CN107750053A - Based on multifactor wireless sensor network dynamic trust evaluation system and method - Google Patents

Based on multifactor wireless sensor network dynamic trust evaluation system and method Download PDF

Info

Publication number
CN107750053A
CN107750053A CN201710380752.9A CN201710380752A CN107750053A CN 107750053 A CN107750053 A CN 107750053A CN 201710380752 A CN201710380752 A CN 201710380752A CN 107750053 A CN107750053 A CN 107750053A
Authority
CN
China
Prior art keywords
trust
node
value
factor
direct
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
CN201710380752.9A
Other languages
Chinese (zh)
Inventor
李晓红
王江娟
宋姣娇
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.)
Tianjin University
Original Assignee
Tianjin 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 Tianjin University filed Critical Tianjin University
Priority to CN201710380752.9A priority Critical patent/CN107750053A/en
Publication of CN107750053A publication Critical patent/CN107750053A/en
Pending legal-status Critical Current

Links

Classifications

    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04W—WIRELESS COMMUNICATION NETWORKS
    • H04W24/00—Supervisory, monitoring or testing arrangements
    • H04W24/08—Testing, supervising or monitoring using real traffic
    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04W—WIRELESS COMMUNICATION NETWORKS
    • H04W40/00—Communication routing or communication path finding
    • H04W40/02—Communication route or path selection, e.g. power-based or shortest path routing
    • H04W40/04—Communication route or path selection, e.g. power-based or shortest path routing based on wireless node resources
    • 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

Abstract

The invention discloses a kind of based on multifactor wireless sensor network dynamic trust evaluation system and method, characterized in that, the system includes trust factor acquisition module (11), trust metrics module (12), trusts computing module (13) and trust evaluation module (14)；Step 1: obtain each trust factor that may influence node trust value, Step 2: define the metric form of each trust factor, Step 3: realize it is each trust calculate, trust the obtained synthesis trust value of destination node of calculating process and the magnitude relationship of trust threshold Step 4: comparing to judge whether destination node credible.Compared with prior art, the present invention has considered the behavior when node transmits sensing data, so that the behavior of monitoring node more fully hereinafter, ensures in time, effectively to handle the insincere node in network；In addition, using the direct trust value trusted and weigh node by the way of trust synthesis indirectly, there is the good effect of offset issue in evaluation result between solving node.

