阅读说明：本技术 水声传感网中基于冲突裁决的信任管理方法 (Trust management method based on conflict resolution in underwater acoustic sensor network ) 是由 江金芳 华善善 韩光洁 于 2021-09-27 设计创作，主要内容包括：本发明公开了一种水声传感网中基于冲突裁决的信任管理方法,用于处理信任管理机制中,多个第三方节点推荐信息的相互冲突问题。该方法首先结合节点交互的特征以及水声传感网的特征,提出基于节点可靠性和链路可靠性的信任决策方法。此外,提出一种基于囚徒困境的激励机制,平衡信任推荐过程的能耗和信任推荐结果的准确性。(The invention discloses a trust management method based on conflict resolution in an underwater acoustic sensor network, which is used for processing the problem of mutual conflict of recommendation information of a plurality of third-party nodes in a trust management mechanism. The method firstly combines the characteristics of node interaction and the characteristics of the underwater acoustic sensor network to provide a trust decision method based on node reliability and link reliability. In addition, an incentive mechanism based on prisoner predicament is provided, and energy consumption in the trust recommendation process and accuracy of the trust recommendation result are balanced.)

1. The trust management method based on conflict resolution in the underwater acoustic sensor network is characterized by comprising the following steps:

(1) trust evidence collection

Storing a trust value based on sliding windows, wherein each sliding window comprises s time slots, and the number s of the time slots is set according to the storage capacity of the node and the data arrival rate; the sliding window slides forward along the time axis, i.e. one time slot is added, and the oldest time slot is discarded; recording trust evidence for trust calculation, namely delivery rate trust, delay trust and energy consumption trust in each time slot; recording channel information, namely received signal strength, signal-to-noise ratio and node moving distance in each time slot;

(2) incentive mechanism based on prisoner trapping situation

In the trust recommendation process, only the neighbor nodes with the trust values larger than the threshold value participate in the recommendation process, and the value interval of the trust values is set to be [0,1]]The threshold is set to 0.5; the number of neighbor nodes which can participate in the recommendation process is represented as n, and the n neighbor nodes are called as recommendation nodes; in order to ensure that at least k recommendation nodes participate in the trust recommendation process, the process of uploading recommendation information by the n recommendation nodes is described as prisoner predicament; weighing accuracy of trust evaluation resultsAnd energy consumption of the recommendation processDetermining the value of k, wherein k is more than or equal to 1 and less than or equal to n, and the value of k needs to meet the target

(3) Trust decision based on node reliability and data reliability

Defining the node reliability of the recommended node according to the time-frequency characteristics of the interactive history of the recommended node and the target node and the preferences of different trust evidences; and defining the data reliability of the recommendation information based on a grey theory according to the recorded channel information, and finishing a trust decision based on the node reliability and the data reliability.

2. The trust management method based on conflict resolution in the underwater acoustic sensor network according to claim 1, wherein the specific setting method of delivery rate trust, delay trust and energy consumption trust in the step (1) is as follows:

the delivery rate trust is described by using the number of interactive failures and the number of interactive successes:

α + s +1 β + f +1, s represents the number of successful interaction times, and f represents the number of failed interaction times; eta represents a penalty factor for penalizing an attacker who initiates on-off attack; η is calculated as follows:

the delay trust is related to the inter-node distance:

wherein, delay _ time_iIndicating the actual transmission delay of the ith packet, distance_iIndicating the distance between the transmission node and the sending node when the ith data packet is sent;

the energy consumption trust is as follows:

whereinRepresenting the consumed energy of the receiving node, and E is the initial energy of the receiving node.

3. The trust management method based on conflict resolution in the underwater acoustic sensor network according to claim 1, wherein in the step (1), when recording the channel information, in order to quantify the influence of the underwater acoustic communication process on the trust calculation, the following factors are considered:

(i) link connection capability: for a mobile node, link fading is a statistical process with unknown parameters, and from a trust perspective, a trust-related metric is whether a channel supports successful delivery of a data packet; if the received signal strength value between the two nodes is less than the carrier sense threshold, the data packet will not be received; decoding failure may result if the received signal strength value is greater than the carrier sense threshold but less than the detection threshold;

(ii) link stability capability: an important factor influencing the quality of an underwater acoustic channel is that various complex noises exist in an underwater environment, and similarly, the noises are a statistical process with unknown parameters; from a trust perspective, the trust-related metric is whether the signal-to-noise ratio supports successful delivery of the data packet;

(iii) water flow influence: the faster the distance between nodes changes over time, the greater the water flow mobility impact.

