CN101242431B - Layer-cross mobile Ad Hoc network service discovery method - Google Patents
Layer-cross mobile Ad Hoc network service discovery method Download PDFInfo
- Publication number
- CN101242431B CN101242431B CN2008100565196A CN200810056519A CN101242431B CN 101242431 B CN101242431 B CN 101242431B CN 2008100565196 A CN2008100565196 A CN 2008100565196A CN 200810056519 A CN200810056519 A CN 200810056519A CN 101242431 B CN101242431 B CN 101242431B
- Authority
- CN
- China
- Prior art keywords
- node
- layer
- bunch
- network
- service
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Expired - Fee Related
Links
Images
Landscapes
- Mobile Radio Communication Systems (AREA)
Abstract
The invention relates to a cross-layer-based service discovery method in mobile Ad Hoc network environment. The invention is based on cross-layer design, aims to improve the utility rate of network resource, and solves the service discovery problem in mobile Ad Hoc network environment. By the integration of link layer and network layer, self-adaptive network-mobile optimized cluster structure is formed. When the node mobility is so insufficient to severely affect the stability of network topology, the clustering method actively expands the cluster structure to organize large-scale cluster structure, so that the operation of top-layer protocol on the whole network is simplified; when the fast shift of the node results in frequent change of network topology, the clustering method actively contracts the cluster structure to control the network cost for maintaining the cluster structure. By the integration of network layer and application layer, cross-layer service discovery in mobile Ad Hoc network environment is realized. When a target service is hit in a cluster, the cost of service discovery is controlled in the network layer sub-tree where service requester and target service provider coexist; when the service is dispersed to the whole network, the operation of service discovery is simplified to an operation on application inter-cluster organization, thus effectively controlling the network complexity of service discovery.
Description
Technical field
The present invention relates to the service discovery field of mobile Ad Hoc network, particularly relate to a kind of based on the mobile Ad Hoc network service discovery method of striding layer.
Background technology
Multi-hop, provisional autonomous system that mobile Ad Hoc network is made up of one group of portable terminal that has a wireless transmitter.When node joins in the new Ad Hoc network, the service that network provided is known nothing.Owing to can't serve necessary information on services to the new network of node configuration use in advance, so be even more important for the mobile Ad Hoc network efficient, that service discovery protocol is poor for network reliability, the node mobility is big reliably.
In recent years, in order to solve the service discovery problem under the mobile Ad Hoc network environment, the researcher launches research from many aspects, and obtained achievement mainly contains:
(1) based on the service discovery method of broadcast type service request packet forward:
When this method of employing, service requester is encapsulated as service request messages with the description to destination service of oneself, broadcasts in network.In order to guarantee the coverage of service request messages, all nodes in the network all will continue to transmit the non-duplicate services request message of being received, up to the hit service.
(2) service discovery method of transmitting based on border broadcast type service request:
When this method of employing, mobile node is responsible for all its network distances of buffer memory and is jumped with the interior service description file that node was provided at R.And the node that the network distance between the node is just jumped for R is called as the boundary node of this node.In service discovery process, the service request node is searched R and is jumped all interior information on services of zone according to the cache contents of oneself.If there is not the hit service, then to all boundary node broadcast service request messages.This process repeats until the hit service.
(3) service discovery method of transmitting based on multicast formula service request:
When this method of employing, one group of node in the Ad Hoc network is elected as directory junction.This election process guarantees all mobile nodes in network or becomes directory junction, or is not more than d with the network distance of directory junction and jumps (d is the fixed value that sets in advance); Elected directory junction then is responsible for the service description file that all node provided of maintaining network distance in d jumping scope.After election process finished, directory node was at the application layer tissue and safeguard backbone network, was responsible for the distribution and the processing of service request messages.Service request messages is transmitted in the multicast mode in backbone network, hits until destination service.
According to the difference of the service request messages pass-through mode of being taked, (1) and (2) can be reduced the network layer service discovery method, (3) then belong to the application layer services method.More than the design of three kinds of methods based on the principle independently of each layer in the conventional network architecture, with the interface of other each interlayer be static, and irrelevant with the constraint and the demands of applications of network.Yet mobile Ad Hoc network is specific purpose and interim occasion structure.In different applied environments, the quantity of node, translational speed etc. have nothing in common with each other, and these factors have seriously restricted the availability of above method under mobile Ad Hoc network environment.
Mobility is a key property of mobile Ad Hoc network, and this mobility all has certain influence to each layer in the protocol stack.At link layer, mobility has determined the duration of the rate of change and the connectivity of link of link property.In network layer, the mobility of node is to the efficient decisive role of packet forward.The node mobility to the combined influence final decision of network performance the robustness and the availability of this network.In order to adapt to the dynamic environment of mobile Ad Hoc network, be necessary synthetically each layer of network architecture to be designed, utilize the correlation between each layer to try hard to each layer protocol is integrated in the comprehensive classification framework.Therefore, stride layer design to the overall performance that improves network, use Internet resources to have very important significance rationally and effectively.