Description

Based on multifactor wireless sensor network dynamic trust evaluation system and method
Technical field
The invention belongs to the wireless communication network technology of Internet of Things security fields, more particularly to a kind of based on multifactor Wireless sensor network dynamic trust evaluation system and method.
Background technology
With the popularization that Internet of Things is applied, wireless sensor network is perceived as the most widely used bottom of Internet of Things and set It is standby, it has been successfully applied to environmental monitoring, SS, aerospace station, military and national defense, smart home, smart city, doctor Treat in the key areas such as monitoring.At the same time, due to the limitation of sensing node self-condition and the polytropy of application environment, nothing The safety problem that line sensor network occurs is also increasingly severe.
Wireless sensor network is a dynamic network：Network node can move everywhere；Node may be because Exhausted for the energy content of battery or other reasonses break down or even failed, so as to the operation that exits network；The continuous of new node adds Enter.Wireless sensor network is also an Open Network, and sensing node usually in a manner of launching at random, is deployed in nobody Supervision, in rugged environment, directly contacted between node and environment.On the other hand, because node resource is limited, section Point usually needs to carry out data transmission by way of locally cooperating, and is entered using the communication easily by attack destruction Row communication.Due to the limitation of sensing node self-condition and the polytropy of application environment, it is normal to cause wireless sensor network Chang Rongyi is by various attacks.Including targeted attack, extensive aggression, Wormhole attack, Sybil attack, Sinkhole attacks, false route, ACK impersonation attacks, dos attack, initiation collision attack, unfairness occupy resource attack etc.. Attack can be divided into external attack according to attack pattern and internaled attack.External attack refers to attacking from its exterior Hit, the identity of the promoter of this attack fails to obtain legal checking, and attacker can only take separated network or introduce a large amount of Data flow mode, disturb the proper communication of network.Relative to external attack, it is by capturing network internal section to internal attack Point or the attack for cracking the key structure of network and implementing, this attack pattern is more not easy to discover, to network proper communication Threat it is also bigger.
Traditional security mechanism based on AES, can only solve the external attack that identity not yet obtains certification, And can not solve due to node is captured and what is occurred internals attack problem.Research finds have as to conventional cryptography mechanism Effect supplement, trust evaluation mechanism can solve the problem of internaling attack well.
Wireless sensor network trust evaluation study be concentrated mainly on to influence node trust value factor be modeled with Calculate, and management is controlled it according to node trust value, make every effort to establish a trustable network being made up of trusted node, from And ensure network can long-term safety reliably run.
The content of the invention
In order to reduce influence of the malicious node to network, improve internet security, the present invention propose it is a kind of based on it is more because The wireless sensor network dynamic trust evaluation system and method for element, and have devised and embodied the dynamic letter of wireless sensor network Appoint evaluation platform, using the wireless sensor network dynamic trust evaluation platform of design, realize and wireless sensor network is carried out Formalization analysis, and the general wireless sensor network security of various protocols is detected.
The present invention's is a kind of based on multifactor wireless sensor network dynamic trust evaluation system, and the system includes trusting Factor acquisition module 11, trust metrics module 12, trust computing module 13 and trust evaluation module 14, wherein
The trust factor acquisition module 11, the trust factor that trust evaluation process is related to is obtained, including directly trusted Factor, indirect trust factor and comprehensive trust factor；
The trust metrics module 12, definition include direct trust factor, indirect trust factor and comprehensive trust factor and existed The metric form of interior each trust factor；
The trust computing module 13, is realized according to direct trust value defined in above-mentioned trust metrics module, indirectly Trust value and comprehensive trust value metric form carry out the calculating of each trust value；
The trust evaluation module 14, compare the synthesis trust value for trusting the destination node that calculating process obtains with trusting threshold The magnitude relationship of value, judge whether destination node is credible according to this.
The a kind of of the present invention includes following based on multifactor wireless sensor network dynamic trust evaluation method, this method Step：
Step 1: obtaining each trust factor that may influence node trust value, i.e., direct trust factor includes current straight Trust factor and the direct trust factor of history are connect, current directly trust factor is believed by the communication trust factor and sensing of current period Appoint factor composition, and both direct interaction processes by being operated between the node monitoring of promiscuous mode and other nodes And obtain；Indirect trust factor comes from the recommendation of other adjacent nodes；Direct trust factor and indirect trust factor are formed Comprehensive trust factor, using the final foundation that comprehensive trust value is credible as evaluation node；
Step 2: definition including direct trust factor, indirect trust factor and integrate trust factor including each trust because The metric form of element；
Step 3: realize that each trust calculates, including：
Directly trust and calculate, to the communication trust-factor w used in calculating process1, sensing trust value factor w2, the time declines Subtracting coefficient f (t-t0), historical influence factor θ analyzed and defined；
Trust indirectly and calculate, analyze and define recommendation trust sorting algorithm；Using the algorithm by the recommendation of adjacent node Opinion be divided into determination recommend and it is uncertain recommend two classes, and respectively define subjective assessment algorithm and irrelevance testing algorithm is used for Trust the weight distribution of calculating process；
It is comprehensive trust calculate, define the dynamic weights assignment algorithm based on interaction times, by synthesize direct trust value and The synthesis trust value that indirect trust values obtain carrys out the credibility of evaluation node；
Step 4: compare the synthesis trust value of destination node and the magnitude relationship of trust threshold for trusting that calculating process obtains To judge whether destination node is credible；
If the trust value of node is more than or equal to trust threshold, predicate node is believable, can continue to interact； If the trust value of node is less than trust threshold, node is incredible, is no longer interacted, and carries out deleting related route letter Cease, identify the work of the subsequent treatments such as insincere node.
Compared with prior art, the present invention only focused on different from former wireless sensor network dynamic trust evaluation method Behavior of the node in route finding process；The behavior when node transmits sensing data is considered, so as to more comprehensive The behavior of ground monitoring node, ensure in time, effectively to handle the insincere node in network；In addition, the present invention is using directly letter Appoint and trust the trust value that comprehensive mode weighs node indirectly, serve and solve between node because intersection record is few, evidence It is insufficient and cause evaluation result the good effect of offset issue occur.
Brief description of the drawings
Fig. 1 is the present invention based on multifactor wireless sensor network dynamic trust evaluation system frame diagram；
Fig. 2 is the acquisition modes figure of trust factor；
Fig. 3 is indirect trust values calculation flow chart；
Fig. 4 is recommendation trust graph of a relation；
