数据驱动的无线信道可用吞吐量估计与预测方法

doi:10.11959/j.issn.2096-3750.2023.00341

摘要/Abstract

摘要：

无线局域网络技术正蓬勃发展，但随之而来的新问题严重影响了无线信道的通信质量。无线信道质量对指导路由器应对突发拥塞和选择合适信道具有重大意义。以信道可用吞吐量为指标设计了一套解决方案：首先，采用入侵式数据采集方法收集信道数据，使用人工神经网络训练并估计当前时刻的信道可用吞吐量；然后，采用非入侵式数据采集方法收集信道数据，使用改进的递归神经网络模型预测未来一段时间的信道可用吞吐量。在真实数据上的实验表明，该方案可以有效地对信道可用吞吐量进行估计与预测，对路由器的决策有着指导意义。

关键词: 无线信道, 数据驱动, 神经网络, 吞吐量估计与预测

Abstract:

The rapid development of wireless local area network technology has brought about new challenges that significantly affect the communication quality of wireless channels.Wireless channel quality is crucial for guiding routers in managing sudden congestion and selecting appropriate channels.A set of solutions using channel available throughput as an indicator was designed.Firstly, invasive data collection methods were used to collect channel data, and an artifical neural network was trained to estimate the available throughput of the channel at the current time.Subsequently, non-invasive data collection methods were utilized to collect channel data, and an improved recurrent neural network model was employed to predict the available throughput of the channel for a future period.Experiments on the real data show that the scheme can effectively estimate and predict the available throughput of the channel, providing guidance for router decisions.

Key words: wireless channel, data-driven, neural network, throughput estimation and prediction

中图分类号:

TN925

肖遥, 刘峻铄, 龙智夫, 邱才明. 数据驱动的无线信道可用吞吐量估计与预测方法[J]. 物联网学报, 2023, 7(3): 32-41.

Yao XIAO, Junshuo LIU, Zhifu LONG, Caiming QIU. A data-driven approach to wireless channel available throughput estimation and prediction[J]. Chinese Journal on Internet of Things, 2023, 7(3): 32-41.

