Ryo Nishide

本稿では，環境中に存在するあらゆるBluetooth機器を対象に，ユーザが常時携帯するBluetooth機器から得られる検出履歴を分析しユーザの行動を推定できる可能性について述べる．ユーザが得られる検出履歴の特徴は場所や状況によって異なり，行動分類に有効な分節パラメータの選択が重要になる．

In this paper, we propose method to construct a Spatial MMOG system on a P2P Delaunay Network, which is built from the computer nodes provided by users. In this method, we use the player character's location on MMOG as node location, and construct Delaunay Network and process the task for MMOG distributively. First, we propose construction method of P2P Delaunay Network. In Spatial MMOG, to construct a view for each player, the view requires the surrounding object information from the nearby player. Thus, we divide the whole plane into Voronoi region, and assign the object information to the node managing the Voronoi region. Furthermore, it is necessary to maintain the P2P Delaunay Network connection considering the case of the moving nodes. To handle with this case, we propose method to rebuild the network locally according to the new location of the node, and show method to maintain the P2P Delaunay Network.

In this paper, we propose method to construct a Spatial MMOG system on a P2P Delaunay Network, which is built from the computer nodes provided by users. In this method, we use the player character's location on MMOG as node location, and construct Delaunay Network and process the task for MMOG distributively. First, we propose construction method of P2P Delaunay Network. In Spatial MMOG, to construct a view for each player, the view requires the surrounding object information from the nearby player. Thus, we divide the whole plane into Voronoi region, and assign the object information to the node managing the Voronoi region. Furthermore, it is necessary to maintain the P2P Delaunay Network connection considering the case of the moving nodes. To handle with this case, we propose method to rebuild the network locally according to the new location of the node, and show method to maintain the P2P Delaunay Network.

This paper proposes an incremental algorithm for constructing the Delaunay overlaid network for virtual collaborative space. In our algorithm, every node uses Delaunay triangulation, by using its knowledge of other nodes and cooperatively exchanging information among nodes, which generates and refines overlaid networks. On the resultant Delaunay network, nodes communicate with each other over virtual collaborative space, and employ multihopping communication among distant nodes. This algorithm structure has the following advantages: (1) utilization of only the localized nodes, omitting the use of unnecessary distant nodes for communication, (2) scalability through surveillance of nodes by refining and updating the network continuously, (3) two dimensional space organization which allows construction of super nodes that govern regional areas to process geographical range query, and (4) directional communication through a small routing table. The authors believe that this approach is applicable to large-scale virtual collaborative space and location aware system.