CN107155197A

CN107155197A - Distributed storage method, device and the electronic equipment cooperated based on multi-hop

Info

Publication number: CN107155197A
Application number: CN201710398205.3A
Authority: CN
Inventors: 王莉; 宋梅; 满毅; 吴华清; 魏青
Original assignee: Beijing University of Posts and Telecommunications
Current assignee: Beijing University of Posts and Telecommunications
Priority date: 2017-05-31
Filing date: 2017-05-31
Publication date: 2017-09-12
Anticipated expiration: 2037-05-31
Also published as: CN107155197B

Abstract

The embodiments of the invention provide the distributed storage method cooperated based on multi-hop, device and electronic equipment, the distributed storage method that should be cooperated based on multi-hop is included：When the node failure in first node set, predetermined number of thresholds is obtained；In other nodes in the current area in addition to the node in the first node set, all nodes of efficient communication can be carried out with the node in the first node set more than or equal to the number of thresholds by choosing, and be used as Section Point set；Each node and the correlation degree of the current area in the Section Point set, cache node is chosen in the Section Point set；In the range of efficient communication, the number of thresholds node in the first node set is chosen, the content fragment that the node of the recovery failure is cached to the cache node.With it, the effective percentage of data to be forwarded when being communicated using D2D can be improved, the reliability of communication is improved.

Description

Distributed storage method, device and the electronic equipment cooperated based on multi-hop

Technical field

The present invention relates to technical field of distributed memory, the distributed storage method more particularly to cooperated based on multi-hop, Device and electronic equipment.

Background technology

The limited user of memory capacity and computing capability is related to one by loT (Internet of Things, Internet of Things) Rise, can be in communication with each other between user by loT and realize the shared of resource.The data volume that equipment is collected into loT it is larger and With variation, herein under the premise of caching be considered as ensureing the effective solution of higher information transmission quality.In loT network racks In structure, caching information can by D2D, (Device-to-device, equipment be arrived between mobile terminal or other smart machines Equipment communication) it is transmitted, so that the communications infrastructure is alleviated, such as AP (Access point, access points) or BS The communication pressure of (Base Station, base station), and the utilization rate of frequency spectrum and energy can be effectively improved, reduce access delay And reduce return loss.

In actual loT communication process, the mobile device that the mobility of user may result in user's carrying leaves and worked as Preceding cell, can so cause the data failure being buffered in the mobile device, in order to ensure the validity of data in D2D communications, The method for using distributed storage in the prior art, data to be forwarded be grouped and random cache is to multiple mobile devices In (node), when node failure therein, randomly select a node to replace the node of failure, do not failed by other Node, recovers on data to the node randomly selected in failure node.But, replace losing using a node is randomly selected The node of effect, the stability difference of the node of selection and easily failure, can so cause the failure of data to be forwarded, existing distribution The Loss Rate of data to be forwarded is high in formula storage scheme, the poor reliability of communication.

The content of the invention

The purpose of the embodiment of the present invention is to provide a kind of distributed storage method, device and electronics cooperated based on multi-hop Equipment, to realize the effective percentage improved using data to be forwarded during D2D communications, improves the reliability of communication.Concrete technical scheme It is as follows：

For up to above-mentioned purpose, the embodiments of the invention provide a kind of distributed storage method cooperated based on multi-hop, including：

When the node failure in first node set, predetermined number of thresholds is obtained, wherein, the first node Collection is combined into the set for all nodes that content fragment is cached with the current area, and the content fragment includes data to be forwarded Partial content, the number of thresholds is the nodes needed for the content fragment that the node that recovery is failed is cached；

In other nodes in the current area in addition to the node in the first node set, selection can with it is big All nodes of efficient communication are carried out in or equal to node in the first node set of the number of thresholds, Section Point is used as Set；

Each node and the correlation degree of the current area in the Section Point set, in the second section Cache node is chosen in point set；

In the range of efficient communication, the number of thresholds node in the first node set is chosen, recovers described The content fragment that the node of failure is cached is to the cache node.

For up to above-mentioned purpose, the embodiments of the invention provide a kind of distributed storage devices cooperated based on multi-hop, including：

Number of thresholds acquisition module, for when the node failure in first node set, obtaining predetermined threshold value Quantity, wherein, the first node collection is combined into the set for all nodes that content fragment is cached with the current area, described Content fragment includes the partial content of data to be forwarded, the content fragment that the node that the number of thresholds is failed by recovery is cached Required nodes；

Node set determining module, for its in the current area in addition to the node in the first node set In his node, efficient communication can be carried out with the node in the first node set more than or equal to the number of thresholds by choosing All nodes, are used as Section Point set；

Cache node determining module, for each node in the Section Point set and the current area Correlation degree, cache node is chosen in the Section Point set；

Content fragment recovery module, in the range of efficient communication, choosing the threshold in the first node set It is worth quantity node, the content fragment that the node of the recovery failure is cached to the cache node.

For up to above-mentioned purpose, the embodiments of the invention provide a kind of electronic equipment, including processor, communication interface, storage Device and communication bus, wherein, processor, communication interface, memory completes mutual communication by bus；

Memory, for depositing computer program；

Processor, for performing during the program deposited on memory, realizes that the distribution cooperated based on multi-hop is deposited Method for storing.

It is described computer-readable the embodiments of the invention provide a kind of computer-readable recording medium for up to above-mentioned purpose Storage medium internal memory contains computer program, is cooperated when the computer program is executed by processor described in realization based on multi-hop Distributed storage method.

Distributed storage method, device and the electronic equipment provided in an embodiment of the present invention cooperated based on multi-hop, when first During node failure in node set, predetermined number of thresholds is obtained；The first node is removed in the current area In other nodes outside node in set, be chosen in the range of efficient communication, can with more than or equal to the number of thresholds First node set in node carry out efficient communication all nodes, be used as Section Point set；According to the second section Each node and the correlation degree of the current area in point set, cache node is chosen in the Section Point set； In the range of efficient communication, the number of thresholds node in the first node set is chosen, recovers the section of the failure Cached content fragment is put to the cache node.The mobility of user is taken into full account, when can improve using D2D communications The effective percentage of data to be forwarded, improves the reliability of communication.Certainly, implementing any product or method of the present invention must not necessarily need All advantages of the above are reached simultaneously.

Brief description of the drawings

In order to illustrate more clearly about the embodiment of the present invention or technical scheme of the prior art, below will be to embodiment or existing There is the accompanying drawing used required in technology description to be briefly described, it should be apparent that, drawings in the following description are only this Some embodiments of invention, for those of ordinary skill in the art, on the premise of not paying creative work, can be with Other accompanying drawings are obtained according to these accompanying drawings.

Fig. 1 is the schematic flow sheet of the ancestor node choosing method of the storage content burst of the embodiment of the present invention；

Fig. 2 is the schematic flow sheet of the distributed storage method cooperated based on multi-hop of the embodiment of the present invention；

The first of nodes of the Fig. 3 for base communication device in the communication scenes of the embodiment of the present invention with sending request of data Location diagram；

Second of nodes of the Fig. 4 for base communication device in the communication scenes of the embodiment of the present invention with sending request of data Location diagram；

Fig. 5 for the embodiment of the present invention communication scenes in base communication device with send request of data node the third Location diagram；

The 4th kind of nodes of the Fig. 6 for base communication device in the communication scenes of the embodiment of the present invention with sending request of data Location diagram；

Fig. 7 is the curve map of the data transfer probability of success of the embodiment of the present invention；

Fig. 8 is the curve map of the calculating time of the embodiment of the present invention；

Fig. 9 is the schematic diagram of the distributed storage devices cooperated based on multi-hop of the embodiment of the present invention；

Figure 10 is the schematic diagram of the electronic equipment of the embodiment of the present invention.

Embodiment

Below in conjunction with the accompanying drawing in the embodiment of the present invention, the technical scheme in the embodiment of the present invention is carried out clear, complete Site preparation is described, it is clear that described embodiment is only a part of embodiment of the invention, rather than whole embodiments.It is based on Embodiment in the present invention, it is every other that those of ordinary skill in the art are obtained under the premise of creative work is not made Embodiment, belongs to the scope of protection of the invention.

In the loT network architectures, if a nodal cache in current area has data to be forwarded, due to node motion, Energy consumption, memory space and other reasonses, cause the node to continue as the node in current area and provide data to be forwarded, then Claim the node failure.For example, the mobility of user's (user carries with node) is so that the contact between node becomes complicated, it is impossible to Enough ensure effective transmission of data to be forwarded, when the node for being cached with data to be forwarded leaves current area, the section can be caused Point failure, and the node of data to be forwarded is cached with due to reasons such as memory space and power consumptions, what is stored in the node treats When forwarding data to delete or lose, node failure is also resulted in, to solve this problem, the embodiment of the present invention is cached using redundancy Distributed storage method, after a node failure, pass through other nodes recover failure node in content.

Referring to Fig. 1, Fig. 1 illustrates for the flow of the ancestor node choosing method of the storage content burst of the embodiment of the present invention Figure, including：

S101, the maximum communication radius for obtaining and being cooperated according to the communication coverage of base station in current area, multi-hop, the One numerical value and the predetermined packet number and number of thresholds that there is probability, calculate data to be forwarded, wherein, the first numerical value For the quantity of the node of default caching data to be forwarded, exist probability be still within for the 8th node after the very first time it is current small Probability in area, the very first time is random time, and the 8th node is the arbitrary node in current area.

Current area in the embodiment of the present invention refers to deposit using the distribution cooperated based on multi-hop of the embodiment of the present invention The region that method for storing is communicated, current area can be identical with the communication range that communication infrastructure is covered, can also not Together, it is not specifically limited herein.First numerical value is default numerical value, is determined according to the mode that specific redundancy is cached.Example Such as, when carrying out distributed storage using EC (Erasure Codes, correcting and eleting codes), (at most there are 4 nodes in an EC 6/10 Failure, original document can still recover) distributed storage environment in, the first numerical value is set as 10 or bigger.