Summary of the invention
The present invention is a target to improve utilization rate of network resource, is designed to core concept to stride layer, solves the service discovery problem under the mobile Ad Hoc network environment.Specifically comprise:
1. at link layer, based on network model and the mobility model that extensively adopt in the emulation field, the link stability model of the mobile Ad Hoc network of deriving predicts that given link is at the probability of connection all the time in the section sometime in future.
2. but at the clustering architecture of network layer and application layer tissue and maintenance dynamic retractility.The core principles of cluster-dividing method is, makes probability that bunch member will be communicated with in the section in future sometime all the time to bunch head's path greater than a certain pre-set threshold.That is, bunch radius is directly determined by the link stability model of bottom.After clustering architecture formed, mobile node was jumped broadcasting exchange topology information and adjacent link state by periodic one, and served as network view under the basis is set up bunch with this information, upgrade and safeguard under clustering architecture.
3. bunch head structure between application layer tissue bunch; Bunch member and affiliated bunch of head are in network layer tissue bunch inner structure.Structure adopts the Mesh structure between bunch, adopts existing application layer multicast agreement towards mobile Ad Hoc network to transmit service request.The tree structure that it is root node that bunch inner tissue adopts with affiliated bunch of head adopts mode of unicast to transmit service request.In service discovery process, node is for being that the subtree of root node provides service index (that is, node is responsible for the description document that buffer memory is under the jurisdiction of the service that all nodes provided of its subtree) with it.Especially, bunch head bunch provides the service index for its place.More than all index by service summary realization based on the Bloom filter data structure.
Compared with prior art, innovation part of the present invention is: be designed to core concept and design service discovery method under the mobile Ad Hoc network environment to stride layer, network overhead is effectively controlled.Be embodied in:
1. the link layer design is combined closely with the network layer design, and tissue forms the ambulant optimization clustering architecture of adaptive network.Cause clustering architecture to be destroyed when node moves, repair in cost is controlled in bunch.
2. the network layer design is combined closely with the application layer design, and the network complexity of service discovery is effectively controlled.When destination service bunch in hit, the cost of service discovery is controlled in the network layer subtree of service requester and ISP coexistence.
When service request is diffused into the whole network, the operation of service discovery is reduced to the operation to tissue between application layer bunch.
The present invention proposes a kind of method based on service discovery problem under the mobile Ad Hoc of the solution network environment of striding layer design, comprise based on the sub-clustering of link stability and clustering architecture and safeguard step and based on layer service discovery step of striding of clustering architecture, it is characterized in that: described sub-clustering and clustering architecture based on link stability safeguards that step comprises:
1) node is gathered by Max inundation, Min inundation and a bunch member according to the link stability of calculating in real time, realizes under the mobile Ad Hoc network environment sub-clustering based on link stability;
2) node upgrades by link and node information renewal, node state according to the link stability of calculating in real time, realizes that the clustering architecture based on link stability is safeguarded under the mobile Ad Hoc network environment;
The described layer service discovery step of striding based on clustering architecture comprises:
1) the service request node will be encapsulated as service request messages to the descriptor of destination service;
2) bunch member's node is transmitted service request messages by mode of unicast in network layer in affiliated bunch;
3) bunch head's node makes up the Mesh structure in application layer, uses based on this towards the multicast protocol of mobile ad hoc network and transmits service request messages;
Wherein, in the step 1) of safeguarding step based on the sub-clustering and the clustering architecture of link stability, in the Max inundation stage, node uses local variable WINNER buffer memory intermediate object program; When the Max inundation started, node was its ID with the WINNER variable assignments; In the Max inundation stage, all nodes are broadcasted the WINNER value in a jumping scope, and WINNER value that node receives that and if only if is designated W
Recv, greater than the WINNER value of node this locality, and node and ID are W
RecvNode between the stability in path during greater than stability threshold given in advance, node replaces with W with the value of local variable WINNER
Recv
In the Min inundation stage, node uses local variable WINNER to be buffered in the minimum node ID that the Min inundation stage receives; All nodes are broadcasted the WINNER value in a jumping scope, WINNER value that node receives that and if only if is designated W
Recv, during less than the WINNER value of local maintenance, node replaces with W with the value of local variable WINNER
Recv
Gather the stage a bunch member, node is determined affiliated bunch in the following manner; Mode 1: if node is received the node ID of oneself in the Min inundation stage, then this node is elected oneself becomes bunch head's node; Mode 2: if mode 1 described condition does not satisfy, and node receives the ID of same node in a certain round in Max inundation stage and a certain round in Min inundation stage, and then selecting this node is bunch head's node; Mode 3: if mode 1 does not all satisfy with mode 2 described conditions, the maximum node ID that then elects the Max inundation stage to occur is bunch head's node.
In the described step 1) of safeguarding step based on the sub-clustering and the clustering architecture of link stability, each lasting D of Max inundation stage and Min inundation stage
MaxRound, wherein D
MaxBe predefined value.
