Trust Propagation Method

Abstract

The present invention belongs to the technical field of data security, and discloses a trust propagation method. The method is applied to cluster head sensor nodes, including: determining the initial distance between each initial recommended node according to the trust metric parameter of each initial recommended node, determining multiple recommended nodes to be selected and the initial influence of each recommended node to be selected according to the community division result and the number of demand nodes; sending the initial influence of each recommended node to be selected to each recommended node to be selected, determining the propagation path and target influence through each recommended node to be selected, and sending the recommendation information and target influence to the trust subject according to the propagation path, and the trust subject makes a trust propagation decision according to the target influence and recommendation information of each recommended node to be selected. Through the above method, it is possible to select a suitable recommended node, minimize the influence distortion at the recommended node, and at the same time, it is possible to establish a suitable propagation path, minimize the propagation distortion, and improve the accuracy of the trust subject when making a trust decision.

Description

Technical Field

[0001] The present invention relates to the field of financial technology, and in particular to a trust propagation method, a cluster head sensor node, a recommendation node and a trust subject. Background Art

[0002] In the existing underwater acoustic sensor network, when sensor nodes with higher trust values ​​are selected to participate in the trust recommendation process, there will be erroneous and dishonest recommendations from sensor nodes. False and dishonest recommendations will affect the accuracy of trust decisions in the trust recommendation and trust propagation mechanisms, and reduce the security of data transmission. Summary of the invention

[0003] The main purpose of the present invention is to provide a trust propagation method, a cluster head sensor node, a recommendation node and a trust subject, aiming to solve the technical problem of how to improve the accuracy of trust decision-making in an underwater acoustic sensor network, thereby improving the security of data transmission.

[0004] To achieve the above object, the present invention provides a trust propagation method, which is applied to a cluster head sensor node. The trust propagation method comprises:

[0005] Upon receiving a trust request data packet from a trust subject, determining a plurality of initial recommended nodes and a number of required nodes according to the trust request data packet;

[0006] Determine the initial distance between each initial recommendation node according to the trust metric parameter of each initial recommendation node;

[0007] Performing community division according to the initial distances between the initial recommended nodes, and determining a plurality of candidate recommended nodes and the initial influence of each candidate recommended node from the plurality of initial recommended nodes according to the community division results and the number of demand nodes;

[0008] The initial influence of each candidate recommendation node is sent to each candidate recommendation node, and each candidate recommendation node determines a propagation path and a target influence according to the initial influence of each candidate recommendation node, and sends the recommendation information and the target influence to the trust subject according to the propagation path, so that the trust subject makes a trust propagation decision according to the target influence of each candidate recommendation node and the recommendation information of each candidate recommendation node.

[0009] Optionally, determining the initial distance between each initial recommended node according to the trust metric parameter of each initial recommended node includes:

[0010] Calculate the proximity according to the number of propagation hops between each initial recommendation node and the trust subject and the number of target acceptance hops of the trust subject, and determine the spatial proximity of each initial recommendation node;

[0011] The reliability of each initial recommended node is determined by calculating the reliability based on the number of interactions between each initial recommended node and the trusted object and the information of the interaction data packet;

[0012] Calculate the reliability of the link between each initial recommended node and the trusted object according to the link quality, and determine the link reliability of each initial recommended node;

[0013] Distance calculation is performed according to the spatial proximity of each initial recommended node, the node reliability of each initial recommended node, and the link reliability of each initial recommended node to determine the initial distance between each initial recommended node.

[0014] Optionally, performing community division according to the initial distances between the initial recommended nodes, and determining a plurality of candidate recommended nodes and an initial influence of each candidate recommended node from the plurality of initial recommended nodes according to the community division result and the number of demand nodes, includes:

[0015] Determine the adjacent nodes of each initial recommended node according to the initial distance between each initial recommended node;

[0016] Influence calculation is performed based on the initial distances between the initial recommended nodes and the adjacent nodes of each initial recommended node to determine the direct influence value, public influence value and adjacent influence value of each initial recommended node;

[0017] Distance calculation is performed based on the direct influence value, public influence value, and adjacent influence value of each initial recommended node to determine the target distance between each initial recommended node;

[0018] A plurality of candidate recommendation nodes and an initial influence of each candidate recommendation node are determined from the plurality of initial recommendation nodes according to the target distance between the respective initial recommendation nodes and the number of the demand nodes.

[0019] Optionally, determining a plurality of candidate recommended nodes and an initial influence of each candidate recommended node from a plurality of initial recommended nodes according to the target distance between each initial recommended node and the number of demand nodes includes:

[0020] Perform relative distance calculation according to the target distance between each initial recommended node to determine the relative distance of each initial recommended node;

[0021] Sorting the relative distances of the initial recommended nodes, and determining a plurality of candidate recommended nodes from the plurality of initial recommended nodes according to the sorting results and the number of demand nodes;

[0022] The influence is calculated according to the relative distance of each candidate recommendation node to determine the initial influence of each candidate recommendation node.

[0023] In addition, to achieve the above purpose, the present invention also provides a trust propagation method, which is applied to a recommendation node. The trust propagation method includes:

[0024] When receiving the initial influence sent by the cluster head sensor node, the path is determined according to the initial influence and the target generalized maximum flow strategy to obtain the propagation path and the target influence. The initial influence is that the cluster head sensor node determines the number of multiple initial recommended nodes and the number of demand nodes according to the trust request data packet sent by the trust subject, and determines the distance between each initial recommended node according to the trust metric parameter of each initial recommended node, and divides the community according to the initial distance between each initial recommended node, and is calculated according to the community division result;

[0025] The recommendation information and the target influence are sent to the trust subject according to the propagation path, so that the trust subject makes a trust propagation decision according to the target influence and the recommendation information.

[0026] Optionally, determining a path according to the initial influence and the target generalized maximum flow strategy to obtain a propagation path and a target influence includes:

[0027] Defining an initial flow according to the initial influence and determining the initial flow;

[0028] Determine the arc capacity between the candidate recommended node and the candidate recommended node according to the direct trust value;

[0029] The gain factor is calculated based on the number of propagation hops between the trusted subject, the current number of traversal hops, and the arc capacity between the recommended node to be selected, and the arc gain factor is determined;

[0030] Calculate the output flow according to the gain factor and the initial flow to determine the node output flow;

[0031] Determining a target influence according to the node output stream;

[0032] A propagation path is determined according to the target influence, the arc gain factor, and the arc capacity.

[0033] In addition, to achieve the above purpose, the present invention also provides a trust propagation method, which is applied to a trust subject, and the trust propagation method includes:

[0034] Upon receiving a data transmission instruction, determining the number of required nodes and the target number of receiving hops according to the data transmission instruction;

[0035] Encapsulating the required node quantity and the target acceptance hop count to obtain a trust request data packet;

[0036] Sending the trust request data packet to the cluster head sensor node, so that the cluster head sensor node determines a plurality of initial recommended nodes and the number of demand nodes according to the trust request data packet, and determines the distance between each initial recommended node according to the trust metric parameter of each initial recommended node, performs community division according to the initial distance between each initial recommended node, determines a plurality of candidate recommended nodes and the initial influence of each candidate recommended node in the plurality of initial recommended nodes according to the community division result and the number of demand nodes, and sends the initial influence of each candidate recommended node to each candidate recommended node, determines a propagation path and a target influence according to the initial influence of each candidate recommended node through each candidate recommended node, and feeds back the recommendation information and the target influence according to the propagation path;

[0037] Upon receiving the recommendation information and the target influence fed back by each candidate recommendation node, determining the total influence value according to the target influence of each candidate recommendation node;

[0038] Calculate the trust value based on the total influence value, the target influence of each candidate recommendation node, and the direct trust value of each candidate recommendation node to determine the indirect trust value of the trust object;

[0039] A trust propagation decision is made according to the indirect trust value of the trust object.

[0040] In addition, to achieve the above object, the present invention also proposes a cluster head sensor node, the cluster head sensor node comprising:

[0041] A processing module, configured to determine a plurality of initial recommended nodes and a number of required nodes according to a trust request data packet when receiving a trust request data packet from a trust subject;

[0042] The processing module is further used to determine the initial distance between each initial recommended node according to the trust metric parameter of each initial recommended node;

[0043] A division module, used to divide the community according to the initial distance between each initial recommendation node, and determine multiple candidate recommendation nodes and the initial influence of each candidate recommendation node from multiple initial recommendation nodes according to the community division result and the number of demand nodes;

[0044] A sending module is used to send the initial influence of each candidate recommendation node to each candidate recommendation node, determine the propagation path and target influence according to the initial influence of each candidate recommendation node through each candidate recommendation node, and send the recommendation information and the target influence to the trust subject according to the propagation path, so that the trust subject makes a trust propagation decision based on the target influence of each candidate recommendation node and the recommendation information of each candidate recommendation node.

