CN106131182B - Name a kind of cooperation caching method based on Popularity prediction in data network - Google Patents
Name a kind of cooperation caching method based on Popularity prediction in data network Download PDFInfo
- Publication number
- CN106131182B CN106131182B CN201610549740.XA CN201610549740A CN106131182B CN 106131182 B CN106131182 B CN 106131182B CN 201610549740 A CN201610549740 A CN 201610549740A CN 106131182 B CN106131182 B CN 106131182B
- Authority
- CN
- China
- Prior art keywords
- content
- popularity
- node
- cache
- network
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
Classifications
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
- H04L67/00—Network arrangements or protocols for supporting network services or applications
- H04L67/50—Network services
- H04L67/56—Provisioning of proxy services
- H04L67/568—Storing data temporarily at an intermediate stage, e.g. caching
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
- H04L67/00—Network arrangements or protocols for supporting network services or applications
- H04L67/50—Network services
- H04L67/56—Provisioning of proxy services
- H04L67/568—Storing data temporarily at an intermediate stage, e.g. caching
- H04L67/5681—Pre-fetching or pre-delivering data based on network characteristics
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
- H04L67/00—Network arrangements or protocols for supporting network services or applications
- H04L67/50—Network services
- H04L67/56—Provisioning of proxy services
- H04L67/568—Storing data temporarily at an intermediate stage, e.g. caching
- H04L67/5682—Policies or rules for updating, deleting or replacing the stored data
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
- H04L67/00—Network arrangements or protocols for supporting network services or applications
- H04L67/50—Network services
- H04L67/60—Scheduling or organising the servicing of application requests, e.g. requests for application data transmissions using the analysis and optimisation of the required network resources
- H04L67/63—Routing a service request depending on the request content or context
Abstract
A kind of cooperation caching method based on Popularity prediction in name data network is claimed in the present invention; name data network (Named Data Networking; NDN low with caching space utilization rate using net internal segment point data High redundancy is caused along path many places cache policy or in important node cache policy in being netted in) when storage content.There is employed herein the modes of " part collaboration caching ", and after carrying out the following Popularity prediction to content first, the spatial cache that each node separates best proportion serves as local cache space, to store the high content of popularity.Each remaining part of node stores the relatively low content of popularity by neighborhood cooperation mode.Optimal space division proportion is calculated by considering the interest packet hop count that node hit and server end request hit are passed through in a network.This method is compared with traditional cache policy, is increased spatial cache utilization rate in net, is reduced the caching amount of redundancy in network, improves net interior nodes cache hit rate, promote the performance of whole network.
Description
Technical field
The present invention is to belong to the technical field of future network to the research of cache policy in name data network.In particular to
Cooperation caching strategy into network between the forecasting mechanism of content popularit, neighborhood node, the division of nodal cache space proportion
Method.
Background technique
With the rapid development of Internet technology and application, " broadband ", " content " and " personalization " have become net
The theme of network development, people are increasingly strong for the demand of data content, the main body of network application gradually to content requests and
Information service evolution transfer.Predicted according to Cisco company, to internet in 2016 on the relevant flow of all the elements will occupy and be more than
97.5% share, the development that network architecture of the tradition centered on host is difficult to meet current network information service are wanted
It asks.For this purpose, a kind of future of the information centre's network (Information-Centric Networking, ICN) as revolution formula
Internet mentality of designing allows data content itself to become the main unit of network communication, by network communication mode from concern
" where " (address, server) is changed into concern " what ", i.e. the purpose and intention of user and application communication, becomes future
The important model of Internet design.Wherein, name data network (Named Data Networking, NDN) is as typical
ICN network architecture replaces IP in middle layer name data, and data transmission uses " publication-request-response " mode, directly
It connects and is routed with content name, realize point-to-multipoint efficient content distribution.
In the design of NDN, by the way of generally being cached in network, interest packet (interest packet) on the way
Storage hit content on all nodes of forward-path (on-path).In routing forwarding, when node receives interest packet, foundation
Content name is successively in content memorizer (Content Store, CS), interest the package list (Pending Interest to be processed
Table, PIT) and the middle progress matching inquiry of forwarding information list (Forwarding Information Base, FIB).Response
Data packet (data packet) carries request content, according to the record of node PIT list item, along request path carry out it is reversed by
Jump set is defeated.