Fig. 5 is dynamic trust evaluation rubric figure；
Fig. 6 is node deployment exemplary plot；
Fig. 7 is that experiment effect compares figure；(a), 10% malicious node network throughput；(b), 10% malicious node network is being just Normal delivery ratio；(c), 20% malicious node network throughput；(d), the 20% normal delivery ratio of malicious node network；(e), 30% dislike Meaning meshed network handling capacity；(f), the 30% normal delivery ratio of malicious node network；(g), the expense of primitive network；(h), based on letter Appoint the expense of the network of mechanism.
Embodiment
The present invention is described in further detail below in conjunction with the accompanying drawings.
As shown in figure 1, four parts are mainly included based on multifactor wireless sensor network dynamic trust evaluation system：
First, trust factor acquisition module 11：It is mainly used to obtain the trust factor that trust evaluation process is related to, including it is straight Connect trust factor (current directly trust factor, the direct trust factor of history), indirect trust factor and comprehensive trust factor.Such as figure It is the acquisition modes schematic diagram of trust factor shown in 2.Direct trust factor be divided into again communication trust factor and sensing trust because Element, monitoring direct interaction process under promiscuous mode is all operated in by node and obtained.Current directly trust factor is by current The communication trust factor in cycle and sensing trust factor composition, and both by be operated in the node of promiscuous mode monitor with Obtained from direct interaction process between other nodes.The direct trust factor of history be exactly history cycle direct trust because Element.Indirect trust factor comes from the recommendation of other adjacent nodes.Direct trust factor and the comprehensive letter of indirect trust factor composition Appoint factor, comprehensive trust value is the credible final foundation of evaluation node.Using the module analysis and summarize possible influence section The factor and its acquisition modes of point trust value.Foregoing interaction refer to when node down hop node send packet when, Whether the packet that is forwarded by contrasting node with the next-hop node that listens in the packet of local cache unanimously judges Whether next-hop node correctly forwarded packet.It is to set node to be operated in promiscuous mode to monitor the basis realized.It is if correct Forwarding is considered as and once successfully interacted, otherwise unsuccessfully to interact.Trust factor comes from principal and (refers to radio sensing network source section Point) history interbehavior record, the current row of target entity between target entity (referring to radio sensing network destination node) It is characterized and the recommendation opinion of other entities etc..
2nd, trust metrics module 12, for defining the metric form of each trust factor：
Define 1, current direct trust value CDTi,j(t) it is defined as：
Wherein, RSCFi,j(t)、DSCFi,j(t) communications of the host node i on destination node j in current period t is represented respectively Trust value and sensing trust value, w1, w2Communication trust-factor and the sensing trust value factor are represented respectively.
Define 2, history direct trust value HDTi,j(t) it is defined as：
HDTi,j(t)=f (t-t0)*CDTi,j(t0) (2)
Wherein, t CDTi,j(t0) represent history cycle t0Direct trust values of the interior host node i on destination node j, time Decay factor f (t-t0) represent on current period t and history cycle t0Time interval linear function.
Define 3, direct trust value DTi,j(t) it is defined as：
DTi,j(t)=θ * HDTi,j(θ)+(1-θ)*CDTi,j(t) (3)
Wherein, HDTi,j(t)、CDTi,j(t) represent that host node i is straight on destination node j history in current period t respectively Trust value and current direct trust value are connect, θ represents the historical influence factor；
Define 4, indirect trust values ITi,j(t) it is defined as：
Wherein, DTi,k(t) represent direct trust values of the host node i on neighbor node k in current period t, host node i with Neighbor node k interaction times are m, Tk,j(t) comprehensive letters of the destination node j on neighbor node k in current period t is represented It is T to appoint valuek,j(t), neighbor node k Combination of recommendation trust weight is w'k(t)；
Define 5, comprehensive trust value Ti,j(t) it is defined as：
Ti,j(t)=β ' * DTi,j(t)+(1-β′)*ITi,j(t) (5)
DTi,j(t)、ITi,jAnd T (t)i,j(t) direct trusts of the host node i on destination node j in cycle t is represented respectively Value, indirect trust values and comprehensive trust value；β ' expression direct trust values DTi,j(t) in comprehensive trust value Ti,j(t) in building-up process Weight.
3rd, trust computing module 13, for realize according to direct trust value defined in above-mentioned trust metrics module, Trust value and comprehensive trust value metric form are connect to carry out the calculating of each trust value：
(1), directly trust and calculate
Directly trust and calculate, to the communication trust-factor w used in calculating process1, sensing trust value factor w2, the time declines Subtracting coefficient f (t-t0), historical influence factor θ analyzed and defined, the meter of direct trust value is carried out according to formula (1)-(3) Calculate.
Direct trust value is that node is successful by calculating respectively in establishing path process and during transmission sensing data Interactive probability, and what is averagely obtained is weighted to it.
In cycle t, t is represented in path process is established, the probability that node i successfully interacts with node j, is set as work(interaction Number is rsci,j(t), failure interaction times are rfci,j(t), then
In cycle t, DSCFi,j(T) represent during sensing data is transmitted, node i successfully interacts general with node j Rate, it is dsc to be set as work(interaction timesi,j(t), failure interaction times are dfci,j(t), then
The trust-factor that communicates w1With sensing trust-factor w2Value be the malicious act of node by node interaction Coverage caused by possible to network determines.Malicious act of node may influence mulitpath just when establishing path Malicious act when often building, and transmitting sensing data can only influence the normal transmission of the packet of this forwarding, so Generally definition, 0≤w2≤w1≤1。
Direct trust value is with certain by the direct trust value in the direct trust value and current period in history cycle Historical influence combinations of factors forms.Wherein, history direct trust value is by after the direct trust value decay in history cycle Obtain.In order to avoid there is original believable node insincere section is mistaken as into because of not interacting for a long time Point, the situation of the wasting of resources is caused, definition decay herein is bounded.It is assumed that CDTi,j(t0) represent cycle t0Interior nodes i on Node j history direct trust value, time decay factor f (t-t0), distinguish credible and incredible trust threshold Th, CDTi,j (t0) through becoming HDT after overdampingi,j(t), it is expressed as：
Furthermore, it is contemplated that " the volatile rare property " trusted, the historical influence factor θ being defined in formula (22)：
Wherein 0 ＜ θ2The ＜ θ of ＜ 0.51＜ 1.
Above-mentioned θ1,θ2, Th, f value are determined by the application environment of reality.
(2), trust indirectly and calculate
Trust indirectly and calculate, analyze and define recommendation trust sorting algorithm；It is thick in order to solve recommendation trust granularity of classification The problem of rough, divide the recommendation trust of adjacent node into determination using the algorithm and recommend and do not know two classes of recommendation, and it is fixed respectively Justice subjective assessment algorithm and irrelevance testing algorithm are used for the weight distribution for trusting calculating process, enter in the ranks according to formula (4) Connect the calculating of trust value.