In the described step 2 of safeguarding step based on the sub-clustering and the clustering architecture of link stability) in, node use node information table is stored all and is had (α, t) information of Guan Xi node with it; Wherein between node (α, t) contextual definition is: the path between node remains the probability of connection greater than a certain threshold alpha given in advance at following t in the period; Node use link information table is stored all and is had (α, t) information of the link of the node of relation formation with it.
In the described step 2 of safeguarding step based on the sub-clustering and the clustering architecture of link stability) in, node and the exchange of one hop neighbor are also upgraded link information table and node information table.
In the described step 2 of safeguarding step based on the sub-clustering and the clustering architecture of link stability) in, node is according to the content update node state of link-local information table and node information table.
Based on the striding in layer step 1) of service discovery step of clustering architecture, use the Bloom filter that the descriptor of service is made a summary described.
In the described layer step 2 of service discovery step of striding based on clustering architecture) in, all node organization tree structures belonged to, wherein the father node of node in the tree structure upstream node that is this node to place bunch head's the stabilizing path with cluster; When contact was received service request messages, if institute's requested service is hit in node leader's subtree, then node was transmitted this service request messages by mode of unicast to its child node.
In the described step 3) of striding layer service discovery step based on clustering architecture, bunch head's node makes up the Mesh structure as the multicast group membership in application layer, uses towards the application layer multicast protocol PAST-DM of mobile Ad Hoc network and safeguards this Mesh structure, transmits service request messages in the multicast mode.
Description of drawings
Fig. 1 moves under the Ad Hoc network environment based on the service discovery architecture of striding layer design
Fig. 2 is based on the data structure of the cluster-dividing method of link stability
Fig. 3 clustering architecture is safeguarded and the optimization method data structure
Fig. 4 node state machine
Fig. 5 serves index data structure
Fig. 6 service request messages structure
Embodiment
Consult Fig. 1, the present invention organizes the ambulant multi-hop clustering structure based on link stability of adaptive network by the comprehensive Design of link layer and network layer in mobile Ad Hoc network.Based on this, by the comprehensive Design of network layer and application layer, realize layer service discovery of striding under the mobile Ad Hoc network environment.Described stride layer service discovery be divided into network layer bunch in service discovery and application layer bunch between two stages of service discovery.When destination service bunch in the service discovery stage hit, the cost of service discovery is controlled in the network layer subtree of service requester and destination service supplier coexistence; When service request is diffused into the whole network, the operation of service discovery is reduced to the operation to tissue between application layer bunch between bunch, thereby the network complexity of service discovery is effectively controlled.
For in the link stability of accurately predicting under the network environment of mobile ad hoc network complexity between node, adopted the extensively network constraint condition of employing of researcher, promptly suppose: the wireless signal coverage of node is that radius is the circle of R; And if only if two nodes are all in the other side's signal cover the time, and link is communicated with between two nodes.When considering node B relative to the moving at random of node A, can with the wireless signal coverage of node A (O R) discretely turns to the segment that a series of spacings are ε, node B to the distance of node A then for being defined in state space { S
0, S
1..., S
nOn stochastic variable, S wherein
iCorresponding to i spacing fragment.Further can set up Markov-chain model and describe node B moving at random with respect to node A.By deriving and the computing mode transfer matrix, can calculate the end-state vector of Markov-chain model.Based on this, suppose that distance between node A and the node B is at t
0Be d constantly
0, can obtain Radio Link at section (t preset time
0, t
0+ probability A (the d that continue to keep is communicated with in t)
0, t), promptly node A and node B are at t
0+ t constantly apart from the probability of d less than R.
Under mobile Ad Hoc network environment, target based on the cluster-dividing method of link stability is: select one group of node to become bunch head's node in all nodes, this selection should guarantee that other node in the network constitutes (α with bunch of head's node at least, t) relation guarantees that promptly this node remains the probability of connection greater than a certain threshold alpha given in advance to the path between bunch head's node in preset time in the section t.According to the probability theory multiplicative principle, can obtain the computational methods of path stability under the mobile ad hoc network environment.If the path between node A and the node B is made up of n bar link.Make P (t) in time t, remain the probability of connection, then for this path
Wherein, A
k(d
k, t) for constituting the stability of the k bar link in path between node A and the node B, d
kBe two nodes that constitute k bar link distance at initial time.In the process that clustering architecture forms, node all meet (α, t) condition around select in the node node of address maximum to become bunch node of oneself.Simultaneously, in order to guarantee the convergence of cluster-dividing method in static network, method is set D
MaxThe upper limit as the clustering architecture radius.
See also Fig. 2.Fig. 2 has described based on the used data structure of the cluster-dividing method of link stability.Wherein each field meaning of intermediate object program record is as follows:
● Sender_ID: the node ID that sends node;
● WINNER: the intermediate object program in inundation stage is the current known maximum node ID of node in the Max inundation stage wherein; And be the known minimum node ID of node in the Min inundation stage.