4. The trust management method based on conflict resolution in the underwater acoustic sensor network according to claim 1, wherein the model of prisoner's predicament in the step (2) is as follows:

analyzing any node called Current node, and Other nodes called Other nodes; after Nash equilibrium analysis, whenAnd ensuring that k pieces of trust recommendation information are used for finishing trust decision, wherein T represents the reward obtained by the node participating in the recommendation process, C represents the cost of the node participating in the recommendation process, a represents the adjusting factor of the reward degree, and a belongs to (0, 1).

5. The trust management method based on conflict resolution in the underwater acoustic sensor network according to claim 1, wherein in the step (3), the specific setting method of the node reliability is as follows:

(5.1) assume that the preference vector of k different recommendation nodes for different evidence of trust is { P_i ¹,P_i ²,...,P_i ^m}i∈{1,2,...,k}；Represents the ith sectionA preference for the mth trust evidence; the preference similarity of the nodes is defined as follows:

wherein the content of the first and second substances,

(5.2) according to the time-frequency characteristics of the interaction history of the recommended node and the target node, defining the time confidence of the nodes as follows:

wherein c is_tIndicating the number of data packets sent in the T-th time slot, T indicating the total number of time slots in which two nodes have interaction, C_tIndicating the maximum number of packets that can theoretically be sent in a time slot, E (c)_t) The calculation method is as follows:

E(c_t)＝-plog(p)-(1-p)log(1-p)

maximum value is obtained when p is 0.5, and (0, C)_t]Mapping to Interval (0, 0.5)]；

(5.3) defining a reliability measurement parameter of the recommended node according to the preference similarity and the time confidence, namely node reliability as follows:

wherein w₁And w₂Represents a weight, w₁+w₂＝1。

6. The trust management method based on conflict resolution in the underwater acoustic sensor network according to claim 1, wherein in the step (3), the specific setting steps of the data reliability are as follows:

assuming that the received signal strength RSSI acquired by the recommended node through the sliding window, the signal-to-noise ratio and the average node mobility are represented as follows:the optimal reference sequence and the worst sequence are selected as follows

(6.1) carrying out grey correlation analysis on the comparison sequence and a reference sequence to obtain a grey correlation coefficient;the analytical method (2) was as follows:

analysis method andthe same; rho is equal to [0,1]]The resolution coefficient is used for adjusting the difference size of the output result;

(6.2) obtaining the gray correlation degree of the k comparison sequences according to the gray correlation coefficient

Wherein v is₁、v₂、v₃Representing the weight, v₁+v₂+v₃＝1；

(6.3) obtaining data reliability of k comparison sequences by using least squares method

7. The trust management method based on conflict resolution in the underwater acoustic sensor network according to claim 1, wherein in the step (3), the specific setting method of the trust decision based on the node reliability and the data reliability is as follows, and the problem is described as an optimization problem:

wherein, node _ reliability_iRepresenting the node reliability of the recommended node i; data _ reliability_iRepresenting the data reliability of the recommendation information uploaded by the recommendation node i; u. of₁，u₂Representing the weight.

8. The trust management method based on conflict resolution in the underwater acoustic sensor network according to claim 1, wherein in the step (3), the trust decision manner is as follows:

wherein R is_iRepresenting recommendation information uploaded by a recommendation node i; weight_iIs the result of the optimization problem in the formula, determining R_iWeights in trust evidence fusion, i.e.

The technical field is as follows:

the invention belongs to the technical field of radio networks, and particularly relates to a trust conflict solution method in a trust management mechanism in an underwater acoustic sensor network.

Background art:

as an important part for developing marine economy, the Underwater Acoustic Sensor Networks (UASNs) have great application value in various aspects such as environmental monitoring, auxiliary navigation, military combat and the like. UASNs realize real-time monitoring of marine environment through perception data, data acquisition and data transmission. Since the underwater nodes are deployed in an open unmanned environment, the underwater nodes are easy to attack and destroy. Therefore, designing an effective security mechanism for UASNs is crucial to data transmission.

