37 results about "Distributed intrusion detection" patented technology

The invention relates to a distributed intrusion detection method and system for an ad hoc network. Nodes in the ad hoc network are clustered and cluster head nodes are selected, block chain systems are configured at all the cluster head node to form a block chain network, and the method includes the following steps: a node where a log is updated sends to the cluster head node thereof a transaction request containing a log information record which is formatted locally at the node; the cluster head node broadcasts the transaction in the block chain network after verifying that the transaction is legal; each cluster head node performs consensus verification of the transaction, after passing the consensus verification, the transaction is packed to be recorded to the block chain to form a shared log data; and each cluster head node executes intrusion detection based on local information and the shared log data. Since consensus verification and recording are performed on the log informationrecord, authenticity and reliability of detection data are ensured, globally reliable shared information records and safety audit data are provided for the centerless weakly-trusted ad hoc network, so as to support global decision making and response, thereby enabling intrusion detection to be more accurate.

The invention discloses a method for detecting distributed intruding, which is used in a wireless sensor network. A frequent item set detecting method is used in a node of a sensor for detecting abnormality degree of discrete web log data and a method of joint probability distribution calculation is used for detecting the abnormality degree of successional web log data; if the node of the sensor judges that the whole data is abnormal intruding data, the node sends the suspicion data to the neighboring node of the sensor for verification; if the neighboring node also judges that the suspicion data is the intruding data, then the node of the sensor marks the data as the intruding data and starts to detect the next data. The method combines the distributed technology and the intruding detecting technology and reduces the calculation quantity and communication quantity of each node by the cooperation among nodes of the sensor network. An intruding movement is distinguished by analyzing the local web log data and reported to a system as an alarm, so as to realize the security purpose of protecting the wireless sensor network from various new and old attacks.

The invention discloses a distributed optical fiber fence intrusion detection and location method based on a phi-OTDR (Optical Time Domain Reflectometer), and the method comprises the following steps of: firstly, subtracting an acquired ith signal by an acquired (i-1)th signal to obtain a preprocessing monitoring signal; secondly, carrying out at least four-layer wavelet decomposition on the preprocessing monitoring signal by adopting a plurality of wavelet functions; thirdly, judging whether actual intrusion exists according to the consistency of distribution positions of maximum-wavelet mode maximums in all scale signals; fourthly, judging that the actual intrusion exists if the distribution positions of the maximum-wavelet mode maximums in at least three scale signals are consistent; fifthly, corresponding multiplying the scale signals in which the intrusion exists and the distribution positions of the maximum-wavelet mode maximums are consistent to obtain a composite signal of which pseudo maximums are inhibited and mode maximums at a catastrophe point are enhanced, using a position corresponding to the maximum-mode maximum in the composite signal as a position of the determined intrusion point, determining the intrusion generation and the intrusion position by combining various composite signal judgment results of wavelet multi-scale analysis according to the majority voting criterion.

The invention provides a distributed vehicle-mounted intrusion detection system based on federated learning and trust evaluation and a method thereof. The method comprises the following steps: designing a distributed intrusion detection system model based on federated learning; building and pre-training of an intrusion detection model are realized based on federated learning, and broadcasting and distributing a global model by a distributed aggregator; performing edge model training based on intrusion detection data of the edge vehicle by the edge vehicle; selecting an edge representative node as a cluster head based on behavior evaluation to complete an aggregation task of an edge model; adding a mask to the model parameters and then uploading the model parameters to the RSU; performing trust evaluation on the RSU through the quality of a model aggregated by the RSU through the RSU aggregates model parameters uploaded by a cluster head, so that distributed aggregation nodes are stimulated to compete and aggregate a model with higher accuracy, and finally, storing the training model based on a block chain principle to complete model sharing. The invention further creates a defense detection system suitable for automatic driving to resist network intrusion by adopting a safety evaluation mode of information safety.

In large-scale high speed network, the invasion detecting system adopts always hierarchical distributive structure including dispersive acquisition, distributive processing and concentrated management. In the hierarchical modular structure, deputy technology is always adopted and the coordinating module in the coordinating management layer assigns and controls the deputies. In abnormal case that the static deputy in the lower layer can not settle, some request is sent to the coordinating module, and the coordinating module assigns movable deputy with specific function to the static deputy for complete the task before the cancel request is sent and the cancel command is sent separately to finish the assignment mechanism. The mechanism can balance the system load, respond timely, complete specific data processing function and make the movable deputy controllable.

