CN111770499B - Distributed spectrum cooperation detection method - Google Patents
Distributed spectrum cooperation detection method Download PDFInfo
- Publication number
- CN111770499B CN111770499B CN202010586269.8A CN202010586269A CN111770499B CN 111770499 B CN111770499 B CN 111770499B CN 202010586269 A CN202010586269 A CN 202010586269A CN 111770499 B CN111770499 B CN 111770499B
- Authority
- CN
- China
- Prior art keywords
- node
- value
- nodes
- message
- values
- 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.)
- Active
Links
Images
Classifications
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04W—WIRELESS COMMUNICATION NETWORKS
- H04W12/00—Security arrangements; Authentication; Protecting privacy or anonymity
- H04W12/12—Detection or prevention of fraud
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04B—TRANSMISSION
- H04B17/00—Monitoring; Testing
- H04B17/30—Monitoring; Testing of propagation channels
- H04B17/309—Measuring or estimating channel quality parameters
Abstract
The invention belongs to the field of cooperative spectrum sensing in cognitive radio, and particularly relates to a distributed spectrum cooperative detection method. The invention provides a credit value-based idea to solve the problem that nodes are attacked by continuous SSDF (secure persistent distributed distribution) and only needs one-time perception, the historical propagation path of each message is taken as a basis for calculating the credit of each cognitive user, and when the nodes receive different messages from the same cognitive user, malicious users must exist on the propagation path, the credit values of the nodes which appear on different paths for more times are rewarded, and the credit values of the nodes which appear less times are punished. And finally, restoring the true value of the tampered message according to the calculated reputation value. Simulation analysis shows that the method has higher recovery rate to the message and identifies the malicious users in the network.
Description
Technical Field
The invention belongs to the field of cooperative spectrum sensing in cognitive radio, and particularly relates to a distributed spectrum cooperative detection method.
Background
In the past decade, the demand for wireless devices and applications has increased, the available licensed spectrum has not been utilized correctly, spectrum sensing technology has emerged, and by spectrum sensing technology, idle spectrum in a network can be detected, and cognitive users can improve spectrum utilization rate by dynamically accessing spectrum. Common single-user spectrum sensing has poor sensing performance and has the problems of multipath effect, shadow fading and the like, so that the cooperative spectrum sensing is provided on the basis. The cooperative spectrum sensing is divided into a centralized type and a distributed type. The centralized cooperative spectrum sensing needs to have a fusion center, all data needs to be sent to the fusion center for processing, and the fusion center performs judgment. Centralized spectrum sensing has many limitations, and decentralized distributed cooperative spectrum sensing has been proposed. The distributed cooperative spectrum sensing convergence is faster, and the judgment result is more reliable.
Distributed cooperative Spectrum Sensing based on average consensus is common, but the method is vulnerable to malicious nodes in the message transmission process, and malicious Secondary User nodes (SUs) change the Sensing value and forge false Sensing values, namely, a Spectrum Sensing Data Failure (SSDF). Therefore, recently, a scholars proposes a message transmission algorithm which continuously transmits messages in SU nodes without changing perception values, so that each SU node can finally obtain the perception values of h-hop neighbor nodes thereof, the perception values are taken as the approximation of the whole network perception values, and each SU node is taken as a fusion center to perform a centralized algorithm, thereby obtaining the performance similar to the centralized algorithm.
Although the message passing algorithm can effectively resist the SSDF attack, the situation that the content of the message can be tampered by a malicious SU node when the message is passed, namely the SSDF attack is continued, is not considered.
Disclosure of Invention
The invention aims to solve the problem that distributed cooperative spectrum sensing is easily attacked by malicious SU nodes and sensing values are tampered, and provides a distributed spectrum cooperative detection method. According to the invention, messages are mutually transmitted among the SU nodes, after h iterations, each SU node has the sensing result of its h-hop neighbor SU node, and finally, malicious nodes are identified through some centralized SSDF resisting algorithms, such as an abnormal value detection algorithm, and the like, so as to obtain a final sensing result. And (3) introducing a credit value into a message transmission algorithm, giving different credit values to each SU node on the message transmission path by comprehensively analyzing the message transmission path and the message transmission perception value, finally finding out a malicious SU node according to the credit value, and excluding the malicious SU node from consensus fusion. And the correct sensing value of the H-hop neighbor SU node is obtained, so that the reliability of the judgment result is greatly improved. Unless otherwise specified, all nodes in the present invention refer to SU nodes.