Have probability to be calculated according to the correlation degree and elapsed time of user and current area, root user with it is current The correlation degree positive correlation of cell, it is negatively correlated with the time.The correlation degree of one node is identified in the node and current area The communication conditions and social factors of other nodes in addition to the node, consider social factors when calculating has probability, obtain Presence probability it is more accurate.Multi-hop cooperative system calculates packet number and number of thresholds, taken into full account according to there is probability The mobility of user, the packet number and the reliability of number of thresholds calculated is higher.

Optionally, before S101, this method also includes：

Step one, the social networks in current area between user are obtained, it is determined that each user leaves the probability of current area.

During social factors between each user in not considering current area, the probability of user's turnover current area obeys Poisson Process, the probability that some node is left into current area is expressed as μ, and scene interior joint meets M/M/ ∞ queuing models.Work as consideration During to social factors between user, the social networks between node i (user i) and node j (user j) are expressed as S_ij, then may be used It is expressed as with the departure rate that user i is left to current area：Wherein, u is in current area The set of all nodes.S_ijSpan is 0≤S_ij≤ 1, associated and set according to the social activity between user i and user j, used Social activity between family i and user j associates closer, S_ijValue it is bigger.For example there is L good friend in user i address list, user j's Have K good friend in address list, user i with user's j address lists by J identical good friend, then the society between user i and user j Friendship relation can be expressed asWherein, K, L, J are positive integer.

Step 2, the probability of current area is left according to each user, and the presence for calculating the corresponding node of each user is general Rate.

Node i residence time T in current area_iObey exponential distributionThen Δ After time, the probability that node i is still present in current area can be expressed as：

Multi-hop cooperative system can be a kind of equipment, including：Processor, memory, communication interface and bus；Processor, Memory and communication interface are connected by bus and complete mutual communication；Memory storage executable program code；Processing The executable program code that device is stored by reading in memory runs program corresponding with executable program code, for Perform the distributed storage method cooperated based on multi-hop.

Multi-hop cooperative system can also be a kind of application program, for operationally performing the distribution cooperated based on multi-hop Storage method.

Multi-hop cooperative system can also be a kind of storage medium for being applied to terminal or the communications infrastructure, can for storing Code is performed, executable code is used to perform the distributed storage method cooperated based on multi-hop.

Multi-hop cooperative system can also be the said equipment, application program and storage medium in arbitrarily combine.

S102, according to packet number, data to be forwarded is grouped, the data to be forwarded after being grouped.

In embodiments of the present invention, to prevent the failure of data to be forwarded, using distributed storage mode.While in order to save The storage resource of about each node, multi-hop cooperative system does not use the method for directly preserving data to be forwarded copy, but according to pre- If algorithm, data to be forwarded is divided into multiple data slots.For example using EC codes (Erasure Coding, correcting and eleting codes) or its The computational methods of his redundant storage coding, data to be forwarded is grouped.

S103, the data to be forwarded after packet is encoded, and obtains multigroup content fragment.

Data to be forwarded after packet is encoded to by multi-hop cooperative system according to number of thresholds according to default coded system Multigroup content fragment, such as when being encoded using the algorithm of correcting and eleting codes, the is turned into by the data to be forwarded coding after packet The content fragment of the quantity of one numerical value.

S104, the node of the quantity of the first numerical value is chosen in current area, ancestor node is used as.

Multi-hop cooperative system uniform node for choosing the first numerical value quantity in current area, is used as cache contents burst Ancestor node.The node of the first numerical value quantity before surplus ranking can also be chosen according to node storage space surplus, or pressed According to multi-hop collaboration communication scope, the node of the first numerical value quantity before communication range ranking is chosen, ancestor node is used as.Certainly The node of the quantity of the first numerical value can be randomly selected, is not specifically limited herein, specific choosing method can be according to reality Signal intelligence is set.

S105, content fragment is cached in ancestor node respectively, obtains first node set.

Each content fragment is cached in each ancestor node by multi-hop cooperative system respectively at random, is cached with the institute of content fragment There is node to constitute first node set.Each node in first node set only caches a content fragment, if because Two of same data to be forwarded and the content fragment of the above are cached simultaneously in one node, can be to this when the node failure The reply of data to be forwarded is affected greatly.

In embodiments of the present invention, according to there is probability calculation packet number and number of thresholds and then compile data to be forwarded Code is multigroup content fragment, it is contemplated that the mobility of user, can improve the effective percentage of data to be forwarded when being communicated using D2D, Improve the reliability of communication.

Referring to Fig. 2, Fig. 2 is the schematic flow sheet of the distributed storage method cooperated based on multi-hop of the embodiment of the present invention, Including：

S201, when the node failure in first node set, obtains predetermined number of thresholds, wherein, first segment Point set is combined into the set for all nodes that content fragment is cached with current area, and content fragment includes the part of data to be forwarded Content, number of thresholds is the nodes needed for the content fragment that the node that recovery is failed is cached.

In actual applications, due to the mobility of user, the node (terminal) that user carries may leave current area, Or because node electricity exhausts, user actively closes D2D communication functions and node clears up the reasons such as memory space, section can be caused Point failure.When the node failure in first node set, multi-hop cooperative system obtains number of thresholds, to facilitate follow-up determination to lose The substitute node of the node of effect.Number of thresholds goes out according to there is probability calculation, the presence probability stamps time of a node The node is still present in the probability in current area after Δ, it is contemplated that the mobility of user, improves the reliability of communication.

In S202, other nodes in current area in addition to the node in first node set, efficient communication is chosen at In the range of, all nodes of efficient communication can be carried out with the node in the first node set more than or equal to number of thresholds, It is used as Section Point set.

The effective communication distance that D2D communicates between two nodes is communication between devices that are limited, equally being cooperated based on multi-hop Effective communication distance be also limited because with the increase (increase of data forwarding number of times) of hop count, the letter of data Order load, when extend to the factors such as noise and will become unacceptable.Multi-hop cooperative system is according to the communication of node in current area Situation, all sections of efficient communication can be carried out with the node for being cached with content fragment more than or equal to number of thresholds by choosing Point, is used as Section Point set.It ensure that each node in Section Point set can successfully recover on the node of failure Content fragment.

S203, the correlation degree of each node and current area in Section Point set, in Section Point set Middle selection cache node.

Multi-hop cooperative system can be chosen and current area correlation degree highest node in Section Point set, as Cache node.The correlation degree of one node identifies the communication of the node and other nodes in current area in addition to the node The correlation degree of situation and social factors, cache node and current area is high, the content fragment stored in cache node it is effective Property is higher, and the reliability communicated when being communicated using D2D is higher.

S204, in the range of efficient communication, chooses the number of thresholds node in first node set, recovery is left current The content fragment that the node of cell is cached is to cache node.

Multi-hop cooperative system, in the number of thresholds node in choosing first node set, can be in first segment point set Number of thresholds is randomly selected in conjunction can carry out the node of efficient communication with cache node.First segment point set can also be calculated Each node in conjunction transmits the transmission cost of data to cache node, the minimum section of number of thresholds transmission cost before choosing Point, to save transmission cost.

In embodiments of the present invention, number of thresholds comes out according to there is probability calculation, the presence probability of a node Identify after the very first time, the node remains in the probability of current area.The caching section finally determined according to number of thresholds Point replaces the node of failure, improves the effective percentage of data to be forwarded when being communicated using D2D, improves the reliability of communication.And It is determined that cache node replaces the node of failure, it is contemplated that each node in Section Point set associates journey with current area Degree, take into full account can node failure situation, it is determined that cache node node failure rate it is low, the successful probability of data transfer Height, improves the effective percentage of data to be forwarded when being communicated using D2D, improves the reliability of communication.

Optionally, each node in embodiments of the present invention in Section Point set and current area associate journey Degree, chooses cache node in Section Point set, including：

Step one, each node obtained respectively in Section Point set sets up efficient communication with the node in current area The number of link, is used as the centrad of the node in Section Point set.

Between effective range and node that multi-hop cooperative system can cooperate according to the multi-hop of each node in Section Point set The successful probability of data is transmitted, determines what each node in Section Point set can be set up with the node in current area The number of efficient communication link, by the number of efficient communication link, is used as the center of the node in corresponding Section Point set Degree.Efficient communication link refers to that each jump data transfer is satisfied by physical constraint and social constraint in the link, and physical constraint is Communication distance between finger joint point is no more than the maximum transmission distance that D2D communicates, and the signal noise ratio of data transfer have to be larger than Default snr threshold.Social activity constraint refers to that the probability of Successful transmissions data is greater than default transmission success rate between node Threshold value.

Step 2, chooses the maximum node of centrad in Section Point set, is used as cache node.

Directly D2D communications are carried out between node i and node j, it is necessary to meet the distance no more than D2D between node i and node j The maximum communication distance of communication, and the signal noise ratio of data transfer have to be larger than default snr threshold, consider simultaneously Social networks between the mobility of user and user, the probability of internodal communication link Successful transmissions data is greater than default biography Defeated success rate threshold value.Transmission.Wherein, snr threshold is any number for meeting the embodiment of the present invention, for example, be set as 13dB Or bigger, 14dB or bigger.Transmission success rate threshold value is any number for meeting the embodiment of the present invention, be for example set as 0.5 or It is bigger, 0.65 or bigger.

In embodiments of the present invention, it regard the number of efficient communication link as the center of the node in Section Point set Degree, centrad is higher, then the probability that current area interior joint downloads data to be forwarded from cache node is higher.Centrad is maximum Node as cache node, taken into full account the communication conditions in current area, improved the content stored in cache node The utilization rate of burst.

Step one, obtain and according to needed between two nodes that can directly carry out D2D device-to-device communications transmission Size, efficient communication time, transimission power and the channel situation of data, calculate transmission cost and are transmitted probability, wherein, pass The probability of data transfer can be completed between two nodes being totally lost into probability stamps current area.