For NDN network, reasonable content is placed and cache decision, is effective key factor for playing network performance.It closes
In the research of NDN caching, when proposing content center network at the beginning, based on its " caching tape swapping is wide " basic thought, just
Set about cache policy in research network.But since NDN spreads unchecked cache way (the Leave Cache whole on the way of formula
Everywhere, LCE), cause nodal cache content to tend to homogeneity, causes largely to cache redundancy.There is research to carry out in succession
Improvement proposes LCD, MCD, Prob, WAVE cache policy, and LCD is no longer deposited in all nodes when return path stores
Part is laid in, is only stored in the downstream individual node of hit node.MCD is compared with LCD, after downstream node caching backup,
The content of hit node is deleted, although greatly reducing the caching redundancy in net in this way, every request is primary, request
Content backup position will shift once in a network, increase the traffic overhead of network.WAVE considers single file and exists
It is divided into the feature that several chunk data blocks are stored in network, there are relevances between chunk data block, work as user
When making requests as unit of file, with increasing for request number of times, increase the caching in net according to index on the return path
Backup.The content for requesting multiple content, that is, popularity high quickly can be diffused into network edge in this way.Above-mentioned strategy is slow
When depositing, the considerations of shortage for content differentiation feature itself, the optimization storage of content cannot achieve.Therefore, in order to effectively send out
The advantage of content net internal segment point cache is waved, the design of efficient cache algorithm just becomes the critical issue for needing to solve.
Existing policies do not consider that position of the node in network topology structure influences caching bring, Chai etc.
People proposes Betw cache policy, i.e., when interest packet is hit in net interior nodes or server end, on return road after acquisition data packet
Caching backup is carried out at betweenness (betweenness centrality) highest node on diameter.This " less for
More " thought is to allow node few as far as possible to meet user's requests more as far as possible, but there is no in return for above-mentioned strategy
Hold the differentiation for carrying out popularity height, the nodal cache space for causing betweenness high is filled in a short time, high interior of popularity
Having may be replaced in the short time by the low content of subsequent popularity.When user requests again, net interior nodes cannot
Hit, needs route to server, increases network overhead.Based on this, Cui Xiandong et al. is carried out on Betwe policy grounds
It improves, it is contemplated that content replacement rate is excessively high in the node chosen in Betwe strategy, they joined replacement rate factor, chooses and is situated between
The high relatively low node of content replacement rate simultaneously of number is as cache location.Above-mentioned strategy has only carried out content not in a small number of positions
It stores with adding differentiation, the bad node space in remaining position will be idle, and cause the waste of spatial cache in netting.To asking
Ask frequency height and the low content of frequency to be not different when treating, distribution trail long low popularity content tied up need by
The space of the high popularity content of caching.
Summary of the invention
Problem to be solved by this invention is: existing cache policy one kind is that content express delivery is pushed away by universal cache way
It is sent to fringe node, leads to occur netting interior data cached High redundancy.One kind is to choose important node in network topology and concentrate to delay
Content is deposited, causes net memory space most of by vacant waste.In order to be maximally utilized to net interior nodes spatial cache, this
Invention carries out Popularity prediction to the content in network, and the strategy of proposition is both multiple standby to the Content Implementation that popularity in network is high
Part, and the Content Implementation cooperation caching that right pop degree is low, a small amount of backup is saved, so that net interior nodes spatial cache sufficiently quilt
It utilizes.It proposes a kind of high-frequency request that not only can satisfy user to high popularity content, but also can satisfy user in net internal segment
The low content of point hit popularity, is effectively reduced the load pressure of content source server, reduces the request delay of user, simultaneously
Improve a kind of request cooperation caching method based on Popularity prediction in the name data network of the hit rate of net interior nodes.This
The technical solution of invention is as follows:
Name a kind of cooperation caching method based on Popularity prediction in data network comprising following steps:
101, Popularity prediction: according to content in name data network accessed frequency and content to request end away from
From carrying out Popularity prediction to content, obtain the high content of the following popularity and the low content of popularity;The slow of each node
It deposits space and is divided into two parts x and (c-x), (c-x) is used to cache the content of high popularity, and x is used to cache the interior of low popularity
Hold;
102, all nodes all cache a backup to the content high to popularity in step 101 on the return path, and flow
The low content of the row degree node that comprehensive state is optimal on return path neighbor domain of node carries out cooperation caching, only caches single standby
Part.
