viewing.Information above based on FIFO record, it is known that, it is stored in FIFO
Newest and oldest request record the difference of corresponding request total degree, the video block is watched exactly within the past period
Total degree;With the total degree divided by (difference at the oldest record moment in current time and FIFO), the video has just been obtained
The ARR value of block.ARR value reflects video block in the importance of the past period.The length of FIFO will affect the steady of ARR numerical value
Qualitative and sensitivity: length is longer, and the observation period is longer, can more react the long-term average statistics feature of video block, and length is shorter,
The observation period is shorter, can more react the short term statistics feature of video block.
Implementation column of the present invention provides different methods, by reflecting video block to the MIRR of online user's importanceonlineValue and
The ARR value of offline user importance is combined, obtains the MIRR value for measuring video block importance.
A kind of Online Video of maximum instantaneous access rate value based on video block provided in an embodiment of the present invention caches pipe
The process flow of reason method is as shown in Figure 1, include following processing step:
Step S110, video cache device receive the request of the video block i of the video v of user's transmission.
Step S120, video cache device respond user's request.Judge whether its internal spatial cache has the view of video v
Frequency block i, (being hereinafter referred to as (v, i)).If there being (v, i), (v, i) is directly returned into user, while triggering update and using
The history access information at family and video block, i.e. y user watch schedule file and video block access count file, then return step
S110, to continue waiting for the request next time of users;Otherwise, video cache device will be downloaded from remote video server (v,
I), user and by (v, i) is returned to, then proceedes to execute step S130.
The user watch schedule file be one with online user ID for key, be with the corresponding four-tuple of the user
The data structure of value, the user of each online user watch the corresponding four-tuple of schedule file, which includes
The starting for the video block that the video block ID and the user that online user ID, video ID, the user are watching are being watched recently
Moment;
The video block access count file is multiple fifo queues for recording the accessed information of video block, each
Fifo queue is used to record the accessed information an of video block, has video block is requested can all produce in corresponding FIFO every time
A raw new record, the corresponding triple of every record, content include: that the video block is requested the moment, by the end of current
Moment accessed total degree, and continuously watched by the end of current time the total degree of user's access of the video.
The renewal process of the history access information of user involved in the above process and video block is as follows:
The renewal process that user watches schedule file is as follows: schedule file more new record (u, v, i, t) is watched in user,
Middle u is the User ID for requesting video block, and v and i are the video and video block number of user's request respectively, t be system it is current when
It carves.
The renewal process of video block access count file is as follows.Increase by one in (v, i) corresponding data fifo structure
Record, records the return moment t of the video block(v, i), by the end of the video block at this time be accessed total degree mt (v, i), and cut
The total degree s of user's access of the video is only continuously watched to the current time video blockt (v, i)。
Step S130, the video block (v, i) itself that video cache device caches its idle spatial cache V and needs
Memory size J is compared: as V >=J, then video block (v, i) is stored in the idle portion in spatial cache by video cache device
Position, while the history access information (the same S120 of update method) for updating user and video block is triggered, process terminates.Then step is returned
Rapid S110, to continue waiting for the request next time of users;As V < J, step S140 is triggered.
Step S140 calculates each video block and (has deposited including all in video block (v, i) to be cached and spatial cache
Video block is indicated with video piece of collection Γ) MIRRonlineValue and ARR value.Then step S150 will be triggered.MIRRonlineValue and
The calculating process of ARR value is as follows:
For any one video block (v ', i ') in set Γ, wherein v ', i ' are that the corresponding video of video block is compiled respectively
Number and video block number, giving a maximum observation window length is [0, i'-1], watched according to online user described in
Video block (v', j in maximum observation window length [0, i'-1] rangek) position, jk≤ i', and it is based on video block access
The video block (v', i') and video block (v', j that accounting file obtainsk) continuously watched total degree, obtain the online user
To the viewing probability and request time of video block (v', i'), using the request time as the maximum observation window length [0,
I'-1] observation window length in range;
In each observation window length, by all users to the sum of viewing probability of video block (v', i') as at this
The total degree that the video block (v', i') measured in observation window length will be watched, by the total degree divided by corresponding observation window
Mouth length, obtains the online user access rate IRR of the corresponding video block of the observation window length (v', i'), by all sights
It surveys the corresponding IRR of length of window to be compared, the maximum value of IRR is taken to request as the maximum online user of video block (v', i')
Rate.
