Human Activity Recognition Using Software Defined Radios For Spinal Cord Patient Post-Surgery Monitoring By Machine Learning Algorithms

Intelligently and timely detection of abnormal activity detection and monitoring are become very essential research due to the human busy and fast life.Researchers are trying hard to ensure timely detection of various diseases at anytime and anywhere.Contactless health care monitoring is one of the uniqueness of the 5G network.The ratio of the number of health care problems and patients are increasing exponentially and burgeoning data.The number of growing applications needs ubiquitous intelligent edge computing(UIEC).This field involves researchers to find artificial intelligence and edge computing resources to health care problems and provide solutions in early diagnosis of vital signs.In post-surgery monitoring of the patient,timely consultation is essential before the further loss.Unfortunately,even after the advice of the doctor to the patient,he/she may forget to perform the activity in the proper way,which may lead to complications in recovery.Therefore,portable and multi-functional software-defined radio(SDR)platform is designed to detect different activities of human life,especially abnormal activities.In this research,the idea is to design a UIEC testbed for experimental data collection and classification of post-surgery activities.Universal software-defined radio peripheral(USRP)is utilized to collect the data of spinal cord operated patients of weight lifting activity.Wireless channel state information(WCSI)amplitude responses to the orthogonal frequency division multiplexing(OFDM).As a result,the data of the frequency domain is extracted and used in classification.The machine learning classification is based on either the proper or wrong way of weight lifting activity that was considered for experimental analysis.Based on our experiments,the accuracy achieved by the proposed testbed by using the fine K-nearest neighbor(FKNN)algorithm is 99.6%and Ensemble boosted trees(EBT)=98.9%and Fine decision tree(FDT)=97.4%.