[0045] In addition, to achieve the above object, the present invention further proposes a recommendation node, the recommendation node comprising:

[0046] A processing module is used for determining a path according to the initial influence and the target generalized maximum flow strategy to obtain a propagation path and a target influence when receiving the initial influence sent by the cluster head sensor node, wherein the initial influence is obtained by the cluster head sensor node determining a plurality of initial recommended nodes and the number of required nodes according to the trust request data packet sent by the trust subject, determining the distance between each initial recommended node according to the trust metric parameter of each initial recommended node, performing community division according to the initial distance between each initial recommended node, and calculating the community division result;

[0047] A sending module is used to send the recommendation information and the target influence to the trust subject according to the propagation path, so that the trust subject makes a trust propagation decision according to the target influence and the recommendation information.

[0048] In addition, to achieve the above object, the present invention further proposes a trust subject, the trust subject comprising:

[0049] A processing module, configured to determine the number of required nodes and the target number of receiving hops according to the data transmission instruction when receiving the data transmission instruction;

[0050] An encapsulation module is used to encapsulate the number of required nodes and the target acceptance hop count to obtain a trust request data packet;

[0051] A sending module, used for sending the trust request data packet to the cluster head sensor node, so that the cluster head sensor node determines multiple initial recommended nodes and the number of demand nodes according to the trust request data packet, and determines the distance between each initial recommended node according to the trust measurement parameter of each initial recommended node, performs community division according to the initial distance between each initial recommended node, determines multiple candidate recommended nodes and the initial influence of each candidate recommended node in the multiple initial recommended nodes according to the community division result and the number of demand nodes, and sends the initial influence of each candidate recommended node to each candidate recommended node, determines the propagation path and the target influence according to the initial influence of each candidate recommended node through each candidate recommended node, and feeds back the recommendation information and the target influence according to the propagation path;

[0052] The processing module is further configured to determine the total influence value according to the target influence of each candidate recommendation node when receiving the recommendation information fed back by each candidate recommendation node and the target influence;

[0053] A calculation module, used to calculate the trust value according to the total influence value, the target influence of each candidate recommendation node and the direct trust value of each candidate recommendation node, so as to determine the indirect trust value of the trust object;

[0054] The decision module is used to make a trust propagation decision according to the indirect trust value of the trust object.

[0055] In addition, to achieve the above purpose, the present invention also proposes a trust propagation system, which includes: a cluster head sensor node, a recommendation node and a trust subject.

[0056] The present invention is applied to a cluster head sensor node, comprising: upon receiving a trust request data packet of a trust subject, determining a plurality of initial recommended nodes and the number of demand nodes according to the trust request data packet; determining the initial distance between each initial recommended node according to the trust metric parameter of each initial recommended node; performing community division according to the initial distance between each initial recommended node, and determining a plurality of candidate recommended nodes and the initial influence of each candidate recommended node in the plurality of initial recommended nodes according to the community division result and the number of demand nodes; sending the initial influence of each candidate recommended node to each candidate recommended node, determining a propagation path and a target influence according to the initial influence of each candidate recommended node through each candidate recommended node, and sending recommendation information and the target influence to the trust subject according to the propagation path, so that the trust subject makes a trust propagation decision according to the target influence of each candidate recommended node and the recommendation information of each candidate recommended node. Through the above method, the initial distance between each initial recommendation node is determined according to the trust measurement parameter of each initial recommendation node, and the community is divided based on the initial distance. The community division result and the number of demand nodes are used to determine multiple candidate recommendation nodes and the initial influence of each candidate recommendation node, and the initial influence of each candidate recommendation node is sent to each candidate recommendation node, so that the candidate recommendation node feeds back the recommendation information and the target influence to the trust subject, so that the trust subject makes a trust propagation decision, selects a suitable recommendation node, minimizes the influence distortion at the recommendation node, and establishes a suitable propagation path to minimize the propagation distortion, thereby improving the accuracy of the trust subject in making trust decisions and the security of data transmission. BRIEF DESCRIPTION OF THE DRAWINGS

[0057] Figure 1 It is a flowchart of the first embodiment of the trust propagation method of the present invention;

[0058] Figure 2 A schematic diagram of an underwater acoustic sensor network according to an embodiment of a trust propagation method of the present invention;

[0059] Figure 3 This is a flow chart of a second embodiment of the trust propagation method of the present invention;

[0060] Figure 4 It is a flowchart of the third embodiment of the trust propagation method of the present invention;

[0061] Figure 5 This is a flow chart of a fourth embodiment of the trust propagation method of the present invention;

[0062] Figure 6 is a structural block diagram of a first embodiment of a cluster head sensor of the present invention;

[0063] Figure 7 It is a structural block diagram of the first embodiment of the recommended node of the present invention;

[0064] Figure 8 It is a structural block diagram of the first embodiment of the trusted subject of the present invention.

[0065] The realization of the purpose, functional features and advantages of the present invention will be further explained in conjunction with embodiments and with reference to the accompanying drawings. DETAILED DESCRIPTION

[0066] It should be understood that the specific embodiments described herein are only used to explain the present invention, and are not used to limit the present invention.

[0067] The embodiment of the present invention provides a trust propagation method, referring to Figure 1 , Figure 1 The figure is a flow chart of a first embodiment of a trust propagation method according to the present invention.

[0068] A trust propagation method is applied to a cluster head sensor node, and the trust propagation method comprises the following steps:

[0069] Step S10: upon receiving a trust request data packet from a trust subject, determining a plurality of initial recommended nodes and a number of required nodes according to the trust request data packet.

[0070] It should be noted that the execution subject of this embodiment is the cluster head sensor node in the trust propagation system, and the trust propagation system includes the cluster head sensor node, a trust subject, multiple recommendation nodes, multiple relay nodes and a trust object.

[0071] It is understandable that the trust management methods in underwater acoustic sensor networks in the past are mostly trust recommendations and trust calculation methods within a one-hop range. However, the dynamic topology and sparse interaction characteristics of underwater acoustic sensor networks usually lead to a small number of single-hop recommendation information. Therefore, the trust subject needs to request multi-hop recommendation information. The recommendation sensor node encapsulates the recommendation information in a data packet and transmits it to the trust subject hop by hop through the relay sensor node. The trust subject calculates the indirect trust of the trust object based on the recommendation information and the attenuated influence, that is, to achieve trust decision on the trust object. This process is called trust propagation. The trust propagation method has a great influence on the accuracy of trust decision.

[0072] In the specific implementation, most trust propagation methods in the prior art only consider the establishment method of the propagation path, ignoring the selection of appropriate recommended sensor nodes. Since wrong recommendations and dishonest recommendations often occur in underwater acoustic sensor networks, in order to avoid unreliable recommendations from being transmitted in the network, it is necessary to select appropriate recommendation nodes. Although some researchers have proposed to select sensor nodes with higher trust values ​​to participate in the trust recommendation process, they ignore the existence of wrong recommendations and dishonest recommendations. At the same time, most current trust propagation methods are committed to solving the establishment of propagation paths. However, when establishing propagation paths, the spatial attenuation of the influence of recommended nodes is often ignored. Even though there are a few studies on the spatial attenuation of the influence of recommended sensor nodes, the spatial attenuation characteristics of the influence of recommended sensor nodes and the dependency of propagation paths are ignored.

[0073] It should be noted that the application scenario network structure corresponding to the trust propagation method proposed in this embodiment is an underwater acoustic sensor network deployed in a three-dimensional space. The network structure diagram is as follows: Figure 2 As shown in the figure, underwater sensor nodes are randomly distributed in the underwater three-dimensional environment. The sensor nodes are responsible for sensing the network environment, collecting monitoring data, and encapsulating the data in data packets and transmitting them hop by hop to the surface control center. In the trust propagation method, due to the large number of recommendation nodes, it is not necessary for all recommendation nodes to upload recommendation information. At this time, the possibility of erroneous and dishonest recommendations being transmitted in the network should be reduced. At the same time, the spatial attenuation characteristics of the influence of the recommendation node and the propagation path are not isolated, and the spatial attenuation characteristics of the influence are not only related to the distance, but also related to the trust attributes of the node. Therefore, when establishing a propagation path, the trust attributes of the relay nodes should also be considered.