Channel conditions include bandwidth, noise intensity, transimission power and the channel gain of channel.In view of the movement of user Property, the node that user carries with may there is a situation where more than D2D effective communication distances when directly carrying out D2D communications. It is that within the efficient communication time, the data that transmission needs to transmit can be completed between the two adjacent nodes that communicate to be transmitted probability Probability.In embodiments of the present invention by being transmitted probability, the possibility of node failure has been taken into full account, data can be increased The probability of transmission success.

Step 2, according to transmission cost and is transmitted probability, respectively using each node in first node set as one Individual summit, builds bipartite graph by another summit of each node in Section Point set respectively.

The structure of bipartite graph, specify that two end points of transmission path, is easy to calculate, facilitates the choosing in subsequent transmission path Select.

Step 3, according to probability and dynamic programming algorithm Dynamic Programming algorithms is transmitted, is determined respectively The transmission path between each node in each node and Section Point set in first node set.

If being transmitted probability more than or equal to default transmission success threshold value between adjacent node, this is transmitted Probability as the path between the adjacent node weights；If being transmitted probability less than transmission success threshold value between adjacent node, Then the weights in the path between the another adjacent node are 0, wherein, transmission success rate threshold value is the Arbitrary Digit for meeting the embodiment of the present invention Value, for example, be set as 0.5 or bigger, 0.55 or bigger.Using DP algorithm, (Dynamic Programming, Dynamic Programming is calculated Method), the path of weights product maximum between two nodes in bipartite graph is determined respectively, is used as the transmission path between the two nodes.

When being transmitted probability less than transmission success threshold value, it is 0 to make weights, it is possible to prevente effectively from because between respective nodes D2D transmission is transmitted that probability is low, and the situation for influenceing whole piece to communicate, and improves the successful probability of data transfer, it is ensured that The reliability of communication.

Step 4, the weights of every transmission paths are determined according to transmission cost respectively, pass through maximum matching algorithm Kuhn- Munkres algorithms obtain the matching of the minimum shortest path of weights sum.

The transmission cost and the transmission cost from BS downloading datas of D2D downloading datas are comprehensively utilized, transmission path is used as Weights, by KM algorithm Kuhn-Munkres algorithms, obtain the matching of the minimum shortest path of weights sum, i.e., with Section Point A node in set is the weights and minimum of the number of thresholds paths of end points.Consider to utilize D2D downloading datas simultaneously Transmission cost and the transmission cost from BS downloading datas, the consideration of transmission cost is more comprehensive, can save data transfer mistake Energy consumption in journey.

Step 5, according to the matching of the minimum shortest path of weights sum, chooses cache node.

The shortest path of the weights minimum on side is obtained using Kuhn-Munkres algorithms in the embodiment of the present invention, it is slow to choose Node is deposited, while transmission cost saving transmission energy consumption is minimized, the situation of node failure has been taken into full account, has improved communication Reliability.

Step one, obtain and according to needed between two nodes that can directly carry out D2D device-to-device communications transmission Size, efficient communication time, transimission power and the channel situation of data, calculate transmission cost and are transmitted probability.

In view of the mobility of user, the node that user carries with may exist super when directly carrying out D2D communications Cross the situation of D2D effective communication distances.It is the energy between the two adjacent nodes that communicate within the efficient communication time to be transmitted probability It is enough to complete the probability that transmission needs the data of transmission.In embodiments of the present invention by being transmitted probability, section has been taken into full account The possibility of point failure, can increase the successful probability of data transfer.

Step 2, according to there is probability and be transmitted probability, is determined between directly two nodes of progress D2D communications Side weights, wherein, the weights on side mark can directly carry out failing between two nodes of D2D communications completing transmission needing to pass The probability of defeated data.

During two adjacent nodes carry out data transmission, either the communication distance between the two nodes exceedes There is the node for leaving current area in effective communication distance, or the two nodes, data transfer fails.Therefore in this hair Considered simultaneously in bright embodiment and there is probability and be transmitted probability, taken into full account the mobility of user, improved data The success rate of transmission.

Step 3, according to dynamic programming algorithm Dynamic Programming algorithms, is determined in first node set respectively Each node and Section Point set in each node between shortest path.

Certainly each node and second in first node set can also be determined according to other routing algorithms herein The shortest path between each node in node set, for example, dijkstra's algorithm, Floyd algorithms and Shortest Path Faster Algorithm algorithms.

, according to shortest path, there is probability, be transmitted probability and transmission cost, first node is determined respectively in step 4 Mutual preference between each node in each node and Section Point set in set.

During data transfer, the node high for transmission success rate of the node preference in Section Point set, and The node low for transmission cost of node preference in first node set.Therefore, saved respectively according in every shortest path Probability is transmitted between the presence probability and node of point, to determine each node in Section Point set to first segment point set The preference of each node in conjunction；First segment point set is determined according to the transmission cost between every shortest path interior joint respectively The preference of each node in conjunction to each node in second set.Respectively according to first segment point set in the present invention is implemented The different demands of each node in conjunction and Section Point set, determine each node in first node set and Section Point set In each node between mutual preference, preference is more accurate.

Step 5, is accordingly to be regarded as number of thresholds node by each node in Section Point set, obtains the 3rd set of node Close.

During content fragment is repaired, substituting the node of failure node needs from number of thresholds first node set Node in downloading data, to recover content fragment, therefore, it is necessary to by second section point set before one is set up Each node in conjunction is accordingly to be regarded as number of thresholds node, obtains the 3rd node set.

Step 6, according to preference by preset algorithm, each node and first node set up in the 3rd node set The one between each node in set.

If the node in the 3rd node set is more than the node in first node set, need to carry out knot removal：With Machine deletes a node in Section Point set, and remaining each node in Section Point set is accordingly to be regarded as into number of thresholds Individual node, obtains the 3rd new node set.If the nodes in the 3rd new node set are still more than first node set In nodes, then repeat above-mentioned knot removal process until the 3rd node set in nodes be less than or equal to described first Nodes in node set.Using the method for such a knot removal, the node being matched in Section Point set ensure that Can be with the node in matching threshold quantity first node set.

Step 7, according to one, determines cache node.

According in the one established, a section for meeting matched rule is chosen in Section Point set Point, is used as cache node.The matched rule refers to that a node in one in Section Point set is corresponding Node in number of thresholds first node set.A node for meeting matched rule is chosen in Section Point set, As cache node, it can be saved to randomly select a node for meeting matched rule in Section Point set as caching Point；Cache node can also be used as to choose the minimum node of transmission cost sum in the node for meet matched rule.

In embodiments of the present invention, by setting up one, the communication resource is saved reducing transmission cost Meanwhile, improve data and download successful probability, improve the effective percentage of data to be forwarded when being communicated using D2D, improve communication Reliability.

Optionally, in embodiments of the present invention according to preference by preset algorithm, set up every in the 3rd node set The one between each node in individual node and first node set, including：

Distributed matcher algorithm Gale-Shapley algorithms or dynamical Matching Algorithm Roth and are passed through according to preference Vande Vate algorithms, a pair set up between each node in each node and first node set in the 3rd node set One corresponding relation.

In embodiments of the present invention, distributed matcher algorithm Gale-Shapley algorithms (Gai Er-sand Puli algorithm) is passed through Or dynamical Matching Algorithm Roth and Vande Vate algorithms, to set up one, give and correspond pass The determination method of system, to determine that cache node provides premise according to one.

Optionally, in embodiments of the present invention according to preference by Roth and Vande Vate algorithms, set up to be selected Each node in node set and each corresponding relation being cached between the node of content fragment are selected, including：

It is initial between each node in each node and first node set in step A, the 3rd node set of acquisition Corresponding relation.

Initial corresponding relation herein is any corresponding relation for meeting the embodiment of the present invention, and initial corresponding relation can be The man-to-man relation set up at random, or pass through Gale-Shapley algorithms, Kuhn-Munkres algorithms or other moneys The man-to-man relation that source Optimum Matching algorithm is set up, wherein, in initial corresponding relation, the node of individualism can be regarded as It is the node of the one-to-one matching of root node itself, its preference is mutually zero.

Optionally, before step A, this method also includes：

When the number of the node in the 3rd node set is more than the number of first node set interior joint, in the 3rd node The node of quantity identical with first node set interior joint number is chosen in set, is built at random with the node in first node set The vertical initial corresponding relation corresponded.

The number of first node set interior joint is represented with c (c is positive integer), number of thresholds is represented with d, in Section Point Arbitrarily chosen in setNode in (c/d is rounded downwards) individual Section Point set, and by thisIndividual node is accordingly to be regarded as D node, as preferred node, chooses preferred node, and randomly select remaining from the 3rd node set Individual node, sets up the corresponding relation with first node set interior joint.

In embodiments of the present invention, the specific method for setting up initial corresponding relation is given, it is ensured that one-to-one between node Correspondence.

Step B, according to preference, judges the node in initial corresponding relation to whether meeting preparatory condition, wherein, initially Node in corresponding relation is to have set up correspond two nodes between first node set and the 3rd node set, in advance If condition, which refers to the 3rd node, is higher than preference of the 3rd node to first node, and fourth node pair to the preference of fourth node The preference of 3rd node is higher than preference of the fourth node to the 5th node, and first node and fourth node are first segment point set Arbitrary node in conjunction, the 3rd node and the 5th node are the arbitrary node in the 3rd node set, and first node and the 3rd Node is the node of the same node centering in initial corresponding relation, and fourth node is with the 5th node in initial corresponding relation The node of same node centering, the node that the 3rd node and fourth node are constituted is to being referred to as destruction pair.

Step C, when all nodes in initial corresponding relation are to being unsatisfactory for preparatory condition, exports initial corresponding relation It is used as one.

Step D, when there is the node pair for meeting preparatory condition in initial corresponding relation, obtains the 3rd node and Section four Point, as destruction to set.

Step E, on the basis of initial corresponding relation, according to destruction to set, builds the second corresponding relation.

