CN114666795A - Node behavior-based underwater acoustic sensing network node reliability evaluation method - Google Patents

Node behavior-based underwater acoustic sensing network node reliability evaluation method Download PDF

Info

Publication number
CN114666795A
CN114666795A CN202210349472.2A CN202210349472A CN114666795A CN 114666795 A CN114666795 A CN 114666795A CN 202210349472 A CN202210349472 A CN 202210349472A CN 114666795 A CN114666795 A CN 114666795A
Authority
CN
China
Prior art keywords
node
trust
degree
nodes
evaluation
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Pending
Application number
CN202210349472.2A
Other languages
Chinese (zh)
Inventor
申晓红
孙霖
刘郑国
王海燕
何轲
员一帆
张之琛
马石磊
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Northwestern Polytechnical University
Original Assignee
Northwestern Polytechnical University
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by Northwestern Polytechnical University filed Critical Northwestern Polytechnical University
Priority to CN202210349472.2A priority Critical patent/CN114666795A/en
Publication of CN114666795A publication Critical patent/CN114666795A/en
Pending legal-status Critical Current

Links

Images

Classifications

    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04W—WIRELESS COMMUNICATION NETWORKS
    • H04W12/00—Security arrangements; Authentication; Protecting privacy or anonymity
    • H04W12/12—Detection or prevention of fraud
    • H04W12/121—Wireless intrusion detection systems [WIDS]; Wireless intrusion prevention systems [WIPS]
    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04W—WIRELESS COMMUNICATION NETWORKS
    • H04W12/00—Security arrangements; Authentication; Protecting privacy or anonymity
    • H04W12/12—Detection or prevention of fraud
    • H04W12/121—Wireless intrusion detection systems [WIDS]; Wireless intrusion prevention systems [WIPS]
    • H04W12/122—Counter-measures against attacks; Protection against rogue devices
    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04W—WIRELESS COMMUNICATION NETWORKS
    • H04W84/00—Network topologies
    • H04W84/18—Self-organising networks, e.g. ad-hoc networks or sensor networks

Landscapes

  • Engineering & Computer Science (AREA)
  • Computer Networks & Wireless Communication (AREA)
  • Signal Processing (AREA)
  • Computer Security & Cryptography (AREA)
  • Information Retrieval, Db Structures And Fs Structures Therefor (AREA)

Abstract

The invention provides a node credibility assessment method of an underwater acoustic sensing network based on node behaviors, which is characterized in that the node behaviors and the interaction success rate in the information interaction process are used as trust factors, then the trust factors are aggregated by using information entropy to obtain direct trust, indirect trust and historical trust, further the node trust is obtained, and a trust assessment model based on the node behaviors is established. Therefore, the malicious nodes can be quickly and accurately detected, effective identification and isolation are carried out, and the safety and reliability of the nodes in the routing path are ensured. The method and the device aim at detecting tampering attack, deception attack and black hole malicious nodes existing in the underwater acoustic sensor network, can quickly and accurately detect the malicious nodes by combining with an opportunistic routing protocol, and effectively identify and isolate the malicious nodes, thereby ensuring the safety and reliability of the nodes in a routing path.

