物聯通訊下考慮資料相關性之群組形成與排程演算法協同設計

在多數的無線物聯通訊研究中，常僅以支援密集部署的網路情境作為優化的目標而忽略了物聯裝置所搜集之資料存在著相關性。由於物聯資料相關特性，無線物聯裝置所要傳送的資料可以透過網路壓縮的技術有效減少，進而增進系統效能。考量到物聯裝置『節省功耗』以及物聯應用『有效資訊』這兩項特性，在本論文中我們提出以有效資訊為基準的群組形成與排程演算法協同設計。我們首先提出兩階層的無線網路傳輸架構並以最小化功率消耗作為目標。接著，我們將此機制拆解成兩個子問題：『最小化功耗之節點群組問題』以及『最小化功耗之節點排程問題』。為解決這兩個子問題，我們提出考慮限制條件的模擬退火演算法(Constrained Simulated Annealing, CSA) 以解決最小化功耗之群組問題。在形成兩階層的網路群組結構後，最小化功耗之節點排程問題則可以被化簡成為混和整數非整數之線性規劃問題。因此我們進一步提出一個有效的演算法以求得此問題之最佳解。我們經由高斯資料模型以及影像資料兩種不同資料來源進行數值效能分析。透過數據驗證，除了可以說明無線群組技術將可以比直接傳輸獲得更好的功耗表現之外，本論文所提出的兩階層網路傳輸機制在相關性資料傳輸情境中也獲得比傳統群組演算法還要好的功耗表現。

關鍵字

無線網路；物聯網；感知無線網路；最佳化

並列摘要

Clustering and transmission power control have been proposed as an e ective way to support massive access in wireless machine-to-machine (M2M) networks. For M2M service, the quality of gathered information is a more realistic factor to evaluate the system performance than the link quality of each machine. However, most of the recent work focuses on enhancing the service quality of individual machines but ignore the nature of data correlation among machines (sensors). In this thesis, we rst formulate a problem for 2-tier minimum power data gather- ing in M2M networks. Then, we decompose the problem into two sub-problems: minimal power consumption data-centric clustering and minimal power consump- tion scheduling. We apply Constrained Simulated Annealing (CSA) to solve the cluster formation problem. After the cluster structure is determined, the minimal power consumption scheduling problem can be transformed into a mixed-integer linear programming (MILP) problem. We then proposed a 0-1 branch-and-bound algorithm to obtain the optimal solution. To evaluate our proposed transmission scheme, we consider both Gaussian data source and real image data. Evaluation results show that the proposed scheme achieves better performance than conven- tional clustering algorithms.

並列關鍵字

Wireless Network ； Machine-to-Machine Network ； Wireless Sensor Network ； Optimization

參考文獻

3GPP Std. TR23.888, Rev. 11.0.0, 09 2012. Online Available at:

Toward intelligent machine-to-machine communications in smart grid,"

Communications Magazine, IEEE, vol. 49, no. 4, pp. 60{65, 2011.

tion in vehicular networks with infrastructure support," in Global Communi-

home energy saving based on energy-prone context," in Intelligent Robots

國際替代計量

物聯通訊下考慮資料相關性之群組形成與排程演算法協同設計

全文下載

主題瀏覽