Fig. 2 is a kind of exemplary diagram of the observation window of video block provided in an embodiment of the present invention.It is assumed that when being scheduled on current
T, video block (v ', i ') ∈ Γ are carved, and its maximum observation window is [0, i'-1], calculates the MIRR of (v ', i ') belowonline。
It is assumed that there is k online users within the scope of the maximum observation window, viewing number is 0 certain for arriving i'-1 respectively
A little video blocks.These specific viewing locations form a set omega={ σk|0≤σk<i'×Tv', k=0,1 ..., | Ω | -1 },
Wherein Tv'It is the play time length of each video block.Each element σ in ΩkFor a video-see position, refer to useful
Family was from video starting position 0 second to current viewing position σkSecond and σ0≤σ1≤…≤σ|Ω|-1。
Without loss of generality, each it is in viewing location σkUser, currently watched video block is
Thereafter centainly to watch probability viewing video block i', which is denoted as extended meetingCalculation formula are as follows:
Wherein, molecule and denominator are that the video block (v', i') that is obtained respectively according to video block access count file is continuous
Watch total degree and video block (v', jk) continuously watched total degree;The user can be in certain time ΔkRequest afterwards (v',
I'),
Δk=i' × Tv'-σk (3)
The time ΔkThe referred to as user is to the request time of video block (v', i'), while ΔkDescribed in being also defined as most
An observation window length in big observation window ranges, corresponding observation scope is section [σk,i'×Tv')。
The present invention is in each observation window length Δk, k=0,1 ..., | Ω | on -1, respectively video block (v', i') is counted
Calculate online user's access rate.It is assumed that being Δ in lengthkObservation window [σk,i'×Tv') on a total of LkIt is a online
User, wherein each user l, l=0,1 ..., Lk- 1, watching different video block εl, jk≤εl<i'.This all LkA use
Family is in the observation window length Δ to the sum of the viewing probability of (v', i')kThe total degree that (v', the i') of upper measurement will be watched
(it is denoted as Ck), it may be assumed that
Use CkDivided by corresponding observation window length Δk, obtain in ΔkThe video block measured on the observation window of length
The average access rate of (v', i'), the present invention are referred to as online user's access rate, are denoted asCount formula are as follows:
By all observation window length Δsk, k=0,1 ..., | Ω | -1 corresponding IRR is compared, and takes its maximum value to be
Maximum online user's request rate of video block (v', i'), is denoted as
The ARR value calculating method of (v ', i ') is as follows next, video block.To any one video block (v ', i ') ∈ Г,
It is assumed that the earliest primary requested moment that the video block records in data fifo structure is ts, corresponding accumulative access
Total degree isThe last requested moment is te, the total degree of corresponding accumulative access isWhen giving current
Carve t, the offline user request rate ARR of video block (v ', i ')(v′,i′)Calculation formula it is as follows:
The ARR value and MIRR of each video block are obtained as a result,onlineThen value triggers step S150.
Step S150, execution is with video block ARR value to video block MIRRonlineThe compensation mechanism of value.
The present invention provides a kind of IRR compensation mechanism based on ARR according to two kinds of access rates of each video block.Often
A video block has an ARR value and multiple IRR values.Wherein, the MIRR of each video blockonlineThe calculating of value all relies on place
The viewing location of online user those of before the video block position.For example, to the last one video block, each online user
Be likely to access it in future, therefore, the video block have enough following access requests to accurate evaluation its
MIRRonlineValue;On the contrary, to first video block, since all online users or are watching the video block, online
User no longer will issue access request to first video block, and therefore, which does not have enough following access requests to use
With its MIRR of accurate evaluationonlineValue.
Due to MIRRonlineValue is that the actual demand based on online user's future calculates, and ARR value is gone through based on user
What history access request calculated, therefore, MIRRonlineMore can reliable reflecting video block actual importance.
Therefore, the present invention is as far as possible according to MIRRonlineTo determine video block importance.Specifically, the present invention is to position
Video block rearward directly uses MIRRonlineAs the MIRR of video block, to determine video block importance;Before contraposition is rested against
Video block, with the ARR of the video block to its MIRRonlineIt compensates, calculates MIRR again then to determine that video block is important
Property.
Compensation method has complete ARR compensation or part ARR to compensate two kinds:
Method 1: complete ARR compensation mechanism
To all N number of video blocks of a video, two parts are classified as according to threshold θ, number is less than or equal to the video of θ
Block and number are greater than the video block of θ.The compensation mechanism is as follows:
According to above formula, the video block uncompensation of θ is greater than for numbering, directly takes MIRRonlineValue is used as video block importance
Angle value MIRR;Otherwise, the larger value in the ARR value and MIRR value of selecting video block is as video block importance angle value MIRR.
