US-12618889-B2 - Method and ethernet physical layer device for detecting, classifying and localizing cable faults of an ethernet physical cable, particularly an automotive ethernet PHY, using a time domain reflectometry algorithm

US12618889B2US 12618889 B2US12618889 B2US 12618889B2US-12618889-B2

Abstract

The invention relates to a method and ethernet physical layer device for detecting, classifying and localizing cable faults of an ethernet physical cable, particularly an automotive ethernet PHY, using a time domain reflectometry algorithm, using a hybrid mode for selectively subtracting sent signals from received signals for fault detection and classification and considering send and received signals during fault localization.

Inventors

Dilip Kumar
Sebastian Höppner
Stephan Hartmann
Emrah ONAT

Assignees

SILICONALLY GMBH

Dates

Publication Date: 20260505
Application Date: 20240208
Priority Date: 20230209

Claims (20)

1 . A method for detecting, classifying and localizing cable faults of an ethernet physical cable, particularly an automotive ethernet PHY, using a time domain reflectometry algorithm, comprising the steps: sending through a transmitter a first finite duration diagnosis pulse signal to the ethernet cable, receiving by a receiver first reflections from the ethernet cable in response to the first diagnosis pulse signal, extracting a first reflected signal by subtracting the send first diagnosis pulse signal from the received first reflections, detecting a fault of the ethernet cable by comparing the first reflected signal to a predetermined threshold, wherein a fault of the ethernet cable is detected when the first reflected signal exceeds the predetermined threshold, classifying the type of fault of the detected fault based on the polarity of the first reflected signal, wherein the fault is an open when the first reflected signal is positive and a short when the first reflected signal is negative, sending through the transmitter a second finite duration diagnosis pulse signal to the ethernet cable, receiving by the receiver second reflections from the ethernet cable in response to the second diagnosis pulse signal, extracting a second reflected signal comprising the send second diagnosis pulse signal and the received second reflections, detecting a first peak in the second reflected signal relating to the send second diagnosis pulse and a second peak in the second reflected signal relating to the received second reflection, and localizing the detected fault of the ethernet cable by determining the time difference between the detected first peak and the detected second peak.
2 . The method for detecting, classifying and localizing cable faults of an ethernet physical cable according to claim 1 , wherein the transmitter and the receiver are clocked by different clock signals from one clock generator.
3 . The method for detecting, classifying and localizing cable faults of an ethernet physical cable according to claim 1 , wherein a matched filtering is used for generating a filter response as an input for comparing with the predetermined threshold.
4 . The method for detecting, classifying and localizing cable faults of an ethernet physical cable according to claim 1 , wherein the predetermined threshold is defined by experiments.
5 . The method for detecting, classifying and localizing cable faults of an ethernet physical cable according to claim 4 , wherein the experiments refer to different fault types and/or different cables length.
6 . The method for detecting, classifying and localizing cable faults of an ethernet physical cable according to claim 4 , wherein the experiments also consider data of a non-faulty ethernet cable and/or random noise.
7 . The method for detecting, classifying and localizing cable faults of an ethernet physical cable according to claim 1 , further comprising the step of performing a pretest analysis of the ethernet link for detecting if the present noise level on the ethernet channel is above a given threshold.
8 . The method for detecting, classifying and localizing cable faults of an ethernet physical cable according to claim 1 , wherein the step of detecting a first peak in the second reflected signal relating to the send second diagnosis pulse and a second peak in the second reflected signal relating to the received second reflection bases on the predefined threshold value of the fault detection step to detect the first peak and the second peak.
9 . The method for detecting, classifying and localizing cable faults of an ethernet physical cable according to claim 8 , wherein predetermined threshold value for detecting the first peak is positive and for detecting the second peak the threshold value is positive for an open fault and negative for a short fault.
10 . The method for detecting, classifying and localizing cable faults of an ethernet physical cable according to claim 1 , wherein the method is implemented in software of an ethernet controller.
11 . An ethernet physical layer device capable of detecting, classifying and localizing cable faults of an ethernet physical cable using a time domain reflectometry algorithm comprising: a media port for connecting to the ethernet physical cable, a transmitter circuit for transmitting signals over the ethernet physical cable, a receiver circuit for receiving signals from the ethernet physical cable, a clock generator for providing different clock signals to the transmitter circuit and receiver circuit, wherein the different clock signals preferably have the same frequency and different phases, a hybrid circuit for subtracting send transmitter signals from received receiver signals, wherein the hybrid circuit can be switched on and off, a programmable pulse generator for generating a first finite duration diagnosis pulse and/or a second finite duration diagnosis pulse, and an ethernet physical cable diagnostic controller for implementing the method according to claim 1 .
12 . The ethernet physical layer device according to claim 11 , wherein the receiver circuit comprises an analog-to-digital converter for converting the received receiver signals to the digital domain.
13 . The ethernet physical layer device according to claim 11 , wherein the clock generator generates clock signals a frequency corresponding to the baud rate of the ethernet physical layer device.
14 . The ethernet physical layer device according to claim 11 , further comprising a memory for storing received receiver signals.
15 . The ethernet physical layer device according to claim 14 , wherein the memory corresponds to existing data registers used during data transfer via the ethernet physical cable.
16 . The method for detecting, classifying and localizing cable faults of an ethernet physical cable according to claim 2 , wherein a matched filtering is used for generating a filter response as an input for comparing with the predetermined threshold.
17 . The method for detecting, classifying and localizing cable faults of an ethernet physical cable according to claim 16 , wherein the predetermined threshold is defined by experiments.
18 . The method for detecting, classifying and localizing cable faults of an ethernet physical cable according to claim 17 , wherein the experiments also consider data of a non-faulty ethernet cable and/or random noise.
19 . The method for detecting, classifying and localizing cable faults of an ethernet physical cable according to claim 18 , further comprising the step of performing a pretest analysis of the ethernet link for detecting if the present noise level on the ethernet channel is above a given threshold.
20 . The method for detecting, classifying and localizing cable faults of an ethernet physical cable according to claim 19 , wherein the step of detecting a first peak in the second reflected signal relating to the send second diagnosis pulse and a second peak in the second reflected signal relating to the received second reflection bases on the predefined threshold value of the fault detection step to detect the first peak and the second peak.