[0074] It can be understood that in this embodiment, the scenario where the trusted subject sensor node needs to request multi-hop recommendation information is considered, and the sensor nodes in the network are clustered according to the communication range, residual energy, storage capacity, computing power, and other factors of the nodes, and the sensor nodes with higher comprehensive capabilities are selected as cluster head sensor nodes. The trusted subject sensor node uses the trust relationship table maintained by the cluster head sensor node to find the recommended nodes outside the multi-hop range. The cluster head sensor node maintains a trust relationship table and a location information table. The trust relationship table maintains the direct trust values ​​between nodes within the cluster and between nodes in adjacent clusters. Adjacent cluster heads regularly exchange intra-cluster trust relationships, which are expressed as , where Node represents the set of nodes in the cluster and the nodes in the adjacent clusters, represents the direct trust value of any node i to any node j; the location information table contains the location information of the nodes in the cluster and the nodes in the adjacent clusters, which is expressed as are the coordinates of the node.

[0075] In the specific implementation, when the trust subject sensor node needs to transmit data with the trust object sensor node, it first needs to make a trust decision on the trust object, calculate the trust value of the trust object, and determine whether it is trustworthy. Due to the sparse interaction of the underwater acoustic sensor network and other reasons, the trust subject lacks direct trust evidence about the trust object, resulting in the trust calculation requiring the recommendation information of the recommended node to calculate the indirect trust value of the trust object for trust decision. The trust subject determines the number of recommended nodes required according to the application requirements. The maximum number of hops of an acceptable recommended node is , and and Encapsulated in a trust request packet, the trust request packet is forwarded to the cluster head sensor node.

[0076] It should be noted that when the cluster head sensor node receives the trust request data packet from the trust subject, it determines the trust object and the number of required nodes according to the trust request data packet. The number of required nodes is the number of recommended nodes determined by the trust subject according to the application requirements. At the same time, the cluster head sensor node selects multiple initial recommended nodes that have direct trust values ​​with the trusted object from the trust relationship table stored in itself.

[0077] Step S20: determining the initial distance between each initial recommended node according to the trust metric parameter of each initial recommended node.

[0078] It should be noted that the trust metric parameters include but are not limited to spatial proximity, node reliability, and link reliability. The initial distance between any initial recommended node i and initial recommended node j can be calculated based on the trust metric parameters of each initial recommended node. ,in, is the link reliability of the initial recommended node i, The node reliability of the initial recommended node i, is the spatial proximity of the initial recommended node i. Similarly, the other parameters involved in the formula are the trust measurement parameters of the initial recommended node j.

[0079] In a specific implementation, in order to accurately calculate the trust metric parameters of each initial recommended node, further, the initial distance between each initial recommended node is determined based on the trust metric parameters of each initial recommended node, including: calculating the proximity based on the number of propagation hops between each initial recommended node and the trusted subject and the target acceptance hops of the trusted subject, and determining the spatial proximity of each initial recommended node; calculating the reliability based on the number of interactions and the interaction data packet information between each initial recommended node and the trusted object, and determining the node reliability of each initial recommended node; calculating the reliability based on the link quality between each initial recommended node and the trusted object, and determining the link reliability of each initial recommended node; and calculating the distance based on the spatial proximity of each initial recommended node, the node reliability of each initial recommended node, and the link reliability of each initial recommended node, and determining the initial distance between each initial recommended node.

[0080] It should be noted that since the recommendation information is transmitted hop by hop from the recommendation node to the trust subject, the spatial attenuation of the influence of the recommendation node occurs in each hop-by-hop transmission process. Therefore, in this embodiment, the number of hops from the node to the trust subject is used to describe the spatial proximity of the recommendation node. The smaller the number of hops from the recommendation node to the trust subject, the smaller the number of times the spatial attenuation is experienced, and the end-to-end delay of transmitting the recommendation information is smaller. The propagation hops from each initial recommendation node to the trust subject and the maximum hops of the recommendation node acceptable to the trust subject are obtained. The maximum hops of the recommendation node acceptable to the trust subject is the target acceptance hops, thereby calculating the spatial proximity of each initial recommendation node. The initial recommendation node i is placed in the spatial proximity as ,in, is the number of propagation hops from the initial recommendation node i to the trust subject, is the target acceptance hop count.

[0081] It can be understood that the node reliability of each initial recommendation node is defined according to the interaction persistence and interaction frequency between each initial recommendation node and the trusted object, and the number of interactions between each initial recommendation node and the trusted object and the interaction data packet information are obtained. The interaction data packet information includes but is not limited to the number of data packets sent by both communicating parties each time the initial recommendation node interacts with the trusted object, the maximum number of data packets that can be sent by both communicating parties, and the ratio of data packets sent by both communicating parties in each interaction. The node reliability of each initial recommendation node can be determined by calculating the reliability based on the number of interactions between each initial recommendation node and the trusted object and the interaction data packet information. The node reliability of initial recommendation node i is ,in , is the number of interactions between the initial recommendation node i and the trusted object, reflecting the frequency of interaction. is the number of packets sent by both parties in the mth interaction, It indicates the maximum number of data packets that can be sent by both parties in a time slot in theory. It reflects the interaction persistence of the two parties in each interaction process. It represents the proportion of data packets sent by a certain communication party in the mth interaction. In this embodiment, the "certain communication party" refers to the initial recommended node.

[0082] In the specific implementation, CATM is used to define the link reliability of the initial recommended node according to the link quality between each initial recommended node and the trusted object. The link reliability of the initial recommended node i is ,in, is the grey relational degree relative to the best reference sequence and the worst reference sequence.

[0083] It should be noted that the distance calculation is performed based on the spatial proximity of each initially recommended node, the node reliability of each initially recommended node, and the link reliability of each initially recommended node to determine the initial distance between any two initially recommended nodes.

[0084] Step S30: performing community division according to the initial distances between the initial recommended nodes, and determining a plurality of candidate recommended nodes and the initial influence of each candidate recommended node from the plurality of initial recommended nodes according to the community division result and the number of demand nodes.

[0085] It should be noted that after determining the initial distance between each initial recommended node, the cluster head sensor node selects the recommended node by using the idea of ​​finding the community structure, that is, dividing the node into multiple communities according to a certain metric parameter, and the selected node has a "higher" value on this metric parameter. This application divides nodes with similar node reliability, link reliability and spatial proximity into the same community. In this embodiment, the Attractor algorithm for dividing the community is used. The Attractor algorithm believes that the factors affecting the community division mainly include three aspects: (i) the influence of direct neighbors (Directed linked Nodes, DN); (ii) the influence of common neighbors (Common Neighbors, CN); (iii) the influence of each neighbor (Exclusive Neighbors, EN), and finally obtains the community division result.

[0086] It can be understood that the initial recommended nodes with the highest number of demand nodes are determined as candidate recommended nodes based on the community division results, and the initial influence of each candidate recommended node is calculated. The initial influence of each candidate recommended node reflects the influence of the candidate recommended node in the current trust evaluation task.

[0087] Step S40: Send the initial influence of each candidate recommendation node to each candidate recommendation node, determine the propagation path and target influence according to the initial influence of each candidate recommendation node through each candidate recommendation node, and send the recommendation information and the target influence to the trust subject according to the propagation path, so that the trust subject makes a trust propagation decision according to the target influence of each candidate recommendation node and the recommendation information of each candidate recommendation node.

[0088] It should be noted that the initial influence of each candidate recommendation node is sent to each candidate recommendation node. Each candidate recommendation node establishes a propagation path by solving the generalized maximum flow and determines the target influence of each candidate recommendation node. Each candidate recommendation node sends the recommendation information to the trust subject according to the propagation path. The trust subject calculates the indirect trust value of the trust object according to the target influence of the candidate recommendation node, realizes the trust decision on the trust object, and judges whether the trust object is trustworthy and whether network data transmission can be carried out with it.