● the stable P (t) in path between the node that node and WINNER identified that Stability:Sender_ID identified;
● Hops_Count: the quantity of links of forming above-mentioned path;
● RelayNode_n: the relay nodes ID that constitutes above-mentioned path.
The election scoring plug is used to preserve the intermediate object program of Max inundation stage and Min inundation all rounds of stage.Table empty item is used to fill in the WINNER value of a certain stage round.
Cluster-dividing method based on link stability is made of three phases: in the Max inundation stage, a Min inundation stage and a bunch member gather the stage.
1. in the Max inundation stage, all nodes are broadcasted the floodMax message in a jumping scope, and neighbours circulate a notice of bunch head's node address of its temporary transient approval towards periphery.The floodMax message is identical with intermediate object program record format, and wherein the value of WINNER is filled in known maximum node address.This process continues D
MaxRound.
2. in the Min inundation stage, all nodes are broadcasted the floodMin message in a jumping scope, and neighbours circulate a notice of its known tuftlet head's node address towards periphery.The floodMin message is made up of an independent WINNER field.Wherein fill in a known tuftlet node address.The intermediate object program of all rounds is by the WINNER field buffer memory of intermediate object program record.The Min inundation stage continues D
MaxRound.
3. gather the stage a bunch member, utilize the election information that scoring plug write down, node is determined affiliated bunch with following principle, thereby gathers to a bunch node.
● principle 1: if find own node address in the floodMin message that node was received in the Min inundation stage, then this node is elected oneself becomes bunch head's node.
● principle 2: the recorded information of each round of node inspection election scoring plug.If there is the address of node to occur simultaneously, then in these nodes, select the node of node address minimum to become bunch head's node in Max inundation stage and Min inundation stage.
● principle 3: the maximum node address that the election Max inundation stage occurs is bunch head's node.
The dynamic topology of mobile Ad Hoc network makes in the middle of clustering architecture is in an unsure state all the time.Along with the enhancing of network stabilization, redundant bunch head's node might appear in the network; Equally, when stability of network decline, clustering architecture is necessary division, contraction or reconstruct, and is downsizing to adapt to the ambulant enhancing of node.In order to safeguard under dynamic network environment and optimize clustering architecture that the present invention has designed based on the clustering architecture of link stability and safeguarded and optimization method.
At first, method is that any node A has defined four class states in the mobile Ad Hoc network:
(1) loyal bunch member condition: if (α t) in all neighbouring nodes of relation (definition sees the 4th page for details), has only bunch of head's node, and then node A is subordinated to this bunch head, and is in loyal state satisfying with node A.
(2) disloyalty bunch member condition: if (α t) in Guan Xi all neighbouring nodes, has a plurality of bunches of head's nodes, and then node A is subordinated to one of them bunch head, but is in the disloyalty state satisfying with node A.
(3) healthy bunch head's state: if node A is bunch head's node, and have at least a node to be loyal bunch member among its bunch member, then node A is in health status.
(4) dying bunch of head's state: if node A be bunch head's node, and its all bunches member all is in the disloyalty state, and then node A leader's bunch is redundancy bunch, and node A is in moribund condition.
See also Fig. 3.Safeguard and temporary this local topology information of optimization method use data structure as shown in Figure 3 based on the clustering architecture of link stability.Each data structural field detailed meanings is as described below among Fig. 3:
(1) node information table:
Each field implication of node information table is as follows:
● Node_Addr: node address;
● CH_Addr: bunch head's node address of clustering architecture under this node;
● Pub_Date: the issuing time of this node information;
● Status: the current state of living in of this node;
● Stability: the path stability that arrives the shortest path of this node;
● Next_Hop: next from this locality to this node jumped node address.
(2) link information table:
Each field implication of link information table is as follows:
● Node_A and Node_B: the index of two nodes in the node information table that constitutes this link;
● Stability: the stability of this link;
● Pub_Date: the issuing time of this link information;
● Exp_Date: the expired time of this link information, the stability of establishing this link information is A (d
0, t), d wherein
0Be two nodes constituting this link distance at initial time, then the expired time of this link information is
Exp_Date＝Pub_Date+t
Above-mentioned equation guarantees that all links in the site link information table kept stable probability to be not less than α before Exp_Date.Node is checked effective situation of link by built-in timer.The expired renewal that will trigger site link information table and node information table of link.
1. the renewal of node information table and link information table
In the maintenance and optimizing process of clustering architecture, node is shared topology information (promptly by periodic broadcast hello packet (hello packet comprises whole list items of site link and node information table) with neighbours, all and its satisfy (α, t) information of Guan Xi neighbouring node and link), thus build the localized network topology in this locality.According to the information of being grasped, node periodically updates the state of oneself, and carries out corresponding actions according to the variation of state, thus Dynamic Maintenance and optimization clustering architecture.
(1) when receiving the hello packet that neighbours broadcast, node at first judge and the hello packet sender between link whether satisfy (α, t) relation.Then in the link information table, insert in this way or upgrade this link information.Pub_Date with the link information table during renewal is a standard, replaces old information with fresh information, timer is set simultaneously to upgrade link table when link is expired.