Description

Node behavior-based underwater acoustic sensing network node reliability evaluation method
Technical Field
The invention relates to the technical field of underwater acoustic sensor network security, in particular to a method for evaluating node reliability in a sensor network through abnormal node behaviors.
Background
The underwater acoustic sensor network is a typical application of a wireless sensor network, can be used as an ideal medium in marine environment, can monitor a target sea area in a large range in real time, and has wide application prospects in the fields of marine resource survey, marine environment monitoring, sea area safety guarantee and the like.
The underwater acoustic sensor network is used as a wireless self-organizing network, has the characteristics of open transmission medium, cooperative algorithm among nodes, fuzzy defense boundary and the like, and is easy to be attacked by various attacks, such as Wormhole, Hello-Flood and Selective Forward attacks. Meanwhile, due to the characteristics of low transmission rate of an underwater acoustic channel, high error rate, time delay, narrow bandwidth, high node energy consumption and the like, a certain node is more likely to be selfish or captured in the network, so that the nodes of a link cannot be guaranteed to be credible in the network communication process, and the network robustness cannot be guaranteed.
In order to ensure that the underwater acoustic sensor network can normally operate under the condition that malicious nodes exist, many scholars at home and abroad propose different methods for node trust evaluation to ensure the basic functions of the network. The Jiang Jinfang et al propose an attack resistance trust model based on multidimensional trust measurement (ARTMM) and consider the characteristics of underwater acoustic channels and the mobility of nodes, but the algorithm has high computational complexity, the trust evidence generation process does not consider the influence of malicious nodes, the fuzzy set definition is subjective, and the algorithm cannot adapt to a dynamic underwater acoustic sensor network. Han Guangjie et al propose a trust model based on cloud Theory (TMC) to solve the above problem. The cloud model is based on the traditional fuzzy set and probability statistical theory, and can better evaluate the uncertainty of the trust relationship. However, many assumptions are made in the model design process, and the requirement on the position information of each node in the underwater acoustic sensor network is high. Lei Shu et al propose a collaborative trust model (STMS) based on SVM, which divides the network into clusters, divides nodes in the clusters into MCH, SCH and CM, and applies k-means and SVM algorithm to generate a trust evaluation model. However, it is assumed that the probability of simultaneous attacks on the MCH and SCH in the same cluster is 0, and the multi-node joint attack cannot be defended.
From the research, the trust model of the existing underwater acoustic sensor network mostly does not consider the uniqueness of the underwater acoustic environment and the limitation of network resources in the design process. Therefore, the trust models have certain limitations in application in the underwater acoustic sensor network. The invention provides a node credibility evaluation method of an underwater acoustic sensing network based on node behaviors by combining the characteristics of complex underwater acoustic channels, updating node credibility evaluation in real time along with network communication, taking the node behaviors and the interaction success rate in the information interaction process as credibility factors and combining space and time dimensions.
Disclosure of Invention
In order to overcome the defects of the prior art, the invention provides a node credibility evaluation method of an underwater acoustic sensing network based on node behaviors. In order to defend malicious nodes which become the best next hop nodes of a candidate set through attack modes such as disguise, monitoring, deception and the like, and to obtain the credibility of the underwater acoustic sensing network nodes, the method for evaluating the credibility of the underwater acoustic sensing network nodes based on the node behaviors is provided. According to the invention, node behaviors and interaction success rate in the information interaction process are used as trust factors, and then the trust factors are aggregated by using the information entropy to obtain direct trust, indirect trust and historical trust, so that the node trust is obtained, and a trust evaluation model based on the node behaviors is established. Therefore, the malicious nodes can be quickly and accurately detected, effective identification and isolation are carried out, and the safety and reliability of the nodes in the routing path are ensured.
Aiming at the position spoofing vulnerability existing in the underwater sound sensor network, the network selfish node and the malicious node are utilized, so that the network is broken down, the service life of the network is shortened, and the reliability evaluation method of the underwater sound sensor network node based on the node behavior is further provided.
The technical scheme adopted by the invention for solving the technical problem comprises the following steps:
firstly, calculating the direct trust of the node;
the direct trust mainly comprises two aspects of information interaction trust and node behavior trust in the node network communication process; direct degree of trust
Figure BDA0003578889340000021
The calculation formula of (a) is as follows:
Figure BDA0003578889340000022
wherein r isijAnd bijRespectively refer to information interaction trust and node behavior trust,
Figure BDA0003578889340000023
and
Figure BDA0003578889340000024
respectively indicating self-adaptive weights obtained by information interaction trust and node behavior trust through an information entropy concept;
(1) information interaction trust
The information interaction trust degree depends on the success times and the failure times of directly carrying out information interaction between the sending node i and the neighbor node j, so that the expectation of the success of the information interaction can be regarded as the information interaction trust degree between the nodes, and the information interaction trust degree obeys beta distribution;
therefore, the information interaction trust degree rijExpressed as:
Figure BDA0003578889340000025
wherein r isijInformation interaction trust representing that an evaluation node i considers to be an evaluation node jDegree; alpha is alphaijAnd betaijRespectively representing the past successful interaction times and the past failure interaction times of an evaluation node i and an evaluated node j; function of probability density
Figure BDA0003578889340000031
(2) Degree of trust of node behavior