The acquisition process trusted indirectly is periodically to be broadcasted from host node to adjacent node single-hop on some destination node Recommendation trust request starts.Credible threshold value Th is set, and determines to recommend n'min, step 301；Node i is broadcasted on target section Point j recommendation request bag RequestTrust, step 302；Neighbor node k receives the recommendation request bag that node i is sent RequestTrust, step 303；Search the trust table Trust of self maintained, step 304；Judge to trust in table Trust whether Node j synthesis trust value, step 305 be present；If it does not exist, then abandon recommendation request bag RequestTrust, flow knot Beam, step 306；Recommend to reply ReplyTrust, step 307 if it is present neighbor node k is sent to node；Node j is received ReplyTrust, step 308 are replied in recommendation to neighbor node k；Search the trust table Trust of self maintained, step 309；Sentence The disconnected direct trust value that neighbor node k be present and trust value is more than Th, step 310；If it does not exist, then abandon ReplyTrust, step 311；If it does, determine whether be more than n' with node k interaction timesmin, step 312； If it is greater, then neighbor node k is determination recommendation trust class, step 313；Perform subjective assessment algorithm and determine synthetic weight, walk Rapid 314；It is determined that the letter of class synthesis is recommended to connect trust value, step 315；If being not more than, neighbor node k is uncertain recommendation trust Class, step 316；Perform irrelevance testing algorithm and determine synthetic weight, step 317；The uncertain indirect trust for recommending class synthesis Value, step 318；Weighting obtains final indirect trust values, step 319.Above-mentioned calculation process is sketched：Adjacent node receives master After the recommendation request packet that node is sent, the trust table itself safeguarded is searched, judges whether that destination node is trusted Value.If in the presence of to host node transmission recommendation reply packet；Otherwise, disregard.Host node receives in the delay of permission Recommend after replying packet, search the recommendation request packet of caching, judge whether to be transmitted across the recommendation request.If in the presence of, Then search and trust table, continue to judge whether the trust value of recommended node is more than recommendation trust threshold value, if being more than, receive this and push away Recommend, otherwise do not receive.
In order to which the calculating process for solving to be triggered by trust chain is complicated, convergence rate is slow, can not adapt to asking for large scale network Topic, herein, common neighbor node ks of the host node i only with reference to host node i and destination node j1,k2,...,kn-1,knPush away Opinion is recommended to calculate destination node j indirect trust values.
As shown in figure 4, the recommendation trust graph of a relation between node i and node j.Wherein, unidirectional solid line represents directly to believe The relation of appointing, two-way solid line represent comprehensive trusting relationship, and dotted line represents recommendation trust relation.In order to solve Malicious recommendation and collusion Attack problem, definition node i are direct trust value to judge whether to receive according to itself trust evaluation result to adjacent node k Neighbor node k recommendation opinion.If the direct trust value of neighbor node is less than trust threshold, neighbor node is insincere section Point, node i do not receive the recommendation trust of its offer.For acceptable recommendation trust, in order to avoid node i and adjacent segments Point k is because interaction times are few, evidence is insufficient and cause secondary effect to node j synthesis trust evaluation results, and synthesis is examined herein Consider interaction times and determine Combination of recommendation trust weight with two factors of direct trust value.
If it is m that node i, which interacts total degree with neighbor node k, node i is DT on node k direct trust valuei,k(t), Then from formula (3-4), synthetic weight w 'kFor：
w′k(t)=f " (DTi,k(t),m) (10)
Recommendation trust is divided into determination according to the interaction times between node and recommends and do not know two major classes of recommendation, and is divided The weight distribution that subjective assessment algorithm and irrelevance testing algorithm are used for during indirect trust combination is not defined.
As described in Hoeffding ' the s inequality of extension：" in acceptable error range t, and if only if saves N when point interaction times are at leastmin, the confidence level of the trust value obtained by monitoring direct interaction process at least meets α ".Assuming that It is n ' in some acceptable error range t ' and minimum interaction times interior confidence level α 'min, n 'minAs determine to recommend threshold Value, and division determines to recommend class to recommend class with uncertain accordingly.
If m >=n 'min, then it is assumed that neighbor node k recommendation trust, which belongs to, determines recommendation trust class, is taken based on subjectivity and comments The weight distribution algorithm synthesis of valency determines recommendation trust, and (node i is according to neighbor node k direct trust value DTi,k(t) big It is small to determine synthetic weight), otherwise it is uncertain recommendation trust class, while using the weight distribution algorithm synthesis based on irrelevance Uncertain recommendation trust (the recommendation that node i provides according to the recommendation trust that neighbor node k is provided with other neighbor nodes Appoint value DTi,k(t) departure degree between determines synthetic weight).
Finally pair determination recommendation trust and uncertain recommendation trust are weighted average calculating operation and produce node i on section Point j indirect trust values.
Algorithm 1：Subjective assessment weight distribution
Input：Node i is on neighbor node k1,k2,...,kmDirect trust value DTi,1(t),DTi,2(t),...,DTi,m (t)。
Output：Neighbor node k1,k2,...,kmWeight.
1. calculate node i is on all direct trust value sums for belonging to the neighbor node for determining to recommend class.
Wherein, DTi,k(t) represent that input node is i, neighbor node kkDirect trust value；K represents of neighbor node Number；
2. calculate neighbor node k weight w 'k(t)；
w'k(t)=DTi,k(t)/S'(t) (12)
Wherein, DTi,k(t) represent that input node is i, neighbor node kkDirect trust value；
Algorithm 2：Irrelevance tests weight distribution
Input：Neighbor node km+1,km+2,...,knSynthesis trust value T on destination node jm+1,j(t), Tm+2,j ..., T (t)n,j(t)；
Output：Neighbor node km+1,km+2,...,knWeight.
1. the deviation journey between calculating the recommendation trust that the recommendation trust of neighbor node k offers and other nodes provide Degree.
Wherein, Tk,j(t) represent to be calculated by formula (5) is synthesis of the host node k on destination node j in cycle t Trust value；Tr,j(t) represent to be calculated by formula (5) is that host node r trusts on the comprehensive of destination node j in cycle t Value；N represents the number of neighbor node；
2. calculate total deviation value of uncertain trust recommendation class.
Wherein, sk(t) represent to calculate recommendation trust and the offer of other nodes that neighbor node k is provided by formula (13) Departure degree between recommendation trust；
3. calculate the deviation value s of total deviation value and neighbor node kk(t) proportionate relationship.
dk(t)=S (t)/sk(t) (15)
Wherein, sk(t) represent to calculate recommendation trust and the offer of other nodes that neighbor node k is provided by formula (13) Departure degree between recommendation trust；S (t) represents the total deviation for the uncertain trust recommendation class being calculated by formula (14) Value；
4. to dk(t) it is normalized, produces node k weight w 'k(t)。
Wherein, dk(t) the total deviation value being calculated by formula (15).
(3) it is, comprehensive to trust computational methods
It is comprehensive to trust calculating, the dynamic weights assignment algorithm based on interaction times is defined, by synthesizing direct trust value The synthesis trust value obtained with indirect trust values carrys out the credibility of evaluation node；In addition, trust value needs to carry out cycle renewal, with Just the Behavioral change of node is observed in time.
When host node and destination node interaction times are less, it may make it that host node can not because evidence is insufficient The understanding of itself is fully relied on accurately to evaluate destination node.In order to ensure the accuracy of evaluation result, use herein Directly trust the trust value for carrying out evaluation node with trusting comprehensive mode indirectly.In addition, trust value needs the cycle to update, so as to reach The purpose of real-time monitor node behavior.
