本論文提出一種具有容錯(fault tolerant)方法可以增進通訊能力，改善在智慧電網訊號傳輸上的可靠度及效率。在任意拓樸(topology)電表用PLC或ZigBee佈線後，電表透過最小生成樹演算法(Minimum Spanning Tree algorithm)，整體電表通訊距離能夠最短，減少PLC佈線上的成本；當負載(device)擁有無線接收器時，電表(meter)透過匈牙利演算法(Hungarian algorithm)能夠搜尋離本身最近的負載，接收負載資訊。當電表通訊產生錯誤時採用fault tolerant方法，在此論文提出兩種型態fault tolerant方法；靜態(static)為某個電表傳至另一個電表傳輸連線固定，即使使用無線ZigBee傳輸；而動態(dynamic)為某個電表傳至另一個電表，無線ZigBee則依照cost function有效通訊範圍內皆可以，且依照cost較低優先考慮。
本論文實驗在任意拓樸下，在100平方公尺範圍內，隨機產生各種不同電表數量與座標的13個實驗對象(SmartHomeUnit)，整體傳輸發生單一錯誤(single link fault)情形，在兩種型態fault tolerant方法均可達到100%的錯誤覆蓋率(fault coverage)；但當整體傳輸發生多重錯誤(multiple link faults)情形，靜態透過FTGDB(Fault Tolerant Generalized De Bruijn algorithm)，隨著通訊模式個數(d)變多且設定有效通訊範圍下，FTGDB 多重錯誤覆蓋率(multiple fault coverage)在d = 4情況下可達到平均43% ( d = 2情況下可達到平均19%)的容錯能力，並且隨著d變多，則有效最短路徑長度(#hop)也跟著變小；動態透過Kth shortest path algorithm，當傳輸發生錯誤時，則透過無線通訊傳輸進行容錯並且找尋最佳路徑，透過ZigBee或Wi-Fi等在有效通訊範圍下進行容錯，多重覆蓋率(multiple fault coverage) 在d = 4情況下可達到平均61% (在d = 2情況下可達到42%)的容錯能力。針對8個實驗對象在靜態FTGDB在d = 4情況下，平均在100條通訊連線可容許平均43條同時阻斷，同理針對動態Kth shortest path algorithm在d = 4情況下，平均100條通訊連線可容許平均61條同時阻斷。
若容錯完成後，即可計算透過PLC or ZigBee傳輸所需的能量(energy)和延遲時間(latency)，可以給予電力公司資訊，讓電力公司可以評估電力成本、電價評估、即時電價(real time price)等等；本論文提出一種線上需量反映電價優先權樹狀等級演算法(Online Priority Tree algorithm)，經由此演算法計算用戶連接負載的電表需量等級(rank)，用戶之智慧電表可以動態根據電器重要性之優先權，針對待機模式(standby mode)轉態為自動關閉(automatic cut-off mode)，而電力之優先權即為cutoff sequence。

This paper presents a fault-tolerant (fault tolerant) can enhance the communications capabilities, improve the reliability and efficiency of in smart grid signal transmission. Meter layout with PLC or ZigBee in any topology, meter adopt Minimum Spanning Tree algorithm to achieve shortest distance and lost cost in PLC; when device contain wireless receiver, meter adopt Hungarian algorithm can search nearest itself’s device to receive device information. The paper propose two fault tolerant methods: static and dynamic methods. Static method is a meter transmit to another one with regular communication even if a meter tranfmit with ZigBee; dynamic method is a meter transmit another one, ZigBee has low priority according to cost function in effective communication range.
The paper simulation in any 100m2 topology, randon produce 13 SmartUnit with different number of meters and coordinate, two fault tolerant method can achieve 100% fault coverage in single link fault case; but static method use FTGDB(Fault Tolerant Generalized De Bruijn algorithm) multiple fault coverage can achieve 43% with d=4 case; dynamic method use Kth shortest path algorithm multiple fault coverage can achieve 61% with d=4 case. In other words FTGDB has average 100 communication line allow average 43 communication line fault tolerant ability with d=4 case in 13 SmartHomeUnit; Kth shortest path algorithm has average 100 communication line allow average 61 communication line fault tolerant ability with d=4 case in 13 SmartHomeUnit.
If after fault tolerant achievement, count to demand energy and delay time with PLC and ZigBee, then it can offer electric company information. Electric company evaluate electric cost、real time price etc…
The paper propose a online demand response method, the method is Online Priority Tree algorithm to be counted end device’s rank priority according to device importance.

