Sensor device with sensing signal selection

Abstract

A sensor device (1) with sensing signal selection comprises a signal transmitter-receiver module (10) to provide a plurality of signal paths (SP1, SP2, SP3, SP4) for respectively transmitting a sensing signal (SS11, SS22, SS12, SS21) between a transmitter side (11) and a receiver side (12) of the signal transmitter-receiver module (10). The sensor device comprises an accelerometer (20) to provide an accelerometer output signal (AS) representing an acceleration of the sensor device (1). The sensor device (1) further comprises a controller (30) being configured to select at least one of the sensing signals (SS11, SS22, SS12, SS21) for further processing in dependence on a signal comparison of each of the sensing signals (SS11, SS22, SS12, SS21) with the accelerometer output signal (AS).

Description

[0001] 2024PF01209 January 9, 2026

[0002] P2024, 0920 WO N - 1 -

[0003] Description

[0004] SENSOR DEVICE WITH SENSING SIGNAL SELECTION

[0005] Technical Field

[0006] The disclosure relates to a sensor device with selection of a sensing signal from a plurality of received sensing signals for further processing.

[0007] Background

[0008] In the measurement of sensor signals by sensor devices, motion often introduces unwanted artifacts and effects that can significantly degrade the quality and reliability of data obtained by analyzing and evaluating the received sensor signals by subsequent signal processing. These artifacts, resulting from both regular and irregular motion, can complicate the subsequent signal processing steps, potentially leading to erroneous interpretations and decisions .

[0009] Regular motion, such as walking, running, and cycling in wearable applications, introduces periodic signal components in the frequency domain. These components can overlap with the frequencies of interest in the sensor signals, such as photoplethysmography (PPG) signals in optical sensing. The spectral overlap can result in the motion artifacts being misinterpreted as part of a physiological signal measured by the sensor device, thereby affecting the accuracy of the calculated data derived from the measured physiological signal, for example the accuracy of heartrate, oxygen2024PF01209 January 9, 2026

[0010] P2024, 0920 WO N - 2 -

[0011] saturation, and other vital sign measurements derived from a PPG signal .

[0012] Moreover, singular motion events, such as sudden j erks, impacts or shifts in the sensor' s position, can produce transient artifacts . These abrupt changes can mimic or obscure the true signal, leading to false positives or missed detections . For instance, a sudden movement might be mistakenly identified as an abnormal heart rhythm in PPG monitoring .

[0013] It would be welcome in the art to provide a sensor device that enables data to be obtained by processing sensing signals received by the sensor device, whereby the data is less likely to be distorted by the negative influence of movement on the sensor device .

[0014] Summary

[0015] An embodiment of a sensor device for providing data by evaluating measured sensing signals, whereby a negative influence of a motion of the sensor device on the data is to a certain degree avoided, is specified in claim 1.

[0016] According to an embodiment, the sensor device comprises a signal transmitter-receiver module to provide a plurality of signal