Existing wireless sensor network trust evaluation method is closed using direct trust value and indirect trust values are static mostly Into method carry out the synthesis trust value of calculate node, this does not meet the dynamic change characterization of wireless sensor network.Weighed in synthesis In the determination problem of weight, the dynamic corresponding relation established based on Hoeffding ' s inequality between interaction times and synthetic weight, Solve the weight static allocation in conventional wireless sensor network trust evaluation method, do not meet wireless sensor network dynamic The problems such as variation characteristic；Therefore, a dynamic weights assignment algorithm for being based on Hoeffding ' s inequality is defined herein： In error allowed band t, when node interaction times are n, the Minimum support4 of gained evaluation result is comprehensive trust combination mistake The synthetic weight of direct trust value in journey.
In addition, the behavior of monitor node, the synthetic weight that we define direct trust value are bounded for maximum magnitude , that is to say, that node i more or less should all refer to the recommendation opinion of other adjacent nodes.Direct trust value is trusted comprehensive Weight span in building-up process is determined by actual application environment, it is assumed that span is [c, d], wherein 0 ＜ c ＜ d ＜ 1。
Algorithm 3：Comprehensive trust weight distribution, i.e. dynamic weights assignment algorithm：
Input：Current environment acceptable error range e, interaction times m.
Output：The synthetic weight β of direct trust value.
1. weight beta, error range e and interaction times m meet equation.
2exp(-2t2M)=1- β (17)
Wherein, weight beta represents the synthetic weight of direct trust value；M represents interaction times；
2. convert equation.
β=1-2exp (- 2t2m) (18)
Wherein, weight beta represents the synthetic weight of direct trust value；M represents interaction times；
3. bounded processing is carried out to β.
Wherein, weight span of the direct trust value during comprehensive trust combination is determined by actual application environment, It is assumed that span is [c, d], wherein 0 ＜ c ＜ d ＜ 1.
Before this, node i calculates the direct letter for obtaining node j by directly trusting to calculate and indirectly trust respectively Appoint value and indirect trust values, the dynamic weights assignment algorithm defined in conjunction with this section, you can in cycle t, node i on Node j synthesis trust value.
Definition and extended description about Hoeffding ' s inequality are as follows：
In probability theory, Hoeffding ' s inequality is used for defining between the average and its desired value of one group of stochastic variable Absolute difference the probability upper bound.
Provided with a series of pairwise independent stochastic variable X1,X2,...,Xn-1,Xn, it is assumed that to all XiAll it is almost bounded Variable, that is, meet
P(Xi∈ [a, b])=1,1≤i≤n (20)
The experience for defining this n stochastic variable it is expected that (estimated value) is：
ForExpectation (actual value) and meet following inequality：
Define herein, stochastic variable XiThe result that ith interacts between expression node, if successfully interacting, Xi=1, otherwise Xi =0.
Therefore, P (r*X are met hereini∈ (a, b))=1, and a=0, b=1.I.e.：
Wherein, r is Dynamic gene.Herein, Dynamic gene declines with the trust-factor that communicates, sensing trust-factor, time Subtracting coefficient, historical influence are factor-related.
On the premise of given quantity n customer transaction evaluation result, the estimation error of privately owned prestige is given more than some Determine the threshold value t probability upper boundMeet：
Produce：
Namely：
Herein, formula (26) is described as：In the range of acceptable error t, node interaction times that and if only if are most It is n lessminWhen, the confidence level of the trust value obtained by monitoring direct interaction process at least meets α.
4th, trust evaluation module 14, for comparing the synthesis trust value and letter of trusting the destination node that calculating process obtains The magnitude relationship of threshold value is appointed to judge whether destination node is credible.
If the trust value of node is more than or equal to trust threshold, predicate node is believable, can continue to interact； If the trust value of node is less than trust threshold, node is incredible, is no longer interacted, and carries out deleting related route letter Cease, identify the work of the subsequent treatments such as insincere node.
As shown in fig. 6, the node deployment exemplary plot for this example.Wherein with the 6th calculating cycle, the calculate node of node 27 Exemplified by the process of 9 synthesis trust value, this paper dynamic trust evaluation method is illustrated.
In existing trust evaluation method, the initial trust value of node is set to be divided into three kinds of situations：Maximum, median And minimum value.In order to prevent the wasting of resources, it is believed that in the case where not monitoring node performance malicious act, node should This is considered as believable.Therefore, in an experiment, the initial trust value of node is defined as 1 by us, and trust threshold is arranged to 0.5。
The setting of other specification：The trust-factor that communicates and sensing trust-factor are respectively w1=1, w2=0.5, historical influence Factor θ1=0.6, θ2=0.4, time decay factor f (t-t0)=0.9* (t-t0), acceptable error range e=0.1.Close Classify in recommendation, acceptable Minimum support4 α=Th=0.5, that is to say, that it is determined that recommending the minimum interaction times n' of classm =70.During comprehensive trust combination, the span of the weight of direct trust value is defined as c=0.2, d=0.8.Specifically Data are as shown in table 1.
Table 1, calculate data and result
It was found from from upper table, the 6th cycle, the history direct trust value of node 9 is 0.856744, is currently directly trusted It is worth for 0.649879, from formula (22), direct trust value 0.732625.By the interaction for determining to recommend sorting algorithm to determine Number 70 understands that node 40, node 14, node 35, node 16, node 2, node 39, node 22 belong to determination and recommend class.Node 26th, node 43, node 1, node 10, node 33 belong to uncertain recommendation class.Further, since the trust value of node 12 and node 5 Less than trust threshold 0.5, therefore, recommendation opinion of the node 27 without reference to the two nodes.
Trust threshold 0.5 is more than from comprehensive trust value 0.582219：The 6th cycle, node 9 is for node 27 For be believable.
Analysis of experimental results
As shown in fig. 7, (a), (b) are represented when malicious node accounts for the 10% of all nodes, in handling capacity, correct delivery ratio Aspect, wireless sensor network and the original wireless sensor network for adding faith mechanism are essentially identical.Because in reality In testing, network node is random placement, it is possible that when malicious node number is less, malicious node shadow in the entire network It is smaller to ring scope, and causes the DeGrain of faith mechanism.
(c), (d) are shown in Fig. 7, and with the increase of malicious node quantity, the effect of faith mechanism is also more obvious.When When malicious node accounts for the 20% of all nodes, relative to original wireless sensor network, the wireless sensor network based on trust Handling capacity and correct delivery ratio difference balanced growth 15.4%, 27.6%.
(e), (f) are shown in Fig. 7, when malicious node accounts for the 30% of all nodes, the wireless biography based on faith mechanism Sensor network is substantially better than the wireless sensor network for running original AODV agreements, and handling capacity and correct delivery ratio averagely increase respectively Long 30.6%, 54.8%.
In Fig. 7 (g), shown in (h), the network overhead of the wireless sensor network based on faith mechanism is higher than original Wireless sensor network because the wireless sensor network of addition faith mechanism may because handling insincere node and Caused new route finding process and increase some network overheads, and trust calculating process, particularly recommendation trust obtain Process is required for some extra network overheads.It has been found, however, that when malicious node quantity is continuously increased, original nothing Line sensor network expense is increased rapidly, and the wireless sensor network expense based on faith mechanism is then held essentially constant.

