No abstract

SUMMARY Intrusion detection system (IDS) has played a central role as an appliance tooliance toely defend our crucial computer systems or networks against attackers on the Internet. The most widely deployed and commercially available methods for intrusion detection employ signaturebased detection. However, they cannot detect unknown intrusions intrinsically which are not matched to the signatures, and their methods consume huge amounts of cost and time to acquire the signatures. In order to cope with the problems, many researchers have proposed various kinds of methods that are based on unsupervised learning techniques. Although they enable one to construct intrusion detection model with low cost and effort, and have capability to detect unforeseen attacks, they still have mainly two problems in intrusion detection: a low detection rate and a high false positive rate. In this paper, we present a new clustering method to improve the detection rate while maintaining a low false positive rate. We evaluated our method using KDD Cup 1999 data set. Evaluation results show that superiority of our approach to other existing algorithms reported in the literature. key words: intrusion detection system, clustering, detection rate, false positive rate IntroductionIn recent years, considerable attention has been given to intrusion detection on the Internet. Intrusion detection is defined as the process of monitoring the events occurring in a computer system or network and analyzing them for signs of intrusions [1]. Intrusion detection system (IDS) is one of the systems designed to perform such intrusion detection and an integral part of any complete security package of a modern well managed network system.The most widely deployed and commercially available methods for intrusion detection employ signature-based detection. However, IDSs using these methods have several problems, for example, they cannot detect novel intrusions intrinsically which are not matched to the signatures, and their methods consume huge amounts of cost and time to ac-

A new service model of public wireless Internet access, called autonomous distributed public wireless Internet access, is presented. In the service model any volunteer with broadband Internet access lines can provide his access points for public service without any fear of malicious use. A user of such service is assumed to have his own account on a authentication server at home in the Internet, and all the Internet access through any of those access points can be treated as if it is from the home. In this paper, we present how the autonomous distributed Internet access services can be securely provided with the combinations of two aspects: treatment of authentication transactions at access points and data path of communication transaction.

No previous reports have described cardiopulmonary bypass (CPB) techniques involving long aortic cross clamp (ACC) times in small-breed dogs. We previously described open heart surgery under deep hypothermic CPB in small and toy dogs with an approximate ACC time limit of 1 hour, given the low success rate beyond 90 minutes of ACC time. Herein, we describe improvements in cardiac anesthesia, CPB, and cardioplegia that led to a high success rate of open heart surgery with a long ACC time in small dogs. Ten small-breed dogs with severe mitral regurgitation were subjected to mitral valve plasty that necessitated cardiopulmonary bypass (CPB) beyond 90 minutes of ACC time. In the present study, all dogs survived surgery; 7 survived to discharge. In other words, we achieved a high success rate of 70% with mitral valve plasty beyond 90 minutes of ACC time (mean, 105.1 minutes: range, 90-117 minutes). For small dogs, successful open heart surgery with a long ACC time may require cardiac balanced anesthesia, CPB with increased drainage and, importantly, thorough cardioplegia without hemodilution.

No abstract