On the basis of initial corresponding relation, it will destroy to the destruction in set to as man-to-man node, by initially With destroying the node corresponded to interior joint in corresponding relation, matched according to preference：If in initial corresponding relation (can effectively it be led to the two nodes with mutually there is preference destroying the node corresponded interior joint Letter), then matching is the new node corresponded；If with destroying the section corresponded to interior joint in initial corresponding relation Preference (efficient communication can not be carried out between the two nodes) is not present between point, then regard the two nodes as single node (being considered as node itself to be matched with the node) obtains the second corresponding relation.In embodiments of the present invention, it is impossible to carry out effective Between the node of communication without matching, prevent because communication link fails and caused by data transmission fails.

Step F, according to preference, judges the node in the second corresponding relation to whether meeting preparatory condition.

Step G, when all nodes in the second corresponding relation are to being unsatisfactory for preparatory condition, exports the second corresponding relation It is used as one.

Step H, when there is the node pair for meeting preparatory condition in the second corresponding relation, judges full in the second corresponding relation The node of the node centering of sufficient preparatory condition, if be present in destruction in set.

Step I, meets the node in the node pair of preparatory condition in the second corresponding relation, is not present in destruction to collection When in conjunction, the wanton destruction pair in the second corresponding relation is chosen, as new node pair, the second corresponding relation is updated.

Step J, according to preference, judges the node in the second corresponding relation to whether meeting preparatory condition.

Step K, when node in the second corresponding relation is to being unsatisfactory for preparatory condition, exports the second corresponding relation conduct One.

Step L, when there is the node pair for meeting preparatory condition in the second corresponding relation, chooses in the second corresponding relation Wanton destruction pair, as new node pair, updates the second corresponding relation, return to step J is continued executing with.

Step M, the 6th node in the node pair of preparatory condition is met in the second corresponding relation and is present in destruction to collection When in conjunction, the 7th node of selection in the node pair of preparatory condition is met in the second corresponding relation, according to the 6th node and the 7th The node pair of node composition, updates the second corresponding relation, wherein, the 7th node is to the preference of the 6th node, more than destruction pair Node in set is to the preference of the 6th node, and the 6th node is the node in first node set, and the 7th node is the Node in three node sets；Or the 7th node be first node set in node, and the 6th node be the 3rd node set In node.

Step N, destruction is added to in set by the 7th node, and return to step F is continued executing with.

In embodiments of the present invention, give Roth and Vande Vate algorithms implements step, it is ensured that one Foundation to a corresponding relation, on the basis of data effectively transmission is ensured, can save energy.

Coded system in the present invention is any distributed coded system for meeting the embodiment of the present invention.In numerous distributions It is more conventional for correcting and eleting codes in formula coded system.According to different requirements, correcting and eleting codes have different coded systems (such as inner institute Code and fountain codes), but all there is identical coding parameter, (n, k can be expressed as_f, d_f), i.e. data to be forwarded f is divided into k_f (packet number) part is simultaneously encoded into n parts of the first numerical value, exists on n ancestor node, the content fragment having on each node is big Small is α_f.When there is node to ask data to be forwarded f, the node can be by connecting any k in n ancestor node_fIndividual node comes Obtain f, each ancestor node transmission size α_fContent.When there is ancestor node failure, it is necessary to select a new caching section Put and connect number of thresholds d_fThe f cached on individual existing ancestor node/cache node, the node to repair out failure content Burst, the content size of each ancestor node/cache node transmission is β_f.Below by taking correcting and eleting codes as an example, the present invention is elaborated The distributed storage method cooperated based on multi-hop, the specific embodiment encoded using correcting and eleting codes.

It is R in the positive radius of communication of a base communication device (such as AP or BS)_cellSingle honeycomb scene in, own The collection of node is combined intoAll collection for being cached with popular content node are combined into H,For all prevalences cached in current area Content (at least including data to be forwarded).It is H by the set expression of the n data to be forwarded f that are stored with node_f, request f's asks It is R to seek node set_f, and defineFor the set of all the elements having on h-th of cache node.

In practical communication scene, generally the time of node requests data is random, and obedience is independently distributed. The average arrival rate of node request is represented with W, W is made_jThe instantaneous time occurred for jth time request, i.e.,：W_jEqual to j times requesting interval The total time of (exponential distribution).Therefore, W_jIreland distribution is obeyed, i.e.,：Therefore, exist In delta time, the probability that j request occurs for node is：

The popularity of data obeys Zipf distributions, and therefore, the requested probability of data of popularity rankings i-th can be represented ForWherein, γ represents the distribution of the relative popularity of data.γ value is bigger to be represented Distribution is more concentrated, i.e., the number of times that more popular data are asked by node is more.ForThe number of middle element.

The failure node collection after delta time is defined on to be combined intoSo needing the set for the data repaired can be expressed asNow, to data to be forwarded f, the quantity that what is do not failed be cached with the node of f content fragment is expressed as x_f, then not The combination one of the node of the content fragment for being cached with f of failure hasKind may, then the content fragment for being cached with f does not lose The set of the node of effect can be expressed as：

The node of failure is replaced to cache, it is necessary to find a new node when being cached with the node failure of prevalence data Prevalence data.Exemplified by being cached with the data to be forwarded f node of content fragment, during f content fragment is repaired, need Choose d_fThe node of the individual content fragment for being cached with f, and it is β to download size from the node of each selection_fData.By energy Enough the alternate node set expression as new cache node isWherein,In node can be without caching any stream Row data, can also be cached with other prevalence datas in addition to f.

WithRepresent the probability in the content fragment reparation to node i of the f in failure node l, and

Wherein,

WithThe probability of node i downloading data from node h when being repaired in representing, and

Wherein,

The communication feature cooperated from multi-hop, at most content biography can be carried out when carrying out D2D communications by A jumps It is defeated, useWhen representing to repair f content fragment, the transmission path from node h to node i (includes transmitting-receiving node and relaying section Point), andWherein,When being expressed as repairing f content fragment, the set of the via node selected in the transmitting procedure from node h to node i, And work asWhen, represent that data are directly transferred to node i from node h；WhenWhen, represent number According to transmission path be：h→r₁→…→r_q→i。

In practical application scene, it should consider directly to carry out the physical communication bar of D2D communications between node i and node j Part, that is, need to meet the distance between node i and node j no more than d_max, and data transfer signal noise ratio have to be larger than it is pre- If snr thresholdAgain in view of user mobility and user between social networks, internodal communication link into The probability of work(transmission data is greater than default transmission success rate threshold θ_min.WhereinNumerical value to meet implementation of the present invention Any number of example, for example, be set as 13dB or bigger, 14dB or bigger.θ_minNumerical value be meet the embodiment of the present invention appoint Meaning numerical value, for example, be set as 0.5 or bigger, 0.65 or bigger.

Therefore, when repairing f content fragment, node i can be carried out to the collection table of the next-hop node of efficient communication It is shown as：

In repair process, node i can be expressed as from the successful probability of node h downloading datas：

I.e. when from least there is feasible transmission path node h to new cache node i,So thatThen

Therefore, node i is successfully connected to d_fIndividual node is come the probability for repairing f content fragment：

Distance is less than d between node i and node j_maxTime T_ijObedience parameter is τ_ijExponential distribution, its probability distribution letter Number isWhen node i transmits the data that size is M to node j, Successful transmissions probability is：

It can be calculated

Wherein, B is channel width,For noise intensity, P_{I, j}For the transimission power of node i to node j transmission link, g_{I, j}For the channel gain between node i and node j, Rayleigh distributed.

In content fragment repair process, each it is cached with what the node of f content fragment was transmitted to new cache node The size of content is β_f, can be obtained by above-mentioned formula (2), when content fragment is repaired, it is β that node i transmits size to node j_f's The successful probability of data is：

It is β that then node h, which is cooperated by multi-hop and transmits size to node i,_fThe successful probability of data is：

Wherein,

RepresentIn element number,RepresentIn j-th of element.

Content fragment repair process interior joint a is transmitted into data β to node b simultaneously_fTransmission cost be expressed as：

Wherein, P_{A, b}For node a to the transimission power of node b transmission link, g_a,bFor the letter between node a and node b Road gain, Rayleigh distributed, i.e. g_a,bObedience average is γ_a,bExponential distribution, g_a,b~Exp (γ_a,b).However, it is above-mentioned into ThisCalculating be difficult to calculating and obtain closed solutions, therefore, carry out problem relaxation using its lower boundary, i.e.,

WithWhen representing that node i repairs f content fragment, the d of selection_fThe set of node of the individual content fragment for being cached with f Close：

Therefore, d of the node i from selection_fBar, which repairs link, can successfully download size for β_fData probability be：

Then can the succeed probability of f content fragment of node i may finally be expressed as：

When node i downloads data to be forwarded f from the node of content fragment for being cached with f, it is necessary to choose k_fIndividual caching There is the node of f content fragment, and it is α to download size from the node of each selection_fData.With above-mentioned reparation f content The process of burst is similar, and it is α that node a downloads size from node b_fThe costs of data be：

The probability that node i downloads f content fragment from node h is expressed as：

Wherein,

WithWhen representing to download f, the transmission path from node h to node i, then

Its In,When being expressed as node i download f, the set of the via node selected in the transmitting procedure from node h to node i.

When downloading data to be forwarded f, the set expression for the next-hop node that node i can be carried out into efficient communication is：

Hold inside during load, node i can be expressed as from the successful probability of node h downloading datas：

At least there is a feasible transmission path i.e. between from node h to node i：

Make

Then

Node i is successfully connected to k_fIndividual node is come the probability for downloading f：

It can be obtained by above-mentioned formula (2), when data to be forwarded is downloaded, node i transmits data α to node j_fSuccessful probability For：

Therefore, it is α that node h, which is cooperated by multi-hop and transmits size to node i,_fThe successful probability of data be：

Wherein,RepresentIn element number,RepresentIn j-th of element.

