隨著時代的變遷與科技的進展，無線傳輸在自動化與控制工程的應用日益受到重視。許多基於無線傳輸的消費性電子產品，其傳輸訊息的過程可能受到許多不同大小和頻率的雜訊干擾。此雜訊甚至可能是脈衝型的雜訊。這樣的脈衝型雜訊會讓自適應濾波器在進行系統識別時得到較差的收斂結果，甚至會造成發散的現象。既有的自適應濾波演算法可以透過可變步長演算法(或是更進一步的組成步階長度演算法)，搭配著符號函數的計算方式，來降低脈衝雜訊之影響。然而，當輸入訊號以及整體系統變為複雜時，收斂速度以及穩態誤差的表現仍需要加以改善。為了克服上述方式的缺點，在本研究中我們提出了一類新型自適應濾波演算法的構想。透過結合增強式機器學習以及組合步長的自適性設計概念，使其不僅對於各式雜訊的兼容程度可以提升，即使當環境有所改變時，也能夠較快且較佳地收斂。透過結合增強式濾波器(Boosted Adaptive Filter)中所運用到的機器學習之集成學習(Ensemble Learning)概念，將多個弱濾波器結合來得到一個強濾波器。然後，以組成步長的方法(Combined-Step-Size)來優化各個弱濾波器之間的組成權重。更進一步，進行脈衝偵測，再以此結果來執行分段更新(Piecewise Update)。其中，分段更新結合了：(1) 仿射投影符號演算法(Affine Projection Sign Algorithm, APSA)以抑制脈衝雜訊；(2) 針對一般雜訊的最小均方演算法(Least Mean Square, LMS)來降低運算複雜度。增強式架構中的重要參數以及函式都需針對具有脈衝雜訊的情況來進行重新設計；此外，也針對組合步長的公式重新設計推導，將兩個步長延伸至多個步長。模擬結果顯示：本研究所提出的方法可以在具有脈衝雜訊環境下成功提高系統識別的性能。即使在真實系統具有多變化的情況下，估測系統與真實系統間的誤差也能更快速地收斂，且維持低穩態誤差。

With the changes of the times as well as the progress of science and technology, the application of wireless transmission to automation and control engineering has attracted more and more attention. Many consumer electronic products based on wireless transmission may be interfered by noises of different sizes and frequencies in the process of transmitting messages. This noise may even be pulse-shaped noise. Such impulsive noise will cause the adaptive filter to obtain poor convergence results in system identification, and even cause divergence. The existing adaptive filtering algorithm can reduce the influence of impulse noise through variable step-size algorithm (or further combinational step-size algorithm), combined with the calculation of sign function, to reduce the impact of impulse noise. However, when the input signals and entire systems are complicated, the convergence speed and steady-state error performance still need to be improved. In order to overcome the shortcomings of the above methods, in this research we propose a new adaptive filtering algorithm concept. By combining the enhanced machine learning and the concept of combinational step-size, it not only can not improve the compatibility corresponding to various types of noise, but also can converge faster and better even when the environment changes. By combining the Ensemble Learning concept of Machine Learning used in Boosted Adaptive Filter, multiple weak filters are combined to obtain a strong filter. The Combined-Step-Size method is then used to optimize the weight of the composition of each weak filter. Furthermore, pulse detection is performed, and then the Piecewise Update is performed based on the detection result. Therein, the Piecewise Update combines: (1) Affine Projection Sign Algorithm (APSA) to suppress impulse noise, and (2) Least Mean Square (LMS) algorithm for general noise to reduce computational complexity. The important parameters and functions in the Boosted Architecture need to be redesigned for the situation with impulse noise; in addition, the formula for the Combined-Step-Size is redesigned and deduced, and the two step lengths are extended to multiple step lengths. The simulation results show that the method proposed in this research can successfully improve the performance of system identification in an environment with impulse noise. Even when the real system is changeable, the error between the estimated system and the real system can converge faster, and the low steady-state error still can be maintained.

[1] J. Lin, W. Yu, N. Zhang, X. Yang, H. Zhang, and W. Zhao, “A survey on internet of things: architecture, enabling technologies, security and privacy, and applications,” IEEE Internet of Things Journal, vol. 4, no. 5, pp. 1125-1142, Oct. 2017, DOI: 10.1109/JIOT.2017.2683200.