[0089] The trust propagation method of this embodiment is applied to a cluster head sensor node, including: when receiving a trust request data packet from a trust subject, determining multiple initial recommended nodes and the number of demand nodes according to the trust request data packet; determining the initial distance between each initial recommended node according to the trust measurement parameter of each initial recommended node; performing community division according to the initial distance between each initial recommended node, and determining multiple candidate recommended nodes and the initial influence of each candidate recommended node in the multiple initial recommended nodes according to the community division result and the number of demand nodes; sending the initial influence of each candidate recommended node to each candidate recommended node, determining a propagation path and a target influence according to the initial influence of each candidate recommended node through each candidate recommended node, and sending the recommendation information and the target influence to the trust subject according to the propagation path, so that the trust subject makes a trust propagation decision according to the target influence of each candidate recommended node and the recommendation information of each candidate recommended node. Through the above method, the initial distance between each initial recommendation node is determined according to the trust measurement parameter of each initial recommendation node, and the community is divided based on the initial distance. The community division result and the number of demand nodes are used to determine multiple candidate recommendation nodes and the initial influence of each candidate recommendation node, and the initial influence of each candidate recommendation node is sent to each candidate recommendation node, so that the candidate recommendation node feeds back the recommendation information and the target influence to the trust subject, so that the trust subject makes a trust propagation decision, selects a suitable recommendation node, minimizes the influence distortion at the recommendation node, and establishes a suitable propagation path to minimize the propagation distortion, thereby improving the accuracy of the trust subject in making trust decisions and the security of data transmission.

[0090] Reference Figure 3 , Figure 3 This is a flow chart of a second embodiment of a trust propagation method according to the present invention.

[0091] Based on the first embodiment, step S30 in the trust propagation method of this embodiment includes:

[0092] Step S31: determining the adjacent nodes of each initially recommended node according to the initial distance between each initially recommended node.

[0093] It should be noted that before the cluster head sensor node searches for the community structure, it first determines the definition of "neighbors". Intuitively, "neighbors" refer to people with close spatial distances. According to the initial distances between the initial recommended nodes, this embodiment defines the "neighbors" of the initial recommended node i as: ,in , according to the above definition, determine the neighbors of each initial recommended node, that is, the adjacent nodes of each initial recommended node.

[0094] Step S32: performing influence calculation according to the initial distances between the initial recommended nodes and the adjacent nodes of the initial recommended nodes, and determining the direct influence value, the public influence value and the adjacent influence value of each initial recommended node.

[0095] It should be noted that the direct influence value is the influence DN of the nodes directly connected to each initial recommended node, the public influence value is the influence CN of the public neighbor nodes of each initial recommended node, and the adjacent influence value is the influence EN of each neighbor of each initial recommended node. , , , ,in, is the initial distance sum between the initial recommended node i and all adjacent nodes, that is, ; ; Represents the cohesion parameter.

[0096] Step S33: performing distance calculation according to the direct influence value, the common influence value and the adjacent influence value of each initial recommended node to determine the target distance between each initial recommended node.

[0097] It should be noted that the target distance between each initial recommended node is determined by performing distance calculation based on the direct influence value, public influence value, and adjacent influence value of each initial recommended node. For example, the target distance between initial recommended node i and initial recommended node j is expressed as follows: .

[0098] Step S34: determining a plurality of candidate recommended nodes and an initial influence of each candidate recommended node from among the plurality of initial recommended nodes according to the target distance between the respective initial recommended nodes and the number of demand nodes.

[0099] It should be noted that in order to accurately select multiple candidate recommendation nodes and determine the initial influence of each candidate recommendation node, further, the multiple candidate recommendation nodes and the initial influence of each candidate recommendation node are determined among the multiple initial recommendation nodes according to the target distance between the initial recommendation nodes and the number of demand nodes, including: performing relative distance calculation according to the target distance between the initial recommendation nodes to determine the relative distance of each initial recommendation node; sorting the relative distance of each initial recommendation node, and determining multiple candidate recommendation nodes among the multiple initial recommendation nodes according to the sorting result and the number of demand nodes; performing influence calculation according to the relative distance of each candidate recommendation node to determine the initial influence of each candidate recommendation node.

[0100] It can be understood that the relative distance calculation is performed based on the target distance between each initial recommended node to determine the relative distance of each initial recommended node. For example, in a trust assessment task, for any candidate recommended node i, its relative distance in this trust assessment task is ,in, is the sum of the distances to all other initially recommended nodes, is the sum of the distances of any initial recommended node pair in the network.

[0101] In the specific implementation, the relative distances of the initial recommended nodes are sorted in ascending order, and the initial recommended nodes with the highest number of nodes are selected as the candidate recommended nodes. The initial influence of the candidate recommended nodes is calculated based on the target distance of the candidate recommended nodes. The initial influence of the candidate recommended node u is The greater the initial influence, the higher the node reliability and link reliability of the candidate recommendation node, and the smaller the number of hops from the trust subject. Therefore, the candidate recommendation node is less likely to transmit wrong recommendations and dishonest recommendations in the network, and the end-to-end delay for transmitting recommendation information is smaller.

[0102] This embodiment determines the adjacent nodes of each initial recommended node according to the initial distance between each initial recommended node; performs influence calculation according to the initial distance between each initial recommended node and the adjacent nodes of each initial recommended node to determine the direct influence value, public influence value and adjacent influence value of each initial recommended node; performs distance calculation according to the direct influence value, public influence value and adjacent influence value of each initial recommended node to determine the target distance between each initial recommended node; determines multiple candidate recommended nodes and the initial influence of each candidate recommended node among multiple initial recommended nodes according to the target distance between each initial recommended node and the number of demand nodes. Through the above method, it is possible to ensure the selection of accurate candidate recommended nodes and accurately calculate the initial influence of each candidate recommended node.

[0103] refer to Figure 4 , Figure 4 This is a flow chart of a third embodiment of a trust propagation method according to the present invention.

[0104] The trust propagation method of this embodiment is applied to the recommendation node, and includes the following steps:

[0105] Step S01: upon receiving the initial influence sent by the cluster head sensor node, a path is determined according to the initial influence and the target generalized maximum flow strategy to obtain a propagation path and a target influence, wherein the initial influence is obtained by the cluster head sensor node determining a plurality of initial recommended nodes and the number of demand nodes according to a trust request data packet sent by a trust subject, and determining the distance between each initial recommended node according to a trust metric parameter of each initial recommended node, performing community division according to the initial distance between each initial recommended node, and calculating the community division result.

[0106] It should be noted that the execution subject of this embodiment is a recommendation node in the trust propagation system, and the trust propagation system includes a cluster head sensor node, a trust subject, multiple recommendation nodes, multiple relay nodes and a trust object.

[0107] It can be understood that the cluster head sensor node sends the initial influence of each candidate recommended node to each candidate recommended node. When each candidate recommended node receives the initial influence sent by the cluster head sensor, each candidate recommended node determines the path according to its own initial influence and the target generalized maximum flow strategy, thereby obtaining the propagation path and the target influence. The target generalized maximum flow strategy refers to abstracting the propagation path establishment problem into a generalized maximum flow problem, and establishing the propagation path by solving the generalized maximum flow.

[0108] In the specific implementation, the initial influence of each candidate recommended node is that the cluster head sensor node determines multiple initial recommended nodes and the number of demand nodes according to the trust request data packet sent by the trust subject, and determines the initial distance between each initial recommended node according to the trust measurement parameter of each initial recommended node. Then, the initial distance between each initial recommended node is used to divide the community. According to the community division result and the number of demand nodes, multiple candidate recommended nodes are selected from multiple initial recommended nodes, and the initial influence of each candidate recommended node is determined.

[0109] It should be noted that in order to accurately obtain the propagation path and target influence, further, the path is determined according to the initial influence and the target generalized maximum flow strategy to obtain the propagation path and target influence, including: defining the initial flow according to the initial influence to determine the initial flow; determining the arc capacity between the candidate recommended node according to the direct trust value; calculating the gain factor according to the number of propagation hops between the trusted subject, the current traversal hops, and the arc capacity between the candidate recommended node to determine the arc gain factor; calculating the output flow according to the gain factor and the initial flow to determine the node output flow; determining the target influence according to the node output flow; determining the propagation path according to the target influence, the arc gain factor and the arc capacity.