The invention relates to the field of network communication, in particular to a distributed intrusion detection method, a distributed intrusion detection device and a distributed intrusion detection system. The method comprises the following steps of: acquiring own flow information and generating a flow table; calculating early warning information according to the flow table; when the flow table only comprises own flow information or when the early warning information meets neighbor communication conditions, acquiring the flow information of a neighbor node and flow information acquisition time, and updating the flow table and the early warning information; when the early warning information meets global communication conditions, acquiring the flow information of a global node and the flow information acquisition time, and updating the flow table and the early warning information; and when the early warning information meets early warning conditions, performing intrusion early-warning. In the technical scheme provided by the embodiment of the invention, only the flow information of the neighbor node needs acquiring when the early warning information meets the neighbor communication conditions, so the data traffic in distributed intrusion detection is reduced and the influence of the distributed intrusion detection on network performance is reduced.

The invention discloses an element identifier distribution method and an application layer routing method based on element identifier, which is applied in centralized-management multilayer distributedsystems. In the method, node elements and element identifier of each level of a distributed system are established layer by layer; a routing table item based on the element identifier is generated inthe process of generating and distributing the element identifier; and addressing is performed through the element identifier during communication so as to realize application layer routing based onthe element identifier, which is needed by the communication of the centralized-management distributed system. The invention also provides a plurality of default routing rules of the routing method. The application layer routing method which is generated in the invention and based on element identifier is unrelated to TCP/IP (Transmission Control Protocol/Internet Protocol) routing methods, has the characteristics of simplicity, practicability, stability, reliability and resource conservation, and can meet special needs of distributed intrusion detection systems, distributed network auditing systems and various centralized-management multilayer distributed systems.

The invention discloses a method for balancing a low load and a packet loss rate in DIDS based on reinforcement learning. The method comprises the following steps: calculating the immediate detectionrate, calculating the theoretical packet loss rats, calculating an average data packet arrival number; calculating the average waiting number of the data packets; calculating the packet latency probability, calculating the number of tasks allocated to the detection engine, calculating the probability that the detection engine is allocated to the tasks, calculating the working efficiency of the detection engine, calculating the overall working efficiency of the distributed intrusion detection system, and adjusting the balance of two contradictory indexes of low load and packet loss rate by detecting the probability that the detection engine is allocated to the detection tasks. The problem that in a task scheduling algorithm of a distributed intrusion detection system with a low load as a target disclosed in the prior art, the packet loss rate is possibly increased due to the fact that the low load is purely emphasized is solved.

The invention discloses a data distribution method based on node data processing capability and node operation load. The data distribution method comprises the following steps: step 1, measuring the processing capability of cluster data processing nodes; step 2, measuring the load of the cluster data processing nodes; step 3, calculating the load capacity weight of the cluster data processing nodes; and step 4, segmenting and distributing data. Under a high-speed distributed network environment, the parallel detection and processing efficiency of mass data is improved, the system detection time is shortened, the response efficiency of an intrusion detection system is improved, and the distributed intrusion detection capability is improved. According to the method, data segmentation and distribution can be matched with the working conditions of the nodes in the cluster in real time, and good load balance of the system is guaranteed. The parallel processing overhead of a distributed intrusion detection system is effectively reduced, the response efficiency of the intrusion detection system is improved, and the distributed intrusion detection capability is improved.

The invention discloses a distributed intrusion detection system and a connecting method of centralized management in the same, improving the security of a detection engine in the distributed intrusion detection system. The method mainly comprises the following steps of: starting a monitoring opening by a management control center; detecting the management control center by the detection engine; sending an identity authentication request to the management control center by the detection engine for identity authentication; and establishing communication connection between the detection engine and the management control center after the management control center is authenticated. Compared with the prior art, the technical scheme provided by the invention greatly lowers the communication management complexity of the distributed intrusion detection system, improves the security of the detection engine and reduces the workload of a system administrator.

The invention relates to a correlation-based hierarchical distributed intrusion detection system, which is technically characterized by comprising a plurality of intrusion detection systems; each intrusion detection system consists of a detection agent subsystem, a distributed cooperation subsystem and an IDXP communication protocol module; intrusion detection information interaction is carried out between the detection agent subsystem and the distributed cooperation subsystem and between the intrusion detection systems through the IDXP communication protocol modules; and the distributed cooperation subsystem is provided therein with a correlation-based distributed cooperation model. The system is reasonable in design, constructs a cooperation model based on the correlation between agents, uses an IDXP communication protocol in intrusion detection standardization, can effectively resist more and more complex attack means, effectively reduces the distributed cooperation cost of the system while guaranteeing the system detection accuracy, and has the characteristics of strong fault-tolerant capability, easy system expansion, wide detection range and the like.

The invention discloses a distributed intrusion detection method based on an ant colony algorithm. The distributed intrusion detection method comprises the following processes: starting, initializinga distributed sensor, setting parameters of the ant colony algorithm based on attribute decision making, constructing a decision tree structure definition rule, setting ants on sensor nodes, determining a next node of data fusion, partially updating the information factors of an ant selection path, trimming the decision tree, and updating all information factors of the ant selection path, if an optimal path is not discovered, returning to the step of partially updating the information factors of the ant selection path; and finding the optimal path, creating the optimal path, and fusing data information. Compared with the prior art, the distributed intrusion detection method has the advantages of improving the detection efficiency of data intrusion in a network system, reducing the energy consumption and effectively guaranteeing the security of the distributed network system.

