Young-Ju Han

In designing the wireless sensor networks, the energy is the most important consideration because the lifetime of the sensor node is limited by the battery of it. To overcome this demerit many research have been done. The clustering is the one of the representative approaches. In the clustering, the cluster heads gather data from nodes, aggregate it and send the information to the base station. In this way, the sensor nodes can reduce communication overheads that may be generated if each sensor node reports sensed data to the base station independently. LEACH is one of the most famous clustering mechanisms. It elects a cluster head based on probability model. This approach may reduce the network lifetime because LEACH does not consider the distribution of sensor nodes and the energy remains of each node. However, using the location and the energy information in the clustering can generate big overheads. In this paper we introduce CHEF - cluster head election mechanism using fuzzy logic. By using fuzzy logic, collecting and calculating overheads can be reduced and finally the lifetime of the sensor networks can be prolonged. To prove efficiency of CHEF, we simulated CHEF compared with LEACH using the matlab. Our simulation results show that CHEF is about 22.7% more efficient than LEACH.

In this paper, we are going to perform performance tests for medical data in ad-hoc protocols. The protocols are: DSR (the dynamic source routing protocol), AODV (the ad hoc on demand distance vector) and DYMO (the dynamic MANET on-demand). An emergency communication means a communication occurred in an un-predictable place. With this peculiarity, ad-hoc network is suitable for an emergency medical communication. However, ad-hoc network was not invented for a medical communication. We need to analysis ad-hoc protocols' appropriateness for medical data. We simulate medical data with node number increase and node speed for an end-to-end transfer rate evaluation.

Computer system is evolving rapidly to ubiquitous computing system. A smart home environment is a typical ubiquitous environment consisting of a network of computerized electronic appliances. Current smart home system provides context information on individual machinery and it responses only to user's passive requirement. Correspondingly, smart home technique does not offer user's authentication access control, and preservation. To emphasize these issues, we study using the environment information by access control method, or user's authentication service requirement. However, current techniques still contains weak points. They cannot establish a large scaled-policy and it cannot also provide various efficient authentication methods. In this research, we propose an integrated preservation system by using the security middleware to solve the problem.

The wireless sensor network is a type of the wireless ad-hoc networks. It is composed of a collection of sensor nodes. Sensor nodes collect and deliver necessary data in response to user's specific requests. It is expected to apply the wireless sensor network technology to various application areas such as the health, military and home. However, because of several limitations of sensor nodes, the routing protocols used in the wireless ad-hoc network are not suitable for the wireless sensor networks. For this reasons, many novel routing protocols for the wireless sensor networks are proposed recently. One of these protocols, the PEGASIS (power-efficient gathering in sensor information systems) protocol is a chain-based protocol. In general, the PEGASIS protocol presents twice or more performance in comparison with the LEACH (low energy adaptive clustering hierarchy) protocol. However, the PEGASIS protocol causes the redundant data transmission since one of nodes on the chain is selected as the head node regardless of the base station's location. In this paper, we propose the enhanced PEGASIS protocol based on the concentric clustering scheme to solve this problem. The main idea of the concentric clustering scheme is to consider the location of the base station to enhance its performance and to prolong the lifetime of the wireless sensor networks. As simulation results, the enhanced PEGASIS protocol using the concentric clustering scheme performs better than the current PEGASIS protocol by about 35%.

A service provider needs a AAA (authentication, authorization, accounting) architecture, because it is necessary to provide services to eligible users with the right authorities. The service provider also needs to charge the users on a commercial purpose. However, the current AAA architecture cannot support a mobile user (MU) sufficiently that moves far away from the AAAH. To provide delay-sensitive services such as VoIP, the delay, which occurred between the MU and the AAAH in the current AAA architecture should be shortened. On this paper, moving AAAH architecture (MAHA) is proposed to decrease the delay, so that the delay-sensitive service is provided stably. MAHA can also be used to prevent the communication delay from being unbearable. In MAHA, an AAAH sends the information of an MU to another AAA server that is closer from the MU when the delay exceeds the critical value. As the result of the decreased delay, no matter how long distance the MU moves, the mobility services and the delay-sensitive services are provided stably.

This paper presents novel buffer overflow countermeasure tool for Windows portable executables at run-time. Our tool enables dynamically detecting and preventing of stack-based buffer overflow attacks for Windows applications, using binary rewriting method. Our solution protects the return address area and the previous frame pointer area of function stack frame in program stack to prevent program control flow from being changed at execution time. Protecting the return address and previous frame pointer, we have used additional stack memory area that is called safe-zone, storing original return address and previous frame pointer values. We has revised function prologue and function epilogue. The revised function prologue stores copies of the return address and the previous frame pointer values to our safe-zone, and the revised function epilogue overwrites the return address and the previous frame pointer on the stack with copies of them. The paper presents performance analysis result of our solution. The result shows that the relative performance overhead is about 1.6-2.6% and additional constant space overhead is about 25 Kbytes.