[0110] It is understandable that in the maximum network flow problem, the flow does not incur cost but is limited by the arc capacity. However, there is an implicit assumption in the maximum network flow problem, that is, except for the source node and the sink node, the flow flowing into each node is equal to the flow flowing out of the node. However, in actual application scenarios, the flow is not necessarily conserved when passing through each node, and the "flow leakage" phenomenon will occur. In order to describe the above phenomenon, the generalized maximum flow problem is generated, which considers the "flow leakage" phenomenon and studies how to transmit as many flows as possible from the source node to the sink node. After the recommended node is selected, the remaining question is how to establish a propagation path from the recommended node to the trusted subject. Since the influence of the recommended node will experience spatial attenuation during the propagation process, it is necessary to establish a suitable propagation path to propagate the influence to the trusted subject with the maximum degree of distortion. The influence is initiated by the recommended node and propagated to the trusted subject through multi-hop relay nodes. This process is similar to the process of flow transmission, and the spatial attenuation characteristics of the influence are similar to the "flow leakage" phenomenon during the flow transmission process. Therefore, this embodiment abstracts the propagation path establishment problem into a generalized maximum flow problem, and establishes the propagation path by solving the generalized maximum flow.

[0111] In the specific implementation, the generalized maximum flow problem is described as follows: Let G = {V, E} represent a directed acyclic graph, where V is the set of nodes in G and E is the set of arcs in G. In network flow theory, "arcs" are no longer used to describe the paths connecting nodes. Assume that u and v represent the source node and sink node in G respectively. The flow flows out from the source node u and finally flows into the sink node v. In the generalized maximum network flow problem, the following basic concepts are involved: 1) Capacity: The capacity of the arc from node i to node j is the following mapping relationship , expressed as , capacity is limited The maximum flow that can pass through; 2) Gain factor: The gain factor of the arc from node i to node j is the following mapping relationship: , expressed as The physical meaning of the gain factor is “the flow through After the flow leakage occurs, there is still a flow of g(i, j) left. The symmetry condition is satisfied, that is ; 3) Generalized flow: The generalized flow from node i to node j is the following mapping relationship: , expressed as . Generalized Flow and The following two constraints are met: (i) Capacity limit: ; (ii) Generalized antisymmetric constraints: ; 4) Residual capacity function: The residual capacity function is the following mapping relationship: , defined as: ; 5) Residual network: The residual network is defined as In the residual network, there may be arcs with a capacity of 0; 6) Residual arcs: Residual edges refer to arcs with a capacity greater than 0. All arcs with a capacity greater than 0 are represented as a set ; 7) Residual path: The residual path is really a path composed entirely of residual arcs; 8) Network flow: The flow of network G is expressed as: , where the generalized maximum network flow problem studies the maximization of F; 9) Augmenting path: The augmenting path is the residual path from the source node u to the node sink v, if If there is an augmenting path in , then F can be improved by sending the flow along this path. When there is no augmenting path in , F reaches its maximum value; 10) Theorem: if and only if When there is no augmenting path, flow F is the maximum flow. If there is an augmenting path, then F can be improved by sending the flow along this path.

[0112] It should be noted that, from the description of the generalized maximum flow problem, it can be seen that in the generalized maximum flow problem: (i) the flow is transmitted from the source node to the sink node, and the arc between any nodes has a capacity limit, and the flow flowing through the two nodes cannot be greater than this capacity limit. In this embodiment, the recommended node to be selected is the source node, and the trusted subject is the sink node; (ii) "flow leakage" occurs when the flow is transmitted on the arc, and the size of the leaked flow is determined by the gain factor of this arc. In order to establish a propagation path based on solving the maximum flow problem, it is necessary to clarify the following three variables, namely the initial flow size , Arc Capacity and gain factor .

[0113] It can be understood that in this embodiment, the size of the initial flow is related to the initial influence of the candidate recommended node in the current trust evaluation task, and the initial flow of each candidate recommended node can be determined according to the initial influence of each candidate recommended node. The initial flow size is set as follows: ,in is the initial influence of each candidate recommendation node. The arc capacity of each candidate recommendation node can be determined according to the direct trust value between the candidate recommendation nodes. When the flow flows from candidate recommendation node i to node j, the flow is limited by the arc capacity, which is defined as follows: ,in is the direct trust value of candidate recommendation node i to candidate recommendation node j.

[0114] In the specific implementation, when calculating the gain factor, since the influence of the candidate recommended node has a spatial attenuation characteristic, that is, before the recommended information reaches the trusted object, the influence of the recommended sensor node will be attenuated to a certain extent after each hop of relay sensor node. This embodiment compares this phenomenon to "flow leakage". In order to describe this phenomenon, it is necessary to define the arc gain factor , the existing gain factor representation method only considers the distance factor of the node. However, in the actual trust assessment task, the gain factor model that only considers the distance is not enough to reflect the differences between the relay nodes. When transmitting data packets, it is more desirable to transmit data packets through trusted relay nodes, so the trust value of the relay node also needs to be considered. Since the relay node is only used to transmit data packets and does not participate in the trust decision-making process, this embodiment only reflects the "flow leakage" size of the current arc based on the trust value of the relay node and its distance from the trusted object. The gain factor model established in this embodiment is shown as follows: , where hop represents the number of propagation hops from the candidate recommendation node u to the trust subject, Indicates the number of hops that have been traversed. represents the trust value between any adjacent nodes. From the above formula, we can see that when At the same time, The bigger, The larger the value, the less the relay node with higher trust value "leaks"; At the same time, The larger the value, the closer it is to the trusted subject and the less "leakage", which is in line with common sense. After hops, the initial flow will decay into the following form, that is, the node output flow size of the candidate recommended node i is: ,

[0115] According to the above formula, the target influence of the candidate recommended node u is:

[0116] .

[0117] It should be noted that in this embodiment, each candidate recommendation node selects a propagation path from the candidate recommendation node to the trust subject with the maximum target influence according to the arc capacity and the arc gain factor, so that the influence is propagated to the trust subject with the maximum degree of undistortion.

[0118] Step S02: sending the recommendation information and the target influence to the trust subject according to the propagation path, so that the trust subject makes a trust propagation decision according to the target influence and the recommendation information.

[0119] It should be noted that the initial influence of the candidate recommended node u undergoes spatial attenuation when propagating along the propagation path, and its attenuated target influence is given by As shown in Figure 2, the recommendation information of the candidate recommendation node is also transmitted to the trust subject along the propagation path. The trust subject calculates the indirect trust of the trust object based on the target influence of the recommendation node after attenuation, that is, it implements the trust decision on the trust object and determines whether the trust object is trustworthy and whether network data transmission can be carried out with it.

[0120] It can be understood that in this embodiment, a trust propagation path is established based on the generalized maximum flow strategy, and the spatial attenuation characteristics of the influence of the candidate recommended node are analogized to the "flow leakage" phenomenon during the propagation process. The trust propagation path is established by finding the generalized maximum flow. In addition, when setting the gain factor of the arc, this embodiment not only considers the distance factor of the sensor node, but also considers the trust value of the relay sensor node, that is, a reliable and short-distance propagation path is established.

[0121] The trust propagation method of this embodiment is applied to the recommendation node, including: upon receiving the initial influence sent by the cluster head sensor node, determining the path according to the initial influence and the target generalized maximum flow strategy to obtain the propagation path and the target influence, wherein the initial influence is the number of multiple initial recommended nodes and demand nodes determined by the cluster head sensor node according to the trust request data packet sent by the trust subject, and the distance between each initial recommended node is determined according to the trust metric parameter of each initial recommended node, and the community is divided according to the initial distance between each initial recommended node, and the result is calculated according to the community division result; according to the propagation path, the recommendation information and the target influence are sent to the trust subject, so that the trust subject makes a trust propagation decision according to the target influence and the recommendation information. In the above manner, the influence distortion at the recommendation node is minimized, and at the same time, a suitable propagation path is established to minimize the propagation distortion, thereby improving the accuracy of the trust subject in making trust decisions and the security of data transmission.

[0122] refer to Figure 5 , Figure 5 This is a flow chart of a fourth embodiment of a trust propagation method according to the present invention.

[0123] The trust propagation method of this embodiment is applied to a trust subject, and includes the following steps:

[0124] Step S11: upon receiving a data transmission instruction, determining the required number of nodes and the target acceptance hop number according to the data transmission instruction.

[0125] It should be noted that the execution subject of this embodiment is a trust subject in the trust propagation system, and the trust propagation system includes a cluster head sensor node, a trust subject, multiple recommendation nodes, multiple relay nodes and a trust object.

[0126] It is understandable that when the trust subject receives the data transmission instruction, it determines the trust object and determines the number of required nodes according to the application requirements attached to the data transmission instruction. and the target receiving hop count .