Method 2: part ARR compensation mechanism.
To all N number of video blocks of a video, two parts are classified as according to threshold θ, number is less than or equal to the video of θ
Block and number are greater than the video block of θ.
It is greater than the video block of θ for numbering, directly uses its maximum online user's request rate as the MIRR of the video block;
It is less than or equal to the video block of θ for numbering, is compensated with viewing demand of the ARR of the video block to online user, is counted
Calculate new balanced user's request rate.It is assumed that the video block for needing to carry out rate compensation is (v, i), a reason is given
The observation window length (it is assumed that be equal to θ) thought, then new balanced user's request rate calculation method are as follows:
Then, with the maximum online user's request rate and IRR of the video block+ (v,i)Maximum value as the video block
MIRR, the compensation mechanism are as follows:
According to above formula, the video block uncompensation of θ is greater than for numbering, directly takes MIRRonlineValue is used as video block importance
Angle value MIRR;It is less than or equal to the video block of θ for numbering, with the maximum online user's request rate and IRR of the video block+ (v,i)
MIRR of the maximum value as the video block.
The ARR value and MIRR of each video block are obtained as a result,onlineThen value triggers step S160.
Step S160 carries out caching replacement according to the importance value MIRR of the video block calculated in step S150.
It (including all in new video block and caching has been deposited according to above-mentioned steps it can be concluded that all video blocks (v, i) ∈ Γ
Video block) importance value, therefrom choose the smallest video block of importance value and eliminated, wherein the smallest video of importance
Block (v', r') is given by,
If (v', r') is exactly then to trigger the history access information for updating user and video block to storage video block (v, r)
(the same S120 of update method) and then return step S110;Otherwise, (v, r) is stored into the corresponding position of (v', r'), is triggered simultaneously
The history access information (the same S120 of update method) of user and video block are updated, then return step S110.
The method of the embodiments of the present invention is suitable for the online Streaming Media such as internet, cable television, video and audio and broadcasts
Place system.
Embodiment two
A kind of structural representation for Online Video cache management system based on instantaneous access rate value that the embodiment provides
Figure is as shown in Figure 3, comprising: video cache device 31 and remote video server 32;
The remote video server 32 is identical and continuous for video to be cut into multiple byte numbers sequentially in time
Video block, all videos of storage and management in a manner of video block pass through communication network and each video cache device and carry out
Data communication;
The video cache device 31 is identical and continuous for video to be cut into multiple byte numbers sequentially in time
Video block, after the request for receiving some video block that user sends, some described view is obtained from remote video server
Some described video block is sent to the user by frequency block;
According to the history access information of the user and video block, calculate each video block in its video piece of collection and
The MIRR of some video block determines the cache way of some video block according to the MIRR of each video block.
Specifically, the video cache device 31 includes video cache module 311, user's viewing video historical data information
Management module 312, video block transmission module 313, video block access rate computing module 314 and video block buffer process module
315。
The video cache module 311, for video is cut into sequentially in time multiple byte numbers it is identical and even
Continuous video block passes through the video block of each video of spatial cache storage and management;
The user watches video historical data information management module 312, the historical data of storage and management video block
Information, the historical data information include: that user watches schedule file, are watched for recording each online user's the last time
Video block and viewing moment;And/or video block access count file, it is tired when each accessed for recording each video block
The access times of meter.The module carries out the update of video block historical data information at the end of each process.