The node behavior trust degree refers to whether the node has abnormal behavior in the network communication process; the method is used for sensing suspicious behaviors of the nodes and adjusting the trust degrees of the nodes in time; calculating the behavior trust degree of the evaluated node according to the abnormal behavior occurrence frequency of the evaluated node counted by the evaluation node in the monitoring process; b is to beijAnd representing the node behavior trust degree of the evaluated node j considered by the evaluation node i, wherein the node behavior trust degree is represented as:
Figure BDA0003578889340000032
wherein the content of the first and second substances,
Figure BDA0003578889340000033
evaluating the node behavior trust degree of the evaluated node j considered by the node i in the last time period;
secondly, counting the trust of the calculation nodes in the spatial dimension;
the indirect trust degree refers to the direct trust degree of the evaluated node fed back by the evaluation node through other adjacent nodes of the evaluated node, and is the trust evaluation of the space dimension of the node; the concrete expression is as follows: when an evaluation node i needs to evaluate the trust degree of an evaluated node j further, the evaluation node i broadcasts query information to neighbor nodes of the evaluation node i to obtain recommendation information of the evaluated node j, and once the query information recommendation node k (a common neighbor node of the evaluation node i and the evaluated node j) is received and the direct trust of the query information recommendation node k on the evaluated node j is taken as the recommendation information to be returned to the evaluation node i, the trust degree of the node j fed back by the recommendation node k by the node i
Figure BDA0003578889340000034
Expressed as:
Figure BDA0003578889340000035
wherein the content of the first and second substances,
Figure BDA0003578889340000036
to evaluate the direct trust of node k to evaluated node j.
The n neighbor nodes of the evaluation node i are taken as recommendation nodes to respectively recommend the evaluated node j, and the trust degrees fed back to the evaluation node i are respectively as follows:
Figure BDA0003578889340000037
the evaluation node i performs weight distribution on the trust level fed back by the recommendation node by using the information entropy, and sums values obtained by multiplying the respective weight by the trust level to obtain the indirect trust level of the evaluation node i on the evaluated node j;
thirdly, counting the trust of the computing node in the time dimension;
the historical trust degree, because of the dynamic change of the underwater acoustic sensor network and the selective attack of the malicious node, the evaluation of the trust degree of the node also changes along with the alternation of network communication and time period, so the historical record trust degree of the node must be used as the trust degree evaluation of the time dimension of the node;
historical confidence level depends on the integrated confidence level of the previous update period
Figure BDA0003578889340000041
And previous historical confidence
Figure BDA0003578889340000042
Therefore, in the invention, the information entropy is used for distributing the comprehensive trust of the previous updating period and the previous historical trust in the weight value occupied by the historical trust, and then the comprehensive trust of the previous updating period and the previous historical trust are aggregated by using the information entropy;
fourthly, comprehensively evaluating the trust of the nodes;
the node trust degree in the trust model based on the node behaviors comprises direct trust degree and the trust degree of the node in space and time dimensions, so that the direct trust degree, the indirect trust degree and the historical trust degree of the node are aggregated for evaluating the comprehensive trust degree of the underwater node; carrying out aggregate calculation on the comprehensive trust of the nodes by using the information entropy;
fifthly, identifying and isolating malicious nodes;
after the node comprehensive trust degree evaluation stage, the evaluation node makes a decision based on the comprehensive trust degree of the evaluated node; setting a threshold value in the trust model based on the node behavior refers to a trust degree interval division process in the trust model based on the information entropy, and the TH value is set to be 0.6; comparing a given preset threshold with the comprehensive trust level of the nodes, wherein the malicious nodes refer to the evaluated nodes with the comprehensive trust level lower than the threshold evaluated by the evaluation nodes; then, the evaluated node with the comprehensive trust degree lower than the threshold is regarded as a malicious node by the evaluated node, and the evaluated node can automatically avoid the node evaluated as the malicious node in the subsequent network communication process.
According to the identification result of the network node, the trusted node is allowed to participate in normal routing operation, and data forwarding and receiving are carried out; untrusted nodes are not allowed to participate in normal routing operations, and data forwarded by them are discarded by normal network nodes.
In the first step, the adaptive weight calculation formula is as follows:
Figure BDA0003578889340000043
Figure BDA0003578889340000044
H(rij)＝-rij log2 rij-(1-rij)log2(1-rij)
H(bij)＝-bij log2 bij-(1-bij)log2(1-bij)
wherein, H (r)ij) And H (b)ij) The information entropy is respectively the information interaction trust and the node behavior trust.
In the second step, first, calculation is performed
Figure BDA0003578889340000051
Entropy of (d):
Figure BDA0003578889340000052
wherein the content of the first and second substances,
Figure BDA0003578889340000053
the direct trust degree of the evaluated node j at the recommended node k is represented when the recommended node k recommends the evaluated node j to the evaluating node i;
weight w for calculating recommendation trust by using information entropyk：
Figure BDA0003578889340000054
Finally, indirect confidence
Figure BDA0003578889340000055
The calculation results are expressed as:
Figure BDA0003578889340000056
wherein, the first and the second end of the pipe are connected with each other,
Figure BDA0003578889340000057
the direct trust degree of the evaluated node j at the recommended node k is represented when the recommended node k recommends the evaluated node j to the evaluating node i; w is akRepresenting the trust weight of the recommendation node k at the evaluation node i.
In the third step, the historical trust level updated with the time period
Figure BDA0003578889340000058
The calculation was performed as follows:
Figure BDA0003578889340000059