Traditional encryption algorithms are not suitable for resource-constrained UASNs, and authentication schemes cannot resist internal attacks. Therefore, the trust management mechanism is a security guarantee mechanism which has low computational complexity and can resist the internal attack of the network, and plays an important role in improving the security of the UASNs. Many researchers have proposed various trust management mechanisms to effectively cope with internal attacks according to the characteristics of underwater acoustic environments, such as high delay, high packet loss, node movement, and the like.

The trust management mechanism is widely researched in multiple fields of wireless sensor networks, recommendation systems, cloud service selection and the like. As a lightweight and efficient safety guarantee mechanism, the method plays an important role in the research of network safety. Most trust management mechanisms focus on the calculation of trust values and the updating of trust values, which take little or no account of the vulnerability of the trust mechanism itself. Particularly, with the development situation of complication and dynamism of the wireless network, the trust calculation process inevitably needs recommendation information of a third-party node, and when the recommendation information conflicts with each other, the trust decision process becomes complicated. There have been many studies on this aspect for the trust recommendation conflict problem, but most of the conventional studies only consider the case of recommendation information conflict caused by dishonest recommendation, and do not distinguish between dishonest recommendation and wrong recommendation. The dishonest recommendation attack is subjective malicious attack behavior of the third-party node, while the wrong recommendation is an objective judgment error of the third-party node. A trust decision method is provided based on node interaction heterogeneity and underwater acoustic channel instability, and the problem of wrong recommendation in UASNs can be solved. In addition, an incentive mechanism based on prisoner predicament is provided, and the energy consumption in the trust decision process and the accuracy of the trust decision result are balanced.

Relevant research literature for the problem of trust recommendation conflicts at present is as follows:

in 1.2015, Antesar M et al proposed a Trust Model and defense mechanism Based on recommendations in "Recommendation Based Trust Model with an efficient defense Scheme for plants", which is Based on a clustering technique, clusters recommended nodes according to the number of interactions between nodes, the consistency with the view of the evaluated nodes, and the physical proximity to the evaluated nodes, and dynamically filters out attacks related to dishonest recommendations within a specific time.

In 2.2015, Jinfang Jiang et al proposed that after receiving the recommendation information, their average values were first calculated, and the degree of outlier of each recommendation information was evaluated based on the average values. And a concept of relation familiarity is provided, and higher weight is given to the recommendation information of the long-term neighbor nodes of the nodes. Recommendation information is fused based on the outlier parameter and the relationship familiarity, and dishonest recommendation attacks are resisted.

In 3.2018, Jia Hu et al proposed a Crowdsourcing Strategy DTCS with enhanced Data reliability in DTCS: An Integrated Stratagy for Enhancing Data trust in Mobile Crowdsourcing, which integrates auction games into MCS, ensures the reliability of collected Crowdsourcing information by exciting participants to provide real Crowdsourcing information, and implements An effective punishment mechanism to effectively prevent internal collusion conflict behavior and spoofing attack.

The invention content is as follows:

aiming at the problems, the invention provides a trust management mechanism based on conflict resolution in an underwater acoustic sensor network, and provides a reliability measurement parameter for measuring recommended nodes, namely node reliability, based on node interaction time-frequency heterogeneity and preferences on different trust evidences; and according to the channel information acquired in the sliding window, a reliability measurement parameter for measuring the recommendation information, namely data reliability, is provided. And then finishing a trust decision based on the node reliability and the data reliability. In addition, in order to save network energy consumption and balance the accuracy of trust decision results, an incentive mechanism based on prisoner predicament is provided.