(2) node of receiving hello packet with oneself link and node information table and hello packet in the information of carrying merge.If information exists in link or node information table, then be that standard is upgraded with Pub_Date.
(3) node of receiving hello packet is new node information table memory allocated space, and uses dijkstra's algorithm, makes up on the basis of link after the merging and node information table and upgrades local topology information.Wherein dijkstra's algorithm is used for finding the solution the shortest path from local node to each neighbours' node on the weighting connected graph model of mobile Ad Hoc network.
(4) after above step is finished, receive that the node of hello packet reclaims the used internal memory of old node information table, and in the link information table, delete all unreachable links.
(5) after timer is out-of-date, node is at first deleted expired link, subsequently from above-mentioned steps (3), upgrades node and link information table.
2. node state is upgraded
After link and node information table were upgraded, node was safeguarded node state according to the local topology information of being grasped by state machine shown in Figure 4.Below each the state transition process that marks among Fig. 4 is explained in detail.
Step T1, for the node that is in bunch member condition, if find a plurality of health bunch head in the process of inquiry node information table, then state transition is disloyalty bunch member.
Step T2, for the node that is in bunch member condition, if when the prevariety head is unique health bunch head in the node information table, perhaps lost efficacy or deteriorate to dying and all network distances (hop count) that belong to one bunch neighbouring node and bunch head with it together are not more than oneself as the prevariety head, then node is initiatively selected bunch head's adding of other health.If only find unique selection in this process, then state exchange is loyal bunch member.
Step T3 for the node that is in bunch member condition, is health bunch head unique in the node information table if work as the prevariety head.
Step T4, for the node that is in bunch member condition, if keep fit or have in the node information table than as the more stable bunch head of prevariety head as the prevariety head, then node is changed place bunch, and keeps the disloyalty state.If lost efficacy or deteriorate to dyingly as the prevariety head, and all network distances (hop count) that belong to one bunch neighbouring node and bunch head with it together are not more than oneself, and then node is initiatively selected bunch head's adding of other health; If in this process, find multiple choices, then after changing affiliated bunch, still be maintained disloyalty bunch member condition.
Step T5 and T6, for the node that is in bunch member condition, if lost efficacy or deteriorate to dyingly as the prevariety head, and all network distances (hop count) that belong to one bunch neighbouring node and bunch head with it together are not more than oneself, and then node is initiatively selected bunch head's adding of other health.If do not find other healthy bunch head in selection course, then node is promoted as healthy bunch head by a bunch member status.
Step T7 and T8, for being in bunch node of head's state, if node finds to be in bunch member of loyal state in the process of Query Information table, or node current be unique visible health bunch head.
Step T9 and T10 for being in bunch node of head's state, if node finds to be in bunch member of loyal state, and find a plurality of health bunch head in the process of Query Information table, then node the leader bunch deteriorate to redundant bunch.If still have unresponsive bunch of member not break away from this bunch, then node is kept a bunch head status, but that state deteriorates to is dying.
Step T11, for being in bunch node of head's state, if the redundancy bunch that has broken away from the node leader as all bunches member, then node will initiatively select other healthy bunch head to add.If in this process, only find unique selection, then add new bunch, and state transitions is loyal bunch member.
Step T12, for being in bunch node of head's state, if the redundancy bunch that has broken away from the node leader as all bunches member, then node will initiatively select other healthy bunch head to add.If selection is not unique in this process, then add new bunch, but state transitions is disloyalty bunch member.
On the basis of above-mentioned steps, can utilize existing clustering architecture, in network, dispose with different levels directory junction, to support layer service discovery of striding under the mobile Ad Hoc network environment.In network layer, belong to all node organization tree structures with cluster, wherein, node is for being that the subtree of root node provides the service index with it; Especially, bunch head's node bunch provides the service index for its place.This index is realized by the service summary based on the Bloom filter data structure.Simultaneously, bunch head's node is organized nerve of a covering Mesh structure on network layer, by integrating existing application layer multicast agreement towards mobile ad hoc network, and the service request packet forward between realizing bunch.
1. service discovery in the network layer bunch
Service discovery in node is searched service by cooperation in affiliated bunch process is called bunch, this process occurs in network layer.For the service discovery protocol of giving the upper strata provides support, being under the jurisdiction of need be in network layer organization tree structure, so that with different levels service index to be provided with cluster head's node.This index is realized by making a summary based on the service of Bloom filter data structure.
(1) makes a summary based on the service of Bloom filter
● during initial condition, with each position 0 of Bloom filter;
● use k hash function separate, given in advance, service description file is mapped among everybody of Bloom filter.To any one element x, the position h of i hash function mapping
i(x) be changed to 1 (1≤i≤k).
● judging a certain given destination service (establishing its service description information is y) when whether in index, hitting, y used k hash function, if in the Bloom filter all h
i(y) position all is 1, and (1≤i≤k) thinks that then this destination service hits, otherwise thinks that destination service is not the element in the index in index.