The video block transmission module 313 after the request for receiving some video block that user sends, is searched simultaneously
It compares with the presence or absence of some described video block in the video piece of collection of video cache device storage, if there is some described video
Some described video block is then returned to user by block, while updating the historical data information of some video block, this process
Terminate;If there is no some described video block, then video cache device to remote video server obtains some described video
Block, and some described video block is returned into the user.It is specific: to receive the request of the video block i of the video v of user's transmission
Afterwards, judge the video block i (hereinafter referred to as (v, i)) for whether having video v in its internal spatial cache, if there being (v, i),
(v, i) is directly returned into user, updates the video block accounting file in the data fifo structure of (v, i), process terminates;It is no
Then, (v, i) will be downloaded from remote video server, and (v, i) is returned into user, in the data fifo structure for updating (v, i)
Video block accounting file, then proceed to the caching process process for executing subsequent (v, i);
Video block accounting file in the data fifo structure for updating (v, i) includes: FIFO number corresponding at (v, i)
According to a record is increased in structure, record is somebody's turn to do the return moment t of (v, i)(v, i), should (v, i) accessed total time by the end of at this time
Number mt (v, i), and be somebody's turn to do (v, i) by the end of current time and continuously watched the total degree s that the user of the video accessest (v, i)。
The video block access rate computing module 314, for being less than the view for needing to cache when idle spatial cache
When the memory size of frequency block (v, i) itself, calculate every in video piece of collection Γ in video block (v, i) and spatial cache to be cached
The maximum instantaneous access rate value MIRR value of a video block,
For any one video block (v ', i ') in set Γ, wherein v ', i ' are that the corresponding video of video block is compiled respectively
Number and video block number, giving a maximum observation window length is [0, i'-1], watched according to online user described in
Video block (v', j in maximum observation window length [0, i'-1] rangek) position, jk≤ i', and it is based on video block access
The video block (v', i') and video block (v', j that accounting file obtainsk) continuously watched total degree, obtain the online user
To the viewing probability and request time of video block (v', i'), using the request time as the maximum observation window length [0,
I'-1] observation window length in range;
In each observation window length, by all users to the sum of viewing probability of video block (v', i') as at this
The total degree that the video block (v', i') measured in observation window length will be watched, by the total degree divided by corresponding observation window
Mouth length, obtains the online user access rate IRR of the corresponding video block of the observation window length (v', i'), by all sights
It surveys the corresponding IRR of length of window to be compared, the maximum value of IRR is taken to request as the maximum online user of video block (v', i')
Rate.
The calculating process of the MIRR value is as follows:
It is assumed that there is k online users within the scope of the maximum observation window, viewing number is 0 certain for arriving i'-1 respectively
A little video blocks.These specific viewing locations form a set omega={ σk|0≤σk<i'×Tv', k=0,1 ..., | Ω | -1 },
Wherein each element σ in ΩkFor a video-see position, refer to user from video starting position 0 second to currently viewing
Position σkSecond, and σ0≤σ1≤…≤σ|Ω|-1, Tv'It is the play time length of each video block.
Without loss of generality, each it is in viewing location σkUser, currently watched video block is
Thereafter centainly to watch probability viewing video block i', which is denoted as extended meetingCalculation formula are as follows:
Wherein, molecule and denominator are that the video block (v', i') that is obtained respectively according to video block access count file is continuous
Watch total degree and video block (v', jk) continuously watched total degree;The user can be in certain time ΔkRequest afterwards (v',
I'),
Δk=i' × Tv'-σk (3)
The time ΔkThe referred to as user is to the request time of video block (v', i'), while ΔkDescribed in being also defined as most
An observation window length in big observation window ranges, corresponding observation scope is section [σk,i'×Tv')。
The present invention is in each observation window length Δk, k=0,1 ..., | Ω | on -1, respectively video block (v', i') is counted
Calculate online user's access rate.It is assumed that being Δ in lengthkObservation window [σk,i'×Tv') on a total of LkIt is a online
User, wherein each user l, l=0,1 ..., Lk- 1, watching different video block εl, jk≤εl<i'.This all LkA use
Family is in the observation window length Δ to the sum of the viewing probability of (v', i')kThe total degree that (v', the i') of upper measurement will be watched
(it is denoted as Ck), it may be assumed that
Use CkDivided by corresponding observation window length Δk, obtain in ΔkThe video block measured on the observation window of length
The average access rate of (v', i'), the present invention are referred to as online user's access rate, are denoted asCount formula are as follows:
By all observation window length Δsk, k=0,1 ..., | Ω | -1 corresponding IRR is compared, and takes its maximum value to be
Maximum online user's request rate of video block (v', i'), is denoted as
The ARR value calculating method of (v ', i ') is as follows next, video block.To any one video block (v ', i ') ∈ Г,
It is assumed that the earliest primary requested moment that the video block records in data fifo structure is ts, corresponding accumulative access
Total degree isThe last requested moment is te, the total degree of corresponding accumulative access isWhen giving current
Carve t, the offline user request rate ARR of video block (v ', i ')(v',i')Calculation formula it is as follows:
Then one of following compensation method is taken, video block importance angle value MIRR is calculated:
Compensation method has complete ARR compensation or part ARR to compensate two kinds:
Method 1: complete ARR compensation mechanism
To all N number of video blocks of a video, two parts are classified as according to threshold θ, number is less than or equal to the video of θ
Block and number are greater than the video block of θ.The compensation mechanism is as follows:
According to above formula, the video block uncompensation of θ is greater than for numbering, directly takes MIRRonlineValue is used as video block importance
Angle value MIRR;Otherwise, the larger value in the ARR value and MIRR value of selecting video block is as video block importance angle value MIRR.