wherein the content of the first and second substances,
Figure BDA00035788893400000510
and
Figure BDA00035788893400000511
respectively representing the comprehensive trust of the previous update period and the historical trust in the latest period, w#Adaptive weights, w, representing the integrated confidence of the previous update period*An adaptive weight representing a previous historical confidence; the calculation formula is as follows:
Figure BDA00035788893400000512
Figure BDA0003578889340000061
Figure BDA0003578889340000062
Figure BDA0003578889340000063
wherein the content of the first and second substances,
Figure BDA0003578889340000064
and
Figure BDA0003578889340000065
and respectively representing the comprehensive credibility of the previous updating period and the information entropy of the previous historical credibility.
And in the fourth step, the comprehensive trust degree comprises the direct trust degree, the indirect trust degree and the historical trust degree of the node. Comprehensive trust level T for nodesijStill adopting information entropy to carry out aggregation calculation, wherein the calculation formula is as follows:
Figure BDA0003578889340000066
wherein, wd，wiAnd whRespectively are self-adaptive weights of the direct trust level, the indirect trust level and the historical trust level of the node,
Figure BDA0003578889340000067
and
Figure BDA0003578889340000068
respectively representing the direct trust, the indirect trust and the historical trust of the node; the calculation formula is as follows:
Figure BDA0003578889340000069
Figure BDA00035788893400000610
Figure BDA00035788893400000611
Figure BDA00035788893400000612
Figure BDA0003578889340000071
Figure BDA0003578889340000072
wherein the content of the first and second substances,
Figure BDA0003578889340000073
and
Figure BDA0003578889340000074
and respectively representing the direct trust degree, the indirect trust degree and the historical trust degree of the node.
The underwater acoustic sensor network node credibility assessment method based on the node behaviors has the advantages that under the complicated and changeable underwater environment which is easy to attack, the model provides corresponding abnormal behavior detection aiming at the potential security threat of the underwater acoustic sensor network, the direct interaction of the nodes, the recommendation of other nodes and the historical behavior of the nodes are integrated, the node credibility is calculated, and further the identified malicious nodes are isolated. Then, the trust model can be introduced into an opportunistic routing protocol, the nodes finish the trust evaluation process by broadcasting a trust probe packet, and the indirect trust of the nodes is updated; meanwhile, in the normal routing operation and data transmission process, the node trust degree is updated in real time by monitoring the behavior of the nodes, and the malicious nodes are identified and isolated in time. Aiming at detecting tampering attack, deception attack and black hole malicious nodes existing in the underwater acoustic sensor network, the invention can quickly and accurately detect the malicious nodes by combining with the opportunistic routing protocol, and effectively identify and isolate the malicious nodes, thereby ensuring the safety and reliability of the nodes in the routing path.
Drawings
Fig. 1 is a schematic diagram of a normal network relay node trust level average value.
Fig. 2 is a schematic diagram of the trust degree of a tampering attack malicious node.
Fig. 3 is a schematic diagram of the trust degree of a malicious node in a spoofing attack.
Fig. 4 is a schematic diagram of trust degree of a malicious node of a black hole attack.
Detailed Description
The invention is further illustrated with reference to the following figures and examples.
The node trust degree in the trust model based on the node behaviors comprises direct trust degree, indirect trust degree and historical trust degree, so the direct trust degree, the indirect trust degree and the historical trust degree of the node need to be aggregated aiming at the calculation of the comprehensive trust degree of the underwater node.
The specific implementation steps are as follows:
the first step is as follows: calculating the direct trust of the node;
the calculation of the direct trust degree comprises the information interaction trust degree rijAnd node behavior confidence bijAnd (4) calculating the aggregation of the two parts. Then, the direct trust is the trust r for exchanging informationijAnd degree of trust of node behavior bijAnd further summing the weight distribution.
In order to enable the given trust model to be suitable for different purpose networks under different scenes and reduce the influence of human subjective factors on the accuracy of the trust model, the invention dynamically distributes the information interaction trust and the weight of the node behavior trust by using the information entropy.
Direct degree of trust
Figure BDA0003578889340000081
The calculation formula of (c) is as follows:
Figure BDA0003578889340000082
wherein r isijAnd bijRespectively refer to information interaction trust and node behavior trust,
Figure BDA0003578889340000083
and
Figure BDA0003578889340000084
the self-adaptive weights are obtained by information interaction trust and node behavior trust through an information entropy concept respectively. H (r)ij) And H (b)ij) The information entropies respectively represent information interaction trust and node behavior trust, and the calculation formula is as follows:
H(rij)＝-rij log2 rij-(1-rij)log2(1-rij)
H(bij)＝-bij log2 bij-(1-bij)log2(1-bij)
Figure BDA0003578889340000085
Figure BDA0003578889340000086
the second step: calculating the indirect trust of the node;
the calculation of the indirect trust degree is mainly the aggregation calculation of the recommendation trust degree given to the evaluated node by the recommendation node. Suppose there are n recommending nodes recommending evaluated nodes j for evaluating node i, and there are n indirect trusts correspondingly
Figure BDA0003578889340000087
Figure BDA0003578889340000088
In the process, whether a certain recommended node is credible or not and the weight of each recommended node is excellent cannot be subjectively determined, so that information entropy is introduced to allocate weights to different recommended nodes. Therefore, the difference of different recommended nodes is taken into account, and the adaptability of the model is enhanced. First, calculate
Figure BDA0003578889340000089
Entropy of (c):
Figure BDA00035788893400000810
wherein the content of the first and second substances,
Figure BDA0003578889340000091
the direct trust degree of the evaluated node j at the recommending node k is represented when the recommending node k recommends the evaluated node j to the evaluating node i;