(2) method of tissue bunch inner structure
Belong to all node organization tree structures of cluster and find that for top service agreement provides index.Wherein father node for its leader's subtree index is provided, and bunch head's node bunch provides index for its place.Its index data structure as shown in Figure 5.Each field meaning of service index data structure is as follows:
● father node address (Parent): the father node address of tree structure in bunch;
● the effective time of expired time (Deadline) father node;
● Bloom filter B it place value (BitValue): the value of a certain position of Bloom filter;
● bit timer: the expired time of corresponding position.
In this data structure, use bit timer that the Bloom filter is expanded.Introduce the position timing phase validity detection is carried out in each position of Bloom filter.When timer arrives constantly with this position 0.The renewal of bit timer is responsible for by the direct descendent of this node in tree structure.
● the service index of node is carried by hello packet, jumps broadcasting by periodicity one and exchanges between adjacent father and son's node of tree structure.Behind the hello packet of receiving neighbours' node, node is handled according to the following steps:
● child node upgrades the father node term of validity: prolong the expired time in the service index data structure.
● update service index: the Bloom filter of child node and the Bloom filter of father node are carried out " with " operation.
● updated space timer: the term of validity that prolongs the corresponding Bit of Bloom filter position.
● the more new element that the father node expired time expires and triggers: for expired father node, link information table that node is safeguarded according to its sub-clustering agreement and node information table bunch in select new father node, and broadcast the change of hello packet immediately with the notice tree structure.
● the more new element of the expired triggering of bit timer: for the Bit position of expired Bloom filter, node puts 0 with it.
(3) bunch the retransmission method of service request messages in
When node is initiated the service discovery operation, at first use the Bloom filter that its requested service is made a summary.Subsequently, encapsulation as shown in Figure 6 service request messages and submit to network.Its each field meaning is as follows:
● source address (SourceAddr): service request node address;
● message identification (RequestID): service request messages ID;
● request service summary (Service Abstract): use the summary of Bloom filter to institute's requested service statement file.
The related step of the retransmission method of service request messages is as described below in bunch:
● the promoter of service request is changed to its node address with LastHop, and should ask to submit to network.
● when initiating or receiving service request messages, node at first carries out necessary record to service request messages, and abandons the message of repetition according to this record.
● whether node hits in the subtree of oneself by the operation judges destination service to the Bloom filter.
● when node oneself was by the supplier of destination service, node was initiated the service Recovery Process, and this process utilizes the network layer route to finish.Otherwise node continues to transmit service request messages in the subtree of oneself, until hitting.
If ● destination service node do not hit in leader's the subtree, then node is submitted service request messages to its father node.
● as bunch in service discovery process failure, service discovery between then initiating bunch by bunch head.
2. service discovery between application layer bunch
In the mobile ad hoc network of clustering architecture, service discovery protocol is operation to the backbone network of being made up of bunch head to the simplified control of the whole network.That is, the service discovery request message is only transmitted in backbone network between bunch, and bunch head then is responsible for bunch filtering services request of its place.Between design application layer bunch during service discovery method, the present invention realizes that by integrating existing mobile ad hoc network application layer Radio Broadcasting Agreements PAST-DM the service request messages in the backbone network transmits, and hits until destination service.
Claims (8)
1. method based on service discovery problem under the mobile Ad Hoc of the solution network environment of striding layer design comprises based on the sub-clustering of link stability and clustering architecture and safeguards step and based on layer service discovery step of striding of clustering architecture, it is characterized in that:
Described sub-clustering and clustering architecture based on link stability safeguards that step comprises:
1) node is gathered by Max inundation, Min inundation and a bunch member according to the link stability of calculating in real time, realizes under the mobile Ad Hoc network environment sub-clustering based on link stability;
2) node upgrades by link and node information renewal, node state according to the link stability of calculating in real time, realizes that the clustering architecture based on link stability is safeguarded under the mobile Ad Hoc network environment;
The described layer service discovery step of striding based on clustering architecture comprises:
1) the service request node will be encapsulated as service request messages to the descriptor of destination service;
2) bunch member's node is transmitted service request messages by mode of unicast in network layer in affiliated bunch;
3) bunch head's node makes up the Mesh structure in application layer, uses based on this towards the multicast protocol of mobile ad hoc network and transmits service request messages;
Wherein, in the step 1) of safeguarding step based on the sub-clustering and the clustering architecture of link stability, in the Max inundation stage, node uses local variable WINNER buffer memory intermediate object program; When the Max inundation started, node was its ID with the WINNER variable assignments; In the Max inundation stage, all nodes are broadcasted the WINNER value in a jumping scope, and WINNER value that node receives that and if only if is designated W
Recv, greater than the WINNER value of node this locality, and node and ID are W
RecvNode between the stability in path during greater than stability threshold given in advance, node replaces with W with the value of local variable WINNER
Recv
In the Min inundation stage, node uses local variable WINNER to be buffered in the minimum node ID that the Min inundation stage receives; All nodes are broadcasted the WINNER value in a jumping scope, WINNER value that node receives that and if only if is designated W
Recv, during less than the WINNER value of local maintenance, node replaces with W with the value of local variable WINNER
Recv
Gather the stage a bunch member, node is determined affiliated bunch in the following manner; Mode 1: if node is received the node ID of oneself in the Min inundation stage, then this node is elected oneself becomes bunch head's node; Mode 2: if mode 1 described condition does not satisfy, and node receives the ID of same node in a certain round in Max inundation stage and a certain round in Min inundation stage, and then selecting this node is bunch head's node; Mode 3: if mode 1 does not all satisfy with mode 2 described conditions, the maximum node ID that then elects the Max inundation stage to occur is bunch head's node.