[2] H. Lee and K. Ke, “Monitoring of large-area IoT sensors using a LoRa wireless mesh network system: design and evaluation,“ IEEE Transactions on Instrumentation and Measurement, vol. 67, no. 9, pp. 2177-2187, Sept. 2018, DOI: 10.1109/TIM.2018.2814082.
[3] A. K. Gupta and R. Johari, “IOT based electrical device surveillance and control system,” in Proc. International Conference on Internet of Things: Smart Innovation and Usages, Ghaziabad, India, April 18-19, 2019, pp. 1-5, DOI: 10.1109/IoT-SIU.2019.8777342.
[4] M. M. Rana and W. Xiang, “IoT communications network for wireless power transfer system state estimation and stabilization,” IEEE Internet of Things Journal, vol. 5, no. 5, pp. 4142-4150, Oct. 2018, DOI: 10.1109/JIOT.2018.2852003.
[5] Y. Yao, Z. Zhu, S. Huang, X. Yue, C. Pan, and X. Li, “Energy efficiency characterization in heterogeneous IoT system with UAV swarms based on wireless power transfer,” IEEE Access, vol. 8, pp. 967-979, Dec. 2020, DOI: 10.1109/ACCESS.2019.2961977.
[6] S. A. Yadav, S. Sharma, L. Das, S. Gupta, and S. Vashisht, “An effective IoT empowered real-time gas detection system for wireless sensor networks,” in Proc. International Conference on Innovative Practices in Technology and Management, Feb. 17-19, 2021, pp. 44-49, DOI: 10.1109/ICIPTM52218.2021.9388365.
[7] S. Sharma, V. Bhatia, and A. K. Mishra, “Wireless consumer electronic devices: The effects of impulsive radio-frequency interference,” IEEE Consumer Electronics Magazine, vol. 8, no. 4, pp. 56-61, July 2019, DOI: 10.1109/MCE.2019.2905538.
[8] J. Cabra, D. Castro, J. Colorado, D. Mendez, and L. Trujillo, “An IoT approach for wireless sensor networks applied to e-health environmental monitoring,” in Proc. IEEE International Conference on Internet of Things and IEEE Green Computing and Communications and IEEE Cyber, Physical and Social Computing and IEEE Smart Data, Exeter, UK, June 21-23, 2017, pp. 578-583, DOI: 10.1109/iThings-GreenCom-CPSCom-SmartData.2017.91.
[9] C. M. Parmar, P. Gupta, K. S. Bharadwaj, and S. S. Belur, “Smart work-assisting gear,” in Proc. IEEE World Forum on Internet of Things, Singapore, Singapore, Feb. 5-8, 2018, pp. 724-728, DOI: 10.1109/WF-IoT.2018.8355176.
[10] D. Manyvone, R. Takitoge, and K. Ishibashi, “Wireless and low-power water quality monitoring beat sensors for agri and acqua-culture IoT applications,” in Proc. International Conference on Electrical Engineering/Electronics, Computer, Telecommunications and Information Technology, Chiang Rai, Thailand, July 18-21, 2018, pp. 122-125, DOI: 10.1109/ECTICon.2018.8620024.
[11] H. Li, L. Liu, Y. Li, Z. Yuan, and K. Zhang, “Measurement and characterization of electromagnetic noise in edge computing networks for the industrial internet of things,” Sensors, vol. 19, no. 14, pp. 3104, July 2019, DOI: 10.3390/s19143104.
[12] M. Ballerini, T. Polonelli, D. Brunelli, M. Magno, and L. Benini, “NB-IoT versus LoRaWAN: An experimental evaluation for industrial applications,” IEEE Transactions on Industrial Informatics, vol. 16, no. 12, pp. 7802-7811, Dec. 2020, DOI: 10.1109/TII.2020.2987423.
[13] K. Banse, L. F. Herrera, M. Nunez, J. C. Navarro, D. Bermudez, and J. Chavarriaga, “ITS development in Colombia: Challenges and opportunities,” in Proc. ICAI Workshops, Bogota, Colombia, Nov. 1-3, 2018, pp. 1-6, DOI: 10.1109/ICAIW.2018.8554991.
