- 专利标题： ROBUST TOA-ESTIMATION USING CONVOLUTIONAL NEURAL NETWORKS (OR OTHER FUNCTION APPROXIMATIONS) ON RANDOMIZED CHANNEL MODELS
-
申请号： EP24158029.9申请日： 2021-12-06
-
公开(公告)号： EP4350573A2公开(公告)日： 2024-04-10
- 发明人: FEIGL, Tobias , EBERLEIN, Ernst , MUTSCHLER, Christopher , KRAM, Sebastian
- 申请人： Fraunhofer-Gesellschaft zur Förderung der angewandten Forschung e.V.
- 申请人地址： DE 80686 München Hansastraße 27c
- 代理机构： Zuccollo, Alberto
- 优先权： DE102020215852 20201214
- 主分类号： G06N3/02
- IPC分类号： G06N3/02
摘要：
There is disclosed a method (300) for inferring a predetermined time of arrival (261) of a predetermined transmitted signal on the basis of channel impulse responses, CIRs, of transmitted signals between a mobile antenna and a fixed antenna, the method comprising:
(310) intermittently obtaining a present channel impulse response condition characteristic, CIRCC (317), which is descriptive of CIRs (315) of transmitted signals associated with mobile antenna positions within a reach of the fixed antenna (140);
(380) checking whether the present CIRCC (317) fits to a predetermined present CIRCC (117, 217) with which a predetermined parametrization (233) of a neural network, or another function approximator, (131) is associated, and
(387) as longs as the check (380) reveals that (381) the present CIRCC (317) fits to the predetermined present CIRCC (117, 217), inferring (260), using the neural network or other function approximator (131), parametrized using the predetermined parametrization (233), the predetermined time of arrival (261), and
as soon as (382) the present CIRCC (317) no longer fits to the predetermined present CIRCC (117), cease (384) the use of the neural network, or other function approximator (131), parametrized using the predetermined parametrization (233), for inferring the predetermined time of arrival, and/or initiate a compensation step in which the neural network, or other function approximator (131) is analyzed and/or retrained.
(310) intermittently obtaining a present channel impulse response condition characteristic, CIRCC (317), which is descriptive of CIRs (315) of transmitted signals associated with mobile antenna positions within a reach of the fixed antenna (140);
(380) checking whether the present CIRCC (317) fits to a predetermined present CIRCC (117, 217) with which a predetermined parametrization (233) of a neural network, or another function approximator, (131) is associated, and
(387) as longs as the check (380) reveals that (381) the present CIRCC (317) fits to the predetermined present CIRCC (117, 217), inferring (260), using the neural network or other function approximator (131), parametrized using the predetermined parametrization (233), the predetermined time of arrival (261), and
as soon as (382) the present CIRCC (317) no longer fits to the predetermined present CIRCC (117), cease (384) the use of the neural network, or other function approximator (131), parametrized using the predetermined parametrization (233), for inferring the predetermined time of arrival, and/or initiate a compensation step in which the neural network, or other function approximator (131) is analyzed and/or retrained.
公开/授权文献
信息查询
