A multiagent based framework secured with layered SVM-based IDS for remote healthcare systems
This work addresses security concerns for remote healthcare systems, which are critical for growing elderly and patient populations, but it appears incremental as it builds on existing multiagent and SVM-based methods.
The authors tackled the vulnerability of remote healthcare systems to common network attacks like DoS and U2R by proposing a secure framework that combines multiagent technology for data collection with a layered SVM-based intrusion detection system. They implemented and evaluated the system against attacks such as Smurf and Neptune, reporting efficiency and correctness in the results.
Since the number of elderly and patients who are in hospitals and healthcare centers are growing, providing efficient remote healthcare services seems very important. Currently, most such systems benefit from the distribution and autonomy features of multiagent systems and the structure of wireless sensor networks. On the one hand, securing the data of remote healthcare systems is one of the most significant concerns; particularly recent types of research about the security of remote healthcare systems keep them secure from eavesdropping and data modification. On the other hand, existing remote healthcare systems are still vulnerable against other common attacks of healthcare networks such as Denial of Service (DoS) and User to Root (U2R) attacks, because they are managed remotely and based on the Internet. Therefore, in this paper, we propose a secure framework for remote healthcare systems that consists of two phases. First, we design a healthcare system base on multiagent technology to collect data from a sensor network. Then, in the second phase, a layered architecture of intrusion detection systems that uses Support Vector Machine to learn the behavior of network traffic is applied. Based on our framework, we implement a secure remote healthcare system and evaluate this system against the frequent attacks of healthcare networks such as Smurf, Buffer overflow, Neptune, and Pod attacks. In the end, evaluation parameters of the layered architecture of intrusion detection systems prove the efficiency and correctness of our proposed framework.