WithWhen representing that node i downloads f, the k of selection_fThe node set of the individual content fragment for being cached with f：

Therefore, k of the node i from selection_fBar link can successfully download size for α_fThe probability of data be：

Then node i can successfully download f probability and may finally be expressed as：

It is an object of the invention to maximize data to be forwarded f to download successful probability and reparation f content fragment success Probability.When the popular content of caching all in view of current area, content fragment repairs successful average probability and is：

For new cache node.By popular content setIn popularity rankings r data be expressed as Popular content obeys Zipf distributions, thenIt is requested, and the probability for asking j times altogether by all nodes in delta time can be with It is expressed asWherein, Pr_Δ(x) represent to produce x in interval delta interior nodes The probability of secondary request,The ranking r requested possibility of content is represented, the average probability of success that popular content is downloaded can To be expressed as：

Therefore, objective optimization function can be expressed as：

It is R in the positive radius of communication of a base communication device (such as AP or BS) referring to Fig. 3 to Fig. 6_cellSingle honeybee In nest scene, the efficient communication scope of base communication device is regard as current area.For one apart from honeycomb center O₁Away from From for x (0≤x≤R_cell) node, wherein, position where node is O₂, the positive radius of communication of multi-hop cooperation is Ad_max, A The maximum effectively hop count, d cooperated for multi-hop_maxThe maximum effective communication distance communicated for D2D.Then the node multi-hop cooperates Communication range and the overlapping region of current area there is following situation.

As shown in figure 3, working as Ad_max≥R_cell, and Ad_max-R_cellDuring >=x, the area of overlapping region can be expressed as：Wherein, S_com(x) it is the area of overlapping region.

As shown in figure 4, working as Ad_max＜ R_cell, and R_cell-Ad_maxDuring >=x, the area of overlapping region can be expressed as：S_com (x)=π (Ad_max)², wherein, S_com(x) it is the area of overlapping region.

As shown in figure 5, working as Ad_max＜ R_cell, and R_cell-Ad_max＜ x≤R_cellWhen, the area of overlapping region can be represented For：

Wherein, S_com(x) it is the area of overlapping region.

As shown in fig. 6, working as R_cell＜ Ad_max, and Ad_max-R_cell＜ x≤R_cellWhen, the area of overlapping region can be represented For：

Wherein, S_com(x) it is the area of overlapping region.

In order to ensure the validity of data to be forwarded in scene, it is necessary to ensure the scope in node multi-hop collaboration communication with working as When the overlapping area of preceding cell is minimum, still can meet the individual node for being cached with content fragment of at least number of thresholds can be with the section Point carries out efficient communication.I.e. when node obeys random distribution in the scene, it is necessary to which meet has in overlapping rangeThe individual node for being cached with data to be forwarded.

In view of the social networks between user, the probability that each node leaves current area is differed, therefore for section Point i, after delta time, the number that what is existed in the range of its efficient communication be cached with the node of data to be forwarded can be expressed as：Wherein, p_h(Δ) is that delta time posterior nodal point h is still present in current area Probability, by above-mentioned formula (1) obtain.H_fTo be cached with the node of data to be forwarded in the range of now node i efficient communication Set.

OrderBy the first preset algorithm, obtained k_fWith d_f, wherein, R_fTo ask Ask the set of the node of data to be forwarded.

Optionally, the first preset algorithm includes：

The first step, initializes k'_f。

ByInitialize k'_f, wherein,

Second step, according to current k'_f, calculate k_fWith d_f。

When application MDS (Maximum Distance Separable) encoding scheme, k_f=k'_f, d_f=k_f。

In application MSR (Minimum Storage Regenerating) encoding scheme, whenWhen,d_f =2k_f- 1=n-1；WhenWhen, k_f=k'_f, d_f=2k'_f-1。

In application MBR (Minimum Bandwidth Regenerating) encoding scheme, k_f=k'_f, d_f=n-1.

3rd step, according to current k_fWith d_f, calculate α_fWith β_f。

When application MDS encoding schemes, α_f=β_f=f/k_f。

In application MSR encoding schemes, whend_f≥2k_fWhen -1, α_f=f/k_f,

In application MBR encoding schemes,

4th step, it is determined that in the range of efficient communication, can succeed and data to be forwarded f content fragment is transmitted to node i Node set.

Content download procedure (download in data to be forwarded f), can be with node i in A to any one requesting node i In jump (comprising it is all less than or equal to A jump situations) Successful transmissions f content fragment node number (so that Teamworker number) be expressed as k_{F, i}, therefore after delta time, the distance apart from honeycomb center exists for x requesting node i The node number of the content fragment for being cached with f in the connection of reliable communication range interior energy is：

Therefore, k'_fThe minimum of requesting node equal to fUpdateIf (thering is request to save Point can not be attached with the node of any content fragment for being cached with f, then make the requesting node download f from base station, do not consider Its)。

5th step, if the k' after updating_fLess than k_f, then the rear k' of renewal is made_fFor initialization k'_f, return to second step and continue to hold OK, if update after k'_fMore than or equal to k_f, then k is exported_fAnd d_f。

According to the k of output_fAnd d_f, data to be forwarded f is encoded into n group content fragments, and distributed storage using correcting and eleting codes Into n node.

In embodiments of the present invention, k is being calculated_fAnd d_fWhen, it is contemplated that the social connection between the mobility of node and each node System, can reduce request f node before f has been downloaded, be cached with the probability of the node failure of f content fragment.

When being cached with the f node failure of content fragment, it is necessary to carry out content fragment reparation, that is, utilize new caching section Put to replace the node of the failure.In order to ensure on data to be forwarded f stability and reliability, the cache node newly selected not Can be stored with f content fragment, i.e.,：The new cache node i of selection must is fulfilled forWherein,For all sections The set of point,For the set of all nodes of the content fragment that is cached with f.Because, if f multiple content fragments delay Exist on same node, if the node failure, stability and reliability to f are affected greatly.

In addition, on the premise of allowing to carry out multi-hop data transmission, when at most allowing to carry out A jump data transfers, new choosing The cache node selected must is fulfilled for：In the presence of at least d_fThe individual node of f content fragment that is cached with can be within A jumps by active chain Road is by the data transfer in f to the new cache node.

Therefore, it is to be selected when the content fragment for repairing the f on a failure node, the collection of new node be combined into：

WhenIn when there is multiple nodes, new caching section can be selected according to the centrad of current area interior joint Point, wherein, in the case that the centrad of node is considers that physical factor is associated with social activity at the same time, the node can be with other nodes Set up the number of efficient communication link.

The maximum node of middle Selection Center degree is as new cache node, i.e.,：

When being cached with the node of content fragment of data to be forwarded between new cache node, there is a plurality of feasible transmission During path, the minimum path of transmission cost will be selected.

In embodiments of the present invention, it regard the number of efficient communication link as the content fragment for being cached with data to be forwarded The centrad of node, the probability that the higher current area interior joint of centrad downloads data to be forwarded from cache node is higher.By in The maximum node of heart degree has taken into full account the communication conditions in current area, has improved in cache node and deposit as cache node The utilization rate of the content fragment of storage.

, can be using centralization matching in embodiments of the present invention in order to solve the select permeability in multi-hop cooperation transmission path Algorithm is optimized, and is matched using bipartite graph.Because content download procedure and content fragment repair process are the power on side The definition of value and the number difference for carrying out data transmission the side that the match is successful using D2D, other mathematical procedure all sames, therefore By taking the process that content fragment is repaired as an example.

Step one, bipartite graph is built.

Two node sets of bipartite graph are the first segment for all nodes that content fragment is cached with current area respectively Point setAnd all sections of efficient communication can be carried out with the node in the first node set more than or equal to number of thresholds The Section Point set of point(In arbitrary node i can at least andIn number of thresholds node A jump in by having Effect link is communicatively coupled, i.e.,)。

Step 2, weights are assigned to bipartite graph side.

In the bipartite graph of structure, the weights on side are that node i passes through multi-hop cooperation transmission data between node i and node j To node j transmission cost w_ij。

To determine w_ij, first have to determine the transmission path of the multi-hop cooperation between node i and node j.Construct a weighted graph , comprising all nodes in current area, there is side, the weights on the side between node a and node b between any two node in G It is expressed as Z_ab, order(For when content fragment is repaired, node a transmits the data success in f to node b Probability), if the weights on the side between node a and node bThen make the weights on side between node a and node b z_ab=0.Then according to the weights on side between any userNode i is found using Dynamic Programming algorithms Optimal path between node jSo thatIn all side weights product it is maximum.

It is determined that after transmission path between bipartite graph two ends node i and node j, by node i and the weights on node j side It is changed toWherein,To be passed through from node-to-node i The probability of success of data in multi-hop cooperation transmission f is crossed,Cooperated for node h by D2D multi-hops and transmit number in f to node i According to transmission cost,Represent the cost of node i data from BS acquisitions f.So far, bipartite graph builds and finished.

Step 3, the matching of minimum weights sum is obtained using KM algorithms (Kuhn-Munkres algorithms).

The matching for obtaining minimum weights sum is correspondingIn node, be used as new cache node.

In embodiments of the present invention, according to the optimal path between the successful determine the probability node of data, it is ensured that communication It is reachable, using total transmission cost as the weights on side between bipartite graph two-end-point, of minimum weights sum is obtained using KM algorithms Match somebody with somebody, and then determine new cache node, transmission cost can be reduced.

In embodiments of the present invention in order to solve the select permeability in multi-hop cooperation transmission path, distributed matcher is additionally provided Algorithm.

Step one, matching for first node set interior joint and the 3rd node set interior joint is built.

For content fragment repair process, it is by first node set expressionDue to cache node to be chosen Repair content repair text burst needs to connect d_fThe node of the individual content fragment for being cached with f, therefore, cache node to be chosen is with being cached with Matching between the node of f content fragment is a 1-d_fMatching problem.So, by each section in Section Point set Point expands to d_fIt is individual, obtain the 3rd node setSo that the matching is changed into a 1-1 matching.Accordingly, load is held inside During,For each request of data node is accordingly to be regarded as into k_fThe set of individual node,To be cached with the node of content fragment Set.For content fragment repair process, or for content download procedure, the selection in multi-hop cooperation transmission path It can be equivalent to be based onWithBetween 1-1 matching problems.