Method 2: part ARR compensation mechanism.
To all N number of video blocks of a video, two parts are classified as according to threshold θ, number is less than or equal to the video of θ
Block and number are greater than the video block of θ.
It is greater than the video block of θ for numbering, directly uses its maximum online user's request rate as the MIRR of the video block;
It is less than or equal to the video block of θ for numbering, is compensated with viewing demand of the ARR of the video block to online user, is counted
Calculate new balanced user's request rate.It is assumed that the video block for needing to carry out rate compensation is (v, i), a reason is given
The observation window length (it is assumed that be equal to θ) thought, then new balanced user's request rate calculation method are as follows:
Then, with the maximum online user's request rate and IRR of the video block+ (v,i)Maximum value as the video block
MIRR, the compensation mechanism are as follows:
According to above formula, the video block uncompensation of θ is greater than for numbering, directly takes MIRRonlineValue is used as video block importance
Angle value MIRR;It is less than or equal to the video block of θ for numbering, with the maximum online user's request rate and IRR of the video block+ (v,i)
MIRR of the maximum value as the video block.
Video block cache module 315, for by idle spatial cache and needing video block (v, i) to be cached itself
Memory size is compared, when idle spatial cache is more than or equal to the interior of the video block (v, i) to be cached itself
When depositing, then by the spatial cache of some described video block deposit free time;When idle spatial cache is less than some described video block
When the memory of itself, obtain each video block in the calculated video piece of collection of video block access rate computing module and
The MIRR value of some video block, using the MIRR of each video block as the importance angle value of each video block, by each view
The importance value of frequency block is ranked up, when the smallest video block of importance value is not video block (v, i) to be cached, from
The smallest video block of the importance value is deleted in the spatial cache of the video cache device, by the video block to be cached
(v, i) is stored in the spatial cache of the video cache device;When the smallest video block of importance value is the video to be cached
When block (v, i), it is not stored in the video block (v, i) to be cached.
The tool of the Online Video caching of the instantaneous access rate value based on video block is carried out with the device of the embodiment of the present invention
Body treatment process is similar to the previous method embodiment, and details are not described herein again.
In conclusion user's watching behavior characteristic of video-on-demand service system is utilized in the method for the embodiment of the present invention,
By video piecemeal, use each video block possible requested maximum instantaneous rate within following a period of time important as video block
Angle value, and caching replacement is carried out accordingly.The foundation of the instantaneous access rate of calculating in the embodiment of the present invention is from online user
The following watching behavior, so that the importance value of video block and the prediction accuracy of video-see are very high, so as to promote video
The efficiency of caching improves QoS of customer.
Those of ordinary skill in the art will appreciate that: attached drawing is the schematic diagram of one embodiment, module in attached drawing or
Process is not necessarily implemented necessary to the present invention.
All the embodiments in this specification are described in a progressive manner, same and similar portion between each embodiment
Dividing may refer to each other, and each embodiment focuses on the differences from other embodiments.Especially for device or
For system embodiment, since it is substantially similar to the method embodiment, so describing fairly simple, related place is referring to method
The part of embodiment illustrates.Apparatus and system embodiment described above is only schematical, wherein the conduct
The unit of separate part description may or may not be physically separated, component shown as a unit can be or
Person may not be physical unit, it can and it is in one place, or may be distributed over multiple network units.It can root
According to actual need that some or all of the modules therein is selected to achieve the purpose of the solution of this embodiment.Ordinary skill
Personnel can understand and implement without creative efforts.
The foregoing is only a preferred embodiment of the present invention, but scope of protection of the present invention is not limited thereto,
In the technical scope disclosed by the present invention, any changes or substitutions that can be easily thought of by anyone skilled in the art,
It should be covered by the protection scope of the present invention.Therefore, protection scope of the present invention should be with scope of protection of the claims
Subject to.
Claims (10)
1. a kind of Online Video buffer memory management method based on instantaneous access rate value, which is characterized in that by video according to the time
Sequence is cut into that multiple byte numbers are identical and continuous video block, which comprises
After video cache device receives the request for some video block that user sends, from remote video server described in acquisition
Some described video block is sent to the user, while recording the history access information of user and video block by some video block;
Video cache device calculates each video in its video piece of collection according to the history access information of user and video block
The maximum instantaneous access rate MIRR of block and some video block determines some described video according to the MIRR of each video block
The cache way of block.
