由於無線通訊技術的成熟與人們對網路需求量的增加，下一代行動通訊技術5G預期將要比現今的4G行動通訊技術有更高的資料量、更低的延遲、更低的成本與涵蓋更多的使用者，因此，如物聯網或觸覺互聯網(tactile Internet)等新興技術得以蓬勃發展。然而，越多的服務將提高系統的複雜度，與多個簡單的同質網路(homogeneous network)相比，不同的服務或協議組成的異質網路(heterogeneous network)提供更高的靈活性及成本效益，因此，近年來，異質網路成為增加資料容量與覆蓋範圍的方案之一。在本論文中，以偏振態多工(polarization division multiplexing, PDM)與密集分波多工（dense wavelength division multiplexing, DWDM）來實現多服務的毫米波異質網路。在分波多工中，相同的波長通道內可再透過兩個正交偏振態傳輸不同的信號。因此，數據容量增加了一倍。但是，偏振態解多工技術通常採用複雜的技術，例如偏振態追蹤(polarization tracking)或多輸入多輸出（multi-input multi-output, MIMO）偏振態估計，這些技術需要訓練來自兩種偏振態的訓練數據進行解多工，這降低了系統的靈活性也增加了接收端的硬件成本。基於實驗至之前的研究，提出了偏振態無感的PDM/DWDM整合系統，並使用基於人工神經網路(artificial neural network, ANN)的等化器改善其表現。實驗分別在back-to-back、25公里單模光纖與3米無線通道下傳輸，比較了PDM、DWDM、PDM/DWDM整合系統的位元錯誤率(bit error rate, BER)的表現，驗證了此PDM/DWDM光纖無線整合系統在未來多服務異質網路架構下的可行性。

Due to the maturity of wireless communication technology and the increasing demands of network, the next generation mobile communication technology, 5G is expected to have higher data capacity, higher coverage, lower latency and lower cost than 4G mobile communication technology. Therefore, emerging technologies such as Internet of Things or tactile Internet have evolved. However, multi-services will increase the complexity of the system. Compare to multiple simple homogeneous networks, heterogeneous network that consists of different services or protocols provide greater flexibility and cost efficiency. Therefore, in the recent years, heterogeneous network (HetNet) has emerged as a new trend to enhance the data capacity and the coverage. A HetNet is a network consisting of different services or protocols. In this paper, we have implemented polarization division multiplexing (PDM) and dense wavelength division multiplexing (DWDM) to realize a multi-service mm-wave heterogeneous network. In a DWDM system, different PDM signals can be transmitted through two orthogonal polarization states in the same wavelength channel. Therefore, data capacity has doubled. However, PDM demultiplexing technologies typically employ laborious technologies, such as polarization tracking or multi-input multi-output (MIMO) polarization estimation. These demultiplexing techniques require training sequences from two polarization states, which reduces the flexibility of the system and increases the hardware cost of the receiver. Based on our previous researches, we proposed a polarization-insensitive PDM/DWDM system with neural network-based equalizer. The experiments transmitted over back-to-back, 25km single-mode fiber and 3m wireless link, and the performances of PDM, DWDM and PDM/DWDM systems are discussed. Results shows that the proposed system could possibly be used in future multi-service heterogeneous network.

[1] E. Ezhilarasan and M. Dinakaran, "A Review on Mobile Technologies: 3G, 4G and 5G," 2017 Second International Conference on Recent Trends and Challenges in Computational Models (ICRTCCM), Tindivanam, 2017, pp. 369-373.
[2] P. Pirinen, "A brief overview of 5G research activities," 1st International Conference on 5G for Ubiquitous Connectivity, Akaslompolo, 2014, pp. 17-22.
[3] A. Damnjanovic et al., "A survey on 3GPP heterogeneous networks," in IEEE Wireless Communications, vol. 18, no. 3, pp. 10-21, June 2011.
[4] S. Sun, K. Adachi, P. H. Tan, Y. Zhou, J. Joung and C. K. Ho, "Heterogeneous network: An evolutionary path to 5G," 2015 21st Asia-Pacific Conference on Communications (APCC), Kyoto, 2015, pp. 174-178.
[5] S. J. B. Yoo, "Wavelength conversion technologies for WDM network applications," in Journal of Lightwave Technology, vol. 14, no. 6, pp. 955-966, June 1996.
[6] L. Cheng et al., "8×200-Gbit/s polarization-division multiplexed CS-RZ-DQPSK transmission over 1200 km of SSMF," 2009 14th OptoElectronics and Communications Conference, Vienna, 2009, pp. 1-2.
[7] X. Steve Yao, L. Yan, B. Zhang, A. Willner, and J. Jiang, "All-optic scheme for automatic polarization division demultiplexing," Opt. Express 15, 7407-7414 (2007).
