臺灣博碩士論文加值系統

分散式系統的潛在優點是很重要的, 包括好的性能, 可靠度, 資源分享和
可擴大性. 在這篇論文中, 我們只考慮在分散式網路中趨近最佳可靠度的
問題. 在實際的例子中, 大部分的分散式網路是很大的. 並且隨著節點數
目的增加, 計算的時間會呈指數的增加. 為了減少計算分散式網路的時間
和複雜度, 我們在這篇論文中發展兩種演算法, 在第一種方法中, 我們提
出一些新的規則, 使在演算過程中能省略不必要的計算. 和其它的方法比
較能減少計算時間, 並節省更多儲存空間. 在第二種方法中, 為了在非常
大的分散式通訊網路中減少計算的時間, 我們提出一種啟發式演算法. 用
一些例子和其它演算法比較, 結果顯示我們的演算法可得到較佳的解答.

Potential advantages of a distributed computing system are
significant, including good performance, reliability, resource
sharing, and extensibility. In this paper, we consider only
thereliability in the approximating optimal distributed
communication network (DCN). In real case, most DCN is very
large, and as the number of nodes growing, it increases the
execution time of solution exponentially. In order to reduce
the total execution time and complexity, we develop two method
in this paper. In the first method, we provide some new rules
which can skip more unnecessary computations, and compared
withother methods, it can reduce the execution time, and save
more storage space. In the second method, in order to reduce the
execution time in a very large DCN, we develop a heuristic
method. Compare some examples with other algorithms, the
results shows that our algorithm can get better solution.