1 |
Efficient Streaming of Medical Sensor DataWang, Zhaoyu January 2012 (has links)
Telemedicine extends medical services to patients in remote areas. Due to the trend of population ageing, telemedicine becomes more and more popular. The Internet provides great opportunity for transmitting medical data. However the Internet is unreliable, some uncontrollable network behaviors (such as packet loss, packet delay) influence the quality of medical services greatly. Additionally,most existing medical data formats are designed for storing medical records not for streaming medical data over an unreliable medium. In order to promote the efficiency and achieve the error control of medical sensor data streaming service, this project aims to 1) investigate different streaming protocols and encoding strategies;2) empirically evaluate these encoding strategies and find an optimum choice.In this project, we focus on a home-based electrocardiograph(ECG) sensor monitoring service which requires little overhead, low latency and constant data rate. We develop a framework for testing the efficiency of medical streaming service, and empirically evaluate four forward error correction (FEC) encoding strategies with different erasure codes (XORcod,RS code) and block interleavers. The performances of the four encoding strategies are measured by calculating the Repair Rate &Peak Signal Noise Ratio (PSNR), and the processing times of these encoders are measured as well. Network simulation is used to establish the unreliable network and Network Simulator3 (NS3) is employed as the simulation tool. From the simulation results,we come up the conclusions: Firstly, the average burst error length of a network is the key factor which influences the performance comparison of the four encoding strategies.Secondly, interleaving provides positive impact when packet losses are bursty; while it results in negative impact when packet losses are scattered.Based on the conclusions above, FEC-XOR-Interleaving encoding strategy is the optimum choice for the home-based ECG sensor monitoring scenario,
|
2 |
Détection d'attaques dans un système WBAN de surveillance médicale à distance / Attacks detection in a WBAN system for remote medical monitoringMakke, Ali 30 May 2014 (has links)
L'un des défis majeurs du monde de ces dernières décennies a été l'augmentation continue de la population des personnes âgées dans les pays développés. D’où la nécessité de fournir des soins de qualité à une population en croissance rapide, tout en réduisant les coûts des soins de santé. Dans ce contexte, de nombreux travaux de recherche portent sur l’utilisation des réseaux de capteurs sans fil dans les systèmes WBAN (Wireless Body Area Network), pour faciliter et améliorer la qualité du soin et de surveillance médicale à distance. Ces réseaux WBAN soulèvent de nouveaux défis technologiques en termes de sécurité et de protection contre les anomalies et les attaques. Le mode de communication sans fil utilisé entre ces capteurs et l’unité de traitement accentue ces vulnérabilités. En effet les vulnérabilités dans un système WBAN se décomposent en deux parties principales. La première partie se compose des attaques possibles sur le réseau des capteurs médicaux et sur le médium de communications sans fils entre ces capteurs et l’unité de traitement. La deuxième partie se compose des attaques possibles sur les communications à haut débit entre le système WBAN et le serveur médical. L’objectif de cette thèse est de répondre en partie aux problèmes de détection des attaques dans un système WBAN de surveillance médicale à distance. Pour atteindre cet objectif, nous avons proposé un algorithme pour détecter les attaques de brouillage radio (jamming attack) qui visent le médium de communications sans fils entre les capteurs et l’unité de traitement. Ainsi nous avons proposé une méthode de mesure de divergence pour détecter les attaques de type flooding qui visent les communications à haut débit entre le système WBAN et le serveur médical. / One of the major challenges of the world in recent decades is the continued increase in the elderly population in developed countries. Hence the need to provide quality care to a rapidly growing population while reducing the costs of health care is becoming a strategic challenge. In this context, many researches focus on the use of wireless sensor networks in WBAN (Wireless Body Area Network) systems to facilitate and improve the quality of medical care and remote monitoring. These WBAN systems pose new technological challenges in terms of security and protection against faults and attacks. The wireless communication mode used between the sensors and the collection node accentuates these vulnerabilities. Indeed vulnerabilities in a WBAN system are divided into two main parts. The first part consists of the possible attacks on the network of medical sensors and on the wireless communications medium between the sensors and the processing unit. The second part consists of possible attacks on high-speed communications between the WBAN system and the medical server. The objective of this thesis is to meet some of the problems of detecting attacks in a WBAN system for remote medical monitoring. To achieve this goal, we propose an algorithm to detect the jamming attacks targeting the wireless communications medium between the sensors and the processing unit. In addition we propose a method of measuring divergence to detect the flooding attacks targeting the high-speed communications between the WBAN system and the medical server.
|
Page generated in 0.056 seconds