[0127] Step S12: encapsulate the required node quantity and the target acceptance hop count to obtain a trust request data packet.

[0128] It should be noted that the trust subject encapsulates the number of required nodes and the target acceptance hop count in a trust request data packet, and forwards the trust request data packet to the cluster head sensor node.

[0129] Step S13: Send the trust request data packet to the cluster head sensor node, so that the cluster head sensor node determines multiple initial recommended nodes and the number of demand nodes according to the trust request data packet, and determines the distance between each initial recommended node according to the trust measurement parameter of each initial recommended node, performs community division according to the initial distance between each initial recommended node, determines multiple candidate recommended nodes and the initial influence of each candidate recommended node in the multiple initial recommended nodes according to the community division result and the number of demand nodes, and sends the initial influence of each candidate recommended node to each candidate recommended node, determines the propagation path and the target influence according to the initial influence of each candidate recommended node through each candidate recommended node, and feeds back the recommendation information and the target influence according to the propagation path.

[0130] It should be noted that the trust subject sends a trust request data packet to the cluster head sensor node, and the cluster head sensor node determines multiple initial recommended nodes and the number of required nodes according to the trust request data packet sent by the trust subject, and determines the initial distance between each initial recommended node according to the trust measurement parameter of each initial recommended node, and then uses the initial distance between each initial recommended node to divide the community, selects multiple candidate recommended nodes from multiple initial recommended nodes according to the community division result and the number of required nodes, and determines the initial influence of each candidate recommended node. At the same time, the cluster head sensor node sends the initial influence of each candidate recommended node to each candidate recommended node.

[0131] It can be understood that each candidate recommendation node establishes a propagation path by solving the generalized maximum flow and determines the target influence of each candidate recommendation node, and each candidate recommendation node sends the recommendation information and the target influence to the trust subject according to the propagation path.

[0132] Step S14: upon receiving the recommendation information and the target influence fed back by each candidate recommendation node, determining the total influence value according to the target influence of each candidate recommendation node.

[0133] It should be noted that the trust subject receives the recommendation information and target influence from each candidate recommendation node. When , the sum of the target influences of all candidate recommended nodes can be calculated, and the sum of the target influences of the candidate recommended nodes is the total influence value. , .

[0134] Step S15: Calculate the trust value according to the total influence value, the target influence of each candidate recommendation node, and the direct trust value of each candidate recommendation node to determine the indirect trust value of the trust object.

[0135] It should be noted that the trust subject calculates the trust value based on the total influence value, the recommendation information of each candidate recommendation node, the target influence of each candidate recommendation node, and the direct trust value of each candidate recommendation node to determine the indirect trust value of the trust object. ,in is the recommendation information uploaded by the candidate recommendation node u, and its value is equal to the direct trust of the trust object calculated by the candidate recommendation node u, that is, .

[0136] Step S16: making a trust propagation decision according to the indirect trust value of the trusted object.

[0137] It should be noted that the trust subject makes a trust decision on the trust object based on the indirect trust value of the trust object, and determines whether the trust object is trustworthy and whether network data transmission can be carried out with it.

[0138] The trust propagation method of this embodiment is applied to a trust subject, comprising: upon receiving a data transmission instruction, determining the number of required nodes and the target acceptance hop count according to the data transmission instruction; performing data encapsulation on the number of required nodes and the target acceptance hop count to obtain a trust request data packet; sending the trust request data packet to a cluster head sensor node, so that the cluster head sensor node determines a plurality of initial recommended nodes and the number of required nodes according to the trust request data packet, and determines the distance between each initial recommended node according to the trust metric parameter of each initial recommended node, performs community division according to the initial distance between each initial recommended node, determines a plurality of candidate recommended nodes and the initial influence of each candidate recommended node in the plurality of initial recommended nodes according to the community division result and the number of required nodes, and sends the initial influence of each candidate recommended node to each candidate recommended node, determines a propagation path and a target influence according to the initial influence of each candidate recommended node through each candidate recommended node, and feeds back recommendation information and the target influence according to the propagation path; upon receiving the recommendation information and the target influence fed back by each candidate recommended node, determines the total influence value according to the target influence of each candidate recommended node;

[0139] The trust value is calculated based on the total influence value, the target influence of each candidate recommended node, and the direct trust value of each candidate recommended node to determine the indirect trust value of the trusted object; and the trust propagation decision is made based on the indirect trust value of the trusted object. Through the above method, the influence distortion at the recommended node is minimized, and at the same time, a suitable propagation path is established to minimize the propagation distortion, thereby improving the accuracy of the trust subject in making trust decisions and the security of data transmission.

[0140] In addition, refer to Figure 6 The embodiment of the present invention further provides a cluster head sensor node, the cluster head sensor node comprising:

[0141] The processing module 10 is used to determine a plurality of initial recommended nodes and a required number of nodes according to the trust request data packet when receiving a trust request data packet from a trust subject.

[0142] The processing module 10 is further configured to determine the initial distance between each initial recommended node according to the trust metric parameter of each initial recommended node.

[0143] The division module 20 is used to divide the community according to the initial distance between each initial recommendation node, and determine multiple candidate recommendation nodes and the initial influence of each candidate recommendation node among multiple initial recommendation nodes according to the community division result and the number of demand nodes.

[0144] The sending module 30 is used to send the initial influence of each candidate recommendation node to each candidate recommendation node, determine the propagation path and target influence according to the initial influence of each candidate recommendation node through each candidate recommendation node, and send the recommendation information and the target influence to the trust subject according to the propagation path, so that the trust subject makes a trust propagation decision based on the target influence of each candidate recommendation node and the recommendation information of each candidate recommendation node.

[0145] This embodiment determines multiple initial recommendation nodes and the number of demand nodes according to the trust request data packet when receiving a trust request data packet from a trust subject; determines the initial distance between each initial recommendation node according to the trust measurement parameter of each initial recommendation node; performs community division according to the initial distance between each initial recommendation node, and determines multiple candidate recommendation nodes and the initial influence of each candidate recommendation node among the multiple initial recommendation nodes according to the community division result and the number of demand nodes; sends the initial influence of each candidate recommendation node to each candidate recommendation node, determines a propagation path and a target influence according to the initial influence of each candidate recommendation node through each candidate recommendation node, and sends the recommendation information and the target influence to the trust subject according to the propagation path, so that the trust subject makes a trust propagation decision according to the target influence of each candidate recommendation node and the recommendation information of each candidate recommendation node. Through the above method, the initial distance between each initial recommendation node is determined according to the trust measurement parameter of each initial recommendation node, and the community is divided based on the initial distance. The community division result and the number of demand nodes are used to determine multiple candidate recommendation nodes and the initial influence of each candidate recommendation node, and the initial influence of each candidate recommendation node is sent to each candidate recommendation node, so that the candidate recommendation node feeds back the recommendation information and the target influence to the trust subject, so that the trust subject makes a trust propagation decision, selects a suitable recommendation node, minimizes the influence distortion at the recommendation node, and establishes a suitable propagation path to minimize the propagation distortion, thereby improving the accuracy of the trust subject in making trust decisions and the security of data transmission.

[0146] In one embodiment, the processing module 10 is further used to calculate the proximity according to the number of propagation hops between each initial recommendation node and the trust subject and the number of target acceptance hops of the trust subject, so as to determine the spatial proximity of each initial recommendation node;

[0147] The reliability of each initial recommended node is determined by calculating the reliability based on the number of interactions between each initial recommended node and the trusted object and the information of the interaction data packet;

[0148] Calculate the reliability of the link between each initial recommended node and the trusted object according to the link quality, and determine the link reliability of each initial recommended node;

[0149] Distance calculation is performed according to the spatial proximity of each initial recommended node, the node reliability of each initial recommended node, and the link reliability of each initial recommended node to determine the initial distance between each initial recommended node.

[0150] In one embodiment, the partitioning module 20 is further used to determine the adjacent nodes of each initially recommended node according to the initial distance between each initially recommended node;

[0151] Influence calculation is performed based on the initial distances between the initial recommended nodes and the adjacent nodes of each initial recommended node to determine the direct influence value, public influence value and adjacent influence value of each initial recommended node;

[0152] Distance calculation is performed based on the direct influence value, public influence value, and adjacent influence value of each initial recommended node to determine the target distance between each initial recommended node;

[0153] A plurality of candidate recommendation nodes and an initial influence of each candidate recommendation node are determined from the plurality of initial recommendation nodes according to the target distance between the respective initial recommendation nodes and the number of the demand nodes.