2. the method according to claim 1, wherein described be cut into video multiple words sequentially in time
Joint number is identical and continuous video block, comprising:
Video is divided into multiple byte numbers are identical and continuous video block sequentially in time, video cache device is inside it
Storage and management are carried out to each video in a manner of video block in spatial cache, remote video server is in a manner of video block
All videos of storage and management, remote video server carry out data communication by communication network and each video cache device;
The historical data information of video cache device storage and management video block, the historical data information include: user's viewing
Schedule file and/or video block access count file, it is nearest for recording each online user that the user watches schedule file
The video block once watched and viewing moment, the video block access count file is for recording each video block by quilt every time
The access times added up when access.
3. the method according to claim 1, wherein the video cache device receives certain that user sends
After the request of a video block, some described video block is obtained from remote video server, some described video block is sent to
The user, comprising:
After video cache device receives the request for some video block that user sends, searches and compare the storage of video cache device
Video piece of collection in the presence or absence of some described video block, then will some described video if there is some described video block
Block returns to user, while updating the historical data information of some video block, this process terminates;
If there is no some described video block, then video cache device to remote video server obtains some described video
Block, and some described video block is returned into the user, then sequence executes going through for some video block described in step S2, update
History data information, this process terminate.
4. the method according to claim 1, wherein it is described according to the MIRR of each video block determine it is described certain
The cache way of a video block, comprising:
The memory size for some video block that the video cache device caches its idle spatial cache and needs itself into
Row compares, when idle spatial cache is more than or equal to the memory of some video block itself, then the video cache
The device spatial cache that the deposit of some described video block is idle;
When idle spatial cache is less than the memory of some video block itself, the video cache device is according to its storage
The historical data information of each video block calculate each video block in its video piece of collection and some video block
MIRR value;When the smallest video block of MIRR value is not some described video block, then the MIRR value is deleted in spatial cache
The smallest video block, and some described video block is stored on corresponding spatial cache;When the smallest video block of MIRR value is described
When some video block, then some described video block is stored not in local cache.
5. method according to claim 1 or 4, which is characterized in that the video cache device is according to user and video
The history access information of block, the maximum instantaneous for calculating each video block and some video block in its video piece of collection are visited
Ask rate MIRR, comprising:
The collection of video block and some video block that setting video buffer storage has cached is combined into Γ, to each video in Γ
Block executes following operation:
According to the viewing demand of online user, maximum online user's request rate of video block is calculated;
According to the viewing demand of offline user, the offline user request rate of video block is calculated;
According to the position of video block in video, with certain strategy in conjunction with maximum online user's request rate of the video block and
The offline user request rate of the video block, forms the MIRR of the video block.
6. according to the method described in claim 5, it is characterized in that, the viewing demand according to online user, calculates view
Maximum online user's request rate of frequency block, comprising:
For any one video block (v ', i ') in set Γ, wherein v ', i ' be respectively the corresponding video number of video block and
Video block number, giving a maximum observation window length is [0, i'-1], the maximum watched according to online user
Video block (v', j in observation window length [0, i'-1] rangek) position, jk≤ i', and it is based on video block access count
The video block (v', i') and video block (v', j that file obtainsk) continuously watched total degree, obtain the online user to view
The viewing probability and request time of frequency block (v', i'), using the request time as the maximum observation window length [0, i'-1]
An observation window length in range;
In each observation window length, by all users to the sum of viewing probability of video block (v', i') as in the observation
The total degree that the video block (v', i') measured in length of window will be watched, the total degree is long divided by corresponding observation window
Degree, obtains the online user access rate IRR of the corresponding video block of the observation window length (v', i'), by all observation windows
The corresponding IRR of mouth length is compared, and takes maximum online user request rate of the maximum value of IRR as video block (v', i').
7. according to the method described in claim 5, it is characterized in that, the viewing demand according to offline user, calculates view
The offline user request rate of frequency block, comprising:
For any one video block (v ', i ') in set Γ, according to video block (v ', i ') in video block access count file
Earliest primary requested moment, the last requested moment and the accumulative access total degree of middle record, calculate current
The offline user request rate of moment video block (v ', i ').