[14] C. Wang, J. Zhang, L. Xu, L. Li, and B. Ran, “A new solution for freeway congestion: cooperative speed limit control using distributed reinforcement learning,” IEEE Access, vol. 7, pp. 41947-41957, March 2019, DOI: 10.1109/ACCESS.2019.2904619.
[15] C. Chalkiadakis, P. Iordanopoulos, F. Malin, K. Helfert, M. Zangl, M. Flachi, and R. Öörni, “Capacity building strategies for further growth of the ITS sector in europe,” in Proc. International Conference on Models and Technologies for Intelligent Transportation Systems, Cracow, Poland, June 5-7, 2019, pp. 1-8, DOI: 10.1109/MTITS.2019.8883287.
[16] X. Xu, Y. Liu, W. Wang, X. Zhao, Q. Z. Sheng, Z. Wang, and B. Shi, “ITS-frame: a framework for multi-aspect analysis in the field of intelligent transportation systems,” IEEE Transactions on Intelligent Transportation Systems, vol. 20, no. 8, pp. 2893-2902, Aug. 2019, DOI: 10.1109/TITS.2018.2868840.
[17] S. Maaloul, H. Aniss, M. Kassab, and M. Berbineau, “Classification of C-ITS services in vehicular environments,” IEEE Access, vol. 9, pp. 117868-117879, Aug. 2021, DOI: 10.1109/ACCESS.2021.3105815.
[18] D.-Z. Feng, Z. Bao, and L.-C. Jiao, “Total least mean squares algorithm,” IEEE Transactions on Signal Processing, vol. 46, no. 8, pp. 2122-2130, Aug. 1998, DOI: 10.1109/78.705421.
[19] D. P. Mandic, “A generalized normalized gradient descent algorithm,” IEEE Signal Processing Letters, vol. 11, no. 2, pp. 115-118, Feb. 2004, DOI: 10.1109/LSP.2003.821649.
[20] Y. Engel, S. Mannor, and R. Meir, “The kernel recursive least-squares algorithm,” IEEE Transactions on Signal Processing, vol. 52, no. 8, pp. 2275-2285, Aug. 2004, DOI: 10.1109/TSP.2004.830985.
[21] T. Shao, Y. R. Zheng, and J. Benesty, “An affine projection sign algorithm robust against impulsive interferences,” IEEE Signal Processing Letters, vol. 17, no. 4, pp. 327-330, April 2010, DOI: 10.1109/LSP.2010.2040203.
[22] F. Huang, J. Zhang, and S. Zhang, “Combined-step-size affine projection sign algorithm for robust adaptive filtering in impulsive interference environments,” IEEE Transactions on Circuits and Systems II: Express Briefs, vol. 63, no. 5, pp. 493-497, May 2016, DOI: 10.1109/TCSII.2015.2505067.
[23] P. L. D. Santos, M. T. Freigoun, C. A. Martin, D. E. Rivera, E. B. Hekier, R. A. Romano, and T. P. A. Perdicoúlis, “System identification of just walk: using matchable-observable linear parametrizations,” IEEE Transactions on Control Systems Technology, vol. 28, no. 1, pp. 264-275, Jan. 2020, DOI: 10.1109/TCST.2018.2884833.
[24] F. Boeren, A. Lanzon, and T. Oomen, “Iterative identification and control using non-normalized coprime factors with application in wafer stage motion control,” IEEE Transactions on Control Systems Technology, vol. 28, no. 2, pp. 413-424, March 2020, DOI: 10.1109/TCST.2018.2877680.
[25] A. Ayyad, M. Chehadeh, M. I. Awad, and Y. Zweiri, “Real-time system identification using deep learning for linear processes with application to unmanned aerial vehicles,” IEEE Access, vol. 8, pp. 122539-122553, July 2020, DOI: 10.1109/ACCESS.2020.3006277.
[26] S. Haykin, “Adaptive filter theory,” Prentice-Hall, 5th edition, Englewood Cliffs, NJ: Prentice-Hall, 2012.
[27] J. J. Jeong, K. Koo, G. T. Choi, and S. W. Kim, “A variable step size for normalized subband adaptive filters,” IEEE Signal Processing Letters, vol. 19, no. 12, pp. 906-909, Dec. 2012, DOI: 10.1109/LSP.2012.2226153.