The technical scheme adopted by the invention is as follows:
a distributed spectrum cooperation detection method comprises the following steps:
s1, each node broadcasts the message M to the adjacent nodes until the message M is broadcasted to the h-hop neighbor nodes, namely all the messages M reach the h-hop neighbor nodes; the format of the message M is [ SRC, PRE, RD ], where SRC is an originator ID of the message, PRE includes SU node information passing through in a message propagation path, and RD is local raw data of SRC, that is, a locally detected perception value; when the node broadcasts the message, the ID of the path node is added to the PRE; each node has a local message library for storing messages from different nodes;
s2, calculating a reputation value of each node according to all messages stored in the local message library, specifically:
classifying all messages in the local message library according to different SRCs, and if N exists, classifying the different SRCsrVarious RD values, NjIf not, the step S21 is executed, otherwise, the step S22 is executed;
s21, updating the reputation value according to the occurrence times of different RD values of each node existing in the propagation path of the SRC class, and assuming that N appears in the SRC class by the node jjSub, Nr≥NjIf the reputation value of the node i is more than or equal to 1, the reputation value of the node i is updated as follows:
wherein R isijIf the credit value of the node i to the node j is positive, p is a reward and punishment factor, and the step S23 is entered;
s22, the SRC value is the ID of the node i itself, that is, the message is returned to the node through a loop, and for each node j in the loop, the reputation value is:
Rij＝Rij+pdetermin
wherein p isdeterminTo determine the reward factor, go to step S23;
s23, repeating the step S2 until each SU obtains the reputation values of the SU to all path nodes contained in all SRC classes according to a local message library, sequencing the obtained reputation values from high to low, and marking the nodes with the reputation values lower than a set threshold value as malicious nodes;
s3, outlier detection: the reputation value detection can only remove malicious nodes which have tampered with the perception values, but the malicious nodes which send the perception values with larger deviation cannot be removed, so that the outlier detection is used for removing the malicious nodes. After eliminating the malicious nodes obtained in the step S2, for the rest of nodes, each node obtains a group of energy values from h-hop neighbors by using Xi＝(xi0,xi1,xi2,…xir) The vector representing the received energy values, the set X being sorted in ascending orderiIs divided into XLHiAnd XUHiTwo moieties of which XLHiIs the set of energy values, X, of the lower halfUHiIs a set of energy values for the upper half, with the energy value for the lower half represented as (L)1<L2<L3…Lh) The energy value of the upper half is expressed as (U)1<U2<U3…Uh) Calculating the gaps between successive data points of the two parts respectively and from the set X respectivelyLHiAnd XUHiTo find the maximum gap PLHiAnd PUHiPosition P ofLHiAnd the node to its left is labeled low outlier, position PUHiAnd nodes on the right side thereof are marked as high outliers;
s4, after excluding the nodes marked as low outlier and high outlier, the node i calculates the estimation value Y of the main user perception value by averaging the remaining perception valuesiAnd comparing with a set threshold value if the estimated value Y isiIf the value is larger than the threshold value, a master user exists; if the estimated value Y isiAnd if the value is less than the threshold value, the master user does not exist. In general, sensing values of nodes are averaged and can be used as an estimation value of the sensing value to perform sensing judgment.
The invention has the beneficial effects that: the reputation-based non-consensus message transmission algorithm provided by the invention adopts the steps of adding a reputation value on the basis of the message transmission algorithm, giving different reputation values to each node on the message transmission path by comprehensively analyzing the message transmission path and the message transmission perception value, and finally finding out a malicious user according to the magnitude of the reputation value to obtain the correct perception value of the h-hop neighbor node. Under all conditions, the reputation-based message passing algorithm provided by the invention can effectively improve the capability of identifying malicious users under the attack of the malicious users. It is worth mentioning that the performance of the reputation value-based algorithm provided by the invention is not affected by the size of attack launched by a malicious user, and only depends on whether the malicious user launches the malicious attack or not.
Drawings
FIG. 1 is a general reward and punishment diagram;
FIG. 2 is a schematic illustration of a loop reward;
FIG. 3 is a flow chart of an outlier detection algorithm;
FIG. 4 is a malicious user identification performance graph;
fig. 5 is a graph of spectrum sensing performance versus time.