Description

CROSS REFERENCE TO RELATED APPLICATIONS This application is the U.S. Patent Application which claims priority to European Application No.: 23155773.7, filed on Feb. 9, 2023. The contents of this prior application is hereby incorporated by reference herein in its entirety. The invention relates to a method for detecting, classifying and localizing cable faults of an ethernet physical cable, particularly an automotive ethernet PHY, using a time domain reflectometry algorithm. The invention further relates to an ethernet physical layer device capable of detecting, classifying and localizing cable faults of an ethernet physical cable using a time domain reflectometry algorithm. BACKGROUND This invention refers to the technical field of automotive ethernet cable diagnostics. To ensure reliability and functional safety, automotive ethernet PHYs require cable diagnostics which include functionalities to detect, to classify and to locate faults of the physical cable with low hardware effort. The faults are for example an open or a short. With the advent of the technology, there have been different techniques in order to detect the faults and then estimate of its location on the cable. The most popular methods used nowadays are visual and radiographic inspections, low frequency and DC methods, capacitive and inductive methods, medium frequency techniques like tone injection, RF radiation, impedance spectroscopy, Reflectometry etc. (see Layane Abboud, Time Reversal techniques applied to wire fault detection and location in wire networks, Supélec, 2012, English, NNT: 2012SUPL0002ff, tel-00771964). Nevertheless, most of the methods have some disadvantages. The major drawback of visual and radiographic inspections is that they require human interventions. So, these techniques cannot be applied to cables that are not easy to access. The low frequency, capacitive and inductive methods need cables to be disconnected from the circuit which is not feasible for automotive ethernet PHY architecture. Unlike the automotive ethernet standards, shielded twisted pair (STP) wires have required in medium frequency techniques like tone injection and RF radiation for higher accuracies. Moreover, Spectroscopy techniques require the analysis of frequency dependent impedance data which needs high hardware effort. All things considered, reflectometry is the best and most common method to detect and localize the fault, with respect to hardware effort and suitability of the ethernet PHY architecture. In reflectometry technique, a high-frequency signal is transmitted over the cable for analysis. If there is a fault on the cable, the transmitted signal will be reflected back with beneficial information about the fault. It basically uses the voltage reflection coefficient and can be used for different types of cables from telecom networks to aircraft cables. There are several Reflectometry techniques which differ in transmitted signal and algorithms for detection and localization of the fault on the cable. On one hand, there are some complex reflectometry algorithms such as Frequency Domain Reflectometry (FDR), Noise Domain Reflectometry (NDR), Sequence Time Domain Reflectometry (STDR), Spread Spectrum Time Domain Reflectometry (SSTDR), Orthogonal Multi-tone Time Domain Reflectometry (OMTDR), Chaos Time Domain Reflectometry (CTDR), Binary Time Domain Reflectometry (BTDR). Stepped frequency sine waves are used as transmitted signal in FDR. It needs Frequency Modulated Continuous Wave (FMCW) to change or shift in frequency of the signal and it requires phase detection block to detect faults (see C. Furse, Y. C. Chung, R. Dangol, M. Nielsen, G. Mabey, and R. Woodward. “Frequency-domain reflectometry for on-board testing of aging aircraft wiring”. In: IEEE Transactions on Electromagnetic Compatibility 45.2 (2003), pp. 306-315. doi: 10.1109/TEMC.2003.811305.). NDR uses the already existing data or existing noise data as a transmitted signal (see Chet Lo and C. Furse, “Noise-domain reflectometry for locating wiring faults,” in IEEE Transactions on Electromagnetic Compatibility, vol. 47, no. 1, pp. 97-104, February 2005, doi: 10.1109/TEMC.2004.842109.). Similarly, STDR and SSTDR techniques use Pseudo Noise (PN) as the transmitted signal and the reflected signal is correlated with the template signal to analyze the fault (see N. K. T. Jayakumar et al., “Postprocessing for Improved Accuracy and Resolution of Spread Spectrum Time-Domain Reflectometry,” in IEEE Sensors Letters, vol. 3, no. 6, pp. 1-4, June 2019, Art no. 1500204, doi: 10.1109/LSENS.2019.2916636.). OMTDR works based on the principle of orthogonal frequency division multiplexing (OFDM). The transmitted signal is generated using Hermitian symmetry. Reflectogram is generated by performing cross-correlation operation between the transmitted and reflected signal. A lot of hardware effort is required for OMTDR (see E. Cabanillas, M. Kafal and W. Ben-Hassen, “On the Implementation of Embedded Comm