查詢結果分析
相關文獻
- Heuristic Load Balancing Schemes for an Omega Network Embedded in WDMA Photonic Networks
- 捷運聯合開發基地之評選--模糊多準則決策方法之應用
- 自我調適的動態排程系統--限制排程、模糊理論和遺傳演算法的應用
- 無參數試題反應理論的能力群組之模糊分割
- 模糊品質機能展開演算模式構建之研究
- 管理元件化資源應用系統--以遠距教學系統為實例
- Cost-Effective Design of SONET Self-Healing Rings for Asymmetric Video Traffic
- Load Balancing Based on Routing Strategies
- 模糊理論應用於規劃滿足學生需求所需之行政配合
- 以模糊理論和遺傳演算法為基礎的中文文件自動分類之研究
頁籤選單縮合
| 題 名 | Heuristic Load Balancing Schemes for an Omega Network Embedded in WDMA Photonic Networks=Omega網路嵌入在波長分割多工光纖網路中負載平衡問題之探討 |
|---|---|
| 作 者 | 黃依賢; 陳義文; 顧凌峰; | 書刊名 | Journal of the Chinese Institute of Electrical Engineering |
| 卷 期 | 8:3 2001.08[民90.08] |
| 頁 次 | 頁279-289 |
| 分類號 | 448.6 |
| 關鍵詞 | 負載平衡; 光纖網路; 模糊理論; 網路效能; 間隔時間; Load balancing scheme; WDMA photonic networks; Fuzzy theory; Window-time member function; |
| 語 文 | 英文(English) |
| 中文摘要 | 波長分割多工是一種能有效提高目前光纖網路頻寬的技術。在波長分割多工光纖網路中如何達到負載平衡是一個新的研究方向。在本篇論文中,我們提出一種啟發性的動態調整方法以達到光纖網路中的負載平衡,其網路架構為 Omega 網路。 這種啟發性的調整方法的目的是為了使得網路中每個節點的工作分配更分平均更有效率,除此之外,也要縮短傳輸回應時間,增加網路的效能。我們使用模糊理論來決定每次啟動負載平衡機制的間隔時間(window-time)。 我們提出三個演算法來解決負載不平衡的問題,包括Available Algorithm, Scheduling Algorithm, and Window-Time Evaluation。 並且將這我們的結果和三對不同變異值k��=0.01 & k��=0.1, k��=0.2 & k��=0.4, and k��=0.09 & k��=0.9 的 Variance method比較效率之好壞。藉由模擬的結果發現,網路系統效能隨著不同角度的window-time 成員函式而改變。一般來說,window-time 的成員函式所呈現的角度為60度時,比起傳統方法和Variance method,在改變負載不平衡上,有較佳的效能。 |
| 英文摘要 | Wavelength-Division Multiplexing (WDM) is an effective way of utilizing the large bandwidth of an optical fiber to develop high performance computer systems. Load-balancing technology in WDM networks is an emerging topic. In this paper, we propose heuristic load-balancing techniques for an Omega network embedded in WDMA Photonic Networks. The objective of our load-balancing schemes is to distribute the load efficiently, lower the response time, reduce the traffic and increase throughput. We use Fuzzy theories to determine an adaptive window-time to initiate the loading balance scheme to overcome the overhead problem. The proposed heuristic load-balancing algorithms include Availablity Algorithm, Scheduling Algorithm, and Window-Time Evaluation. These algorithms are proposed to solve the imbalance problem and their efficiencies compared using the Variance method with three pairs of different values: k��=0.01 & k��=0.1, k��=0.2 & k��=0.4, and k ��=0.09 & K�� =0.9. Our simulation results show that system performances varied with different angles of the window-time member function. In general, the proposed Fuzzy method with an angle 60° of window-time member function has the better performance than the traditional method and variance method. |
本系統中英文摘要資訊取自各篇刊載內容。