The invention discloses an intrusion detection system based on brain-like computing. The intrusion detection system comprises management and control nodes, distributed detection nodes, an SDN switching network and a brain-like computing platform; the management and control node is used for managing and controlling the distributed intrusion detection node; the distributed detection nodes are used for detecting the distributed network flow and judging whether an intrusion behavior occurs in real time in combination with the support of the brain-like computing platform; the SDN switching network is used for copying and/or distributing network traffic in a distributed manner; and the brain-like computing platform is used for providing computing power support for the management and control node, the distributed detection node and the SDN switching network. According to the invention, intelligent detection strategy management and control are realized, intelligent detection flow management and control are realized, and rapid real-time detection is realized.

The invention provides a distributed intrusion detection method of a vehicle ad hoc network. Local detection is started by adopting a Naive Bayesian classification algorithm by vehicle nodes; when an abnormal condition is detected, data are marked to be abnormal and are stored into a local feature library; response measures are immediately adopted by a local response module; meanwhile, abnormal warning is sent out to adjacent nodes and abnormal features are transmitted to the adjacent nodes through an on-line response module, if the adjacent nodes are also abnormal, the abnormal features are transmitted to the local feature library, and response measures are immediately adopted by the local response module; and meanwhile, the abnormal warning is sent to other adjacent nodes and the abnormal features are transmitted to the other adjacent nodes except source nodes of abnormal warning through the on-line response module; and abnormal warning and abnormal feature transmission are continuously passed through the manner, so that distributed intrusion detection of the whole vehicle ad hoc network is realized. With the adoption of the distributed intrusion detection method of the vehicle ad hoc network, analysis and detection ability of the whole vehicle ad hoc network is improved; and the distributed intrusion detection method has the characteristic of being simple in system structure, low in complexity, strong in learning and high in intelligentization degree.

The invention provides a distributed intrusion detection method and system based on a simulated annealing fusion algorithm. Specifically, a first data packet of a current network behavior is acquiredin real time and initial detection is performed, whether a second data packet after initial detection is abnormal is judged based on a simulated annealing algorithm, and if the second data packet after initial detection is abnormal, the abnormal data packet is analyzed and network behavior detection result data based on different categories is generated. According to the invention, the final detection result data is obtained through dual detection and correlation analysis processing, the global optimization capability of simulated annealing is fully utilized, and the accuracy and execution efficiency of the intrusion detection model are improved.

The invention discloses a DIDS theoretical modeling method for detecting different engine processing capacities. The DIDS theoretical modeling method specifically comprises the following steps: step 1, performing model analysis and parameter setting on a distributed intrusion detection system; 2, performing state space analysis on the distributed intrusion detection system; and 3, calculating theaverage busy detection engine number in the distributed intrusion detection system, the average queue length of the data packets in the waiting queue, the average stay time of the data packets in thesystem and the average queuing waiting time of the data packets in the system. By adopting a method for establishing a model, necessary parameters such as the average busy detection engine number andthe average queue length in the waiting queue in the operation process of DIDS with different detection engine processing capacities can be mastered, and decision making and optimization on the basisare facilitated.

The invention discloses a distributed intrusion detection method based on a multi-layer extreme learning machine in an Internet of Things environment, and aims to move tasks with heavier calculation for realizing automatic attack detection to boundary equipment due to the characteristic that resources of related equipment are constrained, so as to enable a processing function to be close to a data source. The boundary devices can operate a preset classification model. However, when facing a large amount of training data, there is no sufficient storage and processing capabilities to construct and upgrade such models. In order to solve the problem, training operation with dense calculation and large storage capacity is moved to a cloud server to be carried out, and a single-hidden-layer extreme learning machine model and a multi-hidden-layer extreme learning machine model are constructed and trained in the cloud server, so that boundary equipment executes flow classification based on a deep learning model preset in the cloud server; therefore, whether the traffic is normal traffic or network attack is classified, and experimental analysis shows that the multi-hidden-layer extreme learning machine has better performance.

The invention discloses a distributed intrusion detection method based on alarm association in a cloud environment. The method comprises the following steps: collecting intrusion evidences; establishing a distributed behavior diagram template; carrying out anomaly detection; carrying out behavior graph template update. Distributed detection in the cloud environment is achieved in a clustering mode, further intrusion detection judgment is achieved based on the thought of alarm association, and the false alarm rate can be reduced. By means of the method, zombie machines in the cloud environmentcan be effectively detected, and therefore cloud computing resources are protected against malicious utilization.