Further, the number that the prediction of popularity is accessed in individual node in addition to considering content in step 101
Rcount, it is also contemplated that average number of hops H when being requested by a useraverage, cycle T is counted by setting to be accessed in record period
Number Rcount, recorded by interest packet every time request hit hop count, in this way consideration content node local popularity with
Cache location in the entire network is far and near, predicts the popularity of next cycle.
Further, it includes: to calculate in next popularity P for counting week that step 101, which carries out Popularity prediction to content,
(Tn+1):
Wherein P (Tn) indicate the current popularity of request content, HaverageIt is that data are accessed every time in one cycle
Average number of hops, RcountIndicate the requested number of the content in hit node, TnIndicate the hit content current count period, α is
Regulating constant, Δ P indicate the difference of the prediction popularity value and upper Cyclic epidemic degree of current period, Δ P=P (Tn+1)-P
(Tn), if Δ P > 0, show that content is in popularity ascent stage, each node caches on the return path when caching;If
Δ P < 0 shows that the content is in the popularity decline stage, on the return path with other node cooperation cachings.
Further, after prediction calculates content popularit, (1) is when the content k of request is in net interior nodes or server end
After hit, according to itself current popularity P (Tn) and the forwarding hop count that is attached to of interest packet and request content in hit node
Requested number calculates following fashion trend P (Tn+1), then show that Δ P's is positive and negative;
(2) it if all nodes of Δ P > 0, content k on the return path are stored, and is drawn no more than node storage
The c-x capacity separated;If nodal cache space c-x has expired, the cache replacement policy LRU in individual node is enabled;
(3) if data k is in the popularity decline phase, the corresponding node of maximum CoP (v) value on path is found out first, if having
Content k illustrates there has been the requested identical content of other nodes during routing, and data k is deposited according to cache policy
Storage directly abandons at this time in the point;If the not content with the neighbor node CoP (v) in its tablenValue compares, and finds out CoP
(v) maximum value is stored in corresponding node.
Further, there are three kinds of storage modes in the whole network: each node stores depositing for high popularity content
It stores up space (c-x);Each node stores the memory space x of low popularity content;Server stores the storage of initial data backup
Space.
Further, the comprehensive state of the low content of popularity node on return path neighbor domain of node is used in step 102
CoP (v) indicates the comprehensive state of node: setting in present node PIT each content name corresponding interface quantity as Dname1,
Dname2..., Dname n, then total request interface number is Dtotal(v)=Dname1+Dname2+…Dname n, current interface number can be straight
Reversed to reflect the requested rate of node, C (v) states the connectivity of each node, and formula is as follows:
CoP (v) value is bigger, indicates that the rate being accessed under the premise of the switching performance is good is slower.
It advantages of the present invention and has the beneficial effect that:
The present invention is distinguished the popularity height of content, is carried out in caching by the Accurate Prediction to content popularit
Different strategies is taken, the content that popularity can be made high is quickly pushed to fringe node, makes user's quickly hit in net
The content of request.The content for keeping popularity low occupies a small amount of memory space, can satisfy use while saving memory space
Family, even if cannot hit in net, can be routed to the content that server end finds request to its low-frequency request, additional
Expense is smaller to the impacts of whole network also within tolerance.The novel cache policy both can satisfy user couple
The high-frequency of high popularity content is requested, and can satisfy user's content low in net interior nodes hit popularity, is effectively reduced
The load pressure of content source server, reduces the request delay of user, while improving request in the hit rate of net interior nodes.
Detailed description of the invention
Fig. 1 is that the present invention provides preferred embodiment content popularit prediction flow chart;
Fig. 2 is cooperation caching strategic process figure in the present invention;
Fig. 3 is that node space division proportion flow chart is determined in the present invention.
Specific embodiment
Following will be combined with the drawings in the embodiments of the present invention, and technical solution in the embodiment of the present invention carries out clear, detailed
Carefully describe.Described embodiment is only a part of the embodiments of the present invention.