[28] W. M. Elsayed, H. M. El-Bakry, and S. M. El-Sayed, “Data reduction using integrated adaptive filters for energy-efficient in the clusters of wireless sensor networks,” IEEE Embedded Systems Letters, vol. 11, no. 4, pp. 119-122, Dec. 2019, DOI: 10.1109/LES.2019.2902404.
[29] Y.-R. Chien, S.-D. Xu, and S. Lu, “Cyclostationary impulsive noise mitigation algorithm for narrowband powerline communications,” Journal of the Franklin Institute, vol. 357, no. 1, pp. 687-703, Jan. 2020, DOI: 10.1016/j.jfranklin.2019.10.026.
[30] Y. Liu, T. Dillon, W. Yu, W. Rahayu, and F. Mostafa, “Noise removal in the presence of significant anomalies for industrial IoT sensor data in manufacturing,” IEEE Internet of Things Journal, vol. 7, no. 8, pp. 7084-7096, Aug. 2020, DOI: 10.1109/JIOT.2020.2981476.
[31] S. A. Bhatti, Q. Shan, R. Atkinson, M. Vieira, and I. A. Glover, “Vulnerability of ZigBee to impulsive noise in electricity substations,” URSI General Assembly and Scientific Symposium, Istanbul, Turkey, Aug. 13-20, 2011, pp. 1-4, DOI: 10.1109/URSIGASS.2011.6050729.
[32] J. Jia and J. Meng, “A novel approach for impulsive noise mitigation in ZigBee communication system,” in Proc. Global Information Infrastructure and Networking Symposium, Montreal, Canada, Sept. 15-19, 2014, pp. 1-3, DOI: 10.1109/GIIS.2014.6934265.
[33] Y. Cheng, Z. Tan, and X. Wu, “Measurement and modelling of burst impulse noise in power line communication,” in Proc. International Conference on Information Technology in Medicine and Education, Qingdao, China, Aug. 23-25, 2019, pp. 688-692, DOI: 10.1109/ITME.2019.00158.
[34] J. J. Jeong and S. Kim, “Robust adaptive filter algorithms against impulsive noise,” Circuits, Systems, and Signal Processing, vol. 38, no. 12, pp. 5651-5664, Dec. 2019, DOI: 10.1007/s00034-019-01135-9.
[35] P. H. Sangave and G. P. Jain, “Impulse noise detection and removal by modified boundary discriminative noise detection technique,” in Proc. International Conference on Intelligent Sustainable Systems, Dec. 7-8, 2017, pp. 715-719, DOI: 10.1109/ISS1.2017.8389266.
[36] Y. Zhang, “A multi-feature based noise detector for random-valued impulse noise,” in Proc. IEEE International Conference on Artificial Intelligence and Computer Applications, June 27-29, 2020, pp. 1149-1155, DOI: 10.1109/ICAICA50127.2020.9182608.
[37] J. Yoo, J. Shin, and P. Park, “Variable step-size affine projection sign algorithm,” IEEE Transactions on Circuits and Systems II: Express Briefs, vol. 61, no. 4, pp. 274-278, April 2014, DOI: 10.1109/TCSII.2014.2305013.
[38] M. S. E. Abadi, H. Mesgarani, and S. Khademiyan, “Robust variable step-size affine projection sign algorithm against impulsive noises,” Circuits, Systems, and Signal Processing, vol. 39, no. 3, pp. 1471-1488, July 2019, DOI: 10.1007/s00034-019-01209-8.
[39] F. Huang, J. Zhang, and S. Zhang, “Combined-step-size normalized subband adaptive filter with a variable-parametric step-size scaler against impulsive interferences,” IEEE Transactions on Circuits and Systems II: Express Briefs, vol. 65, no. 11, pp. 1803-1807, Nov. 2018, DOI: 10.1109/TCSII.2017.2771430.
[40] G. Li, H. Zhang and J. Zhao, “Generalized correntropy induced metric memory-improved proportionate affine projection sign algorithm and its combination,” IEEE Transactions on Circuits and Systems II: Express Briefs, vol. 67, no. 10, pp. 2239-2243, Oct. 2019, DOI: 10.1109/TCSII.2019.2945783.
