環觀電腦通訊科技的整體環境，無線網路服務存取點與個人行動裝置的數量正在與日俱增中，因此加速了網格技術及行動計算的發展；而利用行動設備在無線或行動網格上擷取資訊或找尋知識衍然成為一個日益重要的議題。在網格環境上，目前大多數的研究都只著重在縮短回應時間，而其它的因素並未被考慮其中，像是能源的消耗和即時的網路頻寬等，而在行動計算上，這些因素都將會影響到行動網格的整體效能與系統可用性。因此在本篇論文中我們提出一個 “行動網格環境上的適應性知識擷取” (Adaptive Knowledge Retrieving on mobile grid, AKR)方法，這個方法是以「行動代理人」和我們所提出的「資源評價模式」(Resource Estimation Model) 為基礎，為了增進行動網格的整體效能及可用性，此適應性擷取方法可以動態地決定知識擷取的過程，根據此資源評價模式，利用一個 Saving Factor (SF) 來決定進行擷取的地點，若在行動節點上擷取所需的回應時間及能源的消耗相較於在行動伺服器端上做擷取來得少，則調適為在行動節點進行擷取；反之則在行動伺服器端進行擷取。而由模擬的結果中可以得知採用我們提出的適應性知識擷取方法是更有效率的，且達到更高的可用性。

Retrieving information or discovering knowledge　from wireless or mobile devices is becoming increasingly　important issues. Most researches in grid environment　only consider knowledge retrieving time. Other constraints, such　as energy consumption and immediate bandwidth that　are more important in the mobile computing will　physically affect the performance and system availability.
In this thesis, we propose an “Adaptive Knowledge Retrieving on mobile grid” (AKR) approach based on mobile agent and　a Resource Estimation Model. An Adaptive Knowledge Retrieving approach on mobile grid can dynamically　determine the processing of knowledge retrieving in order to achieve high　performance and high availability. Based on the Resource Estimation Model, we present a Saving Factor (SF) to　determine the place of retrieving.　If the latency and energy consumption of retrieving in mobile grid nodes are less than retrieving in mobile grid servers, then adopt to retrieve in mobile grid nodes. Otherwise, retrieve in mobile grid servers. The result of simulation shows that the approach is efficient and high availability.

Chapter 1. Introduction.............................................1
1.1. Problems and motivations.......................................1
1.2. Objectives.....................................................2
1.3. Thesis organization............................................3
Chapter 2. Background...............................................4
2.1. Preliminary....................................................4
2.1.1. Mobile grid..................................................4
2.1.2. Knowledge grid...............................................4
2.1.3. Mobile agent.................................................5
2.1.4. JADE.........................................................5
2.2. Related works..................................................8
2.2.1. Distributed Data Mining on Grids: Services, Tools, and Applications........................................................8
2.2.2. Distributed data mining on Agent Grid: Issues, platform and development toolkit.................................................9
2.2.3. A Hybrid Model for Improving Response Time in Distributed Data Mining........................................................10
Chapter 3. Adaptive Knowledge Retrieving Approach..................13
3.1. Overview of the AKR...........................................13
3.1.1. Application Layer...........................................14
3.1.2. Execution Layer.............................................14
3.1.3. Fabric Layer................................................15
3.2. Components of the AKR.........................................15
3.3. Work-flow of the AKR..........................................21
Chapter 4. Resource Estimation Model...............................25
4.1. Time Estimation Model.........................................25
4.1.1. Retrieving on mobile grid servers...........................26
4.1.2. Retrieving on mobile grid nodes.............................29
4.2. Energy Estimation Model.......................................30
4.2.1. Retrieving on mobile grid servers...........................30
4.2.2. Retrieving on mobile grid nodes.............................32
4.3. Saving factor.................................................33
4.3.1. Time-energy tradeoff........................................34
4.3.2. Dispatching algorithm.......................................36
Chapter 5. Implementation..........................................38
5.1. Implementation environment....................................38
5.2. Example.......................................................39
Chapter 6. Simulation result.......................................51
6.1. Latency.......................................................51
6.2. Energy Consumption............................................53
6.3. System Availability...........................................54
Chapter 7. Discussion..............................................58
Chapter 8. Conclusion and future works.............................60
Reference..........................................................61
Appendix A.........................................................63