the information entropy reflects the disorder degree of the information, and the information entropy of each recommendation trust reflects the difference degree between the information entropy and the recommendation trust, namely the degree of deviation of each recommendation trust from the recommendation trust set as a whole. The behavior that the malicious node purposefully devaluates the legal node of the underwater acoustic sensor network or maliciously devaluates other malicious nodes can enable the recommendation trust degree of the node evaluated by the malicious node to deviate from the actual trust degree of the node to a certain extent, and the recommendation node can be identified by utilizing the information entropy, so that the influence of the information entropy on the objectivity and the accuracy of the trust degree of the node is reduced. Generally, the smaller the difference between the recommendation trust degrees of the evaluated nodes given by the recommending nodes is, the more objective the recommendation of each recommending node to the evaluated node is, and therefore, the weight w of the recommendation trust can be calculated by using the information entropyk：
Figure BDA0003578889340000092
Finally, indirect confidence
Figure BDA0003578889340000093
The calculation results are expressed as:
Figure BDA0003578889340000094
wherein the content of the first and second substances,
Figure BDA0003578889340000095
the direct trust degree of the evaluated node j at the recommended node k is represented when the recommended node k recommends the evaluated node j to the evaluating node i; w is akRepresenting the trust weight of the recommendation node k at the evaluation node i;
the third step: historical trust computation for nodes
The calculation of the historical trust degree mainly depends on the comprehensive trust degree of the previous updating period
Figure BDA0003578889340000096
And previous historical confidence
Figure BDA0003578889340000097
In the process, if the weight occupied by the comprehensive trust in the latest updating period is higher, the change rate of the historical trust of the malicious node is the largest in the time period, so that the influence degree of the comprehensive trust of the latest updating period of the node on the historical trust is large, and the performance condition of the node in the whole stage cannot be well reflected. If the weight of the previous historical trust is higher, the condition of the node in the latest updating period cannot be well evaluated, so the weight is distributed by using the information entropy. Historical confidence level for updating over time period
Figure BDA0003578889340000098
The calculation was performed as follows:
Figure BDA0003578889340000099
wherein, the first and the second end of the pipe are connected with each other,
Figure BDA00035788893400000910
and
Figure BDA00035788893400000911
respectively representing the comprehensive trust of the previous update period and the historical trust in the latest period, w#Adaptive weight, w, representing the integrated confidence of the previous update cycle*The self-adaptive weight value represents the previous historical trust degree;
Figure BDA0003578889340000101
and
Figure BDA0003578889340000102
respectively representing the comprehensive trust degree of the previous updating period and the information entropy of the previous historical trust degree, and the calculation formula is as follows:
Figure BDA0003578889340000103
Figure BDA0003578889340000104
Figure BDA0003578889340000105
Figure BDA0003578889340000106
the fourth step: integrated trust computation for nodes
The comprehensive trust degree comprises a node direct trust degree, an indirect trust degree and a historical trust degree. Comprehensive trust level T for nodesijStill adopting information entropy to carry out polymerization calculation, wherein the calculation formula is as follows:
Figure BDA0003578889340000107
wherein, wd，wiAnd whRespectively are self-adaptive weights of the direct trust level, the indirect trust level and the historical trust level of the node,
Figure BDA0003578889340000108
and
Figure BDA0003578889340000109
respectively representing the direct trust, the indirect trust and the historical trust of the node;
Figure BDA00035788893400001010
Figure BDA00035788893400001011
and
Figure BDA00035788893400001012
information entropies respectively representing the direct trust degree, the indirect trust degree and the historical trust degree of the node, and the calculation formula is as follows:
Figure BDA00035788893400001013
Figure BDA00035788893400001014
Figure BDA00035788893400001015
Figure BDA0003578889340000111
Figure BDA0003578889340000112
Figure BDA0003578889340000113
the fifth step: identifying and isolating malicious nodes;
after the node comprehensive trust degree calculation stage, the evaluation node can make a decision based on the comprehensive trust degree of the evaluated node. And identifying the nodes with the comprehensive trust degree lower than the threshold value (0.6) as non-trusted nodes, and identifying the nodes with the comprehensive trust degree greater than or equal to the threshold value as trusted nodes. According to the identification result of the network node, the trusted node is allowed to participate in normal routing operation, and data forwarding and receiving are carried out; untrusted nodes are not allowed to participate in normal routing operations, and data forwarded by them are discarded by normal network nodes.
The average value of the trust degrees of the normal relay nodes of the underwater acoustic sensor network is shown in fig. 1, which shows that under the condition of a normal network, the trust degrees of the network nodes are all stabilized above a threshold value, but the change amplitude of the trust degrees is not obvious, so that the change of the trust degrees of the nodes in the network communication process, which is approved by some malicious attack nodes through short-term good performance, is prevented. The average value of the trust degrees of the malicious nodes for performing tampering attack of the underwater acoustic sensor network is shown in fig. 2, the average value of the trust degrees of the malicious nodes for performing deception attack of the underwater acoustic sensor network is shown in fig. 3, and the average value of the trust degrees of the malicious nodes for performing black hole attack of the underwater acoustic sensor network is shown in fig. 4, which indicates that the trust degrees of the malicious nodes in the attacked network can be immediately reduced below a threshold value, and meanwhile, along with the interaction of information, the trust degrees of the nodes are still reduced and the reduction range is larger than that in a normal case. The invention also verifies that the invention can quickly and accurately detect the malicious nodes, effectively identify and isolate the malicious nodes and ensure the safety and reliability of the network nodes.