[41] P. Song and H. Zhao, “Affine-projection-like M-estimate adaptive filter for robust filtering in impulse noise,” IEEE Transactions on Circuits and Systems II: Express Briefs, vol. 66, no. 12, pp. 2087-2091, Dec. 2019, DOI: 10.1109/TCSII.2019.2897620.
[42] J. J. Jeong, “A robust affine projection algorithm against impulsive noise,” IEEE Signal Processing Letters, vol. 27, pp. 1530-1534, Aug. 2020, DOI: 10.1109/LSP.2020.3018652.
[43] Z. Zheng and Z. Liu, “Steady-state mean-square performance analysis of the affine projection sign algorithm,” IEEE Transactions on Circuits and Systems II: Express Briefs, vol. 67, no. 10, pp. 2244-2248, Oct. 2020, DOI: 10.1109/TCSII.2019.2946782.
[44] J. Kumar, “Comparative performance study of various filters on high density impulse noise,” in Proc. International Conference on Nascent Technologies in Engineering, Navi Mumbai, India, Jan. 27-28, 2017, pp. 1-6, DOI: 10.1109/ICNTE.2017.7947949.
[45] C. Hsieh, P. Huang, and Q. Zhao, “Impulse noise replacement with adaptive neighborhood median filtering,” in Proc. International Conference on Machine Learning and Cybernetics, Chengdu, China, July 15-18, 2018, pp. 491-496, DOI: 10.1109/ICMLC.2018.8527058.
[46] O. M. Shekoni, A. N. Hasan, and T. Shongwe, “Mitigation of impulse noise in powerline systems using ANFIS technique,” in Proc. International Conference on Intelligent and Innovative Computing Applications, Plaine Magnien, Mauritius, Dec. 6-7, 2018, pp. 1-6, DOI: 10.1109/ICONIC.2018.8601270.
[47] D. Kari, A. Mirza, F. Khan, H. Ozkan, and S. Kozat, “Boosted adaptive filters,” Digital Signal Processing, vol. 81, pp. 61-78, July 2018, DOI: 10.1016/j.dsp.2018.07.012.
[48] R. A. Servedio, ‘‘Smooth boosting and learning with malicious noise,’’ Journal of Machine Learing Research, vol. 4, no. 4, pp. 633-648, May 2004, DOI: 10.1162/153244304773936072.
[49] R. Schapire and Y. Freund, “Boosting: foundations and algorithms,” Kybernetes, vol. 42, no. 1, pp. 164-166, Jan. 2013, DOI: 10.1108/03684921311295547.
[50] D. Kari, I. Marivani, I. Delibalta, and S. S. Kozat, “Boosted LMS-based piecewise linear adaptive filters,” in Proc. European Signal Processing Conference, Budapest, Hungary, Aug. 29-Sept. 2, 2016, pp. 1593-1597, DOI: 10.1109/EUSIPCO.2016.7760517.
[51] S. Bharati, P. Podder, and M. R. H. Mondal, ‘‘Diagnosis of polycystic ovary syndrome using machine learning algorithms,’’ in Proc. IEEE Region 10 Symposium, Dhaka, Bangladesh, June 5-7, 2020, pp. 1486-1489, DOI: 10.1109/TENSYMP50017.2020.9230932.
[52] A. Haratiannezhadi, S. Setayeshi, and J. Hatami, ‘‘Boosting model of attention network task,’’ in Proc. Conference on Swarm Intelligence and Evolutionary Computatio, Mashhad, Iran, Sept. 2-4, 2020, pp. 32-36, DOI: 10.1109/CSIEC49655.2020.9237299.
[53] J. Nuhić and J. Kevrić, ‘‘Lung cancer typology classification based on biochemical markers using machine learning techniques,’’ in Proc. International Convention on Information, Communication and Electronic Technology, Opatija, Croatia, Sept. 28-Oct. 2, 2020, pp. 292-297, DOI: 10.23919/MIPRO48935.2020.9245114.