CHAPTER 1 簡介 9
1.1 背景 9
1.2 研究動機 11
1.3 論文架構 11
CHAPTER 2 問題描述 12
2.1 背景 12
2.2 建立圖模型 16
CHAPTER 3容錯智能電網電表系統建置 17
3.1 PHASE I: 通訊距離優化（MST） 17
3.2 PHASE II: 電表和終端負載之間最佳化連接(HM) 18
3.3 PHASE III: 電表系統傳送/接收電力資訊 (PLC & ZIGBEE)所需的延遲時間和能源 19
3.4 PHASE IV: 對予PLC和ZIGBEE路由容錯機制 25
3.5 METER PRIORITY MANAGEMENT 32
CHAPTER 4實驗結果 38
4.1 實驗結果 38
CHAPTER 5結論 52
參考文獻 53

[1] S.M. Amin, “U.S. Electrical Grid Gets Less Reliable,” in Proc. IEEE Spectrum, vol. 48, no. 1, pp. 80, Jan 2011.
[2] T.G. Hovgaard, K. Edlund and J.B. Jorgensen, “The Potential of Economic MPC for Power Management,” in Proc. IEEE Decision and Control, pp. 7533-7538, 2010.
[3] A. Taleb, J. Mathew, and D.K. Pradhan, “An Efficient De Bruijn Graph-Based Fault Tolerant Sensor Networks Design,” in Proc. IEEE Sensor Technologies and Applications, pp. 369-374, 2010.
[4] Y. Agarwal, T. Weng and R.K. Gupta, “Micro-system driving smart energy metering in smart grids,” in DAC.COM Knowledge Center Article, 2010.
[5] ZigBee Alliance Overview.
http://www.ZigBee.org/Specifications/ZigBee/Overview.aspx
[6] B. Zarikoff, D. Malone, “Construction of a PLC test bed for network and transport layer experiments,” in Proc. IEEE Power Line Communications and Its Applications, pp. 135-140, 2011.
[7] S.C. Ergen, “ZigBee/IEEE 802.15.4 Summary,” 2004.
[8] F.L. Lewis, “Wireless Sensor Networks,” in Proc. Smart Environments: Technologies, Protocols, and Applications, 2004.
[9] T. Kim, D. Kim, N. Park, “Shortcut Tree Routing in ZigBee Networks,” in Proc. IEEE International Symposium on Wireless Pervasive Computing, pp. 42-47, 2007.
[10] B.R.A. Kumar, “Route failure tolerant multicast in Mobile Ad Hoc Networks using disjoint minimum spanning trees,” in Proc. IEEE International Conference on Computer Science and Information Technology, pp.396-400, 2009.
[11] H. W. Kuhn, “The Hungarian Method for the Assignment Problem,” in Naval Res. Logistic Quart, pp.83–97, 1955.
[12] F.L. Lewis, “Wireless Sensor Networks,” in Proc. Smart Environments: Technologies, Protocols, and Applications, 2004.
[13] A. Meyerson, “Online Algorithm for Network Design,” in Proc. ACM symposium on Parallelism in algorithms and architectures, pp.275-280, 2004.
[14] Cypress Semiconductor Corporation products interface power line communication datasheet.
http://www.cypress.com/?id=2330&source=header
[15] Hearst electronic products analysis.
http://www2.electronicproducts.com/Powerline_communication_IC_is_fully_integrated-article-HLJH03_Mayt2010-html.aspx
[16] Texas Instruments CC2520 datasheet.
http://www.ti.com/lit/ds/symlink/cc2520.pdf
[17] Y.J. Lin, “Periodic Contention-Free Multiple Access for Power Line Communication Networks,” in Proc. IEEE Conference on Advanced Information Networking and Applications, pp.315-318, 2005.
[18] E. Karapistoli and F.N. Pavlidou, “An overview of the IEEE 802.15.4a standard,” in Proc. IEEE Communication Magazine, pp.47-53, 2010.
[19] D.Z. Du and F.K. Hwang, “Generalized de Bruijn digraphs,” in Proc. IEEE Networks, pp.27-28, 1988.
[20] M. Imase and M. Itoh, “Design to minimize diameter on building-block network,” in Proc. IEEE Trans. Computer, pp.439-442, 1981.
[21] G. Liu and K.Y. Lee, “Optimal Routing Algorithms for Generalized de Bruijn Digraphs,” in Proc. IEEE International Conference on Parallel Processing, pp.167-174, 1993.
[22] D. Eppstein, “Finding the K Shortest Paths,” in Proc. Foundations of Computer Science, pp.154-165, 1994.
[23] J. S. Deogun, T. Kloks, D. Kratsch and H. Muller, “On Vertex Ranking for Permutation and other Graphs,” in Proc. Lecture Notes in Computer Science, pp.747-758, 1994.
[24] K. Benkic, M. Malajner, P. Planinsic and Z. Cucej, “Using RSSI value for distance estimation in Wireless networks based on Zigbee,” in Proc. Systems, Signals and Image Processing, pp. 303-306, 2008.
[25] H. A. Latchman and R. Newman, “Lunch Speech II. HomePlug Standards for Worldwide Multimedia In-Home Networking and Broadband Powerline Access,” in Proc. Power Line Communications and its Applications(ISPLC), pp. 21-22, 2007.
[26] K. Amodt, “CC2431 Location Engine,” in Proc. Texas Instruments Application Note AN042, pp. 1-20, 2006.
[27] C. Y. Hsu, J. L. Wu and S. T. Wang, “Capacity Upgrading in Mobile Ad Hoc Access Networks(MAHANs) Using CSMA/CAPA,” in Proc. ACM Workshop on Mobility Management and Wireless Access(MobiWac), pp. 28-34, 2006.