傳統P2P網路中，搜尋的對象通常是人或檔案，而這些人或檔案的名稱是我們在搜尋的過程中主要的參考依據。在這類的搜尋中，所謂的名稱，並不會隨著時間而改變。然而，在P2P計算及網路串流這一類的應用中，搜尋的目標可能是CPU周期，記憶體大小，和硬碟的空間。上述的這些，就是這類應用中的搜尋目標。但是，這一類資源的狀態，會隨著時間的改變而有所不同。由於這類的資源可能會因為被使用而改變狀態而需要頻繁更新資訊，若我們使用傳統目錄式的P2P，無法有效的解決這類資源的搜尋問題。經由觀察，這一類的應用有一個特色，就是每一個節點，都擁有一組資源(resource)。而這一組資源就稱為該節點的屬性(attribute)。
在本篇論文中，我們提出Attribute-Based Overlay Network (AON)來解決上述的問題。這個解法和傳統的解法有一個極大的不同之處：傳統的結構式P2P網路是經由hash key來連接節點，而我們的解法是經由屬性來連接節點。由於這個原因，屬性相似的節點，會成為覆蓋(overlay)中的鄰居。在這個覆蓋中，為了達到更有效的繞徑(routing)，我們採用了小世界中的想法，在我們的覆蓋中建立捷徑。AON可以輕鬆的掌控資源的變化，並同時回傳一群相似的結點。實驗的結果顯示AON在繞徑的效率中和CAN一樣好。此外，在面對系統節點狀態劇烈變化時，AON比CAN有著更好的穩定度。

The search targets in traditional P2P networks are often files or people, and the names of the files or people are the primary attribute for lookup to locate the target. A distinct feature of these applications is that the name attribute do not change with time. However in applications such as P2P computing and streaming that require quality resources, the search targets may be CPU cycles, memory space, and network bandwidth. The status of these targets may change when they are used and their search attributes may differ with time. Traditional P2P systems which implement a distributed directory may not serve these applications efficiently, due to the frequent updates and the possibility of false positive answers. Fortunately, in these applications peers usually supply only one set of resources and can thus be characterized by one set of attributes.
In this thesis, we propose Attribute-Based Overlay Network (AON) for searching dynamic resources. The idea is to interconnect the peers according to their attributes instead of some hash keys. Nodes with similar attributes will thus be neighbors in the overlay. Shortcuts are created according to the principle of small world for routing efficiency. AON can easily handle dynamic changing resources and answers queries requesting a given number of similar nodes. The simulation experiments show that AON can perform as well as CAN in searching efficiency. When facing churns, AON presents higher stability.

[1]S. Ratnasamy, P. Francis, M. Handley, R.M. Karp and S. Shenker, "A Scalable Content-Addressable Network", Proc. ACM SIGCOMM, 2001.
[2]X. Li, J. Wu, "Searching Techniques in Peer-to-Peer Networks", Handbook of Theoretical and Algorithmic Aspects of Sensor, Ad Hoc Wireless, and Peer-to-Peer Networks, CRC press, 2005.
[3]J. Kleinberg, "The Small-World Phenomenon: An Algorithmic Perspective", Technical Report, 99-1776, Cornell Computer Science, (1999).
[4]I. Stoica, R. Morris, D. Liben-Nowell, D. Karger, M. Kaashoek, F. Dabek, H. Balakrishnan, "Chord: A Scalable Peer-to-peer Lookup Protocol for Internet Applications", IEEE/ACM Transactions on Networking, Vol. 11, No. 1, pp. 17-32, February 2003.
[5]A. Rowstron and P. Druschel, "Pastry: Scalable, Decentralized Object Location, and Routing for Large-scale Peer-to-peer Systems", Proc. Middleware, November 2001.
[6]M. Hefeeda, A. Habib, D. Xu, B. Bhargava, and B. Botev, "Collectcast: A Peer-to-peer Service for Media Streaming", ACM/Springer Multimedia Systems Journal, October 2003.
[7]M. Hefeeda, A. Habib, B. Botev, D. Xu, and B. Bhargava, "PROMISE: Peer-to-peer Media Streaming Using CollectCast", ACM Multimedia, Nov 2003.
[8]D. Liben-Nowell, H. Balakrishnan, D. Karger, "Analysis of the Evolution of Peer-to-Peer Systems", Proc. ACM Conf. on Principles of Distributed Computing (PODC), Monterey, CA, July 2002.
[9]Y. Ren, C. Sha, W. Qian, A. Zhou, B. Ooi, K. Tan, "Explore the'Small World Phenomena in Pure P2P Information Sharing Systems," Proc. 3rd International Symposium on Cluster Computing and the Grid (CCGrid), May 2003, pp. 232-239.
[10]K. Anderson, "Analysis of the Traffic on the Gnutella Network", Final Project, CSE222, University of California, San Diego, March 2001.
[11]S. Rhea, D. Geels, T. Roscoe, J.Kubiatowicz, "Handling Churn in a DHT", Proc. USENIX Annual Technical Conference, 2004.
[12] http://setiathome.berkeley.edu/
[13] http://www-unix.mcs.anl.gov/mpi/