1. predicting flow chart as shown in Figure 1 for content popularit.Node request arriving rate generally uses Poisson distribution, and whole
Content distribution in a network meets zipf distribution.Namely sixteen distributions, 20% content can expire before the rank order in network
80% request amount in sufficient network.It is shown below:
C is constant
P (i) indicates the frequency that i corresponding content occurs, and i indicates the order in all the elements sequence, sorts according to content
The height of the frequency of occurrences, sequential arrangement from high to low.Then sum is the requested probability of content that i is ordered as in N number of content
Are as follows:
Wherein s is bigger, indicates that popular content is more concentrated.The probability distribution of k content before ranking are as follows:
Other than the content of server memory storage has long-term steady-state, the content in net is in high dynamic, 20% stream
The high content of row degree is also constantly substituting, and constantly has high popularity content to be degenerated to the low part of popularity, low interior of popularity
Appearance also has part to evolve to popularity height.Studies have shown that the popularity of dynamic content or temperature all meet from low ebb in network
To reach to peak value, low ebb is fallen back to after a period of time has passed.So we, which set, counts the period, a meter is added in interest packet
Numerical digit HcountWith time Tn, every forwarding is primary, and count is incremented, stops counting after hit and reaches HcountNamely request end to hit
Total hop count of node hits the requested primary also counting of the content in node and once, uses RcountIt indicates, when arrival cycle TnWhen stop
Only count.HcountIt can reflect that the backup location of the content in the entire network illustrates because hop count is bigger from request end indirectly
It is remoter, RcountIt is bigger, illustrate it is more to the content requests, fringe node do not have it is enough backup come meet request.If only
Consider that content can not predict that popularity of the corresponding content in whole network, general content more lean in the requested frequency of node
Nearly server end, requested number is opposite will be fewer, relies solely on the requested number of content in node at this time, mostly
Number content will be identified as the low content of popularity.HaverageIt is the average number of hops that data are accessed every time in one cycle,
Haverage*RcountVariation not only can reflect popularity of the content in net, also reflect the content in the network in the period
Expense is especially sensitiveer to the judgement of the node content popularity close to server.It calculates in next stream for counting week
Row degree P (Tn+1):
Wherein P (Tn) indicate the current popularity of request content, TnIndicate the hit content current count period, α is to adjust often
Number.Above formula is unfolded:
The ratio that counting closer to current period accounts for following Popularity prediction is higher, and influence is bigger, and distance is current
Period is remoter, influences on following popularity of content weaker.
Δ P=P (Tn+1)-P(Tn)
If Δ P > 0, show that content is in popularity ascent stage, each node caches on the return path when caching.
If Δ P < 0, show that the content is in the popularity decline stage.On the return path with other node cooperation cachings.
2. if Fig. 2 is cooperation caching strategic process figure.As Δ P < 0, content is in the popularity decline phase, using cooperation
Caching.The node of cooperation caching is chosen according to connectivity C (v) quality of each node.We calculate each node herein
Internuncial method is as follows:
Wherein π1(n) all nodes for being 1 apart from node n, π are indicated1(v) all nodes for being 1 apart from node v are indicated.β
It is the regulating constant belonged between 0 to 1, wherein
eijIndicate the side that point i is connected with point j, CD(ω) indicates all number of nodes being connected with ω, and C (v) is bigger, indicates v
The compactness of point is better.The number for not only considering the neighbor node of the radius r=2 of node in the above-mentioned methods, is also examined
Consider the Pair Analysis between neighbor node.The better node of compactness can be more accurately filtered out in this way.Only consider node connection
It is too fast that property may cause the good node convergence rate of performance, makes content quickly to these node aggregations, causes node congestion, in order to
Alleviate congestion, mitigate the load of node, we comprehensively consider node connectivity and node is accessed speed.Contain in PIT structure
{ content name;Request interface }, we set in present node PIT each content name corresponding interface quantity as Dname1,
Dname2..., Dname n, then total request interface number is Dtotal(v)=Dname1+Dname2+…Dname n.Current interface number can be straight
It is reversed to reflect the requested rate of node.So it is proposed that indicating the comprehensive state of node with CoP (v):
CoP (v) value is bigger, indicates that the rate being accessed under the premise of the switching performance is good is slower, if slow in the point
Deposit content, not only ensure that subsequent request can rapid requests to the point, but also be not in high replacement rate and request rate, while can
The load of node is effectively relieved.
