TY - JOUR
T1 - DSLM
T2 - a decentralized search for large and mobile networks
AU - Chuang, Yung-Ting
AU - Yu, Chung Yen
AU - Wu, Qian Wei
PY - 2018/2/1
Y1 - 2018/2/1
N2 - As ubiquitous networked devices continue to play an increased role in the daily lives of most people, there is a growing desire to share ever more information and perspectives from across the world. In addition, with the recent technical advancements, the networks have turned to a large-scale wireless networks such as mobile ad hoc networks. However, the problem behind such wireless networks is the difficulty of data distribution and information retrieval in such decentralized and large mobile wireless networks, while still maintaining low overhead. Therefore, we present a system we call decentralized search for large and mobile wireless networks, or DSLM, which: (1) divides the entire network into smaller regions; (2) allows nodes to join, leave, distribute metadata, or make requests; (3) applies the LSH method to map the metadata or request to a geographical region; (4) uses a region-based geographic routing method to route messages between nodes; (5) allows nodes to maintain only a partial view of the network; and (6) applies a relocation method to address mobility issues. We finally demonstrate that our DSLM requires low overhead and could still achieve high retrieval rates and mobility resilience.
AB - As ubiquitous networked devices continue to play an increased role in the daily lives of most people, there is a growing desire to share ever more information and perspectives from across the world. In addition, with the recent technical advancements, the networks have turned to a large-scale wireless networks such as mobile ad hoc networks. However, the problem behind such wireless networks is the difficulty of data distribution and information retrieval in such decentralized and large mobile wireless networks, while still maintaining low overhead. Therefore, we present a system we call decentralized search for large and mobile wireless networks, or DSLM, which: (1) divides the entire network into smaller regions; (2) allows nodes to join, leave, distribute metadata, or make requests; (3) applies the LSH method to map the metadata or request to a geographical region; (4) uses a region-based geographic routing method to route messages between nodes; (5) allows nodes to maintain only a partial view of the network; and (6) applies a relocation method to address mobility issues. We finally demonstrate that our DSLM requires low overhead and could still achieve high retrieval rates and mobility resilience.
KW - Distributed search and retrieval
KW - Mobile ad hoc networks
KW - Network churn
KW - Peer-to-peer
KW - Probabilistic analysis
KW - Wireless sensor networks
UR - http://www.scopus.com/inward/record.url?scp=85031507095&partnerID=8YFLogxK
U2 - 10.1007/s11227-017-2158-8
DO - 10.1007/s11227-017-2158-8
M3 - Article
AN - SCOPUS:85031507095
SN - 0920-8542
VL - 74
SP - 738
EP - 767
JO - Journal of Supercomputing
JF - Journal of Supercomputing
IS - 2
ER -