5G time-varying channel playback and simulation method based on machine learning
A time-varying channel, machine learning technology, applied in neural learning methods, wireless communication, transmission monitoring, etc., can solve problems such as insufficient representation of test scenarios, differences, and poor space-time continuity
Inactive Publication Date: 2019-09-17
NORTH CHINA ELECTRIC POWER UNIV (BAODING)
View PDF2 Cites 5 Cited by
- Summary
- Abstract
- Description
- Claims
- Application Information
AI Technical Summary
Problems solved by technology
The QuaDRiGa model combines the characteristics of both the deterministic model and the statistical random model, but unlike the ray tracing model, QuaDRiGa does not use a detailed map that can reproduce the real environment, but uses random statistical distribution to generate scatterers
At the same time, compared with the traditional GSCM model, QuaDRiGa requires the input of information such as the location of the transceiver and the movement trajectory of each mobile station. Based on these parameters, models such as drifting can be used to improve the poor space-time continuity of traditional GSCM and overcome The ray tracing model is too dependent on the accuracy of the topography database and the shortcomings of high computational complexity
However, this channel model still does not adequately represent the test scenario
Method used
the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
View moreImage
Smart Image Click on the blue labels to locate them in the text.
Smart ImageViewing Examples
Examples
Experimental program
Comparison scheme
Effect test
Embodiment Construction
[0024] The embodiments will be described in detail below in conjunction with the accompanying drawings.
[0025] Such as figure 1 As shown, the present invention proposes a time-varying channel playback and simulation method based on machine learning, including:
[0026] Step S101: Extract small-scale parameters of the channel:
[0027] Based on the measured data, the high-resolution SAGE algorithm is used to extract the small-scale parameters of the channel. The SAGE algorithm is an expansive iterative algorithm of the EM algorithm, which reduces the dimension by sequentially updating the parameter subsets, reduces the amount of calculation and speeds up the convergence speed, thus This makes the parameter estimation more accurate and improves the signal-to-noise ratio of the system. Small-scale parameters include the number of multipath (clusters) corresponding to each snapshot, multipath delay, complex amplitude, AoA and AoD, and Doppler frequency shift.
[0028] Step S1...
the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to View More PUM
Login to View More Abstract
The invention belongs to the technical field of wireless channel modeling, and particularly relates to a 5G time-varying channel playback and simulation method based on machine learning, which comprises the following steps: extracting a small-scale parameter corresponding to each snapshot of a channel by adopting a high-resolution spatial alternation wide expectation maximum algorithm; establishing a joint small-scale parameter neural network time varying model corresponding to the snapshot; replacing the random channel parameters in the random statistical model based on the geometric basis with the small-scale parameters generated by the neural time-varying model to carry out small-scale parameter simulation; and according to the small-scale parameters obtained through simulation, obtaining a time-varying channel matrix coefficient through calculation. The 5G actual scene channel test data can be conveniently replayed, and the method has a very important application value for realizing 5G link and system-level performance simulation evaluation and 5G network design in scientific research institutions and industry.
Description
technical field [0001] The invention belongs to the technical field of wireless channel modeling, and in particular relates to a machine learning-based 5G time-varying channel playback and simulation method. Background technique [0002] 5G mobile communication system is a research hotspot at home and abroad in recent years. 5G application scenarios are mainly concentrated in multi-user hotspot areas, where there are a large number of mobile scatterers, such as urban pedestrians and vehicles, high-speed rail waiting rooms, modern grid-shaped offices, etc. movement of people. Therefore, the time-varying nature of 5G is more prominent, and the channel industry urgently needs to abandon or improve the uncertainty caused by traditional channel-based methods based on probability statistics, find time-varying channel models of 5G candidate frequency bands in typical application scenarios, and carry out channel simulation research , to realistically reproduce the time-varying char...
Claims
the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to View More Application Information
Patent Timeline
Login to View More Patent Type & Authority Applications(China)
IPC IPC(8): H04B17/391H04W24/06G06N3/04G06N3/08
CPCH04B17/3912H04W24/06G06N3/08G06N3/048
Inventor 赵雄文杜飞王忠钰
Owner NORTH CHINA ELECTRIC POWER UNIV (BAODING)