[8] Q. Yang, Y. Ma, and W. Shieh, "107 Gb/s Coherent Optical OFDM Reception Using Orthogonal Band Multiplexing," in Optical Fiber Communication Conference/National Fiber Optic Engineers Conference, OSA Technical Digest (CD) (Optical Society of America, 2008), paper PDP7.
[9] Chia-Chien Wei, Chun-Ting Lin, and Chih-Yun Wang, "PMD tolerant direct-detection polarization division multiplexed OFDM systems with MIMO processing," Optics Express, vol. 20, pp. 7316-7322, 2012.
[10] J. Yan, M. Xu, H. Huang, M. Liao, K. Feng and G. Chang, "A PDM based Spectral Aggregation and Cell Densification for 5G Point-to-Multipoint Mobile Fronthaul with a Polarization-tracking-free RAU Design," 2018 Optical Fiber Communications Conference and Exposition (OFC), San Diego, CA, 2018, pp. 1-3.
[11] T. Hwang, C. Yang, G. Wu, S. Li and G. Ye Li, "OFDM and Its Wireless Applications: A Survey," in IEEE Transactions on Vehicular Technology, vol. 58, no. 4, pp. 1673-1694, May 2009.
[12] Calvin C.K. Chan, Optical Performance Monitoring. Academic Press, 2010, ch.13.
[13] M. I. Hayee, M. C. Cardakli, A. B. Sahin and A. E. Willner, "Doubling of bandwidth utilization using two orthogonal polarizations and power unbalancing in a polarization-division-multiplexing scheme," in IEEE Photonics Technology Letters, vol. 13, no. 8, pp. 881-883, Aug. 2001.
[14] N. Antoniades, G. Ellinas, I. Roudas, WDM Systems and Networks: Modeling, Simulation, Design and Engineering. Berlin, Germany : Springer Science & Business Media, 2011, ch.9.
[15] R. Radziwilowicz, H. Teimoori, S. A. Paredes, V. Tolstikhin, K. Hinzer and T. Hall, "Power budget considerations in CWDM/TDM based Passive Optical Networks," 2008 1st Microsystems and Nanoelectronics Research Conference, Ottawa, Ont., 2008, pp. 173-176.
[16] SURFIBER. (2017, Jun, 23). Difference between WDM, CWDM, DWDM and CCWDM modules [Online]. Available: https://www.surfiber.com/TechnologyZone?article_id=25
[17] H. Chen, "Enhanced Independence among NOMA users through Machine Learning Classification for 5G Downlink IFoF Network", Master Thesis, Nat'l Tsing Hua Univ., Taiwan, 2018. Available: https://140.113.39.130/cgi-bin/gs32/gsweb.cgi?o=dallcdr&s=id=%22G021050665330%22.&searchmode=basic
[18] S. McCartney, ENIAC: The triumphs and tragedies of the world's first computer. Walker & Company, 1999.
[19] Hecht-Nielsen, "Theory of the backpropagation neural network," International 1989 Joint Conference on Neural Networks, Washington, DC, USA, 1989, pp. 593-605 vol.1.
[20] I. c. s. I. am. (2018). Gradient Descent. Available: https://ithelp.ithome.com.tw/articles/10198147
[21] M. A. Jarajreh et al., "Artificial Neural Network Nonlinear Equalizer for Coherent Optical OFDM," in IEEE Photonics Technology Letters, vol. 27, no. 4, pp. 387-390, 15 Feb.15, 2015.
[22] E. Giacoumidis et al., "Intra and inter-channel nonlinearity compensation in WDM coherent optical OFDM using artificial neural network based nonlinear equalization," 2017 Optical Fiber Communications Conference and Exhibition (OFC), Los Angeles, CA, 2017, pp. 1-3.
[23] E. Giacoumidis, S. Le, M. Ghanbarisabagh, M. McCarthy, I. Aldaya, S. Mhatli, M. Jarajreh, P. Haigh, N. Doran, A. Ellis, and B. Eggleton, "Fiber nonlinearity-induced penalty reduction in CO-OFDM by ANN-based nonlinear equalization," Opt. Lett. 40, 5113-5116 (2015).
[24] H. Al-Raweshidy and S. Komaki, Radio over fiber technologies for mobile communications networks. Artech House, 2002.
[25] C. Lim et al., "Fiber-Wireless Networks and Subsystem Technologies," in Journal of Lightwave Technology, vol. 28, no. 4, pp. 390-405, Feb.15, 2010.
[26] U. Gliese, S. Norskov and T. N. Nielsen, "Chromatic dispersion in fiber-optic microwave and millimeter-wave links," in IEEE Transactions on Microwave Theory and Techniques, vol. 44, no. 10, pp. 1716-1724, Oct. 1996.