• Journal of KIISE:Information Networking
• /
• v.32 no.1
• /
• pp.110-122
• /
• 2005

Bluetooth has been reputed as a wireless networking technology supplying ad-hoc networks between digital devices. In particular, bluetooth scatternet is a most essential part for dynamic ad-hoc networks. But past researches on bluetooth scatternet has hardly treated dynamic scatternet environment. In this paper, we proposed a scatternet reformation algorithm for the case that some nodes escape from the scatternet. The proposed algorithm is a general algorithm which can be applied to many types of bluetooth scatternet regardless of the topology. The proposed algorithm has short reformation time delay because the process has only page process (not including inquiry process ). The algorithm is operated based on Recovery Node Vector which is composed of Recovery Master and Recovery Slave. In this paper, we performed the real hardware experiments for evaluating the performance of the proposed algorithm. In that experiments, we measured the reformation time and reformation probability. In comparison with the case including inquiry process, the proposed algorithm had the improvement in reformation time delay and we obtained high success rate over 97%.

• Journal of KIISE:Information Networking
• /
• v.31 no.3
• /
• pp.314-326
• /
• 2004

Bluetooth has been reputed as a wireless ad-hoc networking technology supplying scalable and extensible networks between digital devices. For that kind of networks, scatternet is a most essential part in bluetooth. But past researches on bluetooth scatternet has proposed only possibilities of scatternet algorithm based on simulation results. And many of the researches failed in guaranteeing extensibility and flexibility and had many difficulties in real hardware implementation. In this paper, we proposed node ring scatternet(NRS) algorithm guaranteeing extensible and flexible networks. NRS algorithm is designed for hardware implementation using real commercial bluetooth module. That algorithm is divided into initial formation and reformation. For initial formation, we proposed limited SEEK/SCAN algorithm. For reformation, we proposed DIAC algorithm and Reserved Recovery Node algorithm. And we proposed SFMP(Scatternet Formation ＆ Management Protocol) in protocol stack for real implementation. NRS algorithm is operated in SFMP. Finally, we performed real hardware experiments and evaluated the proposed algorithm. In that experiments, we succeeded in forming scatternet up to 20 nodes. In comparison with other similar algorithm, proposed algorithm have the improvement in scatternet formation delay and success rate.

• Proceedings of the Korean Institute of Information and Commucation Sciences Conference
• /
• 2008.05a
• /
• pp.499-502
• /
• 2008

한정된 용량의 배터리에 의존하는 무선 Ad-hoc 네트워크(MANET)에서는 에너지 효율을 높이기 위한 다양한 클러스터링 기법과 라우팅 알고리즘이 연구되고 있다. 이러한 무선 Ad-hoc 네트워크에서는 에너지 효율이 높은 클러스터 기반의 라우팅 알고리즘이 많이 사용된다. 그러나 일반적인 클러스터 방식에 따른 라우팅 알고리즘에서는 클러스터 헤드 노드에 부하가 집중되어 에너지 소모가 많은 문제점을 가진다. 이 문제를 보완하기 위해서 클러스터 헤드 노드의 재 선출을 통해 에너지 소모를 분산하는 동적 클러스터링 방식이 사용되고 있다. 그러나 동적 클러스터링 방식 또한 높은 빈도의 클러스터 재형성 과정에서 많은 에너지를 소모하는 문제점이 있다. 즉, 지금까지 연구되어온 알고리즘은 클러스터 구성에 대한 효율적인 알고리즘을 제시하고 있지만 불필요한 에너지 소모를 최소화하는 최적의 헤드 노드 선정 방법과 클러스터 관리를 통하여 에너지 효율을 높일 수 있는 해결책을 제시하지 않았다. 따라서 본 논문에서는 위의 클러스터 문제를 해결하기 위해 TICC(Time Interval Clustering Control) 알고리즘 기법을 제안한다. 제안된 TICC은 각 노드의 에너지 속성 값에 따라 에너지 Level을 분류하고 분류된 에너지 Level에 따라 타이밍을 고려한 클러스터링 및 노드 관리방법이다. 이러한 TICC기법을 적용하여 실험을 하였고 결과적으로 클러스터 전체의 에너지 효율을 향상되고 Lifetime이 증가함을 보였다.

• The Transactions of the Korea Information Processing Society
• /
• v.7 no.1
• /
• pp.88-98
• /
• 2000

Cooperative query answering provides neighborhood or associate information relevant to the initial query using the knowledge about the query and data. In this paper, we present an intelligent query answering system for suporting cooperative query answering system presented in this paper performs query relaxation process using hybrid knowledge base. The hybrid knowledge base which is used for relaxation of queries, composes of semantic list and rule based knowledge base for structural approach. Futhermore, this paper proposes the query relaxation algorithm for query reformulation using initial query on the basis of hybrid knowledge base.