图/表 15

图1

表1

图2

图3

表2

图4

表3

图5

表4

图6

图7

图8

图9

图10

表5

参考文献 32

[1]	MOZAFFARIAHRAR E , THEOLEYRE F , MENTH M . A survey of Wi-Fi 6:technologies,advances,and challenges[J]. Future Internet, 2022,14(10): 293.
[2]	OUGHTON E J , LEHR W , KATSAROS K ,et al. Revisiting wireless internet connectivity:5G vs Wi-Fi 6[J]. Telecommunications Policy, 2021,45(5): 102127.
[3]	SUER M T , JOSE P , TCHOUANKEM H . Experimental evaluation of IEEE 802.11ax - low latency and high reliability with Wi-Fi 6?[C]// Proceedings of GLOBECOM 2022-2022 IEEE Global Communications Conference. Piscataway:IEEE Press, 2023: 377-382.
[4]	HYDER Z , HE D , LI D Y ,et al. Improved Wi-Fi indoor localization based on signal quality parameters and RSSI smoothing algorithm[C]// Proceedings of 2021 10th International Conference on Information and Automation for Sustainability (ICIAfS). Piscataway:IEEE Press, 2021: 140-144.
[5]	GHOSH J , GUPTA A , HACI H ,et al. User association,power control and channel access schemes for two-tier macro-femto networks:CDF of SINR analysis[J]. IETE Technical Review, 2022,39(1): 219-230.
[6]	MOHANTY S , AGARWAL A , MALI S ,et al. Design and BER performance analysis of MIMO and massive MIMO networks under perfect and imperfect CSI[C]// Proceedings of 2020 Fourth International Conference on I-SMAC (IoT in Social,Mobile,Analytics and Cloud) (I-SMAC). Piscataway:IEEE Press, 2020: 307-312.
[7]	THANUJA T C , DAMAN K A , PATIL A S . Optimized spectrum sensing techniques for enhanced throughput in cognitive radio network[C]// Proceedings of 2020 International Conference on Emerging Smart Computing and Informatics (ESCI). Piscataway:IEEE Press, 2020: 137-141.
[8]	AFIFI W S , EL-MOURSY A A , SAAD M ,et al. A dynamic clustering approach for increasing user throughput in 5G wireless networks[C]// Proceedings of NOMS 2020 - 2020 IEEE/IFIP Network Operations and Management Symposium. Piscataway:IEEE Press, 2020: 1-8.
[9]	HU N N , STEENKISTE P . Evaluation and characterization of available bandwidth probing techniques[J]. IEEE Journal on Selected Areas in Communications, 2003,21(6): 879-894.
[10]	JAIN M , DOVROLIS C . End-to-end available bandwidth:measurement methodology,dynamics,and relation with TCP throughput[J]. IEEE/ACM Transactions on Networking, 2003,11(4): 537-549.
[11]	BEDI H , SHARMA K K . Analyses of CSMA/CA protocol without using virtual channel sensing in DCF mode[C]// Global Emerging Innovation Summit (GEIS-2021).[S.l.:s.n.], 2021: 484-491.
[12]	SHAH S H , CHEN K , NAHRSTEDT K ,et al. Available bandwidth estimation in IEEE 802.11-based[EB]. 2004.
[13]	AL-GARADI M A , MOHAMED A , AL-ALI A K , ,et al. A survey of machine and deep learning methods for internet of things (IoT) security[J]. IEEE Communications Surveys ＆ Tutorials, 2020,22(3): 1646-1685.
[14]	CAO Q , BANERJEE R , GUPTA S ,et al. Data driven production forecasting using machine learning[EB]. 2016.
[15]	OLIVEIRA L , RODRIGUES J , KOZLOV S ,et al. MAC layer protocols for internet of things:a survey[J]. Future Internet, 2019,11(1): 16.
[16]	TACCHINO F , BARKOUTSOS P , MACCHIAVELLO C ,et al. Quantum implementation of an artificial feed-forward neural network[J]. Quantum Science and Technology, 2020,5(4): 044010.
[17]	UZAIR M , JAMIL N . Effects of hidden layers on the efficiency of neural networks[C]// Proceedings of 2020 IEEE 23rd International Multitopic Conference (INMIC). Piscataway:IEEE Press, 2021: 1-6.
[18]	HANIN B , ROLNUCK D . Complexity of linear regions in deep networks[C]// International Conference on Machine Learning. New York:PMLR, 2019: 2596-2604.
[19]	BERRY M A , LINOFF G . Data mining techniques for marketing,sales and customer relationship[EB]. 2004.
[20]	MAO K Z , TAN K C , SER W . Probabilistic neural-network structure determination for pattern classification[J]. IEEE Transactions on Neural Networks, 2000,11(4): 1009-1016.
[21]	BARBOSH M , SINGH P , SADHU A . Empirical mode decomposition and its variants:a review with applications in structural health monitoring[J]. Smart Materials and Structures, 2020,29(9): 093001.
[22]	SARIKAYA Y , ATALAY I C , GURBUZ O ,et al. Estimating the channel capacity of multi-hop IEEE 802.11 wireless networks[J]. Ad Hoc Networks, 2012,10(6): 1058-1075.
[23]	LI Z W , LIU F , YANG W J ,et al. A survey of convolutional neural networks:analysis,applications,and prospects[J]. IEEE Transactions on Neural Networks and Learning Systems, 2022,33(12): 6999-7019.
[24]	YU Y , SI X S , HU C H ,et al. A review of recurrent neural networks:LSTM cells and network architectures[J]. Neural Computation, 2019,31(7): 1235-1270.
[25]	MA X L , TAO Z M , WANG Y H ,et al. Long short-term memory neural network for traffic speed prediction using remote microwave sensor data[J]. Transportation Research Part C:Emerging Technologies, 2015(54): 187-197.
[26]	TIAN Y X , PAN L . Predicting short-term traffic flow by long short-term memory recurrent neural network[C]// Proceedings of 2015 IEEE International Conference on Smart City/Social Com/Sustain Com (Smart City). Piscataway:IEEE Press, 2016: 153-158.
[27]	CHENG J S , YU D Y , YU Y . Research on the intrinsic mode function (IMF) criterion in EMD method[J]. Mechanical Systems and Signal Processing, 2006,20(4): 817-824.
[28]	AGHABOZORGI S , SHIRKHORSHIDI A S , WAH T . Time-series clustering - a decade review[J]. Information Systems, 2015,53: 16-38.
[29]	HALLAC D , VARE S , BOYD S ,et al. Toeplitz inverse covariance-based clustering of multivariate time series data[C]// KDD:Proceedings of International Conference on Knowledge Discovery＆Data Mining, 2017: 215-223.
[30]	MATSUBARA Y , SAKURAI Y , FALOUTSOS C . AutoPlait:automatic mining of co-evolving time sequences[C]// Proceedings of the 2014 ACM SIGMOD International Conference on Management of Data. New York:ACM Press, 2014: 193-204.
[31]	WANG H H , WU X D , HUANG Z Y ,et al. High-frequency component helps explain the generalization of convolutional neural networks[C]// Proceedings of 2020 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR). Piscataway:IEEE Press, 2020: 8681-8691.
[32]	ZHANG C , BENGIO S , HARDT M ,et al. Understanding deep learning requires rethinking generalization[EB]. 2016.

发送端传输速率/(Mbit·s^-1)	接收端接收速率/(Mbit·s^-1)
35	32.7
32	31.4
31	30.9
30	30.0
25	24.9

特征名称	含义
ant_rssi	接收信号强度指示
ant_snr	下行信噪比
chload	信道同协议占空比
duty_ratio	信道非协议（干扰）占空比
Free_power	底噪

特征名称	特征含义
num_of_frame	该秒内帧的总数
mun_of_mac	该秒内帧发送地址的个数
num_of_QosData	该秒内服务质量数据帧的总数
num_of_Data	该秒内数据帧的总数
num_of_BA	该秒内块确认帧的总数
num_of_CTS	该秒内CTS帧的总数
num_of_RTSdata_length	该秒内RTS帧的总数该秒内帧的总长度
retry_rate	该秒内重传帧占帧总数的比重

多元回归算法	MSE
ANN	0.009 026 81
Sigmoid回归	0.012 897 08
多元线性回归	0.012 692 08

时间序列预测模型	RMSE	MAE	MAPE	R-Square	Time_1/s	Time_2/s
EMD+CNN-LSTM	0.064	0.046	15.044%	0.905	197.99	0.075
CNN-LSTM	0.113	0.084	31.669%	0.705	81.47	0.076
EMD+LSTM	0.068	0.050	17.498%	0.892	184.07	0.080
EMD+RNN	0.069	0.050	17.783%	0.889	146.82	0.081
LSTM	0.115	0.087	33.103%	0.691	80.91	0.080
RNN	0.116	0.088	33.243%	0.686	81.73	0.081