使用行動代理人來支援推播系統的目的在於藉由行動代理者跨平台、可移動之特性，達到減低資訊伺服器的網路負載量，也可讓使用者不用再耗費時間於瀏覽網站找尋有興趣的資料。
寬頻網路時代的來臨，推播系統的應用更加廣泛，不再侷限於電子報等小資料量的資訊傳遞，取而代之的是高資料量的應用，如數位電子廣告看板等。在主從式的推播架構（Client-Server Push Technology）下，高資料量的資訊推播勢必會面臨許多難題。
目前網路流量中的使用率以同儕式檔案傳輸(P2P File Sharing)佔最大宗，其架構的優點在於使用者需貢獻自己的網路上傳能力來換取更佳的下載速度，縮短下載的時間。在此架構下，整體環境的總上傳能力將會比主從式檔案傳輸(Client-Server File Sharing)高上許多。
此論文的目的在探討當行動代理人推播技術使用同儕式檔案傳輸概念，進行高資料量推播之可行性及優缺點為何？系統實做是以MACE （Mobile Agent Carrier Environment）作為行動代理人的運載環境。最後透過實驗的量測結果來證明將主從式行動代理人推播架構改進為同儕式行動代理人推播架構，將可以面對未來高資料量的資訊推播需求。

The purpose of using mobile agent to support push technology is that could migrate between nodes, reduce the network overhead of push server, and client will not waste time to look up the interesting information.
The arrival that should be broadband network, it is more extensive to push technology, not confining low-size message transmission , the substitute is the application of the large-size transmission such as Digital Advertising board. Client-Server Push Technology on the premise of stopping improving the server ability to the limit forever, how reach with limited network ability to handle large-size transmission message ?
The most rate of utilization in the network flow is by the P2P file Sharing, advantage of this structure is that all clients must contribute self network ability to obtain good speed of downloaded. So the utilization of whole network is better than Client-Server file sharing structure.
The purpose of thesis is to probe the mobile agent push technology using P2P file sharing concept, is it better or not to handle large-size transmission message? The implementation of a prototype system is based on MACE (Mobile Agent Carrier Environment). Prove that P2P-like push system using mobile agent will better than Client-Server push system using mobile agent of handling large-size transmission messages.

[1] O. Babaoglu, H. Meling, and A. Montresor, "Anthill: a framework for the development of agent-based peer-to-peer systems," 2002, pp. 15-22.
[2] Bittorrent, "http://www.bittorrent.org/index.html."
[3] B. Cohen, "Incentives Build Robustness in Bit Torrent," Workshop on Economics of Peer-to-Peer Systems, June 2003.
[4] P. Dasgupta, "Agent based peer-to-peer systems," 2002, pp. I-663-6 vol.1.
[5] Gnutella, "http://www.gnutella.com/."
[6] Kazaa, "http://www.kazaa.com/us/index.htm."
[7] D. B. Lange and M. Oshima, Programming and Deploying Java(TM) Mobile Agents with Aglets: Addison-Wesley Longman Publishing Co., Inc. Boston, MA, USA, 1998.
[8] E. P. Markatos, "Tracing a Large-Scale Peer to Peer System: An Hour in the Life of Gnutella," 2002, pp. 65-65.
[9] A. Mihalyi, "Optimizing peer-to-peer networks using mobile agents," 2004, pp. 194-199.
[10] Napster, "http://free.napster.com/."
[11] A. Puliafito, O. Tomarchio, and L. Vita, MAP: Design and implementation of a mobile agents' platform: Elsevier, 2000.
[12] S. Subhabrata, S. Oliver, and W. Dongmei, "Accurate, scalable in-network identification of p2p traffic using application signatures," in Proceedings of the 13th international conference on World Wide Web New York, NY, USA: ACM Press, 2004.
[13] L. Tie-Yan, Z. Zhi-Gang, and Y. Si-Zhen, "A-peer: an agent platform integrating peer-to-peer network," 2003, pp. 614-617.
[14] T.-I. Wang, "A Mobile Agent Carrier Environment with Mobile Computing Facilities," in IIP: International Conference on Intelligent Information Processing, The 16th IFIP World Computer Congress Beijin, 2000.
[15] 鍾昆學, "以行動代理者支援推播技術," 國立成功大學工程科學所碩士論文, 2001.