Detailed Description
The technical solution of the present invention is further described in detail below with reference to the accompanying drawings and simulation examples:
the method of the invention mainly comprises the following steps:
the first step is to generate and forward a message M. The message transfer process is slightly different from the original message transfer algorithm, and each SU broadcasts the message M to the adjacent SU until the h-hop neighbor node is broadcasted. The message M contains [ SRC, PRE, RD ], where SRC is the originator of the message, PRE is the propagation path SU node of the message delivery, and RD is the SRC's local raw data, i.e. locally detected perception value. And when the SU generates the message M, the ID is set to the SRC field, and the local data is set to the RD. When the SU broadcasts the message, the content of the PRE field of the message M is increased by using the ID passing through the SU node.
Each SU has a local message store for storing messages from different SUs. SUjFurther check message M [ SRC, RD]Whether a local buffer exists. If this field already exists, SUjWill put [ P ] in the new message MRE]Fields not in the message library [ SRC, RD]Adding the node ID in the node of the propagation path into the propagation path; if this field does not exist, SUjThe new message M will be stored in the message store. Furthermore, if the propagation path of the message [ PRE ]]If the node is in the state of the message, the message is stopped forwarding when the message passes through a loop, otherwise, the node adds the ID of the node to the history propagation path [ PRE ] of the message]Fields and forwards. In the process of cooperative spectrum sensing, each SU in the network broadcasts a message M. After h times of forwarding, all messages M reach the h-hop neighbor SU node.
The second step is to calculate a reputation value for each node from the message. Is the core of reputation based non-consensus messaging algorithms. Different from the traditional trust-based consensus algorithm, each SU node of the algorithm not only calculates the reputation value of the neighbor SU node, but also calculates the reputation values of all the SU nodes passing through the received messages, and the specific rule is as follows:
(1) the message library is classified according to different SRCs, and for different SRCs, N is assumed to existrAnd updating the reputation value according to the occurrence times of different RD values of each node existing in the SRC type propagation path. For example, assume node j has N present in this SRC classificationjThen, (N)r≥NjEqual to or greater than 1), the reputation value of the node i is updated as:
wherein R isijIs the reputation value of node i to node j. p is a reward and punishment factor. It can be seen that the reputation value update for a node is determined by the number of times the node appears in the propagation path for different RD values. If the occurrence frequency is more, the reward is given, and if the occurrence frequency is less, the punishment is given. This is because different RD values are due to the iRDA attack of a malicious node, and the more times a node appears in different RD propagation paths, the lower the probability that it will cause the RD values to be different, whereas the more times it appears in different RD propagation pathsLess, the greater the likelihood that it is a malicious node.
By way of example to illustrate the calculation of reputation values, fig. 1 is a general reward and punishment diagram, as shown in the figure, there are three paths from node i to node j, and in order to more clearly represent different paths, the same node appearing in different paths is duplicated and represented by the same reference numeral. It can be seen that through different paths, two messages with sending node i but different values RD will be finally received in the message library of node j, the sets of historical propagation nodes of the node are {1,2} and {1,3,4,5}, respectively, and it can be seen that since node 1 appears in the two sets of path nodes, the probability of RD value caused by it is low, according to equation (4-4), the node reputation value will be added by 0.25p, and since all the nodes appear only once in different sets of path nodes, the node reputation will be subtracted by 0.25p, i.e. penalty is made. It is noted that node 3, although appearing in two different paths, appears only once in different pairs of SRC, RD messages and therefore penalizes the node.
(2) In all SRC classes, one class is more specific, that is, the SRC value is the node i itself, and the message is returned to the node through a loop. The reputation value for the propagation path node with RD value equal to the original RD value can be increased substantially, i.e.:
Rij＝Rij+pdetermin j∈Ncorrect (2)
wherein p isdeterminFor a deterministic reward factor, NcorrectTo contain the correct [ SRC, RD]Field pair [ i, Pi]The set of path nodes.
There is a special case where nodes form a loop, as shown in fig. 2. Since node i receives the correct message through a loop, all nodes on the loop are rewarded significantly.
And after the two reward and punishment rules are carried out, finally, each node is sequenced according to the reputation value of the propagation path node, and the nodes with lower reputation values are the detected malicious nodes. Meanwhile, for SRC classes containing different RD values, the sum of the credit values of the propagation path node sets is calculated, the highest value is taken as the RD value of the SRC, and the RD value is a perception value and can be used for subsequent perception judgment.