Claims (7)

  1. It is 1. a kind of based on multifactor wireless sensor network dynamic trust evaluation system, it is characterised in that the system includes letter Appoint factor acquisition module (11), trust metrics module (12), trust computing module (13) and trust evaluation module (14), wherein
    The trust factor acquisition module (11), obtains the trust factor that is related to of trust evaluation process, including directly trust because Plain, indirect trust factor and comprehensive trust factor；
    The trust metrics module (12), definition is including direct trust factor, indirect trust factor and comprehensive trust factor Each trust factor metric form；
    The trust computing module (13), realize according to direct trust value, indirectly letter defined in above-mentioned trust metrics module Value and comprehensive trust value metric form is appointed to carry out the calculating of each trust value；
    The trust evaluation module (14), compare the synthesis trust value and trust threshold for trusting the destination node that calculating process obtains Magnitude relationship, judge whether destination node credible according to this.
  2. It is 2. a kind of based on multifactor wireless sensor network dynamic trust evaluation method, it is characterised in that this method include with Lower step：
    Step 1: obtain each trust factor that may influence node trust value, i.e., the current directly letter that direct trust factor includes Appoint factor and the direct trust factor of history, current directly trust factor by the communication trust factor and sensing of current period trust because Element composition, and both monitored by being operated in the node of promiscuous mode the direct interaction process between other nodes to obtain Arrive；Indirect trust factor comes from the recommendation of other adjacent nodes；By direct trust factor and indirect trust factor composition synthesis Trust factor, using the final foundation that comprehensive trust value is credible as evaluation node；
    Step 2: each trust factor of the definition including direct trust factor, indirect trust factor and comprehensive trust factor Metric form；
    Step 3: realize that each trust calculates, including：
    Directly trust and calculate, to the communication trust-factor w used in calculating process1, sensing trust value factor w2, the time decay because Sub- f (t-t0), historical influence factor θ analyzed and defined；
    Trust indirectly and calculate, analyze and define recommendation trust sorting algorithm；Using the algorithm by the recommendation opinion of adjacent node It is divided into determination to recommend and do not know to recommend two classes, and respectively defines subjective assessment algorithm and irrelevance testing algorithm for trusting The weight distribution of calculating process；
    It is comprehensive trust calculate, the dynamic weights assignment algorithm based on interaction times is defined, by synthesizing direct trust value and indirectly Trust the credibility that the synthesis trust value being worth to carrys out evaluation node；
    The obtained synthesis trust value of destination node of calculating process and the magnitude relationship of trust threshold are trusted to sentence Step 4: comparing Whether disconnected destination node is credible；
    If the trust value of node is more than or equal to trust threshold, predicate node is believable, can continue to interact；If section Point trust value be less than trust threshold, then node is incredible, no longer interacts, and carry out deletion dependent routing information, Identify the work of the subsequent treatments such as insincere node.
  3. It is 3. as claimed in claim 1 a kind of based on multifactor wireless sensor network dynamic trust evaluation method, its feature It is, each trust defined including direct trust factor, indirect trust factor and comprehensive trust factor of the step 2 Factor, specifically include following steps：
    Define 1, current direct trust value CDTi,j(t) it is defined as：
    Wherein, RSCFi,j(t)、DSCFi,j(t) represent that host node i trusts on destination node j communication in current period t respectively Value and sensing trust value, w1, w2Communication trust-factor and the sensing trust value factor are represented respectively；
    Define 2, history direct trust value HDTi,j(t) it is defined as：
    HDTi,j(t)=f (t-t0)*CDTi,j(t0) (2)
    Wherein, t CDTi,j(t0) represent history cycle t0Direct trust values of the interior host node i on destination node j, time decay Factor f (t-t0) represent on current period t and history cycle t0Time interval linear function；
    Define 3, direct trust value DTi,j(t) it is defined as：
    DTi,j(t)=θ * HDTi,j(θ)+(1-θ)*CDTi,j(t) (3)
    Wherein, HDTi,j(t)、CDTi,j(t) represent that host node i directly believes on destination node j history in current period t respectively Appoint value and current direct trust value, θ represents the historical influence factor；
    Define 4, indirect trust values ITi,j(t) it is defined as：
    Wherein, DTi,k(t) direct trust values of the host node i on neighbor node k, host node i and neighbours in current period t are represented Node k interaction times are m, Tk,j(t) synthesis trust values of the destination node j on neighbor node k in current period t is represented For Tk,j(t), neighbor node k Combination of recommendation trust weight is w'k(t)；
    Define 5, comprehensive trust value Ti,j(t) it is defined as：
    Ti,j(t)=β ' * DTi,j(t)+(1-β′)*ITi,j(t) (5)
    DTi,j(t)、ITi,jAnd T (t)i,j(t) respectively represent cycle t in host node i on destination node j direct trust value, Connect trust value and comprehensive trust value；β ' expression direct trust values DTi,j(t) in comprehensive trust value Ti,j(t) power in building-up process Weight.
  4. It is 4. as claimed in claim 1 a kind of based on multifactor wireless sensor network dynamic trust evaluation method, its feature Be, the step 3 realize it is each trust calculate, wherein direct trust value calculate specific steps also include communication trust because Sub- w1With sensing trust-factor w2Value be as node interaction interior joint malicious act to caused by network coverage To determine；
    To time decay factor f (t-t0) analysis process be：Cycle t0History direct trust values of the interior nodes i on node j CDTi,j(t0) through becoming HDT after overdampingi,j(t), it is expressed as：
    Wherein, f (t-t0) time decay factor is represented, Th represents to distinguish credible and incredible trust threshold,
    Historical influence factor θ is defined as：
    Wherein 0 ＜ θ2The ＜ θ of ＜ 0.51＜ 1；θ1,θ2, what Th, f value was all determined by the application environment of reality.
  5. It is 5. as claimed in claim 1 a kind of based on multifactor wireless sensor network dynamic trust evaluation method, its feature It is, the subjective assessment algorithm of the step 3, the algorithm specifically includes following processing：
    Input is node i on neighbor node k1,k2,...,kmDirect trust value DTi,1(t),DTi,2(t),...,DTi,m (t)；
    Export as neighbor node k1,k2,...,kmWeight；
    Calculate node i is on all direct trust value sums for belonging to the neighbor node for determining to recommend class；
    Wherein, DTi,k(t) represent that input node is i, neighbor node kkDirect trust value；K represents the number of neighbor node；
    Calculate neighbor node k weight w 'k(t)；
    Wherein, DTi,k(t) represent that input node is i, neighbor node kkDirect trust value.
  6. It is 6. as claimed in claim 1 a kind of based on multifactor wireless sensor network dynamic trust evaluation method, its feature It is, the irrelevance testing algorithm of the step 3, the algorithm specifically includes following processing：
    Input as neighbor node km+1,km+2,...,knSynthesis trust value T on destination node jm+1,j(t), Tm+2,j(t) ..., Tn,j(t)；
    Export as neighbor node km+1,km+2,...,knWeight；
    Calculate the departure degree between the recommendation trust that the recommendation trust of neighbor node k offers and other nodes provide；
    Wherein, Tk,j(t) represent by formula (5) be calculated be in cycle t host node k on destination node j
    Comprehensive trust value；Tr,j(t) represent by formula (5) be calculated be in cycle t host node r on the comprehensive of destination node j Close trust value；N represents the number of neighbor node；
    Calculate total deviation value of uncertain trust recommendation class；
    Wherein, sk(t) represent to calculate recommendation trust and the recommendation of other nodes offer that neighbor node k is provided by formula (13) Departure degree between trust value；
    Calculate the deviation value s of total deviation value and neighbor node kk(t) proportionate relationship；
    dk(t)=S (t)/sk(t) (15)
    Wherein, sk(t) represent to calculate recommendation trust and the recommendation of other nodes offer that neighbor node k is provided by formula (13) Departure degree between trust value；S (t) represents the total deviation value for the uncertain trust recommendation class being calculated by formula (14)；
    To dk(t) it is normalized, produces node k weight w 'k(t)；
    Wherein, dk(t) the total deviation value being calculated by formula (15).
  7. It is 7. as claimed in claim 1 a kind of based on multifactor wireless sensor network dynamic trust evaluation method, its feature It is, the dynamic weights assignment algorithm of the step 3, the algorithm specifically includes following processing：
    Input as the acceptable error range e of current environment, interaction times m；
    Export the synthetic weight β for direct trust value；
    Weight beta, error range e and interaction times m meet equation；
    2exp(-2t2M)=1- β (17)
    Wherein, weight beta represents the synthetic weight of direct trust value；M represents interaction times；
    Convert equation；
    β=1-2exp (- 2t2m) (18)
    Wherein, weight beta represents the synthetic weight of direct trust value；M represents interaction times；
    Bounded processing is carried out to β；
    Wherein, weight span of the direct trust value during comprehensive trust combination is determined by actual application environment, it is assumed that Span is [c, d], wherein 0 ＜ c ＜ d ＜ 1.
CN201710380752.9A 2017-05-25 2017-05-25 Based on multifactor wireless sensor network dynamic trust evaluation system and method Pending CN107750053A (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN201710380752.9A CN107750053A (en) 2017-05-25 2017-05-25 Based on multifactor wireless sensor network dynamic trust evaluation system and method

