Detection Study Of Node Capture In Wireless Sensor Networks

The capture of a node is the first stage of an internal attack in a wireless sensor network(WSN),and a suitable detection method can better protect the security of the WSN network.In the detection method proposed in this paper,the decision-making node obtains the decision through the exchange of information between neighboring nodes,and sends the obtained decision to the base station by a multi-hop route,and the base station isolates the nodes that are deemed to have been captured.For the sleep mechanism,many detection methods use synchronous sleep instead of asynchronous sleep,and even some methods do not consider sleep.For the communication structure problem,only a few of the existing detection methods design simple communication structures.Aiming at the problem of low detection rate of captured nodes and high false alarm rate and high false alarm rate,this paper proposes two detection schemes for the large difference in node density in the target area based on the asynchronous sleep mechanism.The plan is as follows:1.Aiming at the application scenario with high node density in the target area,a multi-polling-based node capture detection scheme is proposed.This scheme builds a three-level system composed of three types of nodes: bottom node,monitoring node,and decision node,and constitutes a spider web structure.The simulation experiment shows that the cobweb structure combined with the multi-polling communication mechanism is used to achieve two collaborative decision-making,and the false alarm rate of this method is reduced to 6.4%,and the false alarm rate is reduced to4.2% through the asynchronous sleep mechanism,and the final detection rate is As high as 94.6%,the optimal communication model was finally confirmed through comparative experiments.2.Aiming at the application scenario with low node density in the target area,a WSN early detection scheme based on chain communication structure and sequential probability ratio test is proposed.This method uses a simplified communication structure into a chain communication structure to be suitable for application scenarios with low node density.In order to ensure that the method has a small number of nodes,it can also ensure that the detection rate will not drop due to the decrease in the number of uncaptured nodes.The sequential probability ratio resolves the sample size contradiction.Simulation experiments show that this detection method can achieve a detection rate of 94.19% even when the sample size is small,and the values of ? and ? can be controlled at 7.54% and 5.37%,respectively,which is also suitable for small node density scenarios.The other detection methods are better.