We define an adaptation function Φ：WithSo that

It is setIn i-th of element,It is setIn j-th of element, Φ is a man-to-man stable matching.

For arbitraryNodeWant to the effectively reliable success rate for obtaining data, i.e. data transfer Will height.For arbitraryNodeWant to save energy as far as possible, i.e. transmission cost is low.

, will be from exemplified by repairing the f process of content fragmentArriveComplete multi-hop transmission path representation beThen,Obtained according to following steps：

By any two node in current area, the weights on side are expressed as between node i and node jWherein p_i(Δ) is that node i is still present in current area after delta time In probability,The transmission success probability of data in f is transmitted to node j for content fragment repair process interior joint i.

According to the weights z between node_ij, obtained using Dynamic Programming algorithmsArriveBetween optimal road FootpathSo that the weights sum on each jump side is minimum in path.

Obtain fromArriveComplete transmission pathAfterwards,It is rightPreference can weigh and be：

RepresentIn element number,Represent pathIn i-th of user. By by matrixRow arrange in descending order, obtainIn node pairIn node preference list.Accordingly, It is rightFancy grade be represented by：

For repair f content fragment when, node a to node b transmit f in data when one jump transmission cost.Equally By by matrixRow arrange in descending order, obtainIn node pairIn node preference list.

ObtainingInterior joint andAfter the mutual preference of interior joint, Gale-Shapley algorithms can be passed through (the Gale-Shapley algorithm, Gai Er-sand Puli algorithm) is obtainedInterior joint andInterior joint it is one-to-one steady Fixed matching, is used as one.

Certainly obtainingInterior joint andAfter the mutual preference of interior joint, Roth and Vande can be passed through Vate algorithms are obtainedInterior joint andThe one-to-one stable matching of interior joint, is used as one.

Using Roth and Vande Vate algorithms, first have to blocking pair destructions to being defined.For SetIn node i and setIn node j for, if φ (i) ≠ j (φ (i) represent node i current matching pair As), and relative to φ (i), node i more preference node j, relative to φ (j) (φ (j) represents the object of node j current matchings), Node j more preference node is, then node i and node j just constitute destruction pair.

In Roth and Vande Vate algorithms, arbitrary node can be matched with itself, and be to the preference of itself 0.If a matchingIn be not present destruction pair, then the matchingIt is stable.

Will setAnd setBetween initial matching be expressed asm_i, n_jSet is represented respectivelyIn i-th section Point and setIn j-th of node.

Step one, judgeWhether stablize.

Make q=0.

IfIt is not stable, it is determined thatIn a pair of destruction to (m_q,n_q), then based on the destruction to constituting new Match somebody with somebodySo thatMake A (q)=(m_q, n_q)。

IfIt is stable, then outputAlgorithm terminates.

Step 2, whenIt is not stable, then existsIt is middle to find all destructions to (m_q+1,n_q+1)

Step 3, ifIn to it is all destruction to (m_q+1,n_q+1) for, there is a destruction to (m_q+1,n_q+1), its Middle m_q+1It is comprised in A (q), then from all destructions to (m, n_q+1) in, select n_q+1, the object for making its most preference is m_q+1, and destruction is constituted to (m_q+1,n_q+1), obtain new matching setSo thatMake A (q+1) =A (q) ∪ { n_q+1}；

If 1) existedIn, m_q+1It is not matched, then would not be included in A (q+1)In destruction pair

If 2) existedIn, m_q+1It is matched, then in A (q+1), it is understood that there may beDestruction to (m_q+2, n_q+2), and meetAnd m is included in A (q+1)_q+2And n_q+2, then change centering selection from all break n_q+2So that m_q+2It is n_q+2The partner of most preference.By m_q+2And n_q+2Matching, obtains new matching set

The process of repeat step three, until being matchedR ＞ q so thatDestruction in the absence of among set A (r), SoAs stable matching.

Step 4, ifIn to it is all destruction to (m_q+1,n_q+1) for, m_q+1All it is not included in A (q), but deposits In a destruction to (m_q+1,n_q+1), wherein n_q+1It is comprised among A (q), then exchange M and N role, and step 3 again In step.

Step 5, ifIn it is all destruction to (m_q+1,n_q+1) in element be not present A (q) among, then selection Wanton destruction is to (m_q+1,n_q+1) constituteAnd define A (q+1)=A (q) ∪ { m_q+1, n_q+1}.Therefore A is included in A (q+1) (q), and do not includeIn destruction pair.This process is repeated, untilIn destruction pair is not present, that is, reach stable Match somebody with somebody.

In embodiments of the present invention, either by the KM algorithms of centralization, or distributed Gale- is utilized Shapley algorithms and Roth and Vande Vate algorithms, can obtain man-to-man stable matching.Using KM algorithms During with Gale-Shapley algorithms, if wherein any one node failure, it is necessary to match again to all nodes again, Using the unstable matching pair of each iteration elimination of Roth and Vande Vate algorithms, stable matching set is most obtained at last, Compared to KM algorithms and Gale-Shapley algorithms, it is possible to reduce the complexity of calculating.

Referring to Fig. 7, Fig. 7 is the curve map of the data transfer probability of success of the embodiment of the present invention.Wherein, curve 701 is to adopt With the curve of KM algorithms, curve 702 is the curve using Gale-Shapley algorithms, and curve 703 is using Roth and The curve of Vande Vate algorithms.Transverse axis represents run time, and unit is the second, and the longitudinal axis is the successful probability of data transfer.Understand, Slightly above used using data transmission success during KM algorithms and use Gale-Shapley algorithms and Roth and Vande Vate Data transmission success during algorithm.

Referring to Fig. 8, Fig. 8 is the curve map of the calculating time of the embodiment of the present invention.Wherein, curve 801 is to use KM algorithms Curve, curve 802 is the curve using Gale-Shapley algorithms, and curve 803 is to be calculated using Roth and Vande Vate The curve of method.Transverse axis is that run time unit is the second, and the longitudinal axis is the calculating time, and unit is the second.With the increase of running time, three The increase for calculating time stage of algorithm is planted, because increasing over time, the number of the node of current area is left Mesh starts to increase, and computing degree becomes complicated, the number regional stability of the node per second for leaving current area after a period of time, The calculating time also tends towards stability relatively.As shown in Figure 8, the calculating time of Roth and Vande Vate algorithms is significantly less than KM The calculating time of algorithm and Gale-Shapley algorithms, computation complexity is lower.

Referring to Fig. 9, Fig. 9 is the schematic diagram of the distributed storage devices cooperated based on multi-hop of the embodiment of the present invention, including：

Number of thresholds acquisition module 901, for when the node failure in first node set, obtaining predetermined threshold It is worth quantity, wherein, first node collection is combined into the set for all nodes that content fragment is cached with current area, content fragment bag Partial content containing data to be forwarded, number of thresholds is the nodes needed for the content fragment that the node that recovery is failed is cached.

Node set determining module 902, for other sections in current area in addition to the node in first node set In point, all sections of efficient communication can be carried out with the node in the first node set more than or equal to number of thresholds by choosing Point, is used as Section Point set.

Cache node determining module 903, for associating for each node in Section Point set and current area Degree, cache node is chosen in Section Point set.

Content fragment recovery module 904, in the range of efficient communication, choosing the number of thresholds in first node set Individual node, the content fragment that the node that recovery is failed is cached to cache node.

In embodiments of the present invention, number of thresholds comes out according to there is probability calculation, the presence probability of a node Identify after the very first time, the node remains in the probability of current area.The caching section finally determined according to number of thresholds Point replaces the node of failure, improves the effective percentage of data to be forwarded when being communicated using D2D, improves the reliability of communication.And It is determined that cache node replaces the node of failure, it is contemplated that each node in Section Point set associates journey with current area Degree, take into full account can node failure situation, it is determined that cache node crash rate it is low, data transfer successful probability height is improved The effective percentage of data to be forwarded when being communicated using D2D, improves the reliability of communication.

Optionally, in embodiments of the present invention, the distributed storage devices that should be cooperated based on multi-hop are also included：

Numerical Simulation Module, for obtaining and being cooperated most according to the communication coverage of base station, multi-hop in current area Big communication radius, the first numerical value and the predetermined packet number and number of thresholds that there is probability, calculate data to be forwarded, Wherein, the first numerical value is the quantity of the node of default caching data to be forwarded, and there is probability is the 8th node after the very first time The probability in current area is still within, the very first time is random time, and the 8th node is the arbitrary node in current area.

Packet module, it is to be forwarded after being grouped for according to packet number, data to be forwarded to be grouped Data.

Data coding module, for according to number of thresholds, the data to be forwarded after packet to be encoded, is obtained in multigroup Hold burst.

Ancestor node chooses module, and the node of the quantity for choosing the first numerical value in current area is used as mesomere Point.

Cached configuration module, for content fragment to be cached in ancestor node respectively.

Probability determination module is left, for obtaining the social networks in current area between user, it is determined that each user leaves The probability of current area.

There is probability evaluation entity, the probability for leaving current area according to each user calculates each user's correspondence Node presence probability.

In embodiments of the present invention, the circular that there is probability is given, the mobility for having omited user is examined, calculated The presence probability gone out is accurate, can improve the effective percentage of data to be forwarded when being communicated using D2D, improves the reliability of communication.

Optionally, in embodiments of the present invention, cache node determining module 903, including：

Centrad calculating sub module, for obtaining each node in Section Point set and the section in current area respectively Point sets up the number of efficient communication link, is used as the centrad of the node in Section Point set.

First cache node chooses submodule, chooses the maximum node of centrad in Section Point set, is saved as caching Point.