Each node increases a new data structure in original data structure, and structural unit group includes { node name
Claim, contents list in corresponding CS, corresponding PIT interface number, for recording the cooperation that distance around it is node in the range of 2
Cache contents list and access interface number.If neighborhood node has duplicate content item, then only retain CoP (v) highest section
It is backed up at point.
Step:
(1) when the content k of request is after net interior nodes or server end are hit, according to itself current popularity P (Tn)
The forwarding hop count and request content being attached to interest packet calculate following fashion trend P in the hit requested number of node
(Tn+1), then show that Δ P's is positive and negative,
(2) it if all nodes of Δ P > 0, content k on the return path are stored, and is drawn no more than node storage
The c-x capacity separated;If nodal cache space c-x has expired, the cache replacement policy LRU in individual node is enabled.
(3) if data k is in the popularity decline phase, the corresponding node of maximum CoP (v) value on path is found out first, if having
Content k illustrates there has been the requested identical content of other nodes during routing, and data k is deposited according to cache policy
Storage directly abandons at this time in the point.If the not content with the neighbor node CoP (v) in its tablenValue compares, and finds out CoP
(v) maximum value is stored in corresponding node.
3. Fig. 3 is the process that nodal cache space divides.When storing the content of popularity upward period, in return path
(c-x) of upper each node division is partially cached to guarantee that more requests can most be hit fastly.Divide proper ratio
Spatial cache takes corresponding storage strategy that will become particularly significant, if (c-x) that divides is very few, is filled with, is replaced quickly
Change that rate is excessively high, if what is divided is too many, cooperation caching effect will be reduced, and whole cache policy is almost degenerated for LCE caching plan
Slightly, very limited to the promotion of network performance.Made herein using interest packet from the hop count that client is issued between hit node
To request cost, under the premise of node linear processing speed, hop count also reflects request delay indirectly.
In strategy proposed in this paper, there are three kinds of storage modes in content, therefore the content requested is in network in a network
There is also differences for the average number of hops of middle hit.Assuming that c-x partial hit of the interest packet in node, corresponding hop count is h1:
Interest packet is hit in the space cooperation caching x of node, and corresponding hop count is h2:
Assuming that interest packet is routed to server, and corresponding hop count is h without finding corresponding content in network3:
Constraint condition are as follows:
0 < xj< c
0 < c-xj< c
yik∈{0,1}
Wherein n represents number of nodes, and N represents chunk number of data blocks, rikIt indicates that each node i observes to ask content k
Rate is sought, Poisson distribution is obeyed.yikIt indicates whether request content hits, hits yik=1, otherwise yik=0.H is that hit is asked later
Seek the hop count in path.According to the different cache way of content, local cache content is respectively indicated with k ∈ c-x, k ∈ x, k ∈ s, is assisted
Make cache contents and server cache contents.If x takes 0, it is equivalent to absolutely not cooperation caching, is cache policy everywhere, leads
Node space is caused quickly to be filled with, popular content is also replaced quickly, and intermediate node hit rate is very in a network for interest packet
Low, hop count total in this way will necessarily be very big.When x takes c, the cache policy of network is that all spatial caches of all nodes are assisted
It caches, will lead to the backup of the high content of popularity in a network in this way can be reduced, high interior of the popularity of client request
The hop count for holding the content low with popularity is essentially the same, and hop count also can be very big.So making there are such a point of x ∈ (0, c)
Total hop count of network is minimum.Problem can be converted into described below:
Min(h1+h2+h3)
The request hop count of h expression whole network.Solution for x, we enumerate from 0 to c to x, find out most
The corresponding x of small h is used as x*.Be normalized is optimal collaboration depth value ρ:
The above embodiment is interpreted as being merely to illustrate the present invention rather than limit the scope of the invention.?
After the content for having read record of the invention, technical staff can be made various changes or modifications the present invention, these equivalent changes
Change and modification equally falls into the scope of the claims in the present invention.