And thirdly, detecting outliers, eliminating malicious user nodes, fusing perception values of the rest nodes, and comparing the perception values with a threshold value to obtain final judgment. After the message transmission process and the credit value calculation process are finished, each node i obtains the perception values of h-hop neighbor nodes excluding malicious users, and final judgment is made according to the perception values. Since the attack type is launched by forging the estimated energy value, maximum gap bi-directional outlier detection is used to better identify the malicious SU, and the specific algorithm flow is shown in fig. 3. And each node i sorts the received energy values in ascending order and judges the perception value of the outlier according to the difference of the adjacent energy values. Each SUiAn outlier detection algorithm is performed separately to identify outliers.
At the end of the messaging process, each SU gets a set of energy values from the h-hop neighbors, with Xi＝(xi0,xi1,xi2,…xir) And (4) showing. Then each SU pair set XiSorting into X in ascending orderLHiAnd XUHiTwo moieties of which XLHiIs the set of energy values, X, of the lower halfUHiIs the set of energy values of the upper half. Because of the pair XLHiSorted, so the energy value can be represented as (L)1<L2<L3…Lh). Similarly, the upper half can be represented as (U)1<U2<U3…Uh). Calculating the gap between successive data points of the two halves and separately from the set XLHiAnd XUHiTo find the maximum gap PLHiAnd PUHiThe position of (a). Corresponding to position PLHiAnd the nodes to its left are detected as low outliers. In the same way, the position PUHiNodes corresponding to those nodes to the right of it are detected as high outliers. TheThe method is based on the idea that the energy of malicious nodes differs considerably from the energy of honest nodes, which are close to each other. Therefore, when the energy levels are sorted (in ascending order) and the energy differences are calculated, the largest energy difference is observed at the position where the energy value of the malicious node appears. Thus, PLHiAnd all nodes to its left are low energy injection aggressors, PUHiAnd all nodes to its right are high energy injection attackers.
After the abnormal outlier is eliminated, the node i calculates an estimated value Y of the perception value of the main user by averaging the rest perception valuesiAnd compared with a set threshold value. If the estimated value Y isiIf the value is larger than the threshold value, a master user exists; if the estimated value Y isiAnd if the value is less than the threshold value, the master user does not exist.
The performance simulation and analysis are carried out on the matlab platform, the network topological structures of the simulation experiment are all randomly generated, and the network parameters are as follows: the number of the cognitive users is N, wherein the malicious users are randomly generated according to the set malicious user probability. All cognitive users are randomly distributed in a square area with the diameter of 1000m, and cognitive users with the distance less than 130m are neighbor nodes and can reliably communicate with each other. All cognitive users remain stationary during spectrum sensing. 5000m of main user from the cognitive user network, 60dB of transmitting power, 3 of channel loss parameter alpha, -10dB of receiving signal-to-noise ratio, 3dB of shadow fading parameter and relative distance d0Is 1000 m. The reward and punishment factor p takes the value of 1 and the deterministic reward factor pdeterminThe value is 10.
The simulation example simulates the spectrum sensing performance under the iRDA attack under different node numbers. The malicious user ratio is 0.2, and the message transfer hop count h is 3. The malicious node selects the attack launching type according to the local sensing result of the malicious node, namely when a malicious user detects that a master user exists, a low-energy injection attack is launched, and when no master user exists, a high-energy injection attack is launched. The attack constant takes on the value-10 dB. Meanwhile, an outlier detection algorithm and an original message passing algorithm in a centralized mode are simulated for comparison, and the simulation result is shown in fig. 5.
As can be seen from fig. 5, first, compared to the original message passing algorithm, the trust-based message passing algorithm provided by the present invention has significantly improved performance, and under the condition that the false alarm probability is 0.2, the detection probability of the network is improved by about 0.3. Meanwhile, the performance of the algorithm provided by the invention is close to that of an outlier detection algorithm under a centralized condition, when N is 10, the difference between the performance of the network under the false alarm probability 0.2 and the performance of the centralized algorithm is only smaller than 0.05, and when N is 20, the difference is increased to be close to 0.1, because when the number of network nodes is 20, the average diameter of the network is 3.478, which is slightly higher than h 3, and therefore, each node cannot obtain the information of all nodes of the whole network. Nevertheless, the perceptual performance of the proposed algorithm is still much higher than that of the original message passing algorithm, and the effectiveness of the proposed algorithm is proved.