In embodiments of the present invention, it regard the number of efficient communication link as the center of the node in Section Point set Degree, the probability that the higher current area interior joint of centrad downloads data to be forwarded from cache node is higher.Centrad is maximum Node has taken into full account the communication conditions in current area as cache node, improves the content point stored in cache node The utilization rate of piece.

First cost calculation submodule, for obtaining and according to can directly carry out two of D2D device-to-device communications Size, efficient communication time, transimission power and the channel situation of the data of transmission are needed between node, transmission cost and transmission is calculated Probability is completed, wherein, it can complete the probability of data transfer between two nodes being transmitted in probability stamps current area.

Bipartite graph builds submodule, for according to transmission cost and being transmitted probability, respectively with first node set Each node be summit, the structure bipartite graph by another summit of each node in Section Point set respectively.

First path calculating sub module, probability and dynamic programming algorithm Dynamic are transmitted for basis Programming algorithms, respectively between each node in each node and Section Point set in determination first node set Transmission path.

KM algorithm submodules, the weights for determining every transmission paths respectively according to transmission cost pass through maximum matching Algorithm Kuhn-Munkres algorithms obtain the matching of the minimum shortest path of weights sum.

Second cache node chooses submodule, for the matching according to the minimum shortest path of weights sum, chooses caching Node.

Second cost calculation submodule, for obtaining and according to can directly carry out two of D2D device-to-device communications Size, efficient communication time, transimission power and the channel situation of the data of transmission are needed between node, transmission cost and transmission is calculated Complete probability.

Weight computing submodule, for according to there is probability and be transmitted probability, being determined to directly carry out D2D communications Two nodes between side weights, wherein, the weights on side mark can directly carry out having failed between two nodes of D2D communications The probability for the data for needing to transmit into transmission.

Second path computing submodule, it is true respectively for according to dynamic programming algorithm Dynamic Programming algorithms Determine the shortest path between each node in each node and Section Point set in first node set.

Preference determination sub-module, for according to shortest path, there is probability, be transmitted probability and transmission cost, point Mutual preference between each node in each node and Section Point set that Que Ding be in first node set.

3rd node set acquisition submodule, for each node in Section Point set to be accordingly to be regarded as into number of thresholds Node, obtains the 3rd node set.

One setting up submodule, for, by preset algorithm, setting up the 3rd node set according to preference In each node and first node set in each node between one.

3rd cache node chooses submodule, for according to one, determining cache node.

Optionally, in embodiments of the present invention, according to preference by preset algorithm, one sets up submodule Block, specifically for：

Optionally, in embodiments of the present invention, one setting up submodule, including：

Initial corresponding relation acquiring unit, for obtaining in each node and first node set in the 3rd node set Each node between initial corresponding relation.

The identifying unit of preparatory condition first, for according to preference, judging the node in initial corresponding relation to whether full Sufficient preparatory condition, wherein, node in initial corresponding relation between first node set and the 3rd node set to have set up one Two nodes corresponding to one, preparatory condition refers to the 3rd node and is higher than the 3rd node to first node to the preference of fourth node Preference, and fourth node to the preference of the 3rd node higher than fourth node to the preference of the 5th node, first node And fourth node is the arbitrary node in first node set, the 3rd node and the 5th node are any in the 3rd node set Node, and first node and the 3rd node are the node of the same node centering in initial corresponding relation, fourth node and the 5th Node is the node of the same node centering in initial corresponding relation, and the node that the 3rd node and fourth node are constituted is to being referred to as Destruction pair.

First output unit, during for all nodes in initial corresponding relation to being unsatisfactory for preparatory condition, output Initial corresponding relation is used as one.

It is broken to change to determining unit, for existing in initial corresponding relation during the node pair for meeting preparatory condition, obtain the Three nodes and fourth node, as destruction to set.

Second corresponding relation determining unit, on the basis of initial corresponding relation, according to destruction to set, builds the Two corresponding relations.

The identifying unit of preparatory condition second, for according to preference, judging the node in the second corresponding relation to whether full Sufficient preparatory condition.

Second output unit, during for all nodes in the second corresponding relation to being unsatisfactory for preparatory condition, output Second corresponding relation is used as one.

It is broken to change to identifying unit, for existing in the second corresponding relation during the node pair for meeting preparatory condition, judge the The node of the node centering of preparatory condition is met in two corresponding relations, if be present in destruction in set.

Second the first updating block of corresponding relation, for being met in the second corresponding relation in the node pair of preparatory condition Node, when being not present in destruction in set, chooses the wanton destruction pair in the second corresponding relation, as new node pair, Update the second corresponding relation.

The identifying unit of preparatory condition the 3rd, for according to preference, judging the node in the second corresponding relation to whether full Sufficient preparatory condition.

3rd output unit, during for the node in the second corresponding relation to being unsatisfactory for preparatory condition, output second Corresponding relation is used as one.

Second the second updating block of corresponding relation, for there is the node pair for meeting preparatory condition in the second corresponding relation When, the wanton destruction pair in the second corresponding relation is chosen, as new node pair, the second corresponding relation is updated, default bar is returned The identifying unit of part the 3rd is continued executing with.

The updating block of second corresponding relation the 3rd, for being met in the second corresponding relation in the node pair of preparatory condition When 6th node is present in destruction in set, is met in the second corresponding relation and Section seven is chosen in the node pair of preparatory condition Point, the node pair constituted according to the 6th node and the 7th node, updates the second corresponding relation, wherein, the 7th node is to Section six The preference of point, more than destruction to preference of the node in set to the 6th node, the 6th node is in first node set Node, and the 7th node be the 3rd node set in node；Or the 7th node be node in first node set, and the Six nodes are the node in the 3rd node set.

Execution unit is returned, is judged for the 7th node to be added into destruction in set, and returns to preparatory condition second Unit is continued executing with.

Referring to Figure 10, Figure 10 is the schematic diagram of the electronic equipment of the embodiment of the present invention, including：

Processor 1001, communication interface 1002, memory 1003 and communication bus 1004, wherein, processor 1001, communication Interface 1002, memory 1003 completes mutual communication by communication bus 1004,

Memory 1003, for depositing computer program；

Processor 1001, for performing during the program deposited on memory 1003, realizes following steps：

When the node failure in first node set, predetermined number of thresholds is obtained, wherein, first node set For the set for all nodes that content fragment is cached with current area, content fragment includes the partial content of data to be forwarded, Number of thresholds is the nodes needed for the content fragment that the node that recovery is failed is cached；

In other nodes in current area in addition to the node in first node set, selection can with more than or equal to Node in the first node set of number of thresholds carries out all nodes of efficient communication, is used as Section Point set；

The correlation degree of each node and current area in Section Point set, chooses in Section Point set Cache node；

In the range of efficient communication, the number of thresholds node in first node set is chosen, recovers the node institute of failure The content fragment of caching is to cache node.

The communication bus that above-mentioned electronic equipment is mentioned can be Peripheral Component Interconnect standard (Peripheral Component Interconnect, PCI) bus or EISA (Extended Industry Standard Architecture, EISA) bus etc..The communication bus can be divided into address bus, data/address bus, controlling bus etc..For just Only represented in expression, figure with a thick line, it is not intended that only one bus or a type of bus.

The communication that communication interface is used between above-mentioned electronic equipment and other equipment.

Memory can include random access memory (Random Access Memory, RAM), can also include non-easy The property lost memory (Non-Volatile Memory, NVM), for example, at least one magnetic disk storage.Optionally, memory may be used also To be storage device that at least one is located remotely from aforementioned processor.

Above-mentioned processor can be general processor, including central processing unit (Central Processing Unit, CPU), network processing unit (Network Processor, NP) etc.；It can also be digital signal processor (Digital Signal Processing, DSP), it is application specific integrated circuit (Application Specific Integrated Circuit, ASIC), existing It is field programmable gate array (Field-Programmable Gate Array, FPGA) or other PLDs, discrete Door or transistor logic, discrete hardware components.

In the electronic equipment of the embodiment of the present invention, number of thresholds comes out according to there is probability calculation, a node The presence probability stamps very first time after, the node remains in the probability of current area.It is finally true according to number of thresholds The node of fixed cache node replacement failure, improves the effective percentage of data to be forwarded when being communicated using D2D, what raising communicated can By property.And it is determined that cache node replaces the node of failure, it is contemplated that each node in Section Point set with it is currently small The correlation degree in area, take into full account can node failure situation, it is determined that cache node crash rate it is low, data transfer is successful Probability height improves the effective percentage of data to be forwarded when being communicated using D2D, improves the reliability of communication.

The embodiment of the present invention additionally provides a kind of computer-readable recording medium, the computer-readable recording medium memory storage There is computer program, the computer program realizes any one above-mentioned distribution based on multi-hop cooperation when being executed by processor The method and step of storage method.

In the storage medium of the embodiment of the present invention, number of thresholds comes out according to there is probability calculation, a node The presence probability stamps very first time after, the node remains in the probability of current area.It is finally true according to number of thresholds The node of fixed cache node replacement failure, improves the effective percentage of data to be forwarded when being communicated using D2D, what raising communicated can By property.And it is determined that cache node replaces the node of failure, it is contemplated that each node in Section Point set with it is currently small The correlation degree in area, take into full account can node failure situation, it is determined that cache node crash rate it is low, data transfer is successful Probability height improves the effective percentage of data to be forwarded when being communicated using D2D, improves the reliability of communication.

For device/electronic equipment/storage medium embodiment, because it is substantially similar to embodiment of the method, so What is described is fairly simple, and the relevent part can refer to the partial explaination of embodiments of method.

It should be noted that herein, such as first and second or the like relational terms are used merely to a reality Body or operation make a distinction with another entity or operation, and not necessarily require or imply these entities or deposited between operating In any this actual relation or order.Moreover, term " comprising ", "comprising" or its any other variant are intended to Nonexcludability is included, so that process, method, article or equipment including a series of key elements not only will including those Element, but also other key elements including being not expressly set out, or also include being this process, method, article or equipment Intrinsic key element.In the absence of more restrictions, the key element limited by sentence "including a ...", it is not excluded that Also there is other identical element in process, method, article or equipment including the key element.