8. according to the method described in claim 5, it is characterized in that, the position according to video block in video, with one
Fixed strategy is formed in conjunction with maximum online user's request rate of the video block and the offline user request rate of the video block
The MIRR of the video block, comprising:
The number threshold θ for setting video block is greater than the video block of θ for numbering, directly with its maximum online user's request rate
MIRR as the video block;It is less than or equal to the video block of θ for numbering, is requested with the maximum online user of the video block
The MIRR of rate and the maximum value of offline user request rate as the video block;
Alternatively,
The number threshold θ for setting video block is greater than the video block of θ for numbering, directly with its maximum online user's request rate
MIRR as the video block;It is less than or equal to the video block of θ for numbering, with the ARR of the video block to online user's
Viewing demand compensates, and calculates new balanced user's request rate IRR+ (v,i), online with the maximum of the video block
User's request rate and IRR+ (v,i)MIRR of the maximum value as the video block.
9. a kind of Online Video cache management system based on instantaneous access rate value characterized by comprising video cache dress
It sets and remote video server;
The remote video server, for video to be cut into sequentially in time, multiple byte numbers are identical and continuous video
It is logical to carry out data by communication network and each video cache device for block, all videos of storage and management in a manner of video block
Letter;
The video cache device, after the request for receiving some video block that user sends, from long-distance video service
Some described video block is obtained in device, some described video block is sent to the user, while recording user and video block
History access information;
According to the history access information of the user and video block, each video block in its video piece of collection and described is calculated
The maximum instantaneous access rate MIRR of some video block determines the caching of some video block according to the MIRR of each video block
Mode.
10. system according to claim 9 characterized by comprising the video cache device includes video cache mould
Block, user watch video historical data information management module, video block transmission module, video block access rate computing module and view
Frequency block buffer process module;
The video cache module, for video to be cut into sequentially in time, multiple byte numbers are identical and continuous video
Block passes through the video block of each video of spatial cache storage and management;
The user watches video historical data information management module, and the historical data for storage and management video block is believed
Breath, the historical data information include: that user watches schedule file and/or video block access count file, user's viewing
Schedule file is used to record the video block and viewing moment that each online user's the last time is watched, the video block access count
File is used to record each video block access times accumulative when each be accessed;
The video block transmission module after the request for receiving some video block that user sends, searches and compares view
With the presence or absence of some described video block in the video piece of collection of frequency buffer storage storage, if there is some described video block, then
Some described video block is returned into user, while updating the historical data information of some video block, this process terminates;
If there is no some described video block, then video cache device to remote video server obtains some described video block, and
Some described video block is returned into the user;
The video block access rate computing module calculates it for the history access information according to user and video block
The maximum instantaneous access rate MIRR of each video block and some video block in video piece of collection;
The collection of video block and some video block that setting video buffer storage has cached is combined into Γ, to each video in Γ
Block executes following operation: according to the viewing demand of online user, calculating maximum online user's request rate of video block;According to
The viewing demand of offline user calculates the offline user request rate of video block;According to the position of video block in video, with one
Fixed strategy is formed in conjunction with maximum online user's request rate of the video block and the offline user request rate of the video block
The MIRR of the video block;
The video block buffer process module, some video block itself for caching idle spatial cache and needs
Memory size is compared, and when idle spatial cache is more than or equal to the memory of some video block itself, then will
The idle spatial cache of some video block deposit;When idle spatial cache is less than the memory of some video block itself
When, obtain each video block in the calculated video piece of collection of video block access rate computing module and some described view
The MIRR value of frequency block, when the smallest video block of MIRR value is not some described video block, then in spatial cache described in deletion
The smallest video block of MIRR value, and some described video block is stored on corresponding spatial cache;When the smallest video block of MIRR value
