分散式偵測系統是利用空間多樣性的概念，由多個本地偵測器及一個資料融合(收集)中心所組成。基於通訊傳輸的限制，每個本地偵測器需具備信號處理的功能，對信號(目標)存在或不存在，先做初步的判斷後再傳至資料融合中心，融合中心再根據融合法則做最後的總判斷。然而考量本地判斷在無線傳輸過程中，會受到雷利衰減通道及高斯雜訊的影響，若融合法則沒有經過適當的設計，將導致融合中心判斷錯誤的機率提高。文獻所提的概似比融合法則有最佳的系統偵測效能，但所需事前資訊量最多，包含了本地偵測器的效能參數及傳輸通道的即時資訊，且從融合統計量來看計算複雜度最高。另外，低複雜度線性結合技術(如最大訊雜比和等增益法則)，雖然減少了事前資訊量但偵測效能亦大幅降低。
為了能在低複雜度下提高偵測效能，在本篇論文中我們利用線性結合技術，將融合統計量的偏斜係數最大化，以求得一組結合權重來做為我們的融合法則。最大偏斜係數法則的結合權重同時考慮了本地偵測器的效能參數及傳輸通道的即時資訊。模擬結果顯示，最大偏斜係數融合法則在低訊雜比時可近似最佳概似比法則，且在大部分訊雜比下，與最大訊雜比法則和等增益法則相比，有較強健的偵測效能。

A distributed detection system is composed of multiple local sensors and a fusion center. Due to communication constraints, each local sensor is equipped with limited data processing capabilities to make preliminary decisions about the hypotheses under test, and transmits them to the fusion center, which then makes a global decision. When transmitted over wireless channels, local decisions are subjected to fading and noise perturbation, which could cause severe degradation in the overall detection performance if the fusion rule is not appropriately designed.
In the literature, likelihood ratio (LR)-based fusion rule has the optimal detection performance, but it requires the maximum amount of a priori information (e.g., channel state information and local sensor performance indices). LR also needs a higher complexity for computing the fusion statistic. On the other hand, linear combining schemes have a lower complexity at the cost of a degraded detection performance, because a reduced amount of a priori information is used in fusion.
In order to enhance the detection performance while keeping the complexity low, we consider a linear-combining fusion rule whereby the combining weights are obtained by maximizing the deflection coefficient of the fusion statistic. The resulting weights depend on both the local sensor performance indices and complete channel state information. The simulation results show that deflection coefficient maximization (DCM) scheme is near optimal at low signal-to-noise ratios (SNRs), and has a better detection performance over a wide range of SNRs, compared to linear combining schemes such as maximum ratio combining (MRC) and equal gain combining (EGC).

[1] R. Viswanathan and P. K. Varshney, “Distributed detection with multiple sensors: Part I – fundamentals,” Proceedings of the IEEE, vol. 85, pp. 54 – 63, January 1997.

[2] P. K. Varshney, Distributed Detection and Data Fusion. New York:Springer,1997.

[3] C. Cordeiro, K. Challapali, D. Birru, and S. S. Nandagopalan, “IEEE 802.22: An introduction to the first wireless standard based on cognitive radios,” Journal of Communications, vol. 1, pp. 38 – 47, April 2006.

[4] F. Gini, F. Lombardini, and L. Verrazzani, “Decentralised detection stratigies under communication constraints,” Proc. Inst. Elect. Eng., F: Radar, Sonar,Navigat., vol. 145, pp. 199–208, Aug. 1998.

[5] Theodore S. Rappaport, Wireless Communications – Principles & Practice,Prentice Hall Inc., 2002.

[6] B. Chen, L. Tong, and P. K. Varshney, “Channel-aware distributed detection in wireless sensor networks,” IEEE Signal Processing Magazine, vol. 23, pp. 16 – 26,July 2006.

[7] J.-F. Chamberland and V. V. Veeravalli, “Wireless sensors in distributed detection applications,” IEEE Signal ProcessingMagazine, vol. 24, pp. 16 – 25, May 2007.

[8] B. Chen, R. Jiang, T. Kasetkasem, and P. K. Varshney, “Channel aware decision fusion in wireless sensor networks,” IEEE Trans. on Signal Processing, vol. 52,pp. 3454– 3458, December 2004.

[9] Z. Chair and P. K. Varshney, “Optimal data fusion in multiple sensor detection systems,” IEEE Trans. on Aerospace and Electronic Systems, vol. 22, pp. 98–101,January 1986.

[10] S. M. Kay, Fundamentals of Statistical Signal Processing: Detection Theory.Prentice-Hall, Inc., 1998.

[11] B. Picinbono, “On deflection as a performance criterion in detection,” IEEE Trans. Aerosp. Electron. Syst., vol. 31, pp. 1072–1081, July 1995.

[12] A. H. Sayed, Fundamentals of Adaptive Filtering. NJ: Wiley, 2003.

[13] Z. Quan, S. Cui, and A. H. Sayed, “Optimal linear cooperation for spectrum ensing in cognitive radio networks,” IEEE Journal of Selected Topics in Signal Processing, vol. 2, pp. 28 – 40, February 2008.

[14] H. L. Van Trees, Detection, Estimation, and Modulation Theory, Part I. New York: Wiley, 1968.