[0154] In one embodiment, the division module 20 is further used to perform relative distance calculation according to the target distance between each initial recommended node to determine the relative distance of each initial recommended node;

[0155] Sorting the relative distances of the initial recommended nodes, and determining a plurality of candidate recommended nodes from the plurality of initial recommended nodes according to the sorting results and the number of demand nodes;

[0156] The influence is calculated according to the relative distance of each candidate recommendation node to determine the initial influence of each candidate recommendation node.

[0157] Since the cluster head sensor node adopts all the technical solutions of all the above embodiments, it at least has all the beneficial effects brought by the technical solutions of the above embodiments, which will not be described one by one here.

[0158] In addition, refer to Figure 7 The embodiment of the present invention further provides a recommendation node, the recommendation node comprising:

[0159] Processing module 01 is used to determine the path according to the initial influence and the target generalized maximum flow strategy when receiving the initial influence sent by the cluster head sensor node, and obtain the propagation path and the target influence. The initial influence is that the cluster head sensor node determines the number of multiple initial recommended nodes and the number of demand nodes according to the trust request data packet sent by the trust subject, and determines the distance between each initial recommended node according to the trust measurement parameter of each initial recommended node, and divides the community according to the initial distance between each initial recommended node, and is calculated according to the community division result.

[0160] The sending module 02 is used to send the recommendation information and the target influence to the trust subject according to the propagation path, so that the trust subject makes a trust propagation decision according to the target influence and the recommendation information.

[0161] This embodiment determines the path according to the initial influence and the target generalized maximum flow strategy when receiving the initial influence sent by the cluster head sensor node, and obtains the propagation path and the target influence. The initial influence is that the cluster head sensor node determines the number of multiple initial recommended nodes and demand nodes according to the trust request data packet sent by the trust subject, and determines the distance between each initial recommended node according to the trust metric parameter of each initial recommended node, and divides the community according to the initial distance between each initial recommended node, and calculates it according to the community division result; sends the recommendation information and the target influence to the trust subject according to the propagation path, so that the trust subject makes a trust propagation decision according to the target influence and the recommendation information. Through the above method, the influence distortion at the recommended node is minimized, and at the same time, a suitable propagation path is established to minimize the propagation distortion, thereby improving the accuracy of the trust subject in making trust decisions and the security of data transmission.

[0162] In one embodiment, the processing module 01 is further used to define an initial flow according to the initial influence and determine the initial flow;

[0163] Determine the arc capacity between the candidate recommended node and the candidate recommended node according to the direct trust value;

[0164] The gain factor is calculated based on the number of propagation hops between the trusted subject, the current number of traversal hops, and the arc capacity between the recommended node to be selected, and the arc gain factor is determined;

[0165] Calculate the output flow according to the gain factor and the initial flow to determine the node output flow;

[0166] Determining a target influence according to the node output stream;

[0167] A propagation path is determined according to the target influence, the arc gain factor, and the arc capacity.

[0168] Since this recommended node adopts all the technical solutions of all the above embodiments, it has at least all the beneficial effects brought by the technical solutions of the above embodiments, which will not be described one by one here.

[0169] In addition, refer to Figure 8 The embodiment of the present invention further provides a trust subject, wherein the trust subject includes:

[0170] The processing module 11 is used to determine the required node quantity and the target acceptance hop number according to the data transmission instruction when receiving the data transmission instruction.

[0171] The encapsulation module 12 is used to encapsulate the required node quantity and the target acceptance hop count to obtain a trust request data packet.

[0172] The sending module 13 is used to send the trust request data packet to the cluster head sensor node, so that the cluster head sensor node determines multiple initial recommended nodes and the number of demand nodes according to the trust request data packet, and determines the distance between each initial recommended node according to the trust measurement parameter of each initial recommended node, performs community division according to the initial distance between each initial recommended node, determines multiple candidate recommended nodes and the initial influence of each candidate recommended node in the multiple initial recommended nodes according to the community division result and the number of demand nodes, and sends the initial influence of each candidate recommended node to each candidate recommended node, determines the propagation path and the target influence according to the initial influence of each candidate recommended node through each candidate recommended node, and feeds back the recommendation information and the target influence according to the propagation path.

[0173] The processing module 11 is further configured to determine a total influence value according to the target influence of each candidate recommendation node when receiving the recommendation information and the target influence fed back by each candidate recommendation node.

[0174] The calculation module 14 is used to calculate the trust value according to the total influence value, the target influence of each candidate recommendation node and the direct trust value of each candidate recommendation node, so as to determine the indirect trust value of the trust object.

[0175] The decision module 15 is used to make a trust propagation decision according to the indirect trust value of the trust object.

[0176] In this embodiment, when receiving a data transmission instruction, the number of required nodes and the target acceptance hop number are determined according to the data transmission instruction; the number of required nodes and the target acceptance hop number are data encapsulated to obtain a trust request data packet; the trust request data packet is sent to the cluster head sensor node, so that the cluster head sensor node determines a plurality of initial recommended nodes and the number of required nodes according to the trust request data packet, determines the distance between each initial recommended node according to the trust metric parameter of each initial recommended node, divides the community according to the initial distance between each initial recommended node, and determines a plurality of candidate recommended nodes and each candidate recommended node from the plurality of initial recommended nodes according to the community division result and the number of required nodes. Select the initial influence of the recommended node, and send the initial influence of each candidate recommended node to each candidate recommended node, determine the propagation path and target influence through each candidate recommended node according to the initial influence of each candidate recommended node, and feedback the recommendation information and the target influence according to the propagation path; when receiving the recommendation information and the target influence fed back by each candidate recommended node, determine the total influence value according to the target influence of each candidate recommended node; calculate the trust value according to the total influence value, the target influence of each candidate recommended node and the direct trust value of each candidate recommended node, and determine the indirect trust value of the trusted object; make a trust propagation decision according to the indirect trust value of the trusted object. Through the above method, the influence distortion at the recommended node is minimized, and at the same time, a suitable propagation path is established to minimize the propagation distortion, thereby improving the accuracy of the trust subject in making trust decisions and the security of data transmission.

[0177] Since the trust subject adopts all the technical solutions of all the above embodiments, it has at least all the beneficial effects brought by the technical solutions of the above embodiments, which will not be described one by one here.

[0178] It should be noted that the workflow described above is merely illustrative and does not limit the scope of protection of the present invention. In practical applications, technicians in this field can select part or all of them according to actual needs to achieve the purpose of the present embodiment, and no limitation is made here.

[0179] In addition, for technical details not fully described in this embodiment, reference can be made to the trust propagation method provided in any embodiment of the present invention, and will not be repeated here.

[0180] In addition, it should be noted that, in this article, the terms "include", "comprises" or any other variations thereof are intended to cover non-exclusive inclusion, so that a process, method, article or system including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, article or system. In the absence of further restrictions, an element defined by the sentence "comprises a ..." does not exclude the existence of other identical elements in the process, method, article or system including the element.

[0181] The serial numbers of the above embodiments of the present invention are only for description and do not represent the advantages or disadvantages of the embodiments.

[0182] Through the description of the above implementation methods, those skilled in the art can clearly understand that the above-mentioned embodiment methods can be implemented by means of software plus a necessary general hardware platform, and of course by hardware, but in many cases the former is a better implementation method. Based on such an understanding, the technical solution of the present invention is essentially or the part that contributes to the prior art can be embodied in the form of a software product, which is stored in a storage medium (such as a read-only memory (ROM) / RAM, a magnetic disk, or an optical disk), and includes a number of instructions for a terminal device (which can be a mobile phone, a computer, a server, or a network device, etc.) to execute the methods described in each embodiment of the present invention.

[0183] The above are only preferred embodiments of the present invention, and are not intended to limit the patent scope of the present invention. Any equivalent structure or equivalent process transformation made using the contents of the present invention specification and drawings, or directly or indirectly applied in other related technical fields, are also included in the patent protection scope of the present invention.

