臺灣博碩士論文加值系統

毫米波已被視為下世代 (5G) 行動通訊系統的關鍵使用頻段。因此，在5G系統設計之初，理當針對毫米波之傳播通道特性進行徹底的研究。
在本論文的研究中，通道的量測是在發射機端使用全向性的天線，在接收機端使用指向性的號角天線 (horn antenna)，藉以幫助提高測量範圍。透過偽隨機序列 (PN sequence) 的傳送，在接收端執行互相關檢測以獲取多路徑延遲的資訊。不同於低頻帶，用於6 GHz以上之頻帶的量測的天線配置 (antenna configuration)會與實際上的操作設備不同。而量測的結果是測量天線的天線場型 (radiation pattern) 與傳播通道的捲積。因此，重建真實的通道模型需要設計一種有效的演算法，來將量測結果與天線場型分離。在本研究中，路徑延遲的資訊是透過PN碼探測 (PN-code sounding)預先得知，而基因演算法 (genetic algorithm) 是用以提取出多路徑的抵達角及接收功率的資訊。
本論文提出了一種簡單的方法來為基因演算法找到一個很好的初始群體 (initial population)。數值結果顯示，與隨機初始群體相比之下，基因演算法搭配所尋找之初始群體，可以大幅提高5G通道量測之準確性。再者，即使在號角天線之大的量測角度 (measurement angle) 與寬的3-dB波束之下，本論文提出之基因演算法亦能精準地估算通道的特性，這意味著可以顯著地減少通道量測時間，並且使用較小尺寸之號角天線來達到量測之易行性。

The millimeter-wave frequency band has been considered as one of the key components for next generation (5G) mobile cellular communications. While these high frequency bands offer vast swaths of underutilized spectrum, the propagation channel characterization needs to be thoroughly scrutinized prior to system design at these bands.
In this thesis, channel measurement is conducted by using an omni-directional antenna at transmitter and a directional horn antenna at receiver to help improve the range of measurements. With transmission of pseudo noise (PN) sequence, the cross-correlation is performed to acquire the multipath delays at receiver. Unlike the low frequency bands, the antenna configuration used for measurements above 6 GHz is actually not the same as used for the operating equipment. The measurements are the result of the convolution of the measurement radiation patterns and the propaga-tion channel. Thus reconstructing the true channel model needs to devise an effective algorithm to decouple the measurements from the measurement radiation patterns.
In this thesis, the path delays are known in advance by PN-code sounding and the genetic algorithm (GA) is utilized to extract the channel characteristics in direction of arrival and received path power for the multipath components. We propose a simple method to find a good initial population for the GA algorithm.
Numerical results show that the GA algorithm with the proposed initial popula-tion can greatly improve the accuracy of 5G channel measurements in comparison with that with random initial population.
Moreover, the proposed GA can accurately estimate the channel characteristics even for large measurement angle and 3-dB beamwidth of horn antenna, which can significantly reduce the time consumption of channel measurement and enable the use of smaller size horn antenna, respectively.

致謝辭 I
摘要 II
Abstract III
目錄 V
圖目錄 VII
表目錄 XI
第1章 第一章 緒論 1
1.1 研究背景 1
1.2 研究動機 2
1.3 論文架構 5
第2章 第二章 量測配置及天線場型介紹 6
2.1 量測配置問題描述 6
2.2 號角天線場型(horn antenna pattern) 7
第3章 第三章 基因演算法(GA) 12
3.1 定義基本運算因子及變數編碼(encoding) 14
3.2 產生初始群體(initial population) 14
3.3 適存度函數(fitness evaluation) 15
3.4 基因演化(genetic evolutions) 15
3.4.1 擇優(mate selection) 15
3.4.2 交換(crossover) 16
3.4.3 突變(mutation) 18
第4章 第四章 設計適合本論文之基因演算法及效能評估 19
4.1 定義訊號模型(signal model) 20
4.2 定義適存度函數(fitness evaluation) 24
4.3 定義基本運算因子與變數編碼方式(encoding) 25
4.4 產生初始群體新方法(new method of initial population) 27
4.4.1 估算路徑個數第一階段及記錄初始值 28
4.4.2 估算路徑個數第二階段－適合度函數值之判定 33
4.5 改良式基因演化 35
4.5.1 擇優(mate selection) 35
4.5.2 交換(crossover) 36
4.5.3 突變(mutation) 36
4.6 基因演算法系統參數設定 38
4.7 基因演算法之重要性 46
第5章 第五章 GA模擬結果分析 47
5.1 模擬結果分析(初始值據估算誤差) 47
5.2 模擬結果分析(初始值估算正確) 49
5.2.1 ，比較不同天線量測角度 49
5.2.2 ，比較不同天線之 51
5.2.3 比較不同波束寬量測並使用基因演算法之極限 54
5.2.3.1 之極限 55
5.2.3.2 之極限 58
5.2.3.3 之極限 61
5.2.4 統計不同波束寬之極限 65
第6章 第六章 總結 67
第7章 第七章 參考文獻 68

[1]E. Dahlman, S. Parkvall, and J. Sköld, 4G LTE/LTE-Advanced for Mobile Broadband. Published by Elsevier Ltd, 2011.
[2]METIS2020, Deliverable D1.4 v3, “METIS Channel Model,” July 2015.
https://www.metis2020.com/wp-content/uploads/METIS_D1.4_v3.pdf
[3]http://www.cost2100.org/
[4]ETSI, “New ETSI Group on Millimetre Wave Transmission starts work,”
http://www.etsi.org/news-events/news/866-2015-01-press-new-etsi-group-on-millimetre-wave-transmission-starts-work
[5]http://www.nist.gov/ctl/wireless-networks/5gmillimeterwavechannelmodel.cfm
[6]MiWEBA, Deliverable D5.1 “Channel modeling and characterization,” June, 2014, http://www.miweba.eu/wp-content/uploads/2014/07/MiWEBA_D5.1_v1.011.pdf.
[7]https://5g-ppp.eu/mmmagic/
[8]T. S. Rappaport, et. al.,” Wideband Millimeter-Wave Propagation Measurements and Channel Models for Future Wireless Communication System Design,” IEEE Trans. Communications, Vol. 63, No. 9, pp.3029-3056, Sept. 2015.
[9]G. MacCartney, et. al, “Indoor Office Wideband Millimeter-Wave Propagation Meas-urements and Channel Models at 28 GHz and 73 GHz for Ultra-Dense 5G Wireless Networks,” IEEE Access, October 2015.
[10]T. S. Rappaport et al., "Millimeter Wave Mobile Communications for 5G Cellular: It Will Work!," in IEEE Access, vol.1, pp. 335-349, 2013.
[11]M. K. Samimi, T. S. Rappaport, and G. R. MacCartney, Jr., “Probabilistic omnidirection-al path loss models for millimeter-wave outdoor communications,” in IEEE Wireless Communications Letters, vol. 4, no. 4, pp. 357-360, Aug. 2015.
[12]NTT DoCoMo et al., White Paper, 5G Channel Model for bands up to100 GH, 2nd Workshop on Mobile Communications in Higher Frequency Bands (MCHFB) in con-junction with IEEE Globecom’15, 2015.
[13]RPa160012, Report of RAN email discussion "[A.Post-RAN#70-01] Scenarios and Re-quirements for Next Generation Access Technologies," 3GPP TSG RAN ad hoc, Barce-lona, Spain, January, 2016.
[14]3GPP TR 38.913 V0.1.1, Study on Scenarios and Requirements for Next Generation Access Technologies, January 2016-01.
[15]J. H. Holland, Adaptation in Natural and Artificial Systems. Ann Arbor, MI, USA: Univ. of Michigan Press, 1975.
[16]D. E. Goldberg, Genetic Algorithms in Search, Optimization and Machine Learning. Reading, MA,USA: Addison-Wesley, 1989.
[17]http://www.antenna-theory.com/antennas/aperture/horn.php#horn
[18]P. Karamalis, A. Marousis, A. Kanatas, P. Constantinou, “Direction of Arrival Estima-tion Using Genetic Algorithms,” IEEE VTS 53rd Vehicular Technology Conference. Pro-ceedings(Cat. No.01CH37202), vol.1, pp.162-166, Spring 2001.
[19]http://wshnt.kuas.edu.tw/network/s8/new_page_1.htm