Claims (1)
1. A distributed spectrum cooperation detection method is characterized by comprising the following steps:
s1, each node broadcasts the message M to the adjacent nodes until the message M is broadcasted to the h-hop neighbor nodes, namely all the messages M reach the h-hop neighbor nodes; the format of the message M is [ SRC, PRE, RD ], where SRC is an originator ID of the message, PRE includes SU node information passing through in a message propagation path, and RD is local raw data of SRC, that is, a locally detected perception value; when the node broadcasts the message, the ID of the path node is added to the PRE; each node has a local message library for storing messages from different nodes;
s2, calculating a reputation value of each node according to all messages stored in the local message library, specifically:
classifying all messages in the local message library according to different SRCs, and if N exists, classifying the different SRCsrVarious RD values, NjIf not, the step S21 is executed, otherwise, the step S22 is executed;
s21, updating the reputation value according to the occurrence times of different RD values of each node existing in the propagation path of the SRC class, and assuming that N appears in the SRC class by the node jjSub, Nr≥NjIf the reputation value of the node i is more than or equal to 1, the reputation value of the node i is updated as follows:
wherein R isijIf the credit value of the node i to the node j is positive, p is a reward and punishment factor, and the step S23 is entered;
s22, the SRC value is the ID of the node i itself, that is, the message is returned to the node through a loop, and for each node j in the loop, the reputation value is:
Rij＝Rij+pdetermin
wherein p isdeterminTo determine the reward factor, go to step S23;
s23, repeating the step S2 until each node acquires the reputation values of the corresponding path nodes in all SRC classes, sequencing the acquired reputation values from high to low, and marking the nodes with the reputation values lower than a set threshold value as malicious nodes;
s3, outlier detection: after eliminating the malicious nodes obtained in the step S2, for the rest of nodes, each node obtains a group of energy values from h-hop neighbors by using Xi＝(xi0,xi1,xi2,…xir) The vector representing the received energy values, the set X being sorted in ascending orderiIs divided into XLHiAnd XUHiTwo moieties of which XLHiIs the set of energy values, X, of the lower halfUHiIs a set of energy values for the upper half, with the energy value for the lower half represented as (L)1<L2<L3…Lh) The energy value of the upper half is expressed as (U)1<U2<U3…Uh) Calculating the gaps between successive data points of the two parts respectively and from the set X respectivelyLHiAnd XUHiTo find the maximum gap PLHiAnd PUHiPosition P ofLHiAnd the node to its left is labeled low outlier, position PUHiAnd nodes on the right side thereof are marked as high outliers;
s4, excluding the nodes marked as low outlier and high outlier, averaging the nodes i according to the remaining perceptual values to obtain an estimated value Y of the perceptual value of the main useriAnd comparing with a set threshold value if the estimated value Y isiIf the value is larger than the threshold value, a master user exists; if the estimated value Y isiAnd if the value is less than the threshold value, the master user does not exist.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202010586269.8A CN111770499B (en) | 2020-06-24 | 2020-06-24 | Distributed spectrum cooperation detection method |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202010586269.8A CN111770499B (en) | 2020-06-24 | 2020-06-24 | Distributed spectrum cooperation detection method |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN111770499A CN111770499A (en) | 2020-10-13 |
| CN111770499B true CN111770499B (en) | 2022-04-26 |
Family
ID=72722469
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202010586269.8A Active CN111770499B (en) | 2020-06-24 | 2020-06-24 | Distributed spectrum cooperation detection method |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN111770499B (en) |
Citations (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN104618908A (en) * | 2014-12-31 | 2015-05-13 | 重庆邮电大学 | Method and device for resisting spectrum sensing data falsification through distributed cognitive radio network |
| CN104618909A (en) * | 2015-01-08 | 2015-05-13 | 沈阳航空航天大学 | Credibility-based aeronautical radio cooperative spectrum sensing method |
| CN104993890A (en) * | 2015-07-14 | 2015-10-21 | 福建师范大学 | Security cooperative spectrum sensing method based on reputation mechanism and dynamic game |
| CN106850583A (en) * | 2017-01-06 | 2017-06-13 | 南京邮电大学 | A kind of distributed collaborative frequency spectrum perception attack defense method based on effectiveness |