Claims (5)

1. A node credibility evaluation method of an underwater acoustic sensing network based on node behaviors is characterized by comprising the following steps:
firstly, calculating the direct trust of the node;
the direct trust mainly comprises two aspects of information interaction trust and node behavior trust in the node network communication process; direct degree of trust
Figure FDA0003578889330000011
The calculation formula of (a) is as follows:
Figure FDA0003578889330000012
wherein r isijAnd bijRespectively refer to information interaction trust and node behavior trustThe degree of the magnetic field is measured,
Figure FDA0003578889330000013
and
Figure FDA0003578889330000014
respectively indicating self-adaptive weights obtained by information interaction trust and node behavior trust through an information entropy concept;
(1) information interaction trust
The information interaction trust degree depends on the success times and the failure times of directly carrying out information interaction between the sending node i and the neighbor node j, so that the expectation of successful information interaction can be regarded as the information interaction trust degree between the nodes, and the information interaction trust degree obeys beta distribution;
therefore, the information interaction trust degree rijExpressed as:
Figure FDA0003578889330000015
wherein r isijRepresenting the information interaction trust degree of the evaluated node j considered by the evaluation node i; alpha (alpha) ("alpha")ijAnd betaijRespectively representing the past successful interaction times and the past failure interaction times of an evaluation node i and an evaluated node j; function of probability density
Figure FDA0003578889330000016
(2) Degree of trust of node behavior
The node behavior trust degree refers to whether the node has abnormal behavior in the network communication process; the method is used for sensing suspicious behaviors of the nodes and adjusting the trust degrees of the nodes in time; calculating the behavior trust degree of the evaluated node according to the abnormal behavior occurrence frequency of the evaluated node counted by the evaluation node in the monitoring process; b is toijAnd representing the node behavior trust degree of the evaluated node j considered by the evaluation node i, wherein the node behavior trust degree is represented as:
Figure FDA0003578889330000017
wherein the content of the first and second substances,
Figure FDA0003578889330000018
evaluating the node behavior trust degree of the evaluated node j considered by the node i in the last time period;
secondly, counting the trust of the calculation nodes in the spatial dimension;
the indirect trust degree refers to the direct trust degree of the evaluated node fed back by the evaluation node through other adjacent nodes of the evaluated node, and is the trust evaluation of the space dimension of the node; the concrete expression is as follows: when the evaluation node i needs to evaluate the further trust degree of the evaluated node j, the evaluation node i broadcasts query information to the neighbor nodes of the evaluation node j to acquire recommendation information of the evaluated node j, and once the query information recommendation node k (the common neighbor nodes of the evaluation node i and the evaluated node j) is received and the direct trust of the node k on the evaluated node j is taken as recommendation information to be returned to the evaluation node i, the trust degree of the node j fed back by the recommendation node k by the node i is obtained
Figure FDA0003578889330000021
Expressed as:
Figure FDA0003578889330000022
wherein, the first and the second end of the pipe are connected with each other,
Figure FDA0003578889330000023
direct trust of evaluated node j for evaluation node k;
the n neighbor nodes of the evaluation node i are used as recommendation nodes to respectively recommend the evaluated node j, and the trust degrees fed back to the evaluation node i are respectively as follows:
Figure FDA0003578889330000024
the evaluation node i willThe trust degree fed back by the recommendation node utilizes the information entropy to carry out weight distribution, and then the respective weight is summed with the value obtained by multiplying the trust degree, so that the indirect trust degree of the evaluation node i to the evaluated node j can be obtained;
thirdly, counting the trust of the computing node in the time dimension;
the historical trust degree, because of the dynamic change of the underwater acoustic sensor network and the selective attack of the malicious node, the evaluation of the trust degree of the node also changes along with the alternation of network communication and time period, so the historical record trust degree of the node must be used as the trust degree evaluation of the time dimension of the node;
historical confidence level depends on the integrated confidence level of the previous update period
Figure FDA0003578889330000025
And previous historical confidence
Figure FDA0003578889330000026
Therefore, in the invention, the information entropy is used for distributing the comprehensive trust of the previous updating period and the historical trust of the previous updating period in the weight occupied by the historical trust, and then the comprehensive trust of the previous updating period and the historical trust of the previous updating period are aggregated by using the information entropy;
fourthly, comprehensively evaluating the trust of the nodes;
the node trust degree in the trust model based on the node behaviors comprises direct trust degree and the trust degree of the node in space and time dimensions, so that the direct trust degree, the indirect trust degree and the historical trust degree of the node are aggregated for evaluating the comprehensive trust degree of the underwater node; carrying out aggregate calculation on the comprehensive trust of the nodes by using the information entropy;
fifthly, identifying and isolating malicious nodes;
after the node comprehensive trust degree evaluation stage, the evaluation node makes a decision based on the comprehensive trust degree of the evaluated node; setting a threshold value in the trust model based on the node behavior refers to a trust degree interval dividing process in the trust model based on the information entropy, and a TH value is set to be 0.6; comparing a given preset threshold with the comprehensive trust level of the nodes, wherein the malicious nodes refer to the evaluated nodes with the comprehensive trust level lower than the threshold evaluated by the evaluation nodes;
according to the identification result of the network node, the trusted node is allowed to participate in normal routing operation, and data forwarding and receiving are carried out; untrusted nodes are not allowed to participate in normal routing operations, and data forwarded by them are discarded by normal network nodes.
2. The node behavior-based underwater acoustic sensing network node credibility assessment method according to claim 1, characterized in that:
in the first step, the adaptive weight calculation formula is as follows:
Figure FDA0003578889330000031
Figure FDA0003578889330000032
H(rij)＝-rijlog2rij-(1-rij)log2(1-rij)
H(bij)＝-bijlog2bij-(1-bij)log2(1-bij)
wherein, H (r)ij) And H (b)ij) The information entropy is respectively the information interaction trust and the node behavior trust.
3. The node behavior-based underwater acoustic sensing network node credibility assessment method according to claim 1, characterized in that:
in the second step, first, calculation is performed
Figure FDA0003578889330000033
Entropy of (d):
Figure FDA0003578889330000034
wherein the content of the first and second substances,
Figure FDA0003578889330000035
the direct trust degree of the evaluated node j at the recommended node k is represented when the recommended node k recommends the evaluated node j to the evaluating node i;
weight w for calculating recommendation trust by using information entropyk：
Figure FDA0003578889330000041
Finally, the indirect confidence
Figure FDA0003578889330000042
The calculation results are expressed as:
Figure FDA0003578889330000043
wherein the content of the first and second substances,
Figure FDA0003578889330000044
the direct trust degree of the evaluated node j at the recommended node k is represented when the recommended node k recommends the evaluated node j to the evaluating node i; w is akRepresenting the trust weight of the recommendation node k at the evaluation node i.
4. The node behavior-based underwater acoustic sensing network node credibility assessment method according to claim 1, characterized in that:
in the third step, the historical trust level updated with the time period
Figure FDA0003578889330000045
The calculation was performed as follows:
Figure FDA0003578889330000046
wherein the content of the first and second substances,
Figure FDA0003578889330000047
and
Figure FDA0003578889330000048
respectively represents the comprehensive trust level of the previous updating period and the historical trust level in the latest period, w#Adaptive weights, w, representing the integrated confidence of the previous update period*An adaptive weight representing a previous historical confidence; the calculation formula is as follows:
Figure FDA0003578889330000049
Figure FDA00035788893300000410
Figure FDA00035788893300000411
Figure FDA00035788893300000412
wherein, the first and the second end of the pipe are connected with each other,
Figure FDA00035788893300000413
and
Figure FDA00035788893300000414
and respectively representing the comprehensive credibility of the previous updating period and the information entropy of the previous historical credibility.
5. The node behavior-based underwater acoustic sensing network node credibility assessment method according to claim 1, characterized in that:
in the fourth step, the comprehensive trust degree comprises the direct trust degree, the indirect trust degree and the historical trust degree of the node, and the comprehensive trust degree T of the nodeijStill adopting information entropy to carry out polymerization calculation, wherein the calculation formula is as follows:
Figure FDA0003578889330000051
wherein, wd，wiAnd whRespectively are self-adaptive weights of the direct trust level, the indirect trust level and the historical trust level of the node,
Figure FDA0003578889330000052
and
Figure FDA0003578889330000053
respectively representing the direct trust, the indirect trust and the historical trust of the node; the calculation formula is as follows:
Figure FDA0003578889330000054
Figure FDA0003578889330000055
Figure FDA0003578889330000056
Figure FDA0003578889330000057
Figure FDA0003578889330000058
Figure FDA0003578889330000059
wherein the content of the first and second substances,
Figure FDA00035788893300000510
and
Figure FDA00035788893300000511
and respectively representing the direct trust degree, the indirect trust degree and the historical trust degree of the node.
CN202210349472.2A 2022-04-01 2022-04-01 Node behavior-based underwater acoustic sensing network node reliability evaluation method Pending CN114666795A (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN202210349472.2A CN114666795A (en) 2022-04-01 2022-04-01 Node behavior-based underwater acoustic sensing network node reliability evaluation method