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN201710380752.9A CN107750053A (en) 2017-05-25 2017-05-25 Based on multifactor wireless sensor network dynamic trust evaluation system and method

Publications (1)

Publication Number Publication Date
CN107750053A true CN107750053A (en) 2018-03-02

Family

ID=61254708

Family Applications (1)

Application Number Title Priority Date Filing Date
CN201710380752.9A Pending CN107750053A (en) 2017-05-25 2017-05-25 Based on multifactor wireless sensor network dynamic trust evaluation system and method

Country Status (1)

Country Link
CN (1) CN107750053A (en)

Cited By (15)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN109246155A (en) * 2018-12-07 2019-01-18 重庆邮电大学 A method of attack is trusted in the wireless sensor network defence based on trust management
CN109257750A (en) * 2018-11-08 2019-01-22 江南大学 The intrusion detection method of multi-protocol layer based on trust and noise spot detection technique
CN110572796A (en) * 2019-09-19 2019-12-13 广东工业大学 Failure node detection method, device and related equipment
CN111510883A (en) * 2020-04-26 2020-08-07 苏州鸿链信息科技有限公司 Internet of vehicles oriented layered trust model and trust value calculation method thereof
CN111654891A (en) * 2020-05-06 2020-09-11 长春工业大学 Wireless sensor network security routing protocol based on trust value
CN112165694A (en) * 2020-11-02 2021-01-01 上海第二工业大学 Method for establishing trust model of wireless sensor network
CN112437407A (en) * 2020-09-18 2021-03-02 中国电力科学研究院有限公司 Method and system for calculating initialized safety state of power internet of things wireless terminal equipment based on mixed information entropy
CN112733170A (en) * 2021-01-14 2021-04-30 中南大学 Active trust evaluation method based on evidence sequence extraction
CN112751867A (en) * 2020-12-31 2021-05-04 南京航空航天大学 Access control authorization method based on logic unit and trust evaluation
CN113380024A (en) * 2021-05-27 2021-09-10 重庆邮电大学 Reputation updating method and trust calculation method based on Internet of vehicles
CN113672932A (en) * 2021-07-21 2021-11-19 国网天津市电力公司 Electric power Internet of things intelligent terminal trusted computing trust value obtaining method based on self-adaptive entropy value weight
CN113824732A (en) * 2021-10-13 2021-12-21 成都安恒信息技术有限公司 Zero trust-based multi-factor authentication method
CN114374969A (en) * 2021-12-16 2022-04-19 国网河北省电力有限公司信息通信分公司 Trust evaluation method in power Internet of things environment
CN114553458A (en) * 2021-12-16 2022-05-27 国网河北省电力有限公司信息通信分公司 Method for establishing and dynamically maintaining credible group in power Internet of things environment
CN116055216A (en) * 2023-03-06 2023-05-02 睿至科技集团有限公司 Security detection method and system based on Internet of things

Citations (7)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
EP2096513A1 (en) * 2008-02-28 2009-09-02 Sap Ag Trustworthiness assessment of sensor data from wireless sensor networks to business applications
US20090328148A1 (en) * 2008-06-30 2009-12-31 Industry-Academic Cooperation Foundation Of Kyung Hee University Method of trust management in wireless sensor networks
CN103139803A (en) * 2013-02-07 2013-06-05 南京邮电大学 Multifactor-based wireless sensor network trust management model
CN104320823A (en) * 2014-10-24 2015-01-28 西安电子科技大学 Wireless sensor network clustering method based on Sink trust value evaluation
CN104837130A (en) * 2015-04-22 2015-08-12 上海物联网有限公司 Wireless sensor network compromised node identification method under Beta distribution with time-varying credit value
CN105873065A (en) * 2016-03-28 2016-08-17 南京邮电大学 Safe positioning method of wireless sensor network based on trust level evaluation
CN106488393A (en) * 2016-09-30 2017-03-08 天津大学 Cluster wireless sensor network election of cluster head model based on evolutionary Game mechanism

Patent Citations (7)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
EP2096513A1 (en) * 2008-02-28 2009-09-02 Sap Ag Trustworthiness assessment of sensor data from wireless sensor networks to business applications
US20090328148A1 (en) * 2008-06-30 2009-12-31 Industry-Academic Cooperation Foundation Of Kyung Hee University Method of trust management in wireless sensor networks
CN103139803A (en) * 2013-02-07 2013-06-05 南京邮电大学 Multifactor-based wireless sensor network trust management model
CN104320823A (en) * 2014-10-24 2015-01-28 西安电子科技大学 Wireless sensor network clustering method based on Sink trust value evaluation
CN104837130A (en) * 2015-04-22 2015-08-12 上海物联网有限公司 Wireless sensor network compromised node identification method under Beta distribution with time-varying credit value
CN105873065A (en) * 2016-03-28 2016-08-17 南京邮电大学 Safe positioning method of wireless sensor network based on trust level evaluation
CN106488393A (en) * 2016-09-30 2017-03-08 天津大学 Cluster wireless sensor network election of cluster head model based on evolutionary Game mechanism