When being some described video block, then some described video block is stored not in local cache.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201811491430.2A CN109769136B (en) | 2018-12-07 | 2018-12-07 | Online video cache management method and system based on instantaneous access rate value |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201811491430.2A CN109769136B (en) | 2018-12-07 | 2018-12-07 | Online video cache management method and system based on instantaneous access rate value |
Publications (2)
Publication Number | Publication Date |
---|---|
CN109769136A true CN109769136A (en) | 2019-05-17 |
CN109769136B CN109769136B (en) | 2021-02-19 |
Family
ID=66451191
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201811491430.2A Active CN109769136B (en) | 2018-12-07 | 2018-12-07 | Online video cache management method and system based on instantaneous access rate value |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN109769136B (en) |
Cited By (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN111222004A (en) * | 2020-01-08 | 2020-06-02 | 广州小鹏汽车科技有限公司 | Audio playing method, terminal equipment and vehicle |
CN114900732A (en) * | 2022-04-25 | 2022-08-12 | 北京奇艺世纪科技有限公司 | Video caching method and device, electronic equipment and storage medium |
Citations (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN105518539A (en) * | 2013-09-04 | 2016-04-20 | 依视路国际集团(光学总公司) | Methods and systems for augmented reality |
CN105960790A (en) * | 2013-09-27 | 2016-09-21 | 阿尔卡特朗讯公司 | Method for caching |
CN106294827A (en) * | 2016-08-17 | 2017-01-04 | 合网络技术(北京)有限公司 | The caching method of multimedia resource and device |
CN108259929A (en) * | 2017-12-22 | 2018-07-06 | 北京交通大学 | A kind of prediction of video active period pattern and caching method |
US10062414B1 (en) * | 2017-08-22 | 2018-08-28 | Futurewei Technologies, Inc. | Determining a future field of view (FOV) for a particular user viewing a 360 degree video stream in a network |
-
2018
- 2018-12-07 CN CN201811491430.2A patent/CN109769136B/en active Active
Patent Citations (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN105518539A (en) * | 2013-09-04 | 2016-04-20 | 依视路国际集团(光学总公司) | Methods and systems for augmented reality |
CN105960790A (en) * | 2013-09-27 | 2016-09-21 | 阿尔卡特朗讯公司 | Method for caching |
CN106294827A (en) * | 2016-08-17 | 2017-01-04 | 合网络技术(北京)有限公司 | The caching method of multimedia resource and device |
US10062414B1 (en) * | 2017-08-22 | 2018-08-28 | Futurewei Technologies, Inc. | Determining a future field of view (FOV) for a particular user viewing a 360 degree video stream in a network |
CN108259929A (en) * | 2017-12-22 | 2018-07-06 | 北京交通大学 | A kind of prediction of video active period pattern and caching method |
Non-Patent Citations (1)
Title |
---|
肖敬伟: "基于数据挖掘的缓存替换算法研究", 《中国优秀硕士学位论文全文数据库信息科技辑,2015年第9期》 * |
Cited By (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN111222004A (en) * | 2020-01-08 | 2020-06-02 | 广州小鹏汽车科技有限公司 | Audio playing method, terminal equipment and vehicle |
CN111222004B (en) * | 2020-01-08 | 2023-03-31 | 广州小鹏汽车科技有限公司 | Audio playing method, terminal equipment and vehicle |
CN114900732A (en) * | 2022-04-25 | 2022-08-12 | 北京奇艺世纪科技有限公司 | Video caching method and device, electronic equipment and storage medium |
CN114900732B (en) * | 2022-04-25 | 2024-01-12 | 北京奇艺世纪科技有限公司 | Video caching method and device, electronic equipment and storage medium |
Also Published As
Publication number | Publication date |
---|---|
CN109769136B (en) | 2021-02-19 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
US8745262B2 (en) | Adaptive network content delivery system | |
EP2359536B1 (en) | Adaptive network content delivery system | |
US20110047120A1 (en) | Anticipated Query Generation and Processing in a Search Engine | |
Su et al. | Whatnext: A prediction system for web requests using n-gram sequence models | |
EP3111339B1 (en) | Systems and methods for managing content items having multiple resolutions | |
US8332775B2 (en) | Adaptive user feedback window | |
US20100318538A1 (en) | Predictive searching and associated cache management | |
US20100235569A1 (en) | Storage Optimization System | |
CN109769136A (en) | Online Video buffer memory management method and system based on instantaneous access rate value | |
CN110471939A (en) | Data access method, device, computer equipment and storage medium | |
NO313399B1 (en) | Procedure for searching and analyzing information in computer networks | |
WO2006104698A2 (en) | Streaming media content delivery system and method for delivering streaming content | |
CN106202581B (en) | A kind of data search method and device | |
CN102067617A (en) | Methods and apparatus for self-organized caching in a content delivery network | |
CN110275983A (en) | The search method and device of traffic monitoring data | |
CN108270860A (en) | The acquisition system and method for environmental quality online monitoring data | |
Akhtar et al. | Avic: a cache for adaptive bitrate video | |
Abdelkrim et al. | A hybrid regression model for video popularity-based cache replacement in content delivery networks | |
CN108538360A (en) | A kind of concentrating type body building management system and implementation method based on time series database | |
Zhang et al. | Autosight: Distributed edge caching in short video network | |
CN109769135A (en) | Online Video buffer memory management method and system based on joint-request rate | |
Yang et al. | Hybrid‐360: an adaptive bitrate algorithm for tile‐based 360 video streaming | |
CN109511009A (en) | A kind of video inline cache management method and system | |
Zhang et al. | Challenges and chances for the emerging short video network | |
CN108881943B (en) | Node cache replacement method based on elimination index in P2P streaming media on-demand system |
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 |