Claims

1. A trust propagation method, characterized in that: Applied to cluster head sensor nodes, the trust propagation method comprises: Upon receiving a trust request data packet from a trust subject, determining a plurality of initial recommended nodes and a number of required nodes according to the trust request data packet; Determine the initial distance between each initial recommendation node according to the trust metric parameter of each initial recommendation node; Performing community division according to the initial distances between the initial recommended nodes, and determining a plurality of candidate recommended nodes and the initial influence of each candidate recommended node from the plurality of initial recommended nodes according to the community division result and the number of demand nodes; The initial influence of each candidate recommendation node is sent to each candidate recommendation node, and each candidate recommendation node determines a propagation path and a target influence according to the initial influence of each candidate recommendation node, and sends the recommendation information and the target influence to the trust subject according to the propagation path, so that the trust subject makes a trust propagation decision according to the target influence of each candidate recommendation node and the recommendation information of each candidate recommendation node.

2. The trust propagation method according to claim 1, characterized in that: The determining the initial distance between each initial recommended node according to the trust metric parameter of each initial recommended node includes: Calculate the proximity according to the number of propagation hops between each initial recommendation node and the trust subject and the number of target acceptance hops of the trust subject, and determine the spatial proximity of each initial recommendation node; The reliability of each initial recommended node is determined by calculating the reliability based on the number of interactions between each initial recommended node and the trusted object and the information of the interaction data packet; Calculate the reliability of the link between each initial recommended node and the trusted object according to the link quality, and determine the link reliability of each initial recommended node; Distance calculation is performed according to the spatial proximity of each initial recommended node, the node reliability of each initial recommended node, and the link reliability of each initial recommended node to determine the initial distance between each initial recommended node.

3. The trust propagation method according to claim 1, characterized in that: The performing of community division according to the initial distances between the initial recommended nodes, and determining a plurality of candidate recommended nodes and the initial influence of each candidate recommended node from the plurality of initial recommended nodes according to the community division result and the number of demand nodes, includes: Determine the adjacent nodes of each initial recommended node according to the initial distance between each initial recommended node; Influence calculation is performed based on the initial distances between the initial recommended nodes and the adjacent nodes of each initial recommended node to determine the direct influence value, public influence value and adjacent influence value of each initial recommended node; Distance calculation is performed based on the direct influence value, public influence value, and adjacent influence value of each initial recommended node to determine the target distance between each initial recommended node; A plurality of candidate recommendation nodes and an initial influence of each candidate recommendation node are determined from the plurality of initial recommendation nodes according to the target distance between the respective initial recommendation nodes and the number of the demand nodes.

4. The trust propagation method according to claim 3, characterized in that: The determining of a plurality of candidate recommendation nodes and an initial influence of each candidate recommendation node from a plurality of initial recommendation nodes according to the target distance between each initial recommendation node and the number of demand nodes includes: Perform relative distance calculation according to the target distance between each initial recommended node to determine the relative distance of each initial recommended node; Sorting the relative distances of the initial recommended nodes, and determining a plurality of candidate recommended nodes from the plurality of initial recommended nodes according to the sorting results and the number of demand nodes; The influence is calculated according to the relative distance of each candidate recommendation node to determine the initial influence of each candidate recommendation node.

5. A trust propagation method, characterized in that: Applied to the recommendation node, the trust propagation method includes: When receiving the initial influence sent by the cluster head sensor node, the path is determined according to the initial influence and the target generalized maximum flow strategy to obtain the propagation path and the target influence. The initial influence is that the cluster head sensor node determines the number of multiple initial recommended nodes and the number of demand nodes according to the trust request data packet sent by the trust subject, and determines the distance between each initial recommended node according to the trust metric parameter of each initial recommended node, and divides the community according to the initial distance between each initial recommended node, and is calculated according to the community division result; The recommendation information and the target influence are sent to the trust subject according to the propagation path, so that the trust subject makes a trust propagation decision according to the target influence and the recommendation information.

6. The trust propagation method according to claim 5, characterized in that: The path determination is performed according to the initial influence and the target generalized maximum flow strategy to obtain the propagation path and the target influence, including: Defining an initial flow according to the initial influence and determining the initial flow; Determine the arc capacity between the candidate recommended node and the candidate recommended node according to the direct trust value; The gain factor is calculated based on the number of propagation hops between the trusted subject, the current number of traversal hops, and the arc capacity between the recommended node to be selected, and the arc gain factor is determined; Calculate the output flow according to the gain factor and the initial flow to determine the node output flow; Determining a target influence according to the node output stream; A propagation path is determined according to the target influence, the arc gain factor, and the arc capacity.

7. A trust propagation method, characterized in that: Applied to a trust subject, the trust propagation method includes: Upon receiving a data transmission instruction, determining the number of required nodes and the target number of receiving hops according to the data transmission instruction; Encapsulating the required node quantity and the target acceptance hop count to obtain a trust request data packet; Sending the trust request data packet to the cluster head sensor node, so that the cluster head sensor node determines a plurality of initial recommended nodes and the number of demand nodes according to the trust request data packet, and determines the distance between each initial recommended node according to the trust metric parameter of each initial recommended node, performs community division according to the initial distance between each initial recommended node, determines a plurality of candidate recommended nodes and the initial influence of each candidate recommended node in the plurality of initial recommended nodes according to the community division result and the number of demand nodes, and sends the initial influence of each candidate recommended node to each candidate recommended node, determines a propagation path and a target influence according to the initial influence of each candidate recommended node through each candidate recommended node, and feeds back the recommendation information and the target influence according to the propagation path; Upon receiving the recommendation information and the target influence fed back by each candidate recommendation node, determining the total influence value according to the target influence of each candidate recommendation node; Calculate the trust value based on the total influence value, the target influence of each candidate recommendation node, and the direct trust value of each candidate recommendation node to determine the indirect trust value of the trust object; A trust propagation decision is made according to the indirect trust value of the trust object.

8. A cluster head sensor node, the cluster head sensor node comprising: A processing module, configured to determine a plurality of initial recommended nodes and a number of required nodes according to a trust request data packet when receiving a trust request data packet from a trust subject; The processing module is further used to determine the initial distance between each initial recommended node according to the trust metric parameter of each initial recommended node; A division module, used to divide the community according to the initial distance between each initial recommendation node, and determine multiple candidate recommendation nodes and the initial influence of each candidate recommendation node from multiple initial recommendation nodes according to the community division result and the number of demand nodes; A sending module is used to send the initial influence of each candidate recommendation node to each candidate recommendation node, determine the propagation path and target influence according to the initial influence of each candidate recommendation node through each candidate recommendation node, and send the recommendation information and the target influence to the trust subject according to the propagation path, so that the trust subject makes a trust propagation decision based on the target influence of each candidate recommendation node and the recommendation information of each candidate recommendation node.

9. A recommendation node, the recommendation node comprising: A processing module is used for determining a path according to the initial influence and the target generalized maximum flow strategy to obtain a propagation path and a target influence when receiving the initial influence sent by the cluster head sensor node, wherein the initial influence is obtained by the cluster head sensor node determining a plurality of initial recommended nodes and the number of required nodes according to the trust request data packet sent by the trust subject, determining the distance between each initial recommended node according to the trust metric parameter of each initial recommended node, performing community division according to the initial distance between each initial recommended node, and calculating the community division result; A sending module is used to send the recommendation information and the target influence to the trust subject according to the propagation path, so that the trust subject makes a trust propagation decision according to the target influence and the recommendation information.

10. A trust subject, the trust subject comprising: A processing module, configured to determine the number of required nodes and the target number of receiving hops according to the data transmission instruction when receiving the data transmission instruction; An encapsulation module is used to encapsulate the number of required nodes and the target acceptance hop count to obtain a trust request data packet; A sending module, used for sending the trust request data packet to the cluster head sensor node, so that the cluster head sensor node determines multiple initial recommended nodes and the number of demand nodes according to the trust request data packet, and determines the distance between each initial recommended node according to the trust measurement parameter of each initial recommended node, performs community division according to the initial distance between each initial recommended node, determines multiple candidate recommended nodes and the initial influence of each candidate recommended node in the multiple initial recommended nodes according to the community division result and the number of demand nodes, and sends the initial influence of each candidate recommended node to each candidate recommended node, determines the propagation path and the target influence according to the initial influence of each candidate recommended node through each candidate recommended node, and feeds back the recommendation information and the target influence according to the propagation path; The processing module is further configured to determine the total influence value according to the target influence of each candidate recommendation node when receiving the recommendation information fed back by each candidate recommendation node and the target influence; A calculation module, used to calculate the trust value according to the total influence value, the target influence of each candidate recommendation node and the direct trust value of each candidate recommendation node, so as to determine the indirect trust value of the trust object; The decision module is used to make a trust propagation decision according to the indirect trust value of the trust object.

Images