Claims (4)
1. naming a kind of cooperation caching method based on Popularity prediction in data network, which comprises the following steps:
101, Popularity prediction: according to the distance of content in name data network accessed frequency and content to request end,
Popularity prediction is carried out to content, obtains the high content of the following popularity and the low content of popularity;The caching of each node
Space is divided into two parts x and (c-x), and (c-x) is used to cache the content of high popularity, and x is used to cache the content of low popularity;
It includes: to calculate in next popularity P (T for counting week that step 101, which carries out Popularity prediction to content,n+1):
Wherein P (Tn) indicate the current popularity of request content, HaverageIt is that data are accessed every time in one cycle to be averaged
Hop count, RcountIndicate the requested number of the content in hit node, TnIndicate the hit content current count period, α is to adjust
Constant, Δ P indicate the difference of the prediction popularity value and upper Cyclic epidemic degree of current period, Δ P=P (Tn+1)-P(Tn), if
Δ P > 0 shows that content is in popularity ascent stage, and each node caches on the return path when caching;If Δ P < 0,
Show that the content is in the popularity decline stage, on the return path with other node cooperation cachings;
102, all nodes all cache a backup to the content high to popularity in step 101 on the return path, and popularity
The low content node that comprehensive state is optimal on return path neighbor domain of node carries out cooperation caching, only caches single backup, has
Body includes: (1) when the content k of request is after net interior nodes or server end are hit, according to itself current popularity P (Tn) and
The forwarding hop count and request content that interest packet is attached to calculate following fashion trend P in the hit requested number of node
(Tn+1), then show that Δ P's is positive and negative;
(2) it if all nodes of Δ P > 0, content k on the return path are stored, and is marked off no more than node storage
C-x capacity;If nodal cache space c-x has expired, the cache replacement policy LRU in individual node is enabled;
(3) if data k is in the popularity decline phase, the corresponding node of maximum comprehensive state value CoP (v) on path is found out first,
If there is content k, illustrate there has been the requested identical content of other nodes during routing, and according to cache policy data
K is stored in the point, directly abandons at this time;If the not content with the neighbor node CoP (v) in its tablenValue compares, CoP
(v)nThe comprehensive state value for indicating n-th of neighbor node, finds out CoP (v) maximum value, is stored in corresponding node.
2. a kind of cooperation caching method based on Popularity prediction in name data network according to claim 1, special
Sign is that the prediction of popularity is in addition to considering content number R accessed in individual node in step 101count, it is also contemplated that quilt
Average number of hops H when user requestsaverage, cycle T is counted come number R accessed in record period by settingcount, pass through
Interest packet records the hop count of request hit every time, considers content in the local popularity and in the entire network of node in this way
Cache location is far and near, predicts the popularity of next cycle.
3. a kind of cooperation caching method based on Popularity prediction in name data network according to claim 1, special
Sign is that there are three kinds of storage modes in whole network: each node stores the memory space (c-x) of high popularity content;Often
A node stores the memory space x of low popularity content;Server stores the memory space of initial data backup.
4. a kind of cooperation caching method based on Popularity prediction in name data network according to claim 1, special
Sign is, in step 102 the comprehensive state value of the low content of popularity node on return path neighbor domain of node with CoP (v) come
It indicates, if each content name corresponding interface quantity is D in present node PITname1, Dname2..., Dname n, then total request
Number of ports is Dtotal(v)=Dname1+Dname2+…Dname n, current interface number can directly reflect the requested rate of node, C
(v) connectivity of each node is stated, formula is as follows:
CoP (v) value is bigger, indicates that the rate being accessed under the premise of the switching performance is good is slower.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201610549740.XA CN106131182B (en) | 2016-07-12 | 2016-07-12 | Name a kind of cooperation caching method based on Popularity prediction in data network |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201610549740.XA CN106131182B (en) | 2016-07-12 | 2016-07-12 | Name a kind of cooperation caching method based on Popularity prediction in data network |
Publications (2)
Publication Number | Publication Date |
---|---|
CN106131182A CN106131182A (en) | 2016-11-16 |
CN106131182B true CN106131182B (en) | 2019-04-09 |
Family
ID=57283817
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201610549740.XA Active CN106131182B (en) | 2016-07-12 | 2016-07-12 | Name a kind of cooperation caching method based on Popularity prediction in data network |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN106131182B (en) |
Families Citing this family (27)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN106686399B (en) * | 2016-12-22 | 2019-07-02 | 陕西尚品信息科技有限公司 | It is a kind of based on joint cache structure network in video cache method |
CN106982248B (en) * | 2017-03-01 | 2019-12-13 | 中国科学院深圳先进技术研究院 | caching method and device for content-centric network |
CN106899692A (en) * | 2017-03-17 | 2017-06-27 | 重庆邮电大学 | A kind of content center network node data buffer replacing method and device |
CN107404530B (en) * | 2017-07-27 | 2019-12-31 | 北京邮电大学 | Social network cooperation caching method and device based on user interest similarity |
CN107948247B (en) * | 2017-11-01 | 2020-04-10 | 西安交通大学 | Virtual cache channel cache management method of software defined network |
CN107911471B (en) * | 2017-12-01 | 2019-12-03 | 中国联合网络通信集团有限公司 | The distributed caching method and equipment of data |
CN108259929B (en) * | 2017-12-22 | 2020-03-06 | 北京交通大学 | Prediction and caching method for video active period mode |
CN108093056B (en) * | 2017-12-25 | 2020-11-10 | 重庆邮电大学 | Node cache replacement method in information center wireless network virtualization network |
CN108183867B (en) * | 2017-12-25 | 2021-08-10 | 重庆邮电大学 | Information center network node cache replacement method |
CN108307446B (en) * | 2018-01-05 | 2020-01-21 | 华中科技大学 | Wireless network edge cooperative caching system and method based on software definition |
CN108616923A (en) * | 2018-02-27 | 2018-10-02 | 南京邮电大学 | A kind of cooperation caching system based on mobile ad-hoc network |
CN108540569B (en) * | 2018-04-23 | 2020-01-24 | 燕东科技(广东)有限公司 | Software installation package replacement method and device and computer storage medium |
CN108566434B (en) * | 2018-05-03 | 2020-04-17 | 北京邮电大学 | Caching method and device based on popularity and node importance |
CN108900570B (en) * | 2018-05-30 | 2020-11-03 | 南京邮电大学 | Cache replacement method based on content value |
CN108900618A (en) * | 2018-07-04 | 2018-11-27 | 重庆邮电大学 | Content buffering method in a kind of information centre's network virtualization |
CN109413694B (en) * | 2018-09-10 | 2020-02-18 | 北京邮电大学 | Small cell caching method and device based on content popularity prediction |
CN109951875A (en) * | 2019-02-22 | 2019-06-28 | 广州大学 | Caching method, device, equipment and medium based on content popularit prediction |
CN110086857B (en) * | 2019-04-15 | 2022-01-14 | 哈尔滨工程大学 | Information center network cache deployment method based on cache node state and position |
CN110677190B (en) * | 2019-10-09 | 2021-06-22 | 大连大学 | Static processing and caching method for space-ground integrated intelligent network node |
CN110602652B (en) * | 2019-10-15 | 2021-07-27 | 中移信息技术有限公司 | Complaint model training method, and user complaint prediction method, device and equipment |
CN110602653A (en) * | 2019-10-30 | 2019-12-20 | 中国科学技术大学 | Pre-caching method based on track prediction |
CN111294394B (en) * | 2020-01-19 | 2022-09-27 | 扬州大学 | Self-adaptive caching strategy method based on complex network junction |
CN112039784B (en) * | 2020-07-27 | 2022-01-28 | 南京邮电大学 | Popularity-based routing expansion techniques in named data networks |
CN112052198B (en) * | 2020-08-11 | 2022-09-20 | 大连理工大学 | Hash route cooperative caching method based on node betweenness popularity under energy consumption monitoring platform |
CN111935301B (en) * | 2020-08-20 | 2021-10-01 | 北京邮电大学 | Method and device for determining cache placement decision information of named data network |
CN112565437B (en) * | 2020-12-07 | 2021-11-19 | 浙江大学 | Service caching method for cross-border service network |
CN113542349A (en) * | 2021-05-31 | 2021-10-22 | 北京邮电大学 | Cache replacement method and device for named data network |
Citations (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN102523256A (en) * | 2011-11-30 | 2012-06-27 | 华为技术有限公司 | Content management method, device and system |