The technical purpose is achieved, the technical effect is achieved, and the invention is realized through the following technical scheme:

a trust management mechanism based on conflict resolution in an underwater acoustic sensor network comprises the following steps:

the method comprises the following steps: trust evidence collection

Because the storage capacity of the underwater sensor node is limited, based on the sliding window storage trust value, each sliding window comprises a limited number s of time slots, wherein the number s of the time slots is set according to the storage capacity of the node and the data arrival rate. Generally speaking, the larger the value of s, the more accurate the trust calculation result, but the higher the requirement on the storage capacity of the node. The sliding window slides forward along the time axis, i.e. one time slot is added, and the oldest time slot is discarded; recording trust evidence for trust calculation, namely delivery rate trust, delay trust and energy consumption trust in each time slot; recording channel information, namely received signal strength, signal-to-noise ratio and node moving distance in each time slot;

step two: incentive mechanism based on prisoner trapping situation

In the trust recommendation process, all the neighbor nodes with trust values larger than the threshold can participate in the recommendation process, and the trust value of the invention adopts 0,1]Interval, therefore the threshold is set to 0.5. The number of neighbor nodes that can participate in the recommendation process is denoted as n, and these n neighbor nodes are referred to as recommendation nodes. In an actual application environment, because the energy of the sensor nodes is limited, if all the recommendation nodes participate in the process, more energy of the network is consumed, and therefore the n recommendation nodes are not required to actually participate in the recommendation process. However, considering the potential selfish behavior of the nodes, if a sufficient number of recommended nodes do not participate in the recommendation process, the result of the trust decision is inaccurate. It is assumed that in a certain round of trust recommendation process, an evaluation node needs to receive at least k pieces of trust recommendation to complete an accurate trust decision. Therefore, in order to ensure that at least k recommendation nodes participate in the trust recommendation process, the process of uploading recommendation information by the n recommendation nodes is described as prisoner predicament; weighing accuracy of trust evaluation resultsAnd energy consumption of the recommendation processDetermining the value of k, wherein k is more than or equal to 1 and less than or equal to n, and the value of k needs to meet the target

Step three: trust decision based on node reliability and data reliability

Determining the node reliability of the recommended node according to the time-frequency characteristics of the interactive history of the recommended node and the target node and the preferences of different trust evidences; and determining the data reliability of the recommendation information based on a grey theory according to the recorded channel related information. And finishing a trust decision based on the node reliability and the data reliability.

In the first step, the specific setting method of the delivery rate trust, the delay trust and the energy consumption trust is as follows:

(1) the delivery rate trust is described using the number of interaction failures and the number of interaction successes:

assuming that s represents the number of successful interactions and f represents the number of failed interactions, in the traditional trust value calculation based on beta distribution, when there is no interaction, the initial trust value of a node is 0, which is not beneficial to the node to start the interaction in the network. Therefore, the traditional trust value calculation based on beta distribution is improved by using α ═ s +1 β ═ f +1, so that the initial trust value of the node in the network is 0.5, in order to start interaction in the network. Eta represents a penalty factor for penalizing an attacker who initiates on-off attack; η is calculated as follows:

(2) latency trust is related to the distance of two node keys:

wherein, delay _ time_iIndicating the actual delay, distance, of the ith packet_iIndicating the distance between the transfer node and the transmitting node at the time of the ith packet transmission.

(3) The energy consumption is trusted as follows:

whereinRepresenting the consumed energy of the receiving node, and E is the initial energy of the receiving node.

Further, the reason for recording the received signal strength, the signal-to-noise ratio and the node moving distance in the sliding window is as follows:

to quantify the impact of the underwater acoustic communication process on trust calculations. The factors considered are as follows: (i) link connection capability: for a mobile node, link fading is a statistical process with unknown parameters, and from a trust point of view, no actual statistical information of link fading is needed, and a trust-related metric is "whether a channel supports successful delivery of a data packet". If the received signal strength value between two nodes is less than the carrier sense threshold, the data packet will not be received. If the received signal strength value is greater than the carrier sense threshold but less than the detection threshold, decoding failure may result. (ii) Link stability capability: an important factor affecting the quality of the underwater acoustic channel is the presence of multiple noises in the underwater environment, and as such, noise is a statistical process for which parameters are unknown. From a trust perspective, actual noise statistics are not needed, and the relevant metric is "whether the signal-to-noise ratio supports successful delivery of a packet". (iii) Water flow influence: the faster the distance between nodes changes over time, the greater the water flow mobility impact.

