基于DCNN-LSTM负荷预测算法的5G基站节能系统研究

doi:10.11959/j.issn.1000-0801.2023101

Abstract

Abstract:

With the rapid construction of the 5G wireless communication network, the energy consumption pressure of operators, and even the overall communication industry, is simultaneously highlighted.Achieving sustainable development of the industry through energy conservation and consumption reduction has become a new research direction for the current 5G network development.Taking the PRB rate as the load evaluation index, LSTM model was improved by using DCNN to extract the depth feature of the cell’s indicators.A set of DCNN-LSTM deep learning model that could predict the future value of PRB rate was proposed.On the basis of the improved algorithm, the network topology of the current 5G access network was optimized.An additional network element and its working system were designed.An intelligent energy-saving system, which ensured the network experience, of 5G base stations was realized.

Key words: 5G base station energy saving, improved LSTM algorithm, 5G system design

CLC Number:

TN929.5

Jianbin WANG, Shuchun WANG, Shangjin LIAO, Shuyuan SHI. Research on 5G base station energy saving system based on DCNN-LSTM load prediction algorithm[J]. Telecommunications Science, 2023, 39(4): 133-141.

Figures/Tables 12

References 15

[1]	华为技术有限公司. 5G 时代运营商数据和存储架构白皮书[R]. 2020.
	Huawei Technologies Co.,Ltd. White paper on operater data and storage architecture in 5G era[R]. 2020.
[2]	赛迪顾问. 5G 产业发展白皮书（2020）[R]. 2020.
	CCID Consulting. 5G industry development white paper (2020)[R]. 2020.
[3]	张化, 李鹏, 鲁娜 ,等. 5G 基站节能技术性能评估研究[J]. 电子技术应用, 2020,46(10): 2024.
	ZHANG H , LI P , LU N ,et al. Study of performance evaluation on energy saving for 5G base station[J]. Application of Electronic Technique, 2020,46(10): 2024.
[4]	谢泽铖, 徐雷, 张曼君 ,等. 5G 网络共建共享安全研究[J]. 邮电设计技术, 2021(4): 5-9.
	XIE Z C , XU L , ZHANG M J ,et al. Research on security of 5G network co-construction and sharing[J]. Designing Techniques of Posts and Telecommunications. 2021(4): 5-9.
[5]	WU Q Q , LI G Y , CHEN W ,et al. An overview of sustainable green 5G networks[J]. IEEE Wireless Communications, 2017,24(4): 72-80.
[6]	LORINCZ J , CAPONE A , WU J S . Greener,energy-efficient and sustainable networks:state-of-the-art and new trends[J]. Sensors, 2019,19(22): E4864.
[7]	官磊, 丁洋, 李锐杰 ,等. 面向“双碳”的5G网络节能技术[J]. 电信科学, 2022,38(4): 8.
	GUAN L , DING Y , LI R J ,et al. Network energy saving technologies for green 5G[J]. Telecommunications Science, 2022,38(4): 8.
[8]	徐孟强 . 基于AI深度学习的面向业务5G基站节能系统研究[J]. 电信科学, 2021,37(11): 9.
	XU M Q . Research on business oriented 5G base station energy saving system based on AI deep learning[J]. Telecommunications Science, 2021,37(11): 9.
[9]	BARB G , ANDRAS C , BALINT C . Performance analysis of transport layer congestion on 5G systems[C]// Proceedings of 2022 14th International Conference on Communications (COMM).[S.l.:s.n.], 2022: 1-4.
[10]	KALCHBRENNER N , GREFENSTETTE E , BLUNSOM P . A convolutional neural network for modelling sentences[J]. arXiv preprint, 2014.arXiv:1404.2188.
[11]	GRAVES A , MOHAMED A , HINTON G . Speech recognition with deep recurrent neural networks[C]// Proceedings of the 2013 IEEE international Conference on Acoustics,Speechand Signal Processing. Piscataway:IEEE Press, 2013: 6645-6649.
[12]	KINGMA D P , BA J . Adam:A method for stochastic optimization[J]. arXiv preprint, 2014,arXiv:1412.6980.
[13]	BAI Y , YANG E , HAN B ,et al. Understanding and improving early stopping for learning with noisy labels[J]. Advances in Neural Information processing Systems, 2021(34): 24392-24403.
[14]	HAN X Y , PAPYAN V , DONOHO D L . Neural collapse under MSE loss:proximity to and dynamics on the central path[J]. arXiv preprint, 2021,arXiv:2106.02073.
[15]	DATAESATU A , BOONSRIMUANG P , MORI P ,et al. Energy efficiency enhancement in 5G heterogeneous cellular networks using system throughput based sleep control scheme[C]// Proceedings of 2020 22nd International Conference on Advanced Communication Technology (ICACT).[S.l.:s.n.], 2020: 549-553.

Metrics

Recommended 0

No Suggested Reading articles found!

编号	y	x₁	x₂	x₃	x₄	x₅	x₆
指标	PRB利用率	平均用户数	下行流量	上行流量	随机前导的接收次数	专用前导的接收次数	NR 向 LTE 切换出尝试次数

算法	Cell_1	Cell_2	Cell_3
MLP	4.37	6.51	5.11
CNN	2.44	3.66	4.17
SVR	2.77	3.89	3.16
LSTM	2.14	2.25	1.68
DCNN-LSTM	1.46	2.54	1.61

预测PRB利用率门限	小包业务比例	建议节能策略
[0,20%]	[0,100%]	定时载波关断、射频模块深度休眠、MM通道关断、符号关断
(20%,50%]	[0,80%]	MM通道关断、符号关断
(20%,50%]	(80%,100%]	4G/5G协同载波关断、MM通道关断、符号关断、
(50%,100%]	(80%,100%]	符号关断、MM通道关断
(50%,100%]	[0,80%]	不采取节能措施

小区类型	平均能耗/（千瓦时·天^-1）			节电/（千瓦时·天^-1）
小区类型	未开节能	传统节能方案	本系统方案	传统节能方案	本系统方案
64T/64R	12.84	10.44	9.23	2.40	3.61
32T/32R	7.26	5.59	4.73	1.67	2.53
4T/4R	2.34	1.86	1.45	0.48	0.89
整体平均	10.73	8.67	7.60	2.07	3.13

节能方案	符号关断率	MM通道关断率	载波关断率	射频模块深度休眠率
本系统方案	92.55%	49.45%	29.56%	25.28%
传统节能方案	92.43%	47.22%	22.45%	19.37%
差值	0.12%	2.23%	7.11%	5.91%

Research on 5G base station energy saving system based on DCNN-LSTM load prediction algorithm

RichHTML

PDF下载

Knowledge

Abstract

Cite this article

share this article

Figures/Tables 12

References 15

Related Articles 1

Metrics

Recommended 0

本方案	无线接通率	掉线率	切换成功率	CQI优良比
节能前	99.62%	0.182%	99.55%	97.25%
节能后	99.58%	0.211%	99.58%	97.14%