Research On Secure Communication Protocol In Body Area Network

With the continuous development of wireless communication technology in the medical field,various wearable wireless medical monitoring devices have emerged one after another,and wireless body area networks have been implemented and applied in the medical field.Use of intelligent,miniaturized,low-power physiological sensors placed on different parts of the human body to automatically collect various vital signs such as human heart rate,temperature,blood pressure,and electrocardiogram can monitor the status of the human body in real time.Then the data is transferred it to the remote medical institutions to obtain appropriate medical services.This scientific and effective way of monitoring vital signs not only can provide a simple and low-cost method for the elderly health care,but also can effectively prevents patients with chronic diseases from missing the best treatment period.This can also solve the difficult problem of medical treatment in remote areas due to limited availability of medical resources.Although the application of wireless body area network can bring great convenience to people,the user's health data is transmitted through an open wireless environment and lacks physical protection,there are various security risks in the wireless body area network system.We designed a channel feature-based authentication and key extraction scheme to ensure the secure transmission of data in the entire body area network.The channel features are random,fast time-varying and short-time reciprocity,which can not only solve the complex key distribution and management problems in traditional cryptography,avoid the risk of "four handshake",but also reduce the requirements complexity of software in the process of authentication and encryption.However,in practical applications,some problemsneed to be solved in secure communication based on channel features including incomplete symmetry of channel features extracted by both sides of communication,effective improvement of authentication rate and accuracy,further improvement of key strength,and reduction of energy consumption in information coordination and so on.The main contributions of this dissertation are as follows:The whole secure communication process in the body area network includes communication between the control unit and the sensors,and communication between the sensors on the body.First of all,in order to prevent illegal nodes from disguising themselves as legitimate nodes and joining the body area network,we propose that extract the CSI fingerprint features of both sides of the communication,and train RNN model through recurrent neural network to identify the nodes.The CSI fingerprint feature can be used to characterize more fine-grained and rich channel information.It can be used as identity authentication to improve the accuracy of authentication and reduce the false positive rate and false negative rate.At the same time,a recurrent neural network is used to train the fingerprint feature model can increase the rate of authentication.Identity authentication is the key link to realize secure communication in the body area network,and the key extraction and distribution mechanism is the basis for the secure transmission of data in the body area network.To solve the problem of secure data transmission between sensor nodes and control units in the body area network,we introduce the third-party trusted organization Victor to extract the channel characteristics of both sides of communication,which is modified by filtering and quantization as the communication key between sensor nodes and control units,and then distributed.The experimental results show that in the normal working environment,the intervention of the trusted third party Victor can effectively prevent collusion attacks,and also solve the problem of limited energy of legitimate nodes in the body area network.The above work solves the problem of secure communication between the central sensor node and the control unit.However,due to the short transmissiondistance between sensors on the body,it is not easy to introduce a third-party trusted organization,and it is also difficult to find a balance between authentication rate and key strength.The security mechanism proposed above is not suitable for the communication between sensors on the body.Aiming at the authentication and key extraction between sensors on the body,we present an authentication and key agreement protocol based on RSS channel features.It uses the RSS fingerprint signal between communication sensors to start authentication and combines the RSS fingerprint feature with bilinear to complete key negotiation,which is flexible and efficient.In order to facilitate the implementation,the protocol adopts RSS fingerprint feature,which breaks the traditional limitation of relying on physical distance for authentication,and transfers the cost of key negotiation between sensors on the body to the control unit.Under the random oracle model,the security of the protocol is proved based on the DBDH assumption.The implementation performance analysis results show that the protocol is safe and feasible.