In the step (2), the prisoner trapping situation model is as follows:

any node, called Current node, is analyzed, and Other nodes are called Other nodes. Their game benefits are represented in the form shown in the table. Wherein T represents the reward obtained by the node participating in the recommendation process, C represents the cost of the node participating in the recommendation process, a represents the adjustment factor of the reward degree, and a is (0, 1). After Nash equilibrium analysis, whenIn time, k pieces of trust recommendation information can be guaranteed to be used for completing trust decision. Therefore, the method can prevent the potential selfish behavior of the node from influencing the accuracy of the trust decision result, and can not accept excessive recommendation and consume extra energy.

In the step (3), the specific method for setting the node reliability is as follows:

(1) suppose that the preference vectors of k different recommendation nodes with respect to different trust evidences are Indicating the ith node's preference for the mth trust evidence. The preference similarity of the nodes is defined as follows:

wherein the content of the first and second substances,

(2) according to the time-frequency characteristics of the interaction history of the recommended node and the target node, defining the time confidence of the nodes as follows:

wherein c is_tThe number of data packets sent in the T-th time slot is shown, and T shows the total number of time slots with interaction between two nodes. C_tIndicating the maximum number of packets that can theoretically be sent in a time slot, E (c)_t) The calculation method is as follows:

E(c_t)＝-plog(p)-(1-p)log(1-p)

since the above formula takes the maximum value when p is 0.5, (0, C) is added_t]Mapping to Interval (0, 0.5)]And (4) finishing. (3) According to the preference similarity and the time confidence, defining a reliability measurement parameter of the recommended node, namely the node reliability is as follows:

node_reliability_i＝w₁×PS_i+w₂×TC_i

wherein w₁And w₂Represents a weight, w₁+w₂＝1。

Further, in the step (3), the specific setting step of the data reliability is as follows:

assuming that the received signal strength RSSI acquired by the recommended node through the sliding window, the signal-to-noise ratio and the average node mobility are represented as follows:the optimal reference sequence and the worst sequence are selected as follows

(1) And performing grey correlation analysis on the comparison sequence and the reference sequence to obtain a grey correlation coefficient.

The following formulaFor the purpose of example only,the analysis method is similar.

Wherein rho belongs to [0,1], represents a resolution coefficient and is used for adjusting the difference size of an output result;

(2) obtaining the gray correlation degree of the k comparison sequences according to the gray correlation coefficient

Wherein v is₁、v₂、v₃Representing the weight, v₁+v₂+v₃＝1。

(3) Data reliability of k comparison sequences using least squares

Further, in the step (3), a specific setting method of the trust decision based on the node reliability and the data reliability is as follows:

wherein, node _ reliability_iRepresenting the node reliability of the recommended node i; data _ reliability_iRepresenting the data reliability of the recommendation information uploaded by the recommendation node i; u. of₁，u₂Representing the weight.

Further, in the step (3), the trust decision manner is as follows: :

wherein R is_iAnd representing recommendation information uploaded by the recommendation node i. weight_iIs the result of the optimization problem in the formula, determining R_iWeights in trust evidence fusion, i.e.

Drawings

FIG. 1 is a flow chart of the present invention;

FIG. 2 is a schematic view of a sliding window;

fig. 3 is a schematic diagram of the recommendation process.

Detailed Description

In order to make the objects, technical solutions and advantages of the present invention more apparent, the present invention is further described in detail with reference to the following embodiments. It should be understood that the specific embodiments described herein are merely illustrative of the invention and are not intended to limit the invention.

The invention aims to solve the problem of trust recommendation conflict of a trust management mechanism in the existing underwater acoustic sensor network, and provides a trust management mechanism based on conflict resolution in the underwater acoustic sensor network. According to the method, two parameters are provided according to the heterogeneity of the node interaction history and the underwater acoustic communication instability, namely the reliability measurement of the recommended node-node reliability and the reliability measurement of the recommended information-data reliability are measured, and a trust decision method based on the node reliability and the data reliability is provided, so that the problem that the wrong recommendation is neglected in the conventional trust recommendation conflict processing method is solved. In addition, the invention provides an incentive mechanism based on prisoner predicament, which balances the energy consumption of the trust recommendation process and the accuracy of the trust decision result, as shown in fig. 1, the incentive mechanism specifically comprises the following steps:

(1) prior to the recommendation process, the transmitting node maintains a sliding window about the target node, the sliding window containing s time slots for recording information needed for trust decisions. The method comprises the steps of trust calculation required delivery rate trust, delay trust and energy consumption trust to form a trust opinion about a target node; and channel information required to resolve the recommendation conflict, i.e., received signal strength of the channel, signal-to-noise ratio, and node movement distance.