CN104853384A (en) * | 2015-05-14 | 2015-08-19 | 南京邮电大学 | Popularity-based content cache method in fifth-generation mobile communication system |
CN105049254A (en) * | 2015-07-30 | 2015-11-11 | 重庆邮电大学 | Data caching substitution method based on content level and popularity in NDN/CCN |
CN105357246A (en) * | 2015-08-12 | 2016-02-24 | 北京交通大学 | Caching method and system based on information centre network |
CN105407055A (en) * | 2015-12-30 | 2016-03-16 | 深圳先进技术研究院 | Energy consumption control method of content center network |
-
2016
- 2016-07-12 CN CN201610549740.XA patent/CN106131182B/en active Active
Patent Citations (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN102523256A (en) * | 2011-11-30 | 2012-06-27 | 华为技术有限公司 | Content management method, device and system |
CN104853384A (en) * | 2015-05-14 | 2015-08-19 | 南京邮电大学 | Popularity-based content cache method in fifth-generation mobile communication system |
CN105049254A (en) * | 2015-07-30 | 2015-11-11 | 重庆邮电大学 | Data caching substitution method based on content level and popularity in NDN/CCN |
CN105357246A (en) * | 2015-08-12 | 2016-02-24 | 北京交通大学 | Caching method and system based on information centre network |
CN105407055A (en) * | 2015-12-30 | 2016-03-16 | 深圳先进技术研究院 | Energy consumption control method of content center network |
Non-Patent Citations (3)
Title |
---|
Improving Content-centric Networks Performance with Progressive, diversity-load driven Caching;Jason Min Wang, Brahim Bensaou;《Proceedings of the 2012 1st IEEE International Conference on Communications in China.Piscataway: IEEE》;20121231;85-90 |
信息中心网络缓存技术研究综述;张天魁 等;《北京邮电大学学报》;20160630;第39卷(第3期);1-15 |
智慧协同网络服务内容在传输路径上的缓存分配策略;冯博昊 等;《通信学报》;20160331;第37卷(第3期);129-138 |
Also Published As
Publication number | Publication date |
---|---|
CN106131182A (en) | 2016-11-16 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN106131182B (en) | Name a kind of cooperation caching method based on Popularity prediction in data network | |
CN105049254B (en) | Data buffer storage replacement method based on content rating and popularity in a kind of NDN/CCN | |
CN105022700B (en) | A kind of name data network cache management system and management method based on spatial cache division and content similarity | |
CN104811493B (en) | The virtual machine image storage system and read-write requests processing method of a kind of network aware | |
Carofiglio et al. | LAC: Introducing latency-aware caching in information-centric networks | |
Dutta et al. | Caching scheme for information‐centric networks with balanced content distribution | |
CN108900570B (en) | Cache replacement method based on content value | |
CN109905480B (en) | Probabilistic cache content placement method based on content centrality | |
An et al. | An in-network caching scheme based on energy efficiency for content-centric networks | |
CN108366089B (en) | CCN caching method based on content popularity and node importance | |
CN110365801A (en) | Based on the cooperation caching method of subregion in information centre's network | |
CN106899692A (en) | A kind of content center network node data buffer replacing method and device | |
CN105656788B (en) | CCN content buffering method based on popularity statistics | |
CN103905538A (en) | Neighbor cooperation cache replacement method in content center network | |
CN108769252A (en) | A kind of ICN network pre-cache methods based on request content relevance | |
Zhang et al. | A hybrid ICN cache coordination scheme based on role division between cache nodes | |
CN103905539A (en) | Optimal cache storing method based on popularity of content in content center network | |
Ramadan et al. | Big cache abstraction for cache networks | |
CN107302571B (en) | The routing of information centre's network and buffer memory management method based on drosophila algorithm | |
CN108093056A (en) | Information centre's wireless network virtualization nodes buffer replacing method | |
CN108183867A (en) | Information centre's network node buffer replacing method | |
Chao et al. | A game theoretic approach for energy-efficient in-network caching in content-centric networks | |
CN112822275B (en) | Lightweight caching strategy based on TOPSIS entropy weight method | |
Li et al. | HCaching: high-speed caching for information-centric networking | |
Yuan et al. | An optimal fair resource allocation strategy for a lightweight content-centric networking architecture |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
C06 | Publication | ||
PB01 | Publication | ||
C10 | Entry into substantive examination | ||
SE01 | Entry into force of request for substantive examination | ||
GR01 | Patent grant | ||
GR01 | Patent grant |