[54] L. F. O. Chamon, W. B. Lopes, and C. G. Lopes, “Combination of adaptive filters with coefficients feedback,” in Proc. IEEE International Conference on Acoustics, Speech and Signal Processing, Kyoto, Japan, March 25-30, 2012, pp. 3785-3788, DOI: 10.1109/ICASSP.2012.6288741.
[55] M. Lee and P. Park, “Combined regularization affine projection sign algorithm against impulsive noises,” in Proc. International Conference on Control, Automation and Systems, Jeju, South Korea, Oct. 15-18, 2019, pp. 418-422, DOI: 10.23919/ICCAS47443.2019.8971702.
[56] Y. Song, J. Lu, H. Lu, and G. Zhang, “Fuzzy clustering-based adaptive regression for drifting data streams,” IEEE Transactions on Fuzzy Systems, vol. 28, no. 3, pp. 544-557, March 2020, DOI: 10.1109/TFUZZ.2019.2910714.
[57] K. Nongmeikapam, W. K. Kumar, and A. D. Singh, “Fast and automatically adjustable GRBF kernel based fuzzy C-means for cluster-wise colored feature extraction and segmentation of MR images,” IET Image Processing, vol. 12, no. 4, pp. 513-524, March 2018, DOI: 10.1049/iet-ipr.2017.1102.
[58] A. L. Fijri and Z. Rustam, “Comparison between fuzzy kernel C-means and sparse learning fuzzy C-means for breast cancer clustering,” in Proc. International Conference on Applied Information Technology and Innovation, Padang, Padang, Indonesia, Sept. 3-5, 2018, pp. 158-161, DOI: 10.1109/ICAITI.2018.8686707.
[59] P. Lara, F. Igreja, L. D. T. J. Tarrataca, D. B. Haddad, and M. R. Petraglia, “Exact expectation evaluation and design of variable step-size adaptive algorithms,” IEEE Signal Processing Letters, vol. 26, no. 1, pp. 74-78, Jan. 2019, DOI: 10.1109/LSP.2018.2880084.
[60] B. Jalal, X. Yang, X. Wu, T. Long, and T. K. Sarkar, “Efficient direction-of-arrival estimation method based on variable-step-size LMS algorithm,” IEEE Antennas and Wireless Propagation Letters, vol. 18, no. 8, pp. 1576-1580, Aug. 2019, DOI: 10.1109/LAWP.2019.2923700.
[61] H. Zhao, B. Liu, and P. Song, “Variable step-size affine projection maximum correntropy criterion adaptive filter with correntropy induced metric for sparse system identification,” IEEE Transactions on Circuits and Systems II: Express Briefs, vol. 67, no. 11, pp. 2782-2786, Nov. 2020, DOI: 10.1109/TCSII.2020.2973764.
[62] F. Huang, J. Zhang, and S. Zhang, “Combined-step-size affine projection sign algorithm for robust adaptive filtering in impulsive interference environments,” IEEE Transactions on Circuits and Systems II: Express Briefs, vol. 63, no. 5, pp. 493-497, May 2016, DOI: 10.1109/TCSII.2015.2505067.
[63] L. Shi, H. Zhao, and Y. Zakharov, “Generalized variable step size continuous mixed p -norm adaptive filtering algorithm,” IEEE Transactions on Circuits and Systems II: Express Briefs, vol. 66, no. 6, pp. 1078-1082, June 2019, DOI: 10.1109/TCSII.2018.2873254.
[64] H. Zhao, B. Liu, and P. Song, “Variable step-size affine projection maximum correntropy criterion adaptive filter with correntropy induced metric for sparse system identification,” IEEE Transactions on Circuits and Systems II: Express Briefs, vol. 67, no. 11, pp. 2782-2786, Nov. 2020, DOI: 10.1109/TCSII.2020.2973764.
[65] R. Claser and V. H. Nascimento, “On the tracking performance of adaptive filters and their combinations,” in IEEE Transactions on Signal Processing, DOI: 10.1109/TSP.2021.3081045.
[66] F. Huang, J. Zhang, and S. Zhang, “Combined-step-size normalized subband adaptive filter with a variable-parametric step-size scaler against impulsive interferences,” IEEE Transactions on Circuits and Systems II: Express Briefs, vol. 65, no. 11, pp. 1803-1807, Nov. 2018, DOI: 10.1109/TCSII.2017.2771430.