Non-Patent Citations (2)

* Cited by examiner, † Cited by third party
Title
JIAOJIAO SONG 等: "《Dynamic Trust Evaluation of Wireless Sensor Networks Based on Multi-factor》", 《2015 IEEE TRUSTCOM/BIGDATASE/ISPA》 *
王伟龙 等: "《基于信任机制的一种无线传感器网络簇头选举算法》", 《计算机应用》 *

Cited By (21)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN109257750A (en) * 2018-11-08 2019-01-22 江南大学 The intrusion detection method of multi-protocol layer based on trust and noise spot detection technique
WO2020093907A1 (en) * 2018-11-08 2020-05-14 江南大学 Trust and noise point detection technology-based intrusion detection method for multi-protocol layer
CN109246155A (en) * 2018-12-07 2019-01-18 重庆邮电大学 A method of attack is trusted in the wireless sensor network defence based on trust management
CN110572796A (en) * 2019-09-19 2019-12-13 广东工业大学 Failure node detection method, device and related equipment
CN110572796B (en) * 2019-09-19 2022-04-19 广东工业大学 Failure node detection method, device and related equipment
CN111510883A (en) * 2020-04-26 2020-08-07 苏州鸿链信息科技有限公司 Internet of vehicles oriented layered trust model and trust value calculation method thereof
CN111654891A (en) * 2020-05-06 2020-09-11 长春工业大学 Wireless sensor network security routing protocol based on trust value
CN111654891B (en) * 2020-05-06 2023-02-28 长春工业大学 Wireless sensor network secure routing method based on self-adaptive trust mechanism
CN112437407A (en) * 2020-09-18 2021-03-02 中国电力科学研究院有限公司 Method and system for calculating initialized safety state of power internet of things wireless terminal equipment based on mixed information entropy
CN112165694A (en) * 2020-11-02 2021-01-01 上海第二工业大学 Method for establishing trust model of wireless sensor network
CN112751867A (en) * 2020-12-31 2021-05-04 南京航空航天大学 Access control authorization method based on logic unit and trust evaluation
CN112751867B (en) * 2020-12-31 2022-07-05 南京航空航天大学 Access control authorization method based on logic unit and trust evaluation
CN112733170B (en) * 2021-01-14 2024-01-30 中南大学 Active trust evaluation method based on evidence sequence extraction
CN112733170A (en) * 2021-01-14 2021-04-30 中南大学 Active trust evaluation method based on evidence sequence extraction
CN113380024A (en) * 2021-05-27 2021-09-10 重庆邮电大学 Reputation updating method and trust calculation method based on Internet of vehicles
CN113380024B (en) * 2021-05-27 2022-09-02 重庆邮电大学 Reputation updating method and trust calculation method based on Internet of vehicles
CN113672932A (en) * 2021-07-21 2021-11-19 国网天津市电力公司 Electric power Internet of things intelligent terminal trusted computing trust value obtaining method based on self-adaptive entropy value weight
CN113824732A (en) * 2021-10-13 2021-12-21 成都安恒信息技术有限公司 Zero trust-based multi-factor authentication method
CN114553458A (en) * 2021-12-16 2022-05-27 国网河北省电力有限公司信息通信分公司 Method for establishing and dynamically maintaining credible group in power Internet of things environment
CN114374969A (en) * 2021-12-16 2022-04-19 国网河北省电力有限公司信息通信分公司 Trust evaluation method in power Internet of things environment
CN116055216A (en) * 2023-03-06 2023-05-02 睿至科技集团有限公司 Security detection method and system based on Internet of things

Similar Documents

Publication Publication Date Title
CN107750053A (en) Based on multifactor wireless sensor network dynamic trust evaluation system and method
Kumar et al. A Distributed framework for detecting DDoS attacks in smart contract‐based Blockchain‐IoT Systems by leveraging Fog computing
Selvi et al. An energy aware trust based secure routing algorithm for effective communication in wireless sensor networks
Hashemi et al. Dynamic and comprehensive trust model for IoT and its integration into RPL
Caminha et al. A smart trust management method to detect on-off attacks in the internet of things
Zhou et al. Elastic switch migration for control plane load balancing in SDN
CN104618908B (en) The method and apparatus that distributed cognition wireless network is attacked anti-distort perception data
Pasikhani et al. Reinforcement-Learning-based IDS for 6LoWPAN
Raghavendra et al. An Intelligent RPL attack detection using machine learning-based intrusion detection system for Internet of Things
Sert et al. An efficient fuzzy path selection approach to mitigate selective forwarding attacks in wireless sensor networks
Gelenbe et al. IoT Network Cybersecurity Assessment with the Associated Random Neural Network
Huang A data-driven WSN security threat analysis model based on cognitive computing
Jithish et al. A game‐theoretic approach for ensuring trustworthiness in cyber‐physical systems with applications to multiloop UAV control
Shen et al. Prior knowledge based advanced persistent threats detection for IoT in a realistic benchmark
MohanaPriya et al. Restricted Boltzmann machine‐based cognitive protocol for secure routing in software defined wireless networks
Islam et al. Detection of dishonest trust recommendations in mobile ad hoc networks
Du et al. Multiattribute evaluation model based on the KSP algorithm for edge computing
Hang et al. Pervasive hybrid two-stage fusion model of intelligent wireless network security threat perception
Basan et al. Detection of anomalies in the robotic system based on the calculation of Kullback-Leibler divergence
Rani et al. Review of Trust-based Security Models forPacket Routing in Wireless Sensor Networks
Sharma et al. RRAR: Robust recommendation aggregation using retraining in Internet of Things
CN110290520B (en) Access control method of WSN node, related device and computer readable storage medium
SIMHADRI MADHURI A Trusted Node Feedback Based Clustering Model For Detection Of Malicious Nodes In The Network
Puli et al. An Efficient Swift Routing Model with Node Trust Identity Factor (SRM-NTIF) to Perform Secure Data Transmission Among IoT Gadgets
Sivanesan et al. Detecting Distributed Denial of Service (DDoS) in MANET Using Ad Hoc On-Demand Distance Vector (AODV) with Extra Tree Classifier (ETC)

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
WD01 Invention patent application deemed withdrawn after publication
WD01 Invention patent application deemed withdrawn after publication

Application publication date: 20180302