| CN107623554A (en) * | 2017-08-24 | 2018-01-23 | 西安电子科技大学 | A kind of cooperative frequency spectrum sensing method based on trust value, cognition wireless network |
| CN109150623A (en) * | 2018-09-13 | 2019-01-04 | 重庆大学 | Malicious user SSDF attack method and system are resisted based on repeating query credit value |
| CN111193564A (en) * | 2019-12-25 | 2020-05-22 | 南京工业大学 | Broadband weighted cooperative spectrum sensing algorithm for resisting intelligent SSDF attack |
-
2020
- 2020-06-24 CN CN202010586269.8A patent/CN111770499B/en active Active
Patent Citations (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN104618908A (en) * | 2014-12-31 | 2015-05-13 | 重庆邮电大学 | Method and device for resisting spectrum sensing data falsification through distributed cognitive radio network |
| CN104618909A (en) * | 2015-01-08 | 2015-05-13 | 沈阳航空航天大学 | Credibility-based aeronautical radio cooperative spectrum sensing method |
| CN104993890A (en) * | 2015-07-14 | 2015-10-21 | 福建师范大学 | Security cooperative spectrum sensing method based on reputation mechanism and dynamic game |
| CN106850583A (en) * | 2017-01-06 | 2017-06-13 | 南京邮电大学 | A kind of distributed collaborative frequency spectrum perception attack defense method based on effectiveness |
| CN107623554A (en) * | 2017-08-24 | 2018-01-23 | 西安电子科技大学 | A kind of cooperative frequency spectrum sensing method based on trust value, cognition wireless network |
| CN109150623A (en) * | 2018-09-13 | 2019-01-04 | 重庆大学 | Malicious user SSDF attack method and system are resisted based on repeating query credit value |
| CN111193564A (en) * | 2019-12-25 | 2020-05-22 | 南京工业大学 | Broadband weighted cooperative spectrum sensing algorithm for resisting intelligent SSDF attack |
Non-Patent Citations (1)
| Title |
|---|
| 无线传感网中信任安全相关技术研究;尹娜;《中国优秀硕士学位论文全文数据库》;20190215;全文 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN111770499A (en) | 2020-10-13 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Kim et al. | A novel trust evaluation process for secure localization using a decentralized blockchain in wireless sensor networks | |
| Xing et al. | Real-time detection of clone attacks in wireless sensor networks | |
| CN106878995B (en) | Method for identifying abnormal type of wireless sensor network based on perception data | |
| Farzaneh et al. | An anomaly-based IDS for detecting attacks in RPL-based internet of things | |
| US8307459B2 (en) | Botnet early detection using hybrid hidden markov model algorithm | |
| Rmayti et al. | A stochastic approach for packet dropping attacks detection in mobile Ad hoc networks | |
| Banerjee et al. | Fault tolerant multiple event detection in a wireless sensor network | |
| Jamshidi et al. | A precise algorithm for detecting malicious sybil nodes in mobile wireless sensor networks | |
| Khan et al. | A temperature-aware trusted routing scheme for sensor networks: Security approach | |
| Shang et al. | Research on the intrusion detection model based on improved cumulative summation and evidence theory for wireless sensor network | |
| Rafeh et al. | Detecting sybil nodes in wireless sensor networks using two-hop messages | |
| CN106850583B (en) | Distributed cooperative spectrum sensing attack defense method based on utility | |
| CN110213770B (en) | Collusion attack prevention method and system for wireless sensor network | |
| Putrada et al. | Xgboost for ids on wsn cyber attacks with imbalanced data | |
| Gupta et al. | Improved weighted cooperative spectrum sensing algorithm based on reliability in cognitive radio networks | |
| CN111770499B (en) | Distributed spectrum cooperation detection method | |
| Najafi et al. | Decentralized prediction and reputation approach in vehicular networks | |
| Liu et al. | Attacker-exclusion scheme for cooperative spectrum sensing against SSDF attacks based on accumulated suspicious level | |
| Poolsappasit et al. | A secure data aggregation based trust management approach for dealing with untrustworthy motes in sensor network | |
| Bhargava et al. | DEIT: Dempster Shafer Theory‐based edge‐centric Internet of Things‐specific trust model | |
| Nath et al. | Mitigating SSDF attack using K-Medoids clustering in cognitive radio networks | |
| Sandhu et al. | Enhancing dependability of wireless sensor network under flooding attack: a machine learning perspective | |
| Zhang et al. | Dynamic double threshold spectrum sensing algorithm based on block chain | |
| Chaitanya et al. | Defense against PUEA and SSDF attacks in cognitive radio networks | |
| Huo et al. | Three-layer Bayesian model based spectrum sensing to detect malicious attacks in cognitive radio networks |
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 | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |