Research on performance enhancements of mobile computing system using location and mobility analysis위치 및 이동성 분석을 활용한 이동컴퓨팅 시스템의 성능 개선에 관한 연구

Cited 0 time in webofscience Cited 0 time in scopus
  • Hit : 598
  • Download : 0
Today, as mobile communication and computing technologies are significantly improving, many people enjoy the convenience of mobile application services through portable devices, such as smart-phones and tablet computers. Based on these advances, we are approaching the era of the Internet-of-Things (IoT) in which light-weight computing substrates are installed in real world objects so that they become parts of the Internet world. In an effort to realize the vision of the IoT, this dissertation proposes to exploit user location and mobility-related features to improve the performance of wireless mobile computing systems. Specifically, it explores the following two issues. As the first result of this dissertational research, we present a new distributed hash table (DHT) for wireless mobile environments, especially for when a higher level of mobility and membership dynamics occur together. In this DHT paradigm, a publish/look-up process consists of a sequence of spatial motion tracking of the rendezvous node that is responsible for directory information. Thus, we name the protocol Motion-MiX-DHT (MX-DHT). As opposed to conventional topology-based DHT schemes that suffer from frequent disconnections of the logical topology caused by user mobility, MX-DHT provides robust publish/look-up paths exploiting one-hop encounter records as the milestones for motion tracking. This is inspired by two mobility-related features: the mobility capacity effect and motion-mix property. By leveraging the mobility capacity effect, which implies mobility-assisted data dissemination, motion tracking is successfully done with a low-cost search-and-forward process. Join/leave operation also consists of a low-cost one-hop signaling process depending on the motion-mix property. Consequently, in high-speed mobility and dynamic membership environments, MX-DHT achieves a significant reduction in the communication costs of publish/look-up and join/leave operations as compared to conventional mobile DHT schemes. An extensive set of experiments showed that MX-DHT is a cost-effective solution to providing various distributed services (e.g., domain name service, P2P file sharing, distributed file system, cloud storage, and so on) in a wide range of wireless mobile networks with dynamic mobility and membership changes (e.g., vehicular or human-carried device networks). The second focus of this dissertation is developing an analytical method to estimate the mean cell-sojourn time in 2-tier heterogeneous (e.g., macro/pico, macro/femto or macro/Wi-Fi) networks, which have been widely used to increase the capacity density of mobile systems. In heterogeneous networks, accurate estimation of the mean cell- sojourn time is critical for evaluating the performance of the network and its applications. It is especially important to analyze the cell- sojourn times of mobile users, as they reside in different network tiers: either macro-cell-only or small-cell-covered areas. However, because small cells are deployed in an irregular manner, it is difficult to derive the analytical mean cell sojourn time in a macro-cell-only area. We propose a novel approach to resolve this difficulty. Based on the location distribution features of general random walk mobility, we developed a simple but effective trick that approximates a heterogeneous network to a discrete grid so that it becomes tractable, making it possible to derive the analytical mean sojourn time in the macro-cell-only area. Simulation results confirm that the proposed method has excellent accuracy for general random walk mobility models, including random waypoint, Brownian motion, and Levy walk. In this dissertational research, user location and mobility-related properties such as mobility capacity effect, motion-mix property, and uniform location distribution feature, play important roles to achieve the above contributions. Considering that these features are satisfied well in various user mobility models including random waypoint, Brownian motion, Manhattan, urban vehicular, and Levy walk, our proposed works can be applicable to a wide range of mobile computing services. MX-DHT can be used as a data discovery/sharing protocol for autonomous local ad hoc communication. And, our cell sojourn time estimation method may contribute to optimize infrastructured wireless Internet access. By interworking these technologies, flexible pervasive Internet connectivity can be established so that we can realize various IoT services.
Advisors
Yoon, Hyunsooresearcher윤현수researcher
Description
한국과학기술원 :전산학부,
Publisher
한국과학기술원
Issue Date
2016
Identifier
325007
Language
eng
Description

학위논문(박사) - 한국과학기술원 : 전산학부, 2016.8 ,[iv, 67 p. :]

Keywords

location and mobility-related features; wireless mobile network; high-speed mobility; dynamic membership change; last encounter routing; distributed hash table; heterogeneous network; hierarchical cellular mobile network; cell sojourn time; analytical estimation; 위치 및 이동 특성; 무선 이동 네트워크; 고속 이동성; 동적 멤버쉽 변화; 목격 기반 라우팅; 분산 해시 테이블; 이기종 네트워크; 계층화된 셀룰라 네트워크; 셀 체류시간; 해석적 추정

URI
http://hdl.handle.net/10203/222412
Link
http://library.kaist.ac.kr/search/detail/view.do?bibCtrlNo=663209&flag=dissertation
Appears in Collection
CS-Theses_Ph.D.(박사논문)
Files in This Item
There are no files associated with this item.

qr_code

  • mendeley

    citeulike


rss_1.0 rss_2.0 atom_1.0