Fault Detection Technology Of Medical Area Network Based On Massive Machine Communication Of The Fifth Generation

In the Fifth generation,massive Machine Type Communication(mMTC)will be widely used in various vertical industries.The variety of MTC devices,the low complexity and high connection density of MTC can support industry applications that need to deploy a large number of devices,such as medical body area networks.However,the nodes in medical body area networks based on mMTC are generally faulted,because of the small size and limited resources of medical sensing equipment.The fault node not only block the data transmission in the network,but also produces the wrong measurement which deviates from the physiology ground truth values,that is fault data.This will cause extremely bad effects,even threaten the safety of patients which rely on data to judge the patient’s physiological condition.Therefore,the fault detection technology in medical body area network has attracted the attention of scholars in related fields.In medical body area network based on massive Machine Type Communication,this paper focuses on the fault data in medical body area network.This paper presents a cluster scheme of medical body area network based on distributed fault detection.It also proposes a fault detection scheme based on physiological data correlation for fault data prediction and fault data classification.The main contents of this paper are as follows:Firstly,a cluster scheme of medical body area network based on distributed fault detection is proposed,which is expected to be useful for the energy limitation problem of MTC devices.In this paper,the priority division of energy-sensitive wearable sensors and implantable sensors is carried out.In order to reduce the energy consumption of medical sensors,the relay forwarding mode is selected according to the complexity of the forwarding device.Secondly,the fault data detection scheme based on Gaussian regression process is proposed to solve the problem of fault data detection and physiological data missing.Using the correlation of physiological data,the corresponding physiological index values are predicted.According to the comparison between the prediction results and the dynamic threshold,the fault data are classified.The program takes into account the effect of the test results on medical diagnosis,restores the true value of physiological parameters as far as possible.This scheme makes a preliminary judgment on the patient’s physical condition based on physiological data.Simulation results show that the fault detection scheme can improve detection accuracy and reduce false alarm rate.Finally,as for the problems of complex fault data features and insufficient prior knowledge of fault data,the fault data detection scheme based on Gaussian mixture model is proposed.In the scheme,the Gaussian mixture model clustering algorithm is used.In order to improve the detection accuracy,the parameters of the clustering algorithm are adjusted to classify the fault data with different data characteristics.In addition,the probability values of different fault types are determined by data clustering.The vectors classified as fault and non-fault with similar probability values are annotated,and further judgment processing is done to improve the detection accuracy.Simulation results show that the appropriate clustering parameters can increase the probability of positive cases.And the accuracy of system detection is improved compared with the simple prediction algorithm.