Each embodiment in this specification is described by the way of related, identical similar portion between each embodiment Divide mutually referring to what each embodiment was stressed is the difference with other embodiment.It is real especially for system Apply for example, because it is substantially similar to embodiment of the method, so description is fairly simple, related part is referring to embodiment of the method Part explanation.

The foregoing is merely illustrative of the preferred embodiments of the present invention, is not intended to limit the scope of the present invention.It is all Any modification, equivalent substitution and improvements made within the spirit and principles in the present invention etc., are all contained in protection scope of the present invention It is interior.

Claims

1. a kind of distributed storage method cooperated based on multi-hop, it is characterised in that including：

When the node failure in first node set, predetermined number of thresholds is obtained, wherein, the first node set For the set for all nodes that content fragment is cached with the current area, the content fragment includes the portion of data to be forwarded Divide content, the number of thresholds is the nodes needed for the content fragment that the node that recovery is failed is cached；

In other nodes in the current area in addition to the node in the first node set, selection can be more than or Equal to all nodes that the node in the first node set of the number of thresholds carries out efficient communication, second section point set is used as Close；

Each node and the correlation degree of the current area in the Section Point set, in the second section point set Cache node is chosen in conjunction；

In the range of efficient communication, the number of thresholds node in the first node set is chosen, recovers the failure The content fragment that is cached of node to the cache node.

2. according to the method described in claim 1, it is characterised in that it is described when the node failure in first node set when, Obtain before predetermined number of thresholds, methods described also includes：

Obtain and according to the maximum communication radius that the communication coverage of base station, multi-hop cooperate in the current area, the first number Value and the predetermined packet number and the number of thresholds that there is probability, calculate data to be forwarded, wherein, described first Numerical value for the node of the default caching data to be forwarded quantity, it is described to there is probability for the 8th node after the very first time still The probability in the current area is so in, the very first time is random time, and the 8th node is the current area In arbitrary node；

According to the packet number, the data to be forwarded is grouped, the data to be forwarded after being grouped；

According to the number of thresholds, the data to be forwarded after the packet is encoded, multigroup content fragment is obtained；

The node of the quantity of first numerical value is chosen in the current area, ancestor node is used as；

The content fragment is cached in the ancestor node respectively.

3. method according to claim 1 or 2, it is characterised in that described each in the Section Point set Node and the correlation degree of the current area, cache node is chosen in the Section Point set, including：

The each node obtained respectively in the Section Point set sets up efficient communication chain with the node in the current area The number on road, is used as the centrad of the node in the Section Point set；

The maximum node of centrad in the Section Point set is chosen, the cache node is used as.

4. method according to claim 1 or 2, it is characterised in that described each in the Section Point set Node and the correlation degree of the current area, cache node is chosen in the Section Point set, including：

Obtain and according to needed between two nodes that can directly carry out device-to-device D2D communications transmission data size, Efficient communication time, transimission power and channel situation, calculate transmission cost and are transmitted probability, wherein, it is described to be transmitted The probability of data transfer can be completed between two nodes in current area described in probability stamps；

According to the transmission cost and it is described be transmitted probability, respectively using each node in the first node set as one Individual summit, builds bipartite graph by another summit of each node in the Section Point set respectively；

Probability and dynamic programming algorithm Dynamic Programming algorithms are transmitted according to described, described is determined respectively The transmission path between each node in each node and the Section Point set in one node set；

Determine the weights of every transmission path respectively according to the transmission cost, pass through maximum matching algorithm Kuhn- Munkres algorithms obtain the matching of the minimum shortest path of weights sum；

According to the matching of the minimum shortest path of the weights sum, the cache node is chosen.

5. method according to claim 1 or 2, it is characterised in that described each in the Section Point set Node and the correlation degree of the current area, cache node is chosen in the Section Point set, including：

Obtain and according to needed between two nodes that can directly carry out device-to-device D2D communications transmission data size, Efficient communication time, transimission power and channel situation, calculate the transmission cost and described are transmitted probability；

According to it is described exist probability and it is described be transmitted probability, be determined to directly carry out D2D communications two nodes between The weights on side, wherein, fail to complete described in transmission between two nodes that D2D communications can be directly carried out described in the weights mark Need the probability of the data of transmission；

According to dynamic programming algorithm Dynamic Programming algorithms, determine respectively each in the first node set The shortest path between each node in node and the Section Point set；

According to the shortest path, it is described exist probability, it is described be transmitted probability and the transmission cost, determine respectively described Mutual preference between each node in each node and the Section Point set in first node set；

Each node in the Section Point set is accordingly to be regarded as the number of thresholds node, the 3rd node set is obtained；

According to the preference by preset algorithm, each node set up in the 3rd node set and the first node The one between each node in set；

According to a pair of corresponding relation, the cache node is determined.

6. method according to claim 5, it is characterised in that described that preset algorithm, foundation are passed through according to the preference The one between each node in each node and the first node set in 3rd node set, Including：

Distributed matcher algorithm Gale-Shapley algorithms or dynamical Matching Algorithm Roth and are passed through according to the preference Vande Vate algorithms, set up each node in the 3rd node set and each node in the first node set Between a pair of corresponding relation.

7. method according to claim 6, it is characterised in that described that Roth and Vande are passed through according to the preference Vate algorithms, each node set up in the node set to be selected and are each cached between the node of the content fragment Corresponding relation, including：

Between each node in each node and the first node set in step A, acquisition the 3rd node set Initial corresponding relation；

Step B, according to the preference, judges the node in the initial corresponding relation to whether meeting preparatory condition, wherein, Node in the initial corresponding relation between the first node set and the 3rd node set to have set up one-to-one Corresponding two nodes, the preparatory condition refers to the 3rd node and is higher than the 3rd node to first to the preference of fourth node The preference of node, and the fourth node is higher than the fourth node to the 5th node to the preference of the 3rd node Preference, the first node and the fourth node are the arbitrary node in the first node set, the 3rd node And the 5th node is the arbitrary node in the 3rd node set, and the first node and the 3rd node are institute The node of the same node centering in initial corresponding relation is stated, the fourth node is described initial corresponding with the 5th node The node of same node centering in relation, the node that the 3rd node and the fourth node are constituted is to being referred to as destroying It is right；

Step C, when all nodes in the initial corresponding relation are to being unsatisfactory for the preparatory condition, exports described initial Corresponding relation is used as a pair of corresponding relation；

Step D, when there is the node pair for meeting the preparatory condition in the initial corresponding relation, obtains the 3rd node With the fourth node, as destruction to set；

Step E, on the basis of the initial corresponding relation, according to the destruction to set, builds the second corresponding relation；

Step F, according to the preference, judges the node in second corresponding relation to whether meeting the preparatory condition；

Step G, when all nodes in second corresponding relation are to being unsatisfactory for the preparatory condition, output described second Corresponding relation is used as a pair of corresponding relation；

Step H, when there is the node pair for meeting the preparatory condition in second corresponding relation, judges second correspondence The node of the node centering of the preparatory condition is met in relation, if be present in the destruction in set；

Step I, the node in meeting the node pair of the preparatory condition in second corresponding relation is not present in described When destruction is in set, the wanton destruction pair in second corresponding relation is chosen, as new node pair, described second is updated Corresponding relation；

Step J, according to the preference, judges the node in second corresponding relation to whether meeting the preparatory condition；

Step K, when node in second corresponding relation is to being unsatisfactory for the preparatory condition, exports second correspondence Relation is used as a pair of corresponding relation；

Step L, when there is the node pair for meeting the preparatory condition in second corresponding relation, chooses second correspondence Wanton destruction pair in relation, as new node pair, updates second corresponding relation, returns to the step J and continues executing with；

Step M, met in second corresponding relation the 6th node in the node pair of the preparatory condition be present in it is described When destruction is in set, the 7th node of selection, root in the node pair of the preparatory condition are met in second corresponding relation The node pair constituted according to the 6th node and the 7th node, updates second corresponding relation, wherein, described Section seven Point is destroyed to preference of the node in set to the 6th node, institute to the preference of the 6th node more than described It is the node in the first node set to state the 6th node, and the 7th node is the section in the 3rd node set Point；Or the 7th node is the node in the first node set, and the 6th node is the 3rd node set In node；

Step N, is added to the destruction in set by the 7th node, and returns to the step F and continue executing with.

8. method according to claim 7, it is characterised in that each section in acquisition the 3rd node set Before the initial corresponding relation between each node in point and the first node set, methods described also includes：

When the number of the node in the 3rd node set is more than the number of the first node set interior joint, described The node of quantity identical with the first node set interior joint number is chosen in 3rd node set, at random with the first segment Node in point set sets up the initial corresponding relation corresponded.

9. a kind of distributed storage devices cooperated based on multi-hop, it is characterised in that including：

Number of thresholds acquisition module, for when the node failure in first node set, obtaining predetermined number of thresholds, Wherein, the first node collection is combined into the set for all nodes that content fragment is cached with the current area, the content Partial content of the fragment packets containing data to be forwarded, the number of thresholds is for needed for the content fragment that the node that recovery is failed is cached Nodes；

Node set determining module, for other sections in the current area in addition to the node in the first node set In point, all of efficient communication can be carried out with the node in the first node set more than or equal to the number of thresholds by choosing Node, is used as Section Point set；

Cache node determining module, for associating for each node in the Section Point set and the current area Degree, cache node is chosen in the Section Point set；

Content fragment recovery module, in the range of efficient communication, choosing the number of threshold values in the first node set An amount node, recovers content fragment that the node of the failure cached to the cache node.

10. a kind of electronic equipment, it is characterised in that including processor, communication interface, memory and communication bus, wherein, processing Device, communication interface, memory completes mutual communication by bus,

Memory, for depositing computer program；

Processor, for performing during the program deposited on memory, realizes any described method and steps of claim 1-8.