S. Phillips and J. Westbrook, "On-line algorithms: Competitive analysis and beyond," in Algorithms and Theory of Computation Handbook, M. J. Atallah, editor, chapter 10, Boca Raton: CRC Press, 1999.  Home/Search   Document Details and Download   Summary   Related Articles

....is to compare the performance of an algorithm (for, say, caching or retrieval) with the best possible performance if the entire sequence of requests were known in advance, and to attempt to minimize the maximum ratio of the two. A recent overview of this area is given by Phillips and Westbrook [38]. Many such competitive algorithms address the paging problem , in which the memory is divided into two parts, a fast cache and a slower main memory. Requests for pages not in the cache are served in unit time from the main memory. Thus in this problem, access times do not scale with the total ....

....move independently and in parallel. If several read write heads move simultaneously, we are billed only the maximum amount of time taken by the heads. Similar remarks hold in any case if we adopt the rule that we are billed the sum of the times taken; this is the convention in k server problems [38]. In the no redundancy case, we partition the image blocks between the s(n) tapes. Now the problem arises that the alternative experiment given above no longer mirrors the underlying user driven random walk faithfully: if two neighboring blocks i and j on the image are stored on 14 different ....

S. Phillips and J. Westbrook, "On-line algorithms: Competitive analysis and beyond," in Algorithms and Theory of Computation Handbook, M. J. Atallah, editor, chapter 10, Boca Raton: CRC Press, 1999.

Online articles have much greater impact   More about CiteSeer.IST   Add search form to your site   Submit documents   Feedback  

CiteSeer.IST - Copyright Penn State and NEC