(2) After the evaluation node initiates trust recommendation, considering the potential selfish behavior of the node, describing the process of uploading recommendation information by the n recommendation nodes as prisoner-migrant predicament, and ensuring that k recommendation nodes participate in the process. And balancing the energy consumption of the trust decision process and the accuracy of the trust decision result.

(3) After the evaluation node receives the k pieces of recommendation information, determining the node reliability of the recommendation node according to the time-frequency characteristics of the interactive history of the recommendation node and the target node and the preferences of different trust evidences; and determining the data reliability of the recommendation information based on a grey theory according to the recorded channel related information. And finishing a trust decision based on the node reliability and the data reliability.

As shown in fig. 2, the transmitting node maintains the interaction information with the target node using a sliding window, and each time the window moves by one slot, the oldest slot is discarded. And recording trust opinions about the target node in each time slot, wherein the trust opinions comprise delivery rate trust, delay trust and energy consumption trust, and the calculation method comprises the following steps:

(1) the delivery rate trust is described using the number of interaction failures and the number of interaction successes:

where α + s +1 β is f +1, s represents the number of successful interactions, and f represents the number of failed interactions. Eta represents a penalty factor for penalizing an attacker who initiates an on-off attack. η is calculated as follows:

(2) latency trust is related to the distance of two node keys:

wherein, delay _ time_iIndicating the actual delay, distance, of the ith packet_iIndicating the distance between the transfer node and the transmitting node at the time of the ith packet transmission.

(3) The energy consumption is trusted as follows:

whereinRepresenting the consumed energy of the receiving node, and E is the initial energy of the receiving node.

As shown in fig. 3, in n recommendation nodes, there may be a selfish node that does not participate in trust recommendation in order to save energy consumption of itself, which may result in a decrease in accuracy of a trust decision result. Meanwhile, when n is large, if the number of recommended nodes is not limited, a large amount of network energy consumption is caused. Therefore, considering the above two reasons, at least k recommendation nodes are needed to participate in the recommendation process among the n recommendation nodes. This process is therefore described as prisoner's distress:

any node, called Current node, is analyzed, and Other nodes are called Other nodes. Their game benefits are represented in the form shown in the table. Wherein T represents the reward obtained by the node participating in the recommendation process, C represents the cost of the node participating in the recommendation process, a represents the adjustment factor of the reward degree, and a is (0, 1). Assuming that the probability of reply of any node is P, the probabilities of "At most k-1nodes reply" and "At least k nodes reply" are respectively:

according to the formula of Lei-Nenitz:

making the substitution can result in:

p(x≤k-1)＝[p(1-p)]^k-1＝m

p(x≥k)＝1-m

calculating Nash equilibrium:

revenue for Current node selection no reply:

m*0+(1-m)*T

yield of Current node selection reply:

m*(T-C)+(1-m)*a*(T-C)

make the above two formulas equal

T-mT＝mT-mC+aT-aC-maT+maC

Simplifying:

because of the fact thatTherefore, it is not only easy to use

When the above conditions are met, K pieces of trust recommendation information can be ensured to be used for completing trust decision. Therefore, the method not only prevents the potential selfish behavior of the node from influencing the accuracy of the trust decision result, but also can not accept excessive recommendation and consume extra energy

In summary, the following steps:

the invention discloses a trust management mechanism based on conflict resolution in an underwater acoustic sensor network, which is used for processing the problem of mutual conflict of recommendation information of a plurality of third-party nodes in the trust management mechanism. The method firstly combines the characteristics of node interaction and the characteristics of the underwater acoustic sensor network, provides a trust decision method based on node reliability and data reliability, and solves the problem of wrong recommendation of a trust management mechanism in the underwater acoustic sensor network. In addition, an incentive mechanism based on prisoner predicament is provided, and energy consumption in the trust recommendation process and accuracy of the trust recommendation result are balanced.

The above description is only a preferred embodiment of the present invention, and it should be noted that, for those skilled in the art, several modifications and variations can be made without departing from the technical principle of the present invention, and these modifications and variations should also be regarded as the protection scope of the present invention.