[1] Akogrimo project, http://www.mobilegrids.org/
[2] Argo, http://www.argo.ucsd.edu/
[3] Antonio, C., Giuseppe, D. P., "MiPeG: A middleware infrastructure for pervasive grids, " Future Generation Computer Systems," Vol. 24, No. 1, (2008), pp. 17-29.
[4] Bellifemine, F., Caire, G., Poggi, A., Rimassa, G., "JADE: A White Paper," Vol.3, No. 3, 2003.
[5] Cannataro, M., Congiusta, A., Pugliese, A., Talia, D., Trunfio, P., "Distributed data mining on grids: Services, tools, and applications," IEEE Transactions on Systems, Man, and Cybernetics, Vol. 34, No. 6(2004), pp. 2451-2465
[6] Cannataro, M., Talia, D., Paolo, T., "Distributed data mining on the grid," Future Generation Computer Systems, Vol. 18, No. 8, (2002), pp. 1101-1112.
[7] Christopher D. M., Prabhakar R., Hinrich S., Introduction to Information Retrieval, Cambridge University Press. 2008.
[8] Grid Computing Lab, http://grid.deis.unical.it/
[9] H. Zhuge, THE KNOWLEDGE GRID, World Scientific Publishing Co., Singapore, 2004.
[10] Huang, C.-Q., Zhu Z.-T, Wu Y.-H. , and Xiao Z.-H.:"Power-Aware Hierarchical Scheduling with Respect to Resource Intermittence in Wireless Grids, ? 2006 International Conference on Machine Learning and Cybernetics, Aug. 2006, pp. 693-698.
[11] Jade - Java Agent DEvelopment Framework, http://jade.tilab.com/
[12] K* grid project, http://www.gridcenter.or.kr/
[13] Krishnaswamy, S., Loke, S.W., Zaslasvky, A., "A hybrid model for improving response time in distributed data mining," IEEE Transactions on Systems, Man and Cybernetics, Part B, Vo. 34, No. 6, (2004), pp. 2466 -2479.
[14] Litke A., Skoutas D., and Varvarigou T.; "Mobile Grid Computing: Changes and Challenges of Resource Management in a Mobile Grid Environment."
[15] Luo, J., Wang, M., Hu, J., Shi, Z., "Distributed data mining on Agent Grid: Issues, platform and development toolkit," Future Generation Computer Systems Vol. 23, No. 1, (2007), pp. 61-68.
[16] McKnight, L., Howison, J., and Bradner, S.: "Wireless Grids--Distributed Resource Sharing by Mobile, Nomadic, and Fixed Devices, ? IEEE INTERNET COMPUTING, July 2004, pp. 24-31.
[17] Stankovski, et. al., "Grid-enabling data mining applications with DataMiningGrid: An architectural perspective, Future Generation Computer Systems," Vol. 24, No. 4, (2008), pp. 259-279.
[18] The Foundation for Intelligent Physical Agents, http://www.fipa.org/
[19] Wang, G., Wen T., Guo, Q., Ma, X., "A Knowledge Grid Architecture Based on Mobile Agent, Second International Conference on Semantics, Knowledge, and Grid," 2006. SKG '06. (2006), pp. 48 - 51.
[20] Yan, X., Xu, H., Xu, Y., Liu, L.: "A data replica replacement algorithm based on value model in mobile grid environments," 2nd Int'l Conference on Mobile Technology, Applications and Systems, Nov. 2005, pp. 15-17.
[21] Yang, C.-T., Chen C.-J., Hsu C.-H.: "A Peer-to-Peer Resource Sharing System with Data Grid Technology for Mobile Devices," Int'l Conference on Multimedia and Ubiquitous Engineering (MUE '07), April 2007, pp. 723-728.