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN202210349472.2A CN114666795A (en) 2022-04-01 2022-04-01 Node behavior-based underwater acoustic sensing network node reliability evaluation method

Publications (1)

Publication Number Publication Date
CN114666795A true CN114666795A (en) 2022-06-24

Family

ID=82032949

Family Applications (1)

Application Number Title Priority Date Filing Date
CN202210349472.2A Pending CN114666795A (en) 2022-04-01 2022-04-01 Node behavior-based underwater acoustic sensing network node reliability evaluation method

Country Status (1)

Country Link
CN (1) CN114666795A (en)

Cited By (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN117056860A (en) * 2023-08-17 2023-11-14 国网四川省电力公司营销服务中心 Forest fire detection identification method based on evidence system with reduced conflict
CN117313499A (en) * 2023-11-30 2023-12-29 国网山东省电力公司枣庄供电公司 Multi-source sensor arrangement method and system for isolating switch state signals of combined electrical appliance

Cited By (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN117056860A (en) * 2023-08-17 2023-11-14 国网四川省电力公司营销服务中心 Forest fire detection identification method based on evidence system with reduced conflict
CN117313499A (en) * 2023-11-30 2023-12-29 国网山东省电力公司枣庄供电公司 Multi-source sensor arrangement method and system for isolating switch state signals of combined electrical appliance
CN117313499B (en) * 2023-11-30 2024-02-13 国网山东省电力公司枣庄供电公司 Multi-source sensor arrangement method and system for isolating switch state signals of combined electrical appliance

Similar Documents

Publication Publication Date Title
Khan et al. ETERS: A comprehensive energy aware trust-based efficient routing scheme for adversarial WSNs
Yan et al. Vulnerability and protection for distributed consensus-based spectrum sensing in cognitive radio networks
CN114666795A (en) Node behavior-based underwater acoustic sensing network node reliability evaluation method
Sun et al. A comprehensive trust-aware routing protocol with multi-attributes for WSNs
CN108882225B (en) Safe positioning method based on distance measurement in wireless sensor network
Gu et al. k-Nearest Neighbours classification based Sybil attack detection in Vehicular networks
CN112422556B (en) Internet of things terminal trust model construction method and system
Liu et al. A detection framework against CPMA attack based on trust evaluation and machine learning in IoT network
Zheng et al. Dynamic network security mechanism based on trust management in wireless sensor networks
Meng et al. Evaluation of detecting malicious nodes using Bayesian model in wireless intrusion detection
Vosoughi et al. A cooperative spectrum sensing scheme for cognitive radio ad hoc networks based on gossip and trust
CN114553458A (en) Method for establishing and dynamically maintaining credible group in power Internet of things environment
Kim et al. Physical identification based trust path routing against sybil attacks on RPL in IoT networks
Zhang et al. On effective data aggregation techniques in host–based intrusion detection in manet
Su et al. A trust model for underwater acoustic sensor networks based on fast link quality assessment
Gaikwad et al. Survey on secure data aggregation in wireless sensor networks
Vijayakumar et al. A novel jamming detection technique for wireless sensor networks
Karthik et al. Data trustworthiness in wireless sensor networks
CN115802358A (en) Multi-step DDoS prediction poisoning attack based on reinforcement learning and defense method thereof
Hajiee et al. Trust-based routing optimization using multi-ant colonies in wireless sensor network
Saiyed et al. Entropy and divergence-based DDoS attack detection system in IoT networks
KR20220169584A (en) METHOD FOR SELECTING IoT OPTIMIZATION NODES AND REMOVING MALICIOUS NODES
Yuan A Secure Routing Protocol for Wireless Sensor Energy Network Based on Trust Management
Atassi et al. Decentralised malicious node detection in WSN
Gao et al. A trustworthy data aggregation model based on context and data density correlation degree

Legal Events

Date Code Title Description
PB01 Publication
PB01 Publication
SE01 Entry into force of request for substantive examination
SE01 Entry into force of request for substantive examination