2. method according to claim 1 is characterized in that: in the described step 1) of safeguarding step based on the sub-clustering and the clustering architecture of link stability, and each lasting D of Max inundation stage and Min inundation stage
MaxRound, wherein D
MaxBe predefined value.
3. method according to claim 1 is characterized in that: in the described step 2 of safeguarding step based on the sub-clustering and the clustering architecture of link stability) in, node use node information table is stored all and is had (α, t) information of Guan Xi node with it; Wherein between node (α, t) contextual definition is: the path between node remains the probability of connection greater than a certain threshold alpha given in advance at following t in the period; Node use link information table is stored all and is had (α, t) information of the link of the node of relation formation with it.
4. method according to claim 1 is characterized in that: in the described step 2 of safeguarding step based on the sub-clustering and the clustering architecture of link stability) in, node and the exchange of one hop neighbor are also upgraded link information table and node information table.
5. method according to claim 1 is characterized in that: in the described step 2 of safeguarding step based on the sub-clustering and the clustering architecture of link stability) in, node is according to the content update node state of link-local information table and node information table.
6. method according to claim 1 is characterized in that: based on the striding in layer step 1) of service discovery step of clustering architecture, use the Bloom filter that the descriptor of service is made a summary described.
7. method according to claim 1, it is characterized in that: in the described layer step 2 of service discovery step of striding based on clustering architecture) in, belong to all node organization tree structures, wherein the father node of node in the tree structure upstream node that is this node to place bunch head's the stabilizing path with cluster; When contact was received service request messages, if institute's requested service is hit in node leader's subtree, then node was transmitted this service request messages by mode of unicast to its child node.
8. method according to claim 1, it is characterized in that: in the described step 3) of striding layer service discovery step based on clustering architecture, bunch head's node makes up the Mesh structure as the multicast group membership in application layer, uses towards the application layer multicast protocol PAST-DM of mobile Ad Hoc network and safeguards this Mesh structure, transmits service request messages in the multicast mode.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN2008100565196A CN101242431B (en) | 2008-01-21 | 2008-01-21 | Layer-cross mobile Ad Hoc network service discovery method |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN2008100565196A CN101242431B (en) | 2008-01-21 | 2008-01-21 | Layer-cross mobile Ad Hoc network service discovery method |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN101242431A CN101242431A (en) | 2008-08-13 |
| CN101242431B true CN101242431B (en) | 2011-03-16 |
Family
ID=39933639
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN2008100565196A Expired - Fee Related CN101242431B (en) | 2008-01-21 | 2008-01-21 | Layer-cross mobile Ad Hoc network service discovery method |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN101242431B (en) |
Families Citing this family (9)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101741916B (en) * | 2008-11-14 | 2012-10-24 | 复旦大学 | Reliable service discovery and service selection method in self-organization network |
| CN101741917B (en) * | 2008-11-14 | 2012-10-24 | 复旦大学 | Self-adaptive service discovery method in wireless self-organization network |
| WO2012093294A1 (en) | 2011-01-07 | 2012-07-12 | M.S. Ramaiah School Of Advanced Studies | Ad hoc network |
| CN102595314B (en) * | 2012-02-14 | 2014-07-09 | 北京邮电大学 | Position-information-based device-to-device (D2D) clustering multicast method |
| US11265385B2 (en) * | 2014-06-11 | 2022-03-01 | Apple Inc. | Dynamic bloom filter operation for service discovery |
| CN105207804B (en) * | 2015-08-18 | 2018-08-10 | 昆明理工大学 | A kind of real-time estimating method for service path stability |
| CN106850553B (en) * | 2016-12-21 | 2019-11-29 | 大唐高鸿信息通信研究院（义乌）有限公司 | A kind of implementation method of the improvement DSDP service discovery protocol suitable for vehicle-mounted short distance communication network |
| CN108271234B (en) * | 2017-11-28 | 2022-03-11 | 中国电子科技集团公司电子科学研究院 | Service discovery method and server for mobile ad hoc network |
| CN109981808A (en) * | 2019-01-28 | 2019-07-05 | 博频云彩（北京）科技有限公司 | A kind of IP address analytic method of the efficient substitution ARP protocol of low overhead |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN1719803A (en) * | 2005-07-28 | 2006-01-11 | 武汉大学 | Correcting method and application of multi-size router for extensible large-scale sensor network |
