Catálogo de publicaciones - libros

To balance energy consumption among cluster-heads in multi-hop wireless sensor networks, we propose an energy-efficient distance based probabilistic clustering scheme. Our scheme considers a distance from the sink(base station) to each node as well as the residual energy of each node as the criterion of cluster-head election. Through simulation experiments, we show that our scheme is more effective than LEACH and EEUC in prolonging the lifespan of multi-hop wireless sensor networks.

Recently, with development of Internet telephone, we can say that the ENUM could be the most suitable Internet identification system standard in the BcN(NGN). When embody ENUM, there are problems on security for personal information and convergence with existing domain. And it is need that ENUM service provides convenient function to users. In this paper, we design the system that offers Single Sign On without division of user terminal using SAML when it requires URL with a telephone number to ENUM DNS.

Satellite communication environment can be divided into LOS and NLOS environments. While, terminal can get satellite signal in LOS environment, NLOS needs additional function to get the signal such as cooperation between gap filler or terrestrial network (WLAN or WIMAX). In this paper, the two way satellite communication system is based on DVB-RCS specification and terrestrial networks are based on WLAN(IEEE 802.11) or WIMAX(IEEE 802.16). This paper adopt proper algorithms to handover among satellite and wireless network for providing seamless service when moving object enter NLOS environment in collaboration area and estimate the performance of the method by simulations.

The study of DACS Scheme, which manages a whole network system by communication control on a client computer, has been advanced. The problem for applying DACS Scheme to practical network is that, processing load is heavy at the time of controlling communication. In this paper, we propose models for managing a whole network system; Course management type model (COMTM) and Client management type model (CLMTM). Each processing load which occurs at the time of controlling communication in these two models is compared. It is shown that, DACS Scheme which is CLMTM is advantageous in comparison with COMTM from the point of processing load, which occurs at the time of control communication.

This paper describes the testing activities for the maintenance of the NeOSS (New Operations Support System) of Korea Telecom. To ensure the successful maintenance of the NeOSS without any effect on the existing functions and performance, we performed various tests related to functionality, efficiency and others before the added and modified parts were applied to the NeOSS. In this paper, we show the maintenance process, the various tests related to it, the test phases, and the test environment for controlling the quality of the NeOSS maintenance.

We study service problem management and resource trouble management on a telecommunication network with eTOM (Enhnanced Telecom Operations Map). As results of this research, we derive primary necessary functions for process elements of eTOM and compare them with KT NMS functions. In addition, we propose an improvement direction of KT network operations management.

The traditional centralized event correlations are less of simplicity, scalability and robustness. DSEC is presented to solve this problem. It is based on the Divide-and-Conquer strategy that makes parallel local correlation at multiple agents and combines local causes on the global level. It regards a management task as an event probe, thus brings high-level and rich-semantic information to events and achieves the simplicity and scalability of causality graph. Meanwhile, it limits event explosion at each agent and improves robustness to noise.

Neptune attack and Flash Crowd are two typical threats to web servers. These two anomalies have many identical features that make them difficult to distinguish. In this paper, we propose a statistical packet-based method to detect Neptune attacks and Flash Crowds and more importantly, by performing separate analysis by source address aggregation, we also propose additional efficient means to differentiate these two similar anomalies.

Nowadays in communication and network businesses, architectures and systems are independently constructed with existing service or new service deployment without considering common functions and services between them. Consequently, the study of standardized Service Delivery Platform (SDP) structure and its related interfaces is needed to provide the methodology of existing or new services in an efficient way. Therefore, in this paper, we intend to view various solutions about SDP structure with the proposed interfaces which are able to support them, and then recommend various application services which SDP can offer to end users.

In this paper, we discussed methods for recovering from communication failures due to route hijacking. Looking ahead we intend to pursue further work on systemization, authentication, and evaluation.