| CN1731760A (en) * | 2005-08-05 | 2006-02-08 | 武汉理工大学 | Ad Hoc network routing method based on flooding mechanism |
| CN1988509A (en) * | 2006-12-07 | 2007-06-27 | 中国科学院计算技术研究所 | Method for forming fine convergence gradient in sensor network |
| EP1826964A1 (en) * | 2006-02-28 | 2007-08-29 | Nagoya Institute Of Technology | Mobile terminal device, control method, and mobile communication system |
-
2008
- 2008-01-21 CN CN2008100565196A patent/CN101242431B/en not_active Expired - Fee Related
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN1719803A (en) * | 2005-07-28 | 2006-01-11 | 武汉大学 | Correcting method and application of multi-size router for extensible large-scale sensor network |
| CN1731760A (en) * | 2005-08-05 | 2006-02-08 | 武汉理工大学 | Ad Hoc network routing method based on flooding mechanism |
| EP1826964A1 (en) * | 2006-02-28 | 2007-08-29 | Nagoya Institute Of Technology | Mobile terminal device, control method, and mobile communication system |
| CN1988509A (en) * | 2006-12-07 | 2007-06-27 | 中国科学院计算技术研究所 | Method for forming fine convergence gradient in sensor network |
Non-Patent Citations (4)
| Title |
|---|
| Alan D. Amis 等.Max-Min D-Cluster Formation in Wireless Ad Hoc Networks.《NFOCOM 2000. Nineteenth Annual Joint Conference of the IEEE Computer and Communications Societies》.2000,第1卷第32-41页. |
| Alan D. Amis等.Max-Min D-Cluster Formation in Wireless Ad Hoc Networks.《NFOCOM 2000. Nineteenth Annual Joint Conference of the IEEE Computer and Communications Societies》.2000,第1卷第32-41页. * |
| Bosheng Zhou 等.A cross-layer route discovery framework for mobile ad hoc networks.《EURASIP Journal on Wireless Communications and Networking》.2005,第2005卷(第5期),第645-660页. |
| Bosheng Zhou等.A cross-layer route discovery framework for mobile ad hoc networks.《EURASIP Journal on Wireless Communications and Networking》.2005,第2005卷(第5期),第645-660页. * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN101242431A (en) | 2008-08-13 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN101242431B (en) | Layer-cross mobile Ad Hoc network service discovery method | |
| CN101360051B (en) | Energy efficient wireless sensor network routing method | |
| Anupama et al. | Survey of cluster based routing protocols in mobile adhoc networks | |
| CN100417118C (en) | System and method for renewing network mobile node position in wireless net-like network | |
| CN101534520B (en) | Reliable routing method of mobile Ad hoc network | |
| CN102711209B (en) | Dynamic self-organizing hierarchical routing method applied to wireless sensor network | |
| CN101374108B (en) | Routing method suitable for static state self-grouping formation | |
| Khedr et al. | Successors of PEGASIS protocol: A comprehensive survey | |
| Agrawal et al. | GCRP: Grid-cycle routing protocol for wireless sensor network with mobile sink | |
| CN101895954B (en) | Opportunistic network routing method based on incremental transmission of packet index | |
| CN101682579A (en) | Method for operating a mesh-type network, particularly as defined in an IEEE 802.11s standard, formed by a plurality of network nodes | |
| CN101594281A (en) | Collecting network data of wireless sensor method, system and relevant device | |
| CN101262449B (en) | A new wireless self-networking dual-frequency hierarchical routing method | |
| CN103259744A (en) | Method for mapping mobile virtual network based on clustering | |
| CN103118412A (en) | Trust-based on-demand multipath vector routing algorithm of Ad Hoc network | |
| CN105072030A (en) | NDN (Named Data Networking) route system based on content clustering, and clustering query method therefor | |
| Bagula et al. | Ubiquitous sensor network management: The least interference beaconing model | |
| CN101521926A (en) | Cooperative routing method for large-scale wireless distribution network | |
| CN104936249A (en) | Route self-healing method suitable for wireless network | |
| CN101179501A (en) | Local routing discovering method for on-demand routing of mobile ad hoc network | |
| CN101321184B (en) | Method for implementing wireless sensor network query, and data return method | |
| CN101686521A (en) | Network coding based method for searching route of wireless Ad hoc network | |
| CN102006651A (en) | Distributed ad hoc network stable path routing method based on link lifetime | |
| CN105072586A (en) | Management method for forwarding of broadcast message in embedded wireless ad hoc network | |
| Choi et al. | Cell based energy density aware routing: a new protocol for improving the lifetime of wireless sensor networks |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| C14 | Grant of patent or utility model | ||
| GR01 | Patent grant | ||
| C17 | Cessation of patent right | ||
| CF01 | Termination of patent right due to non-payment of annual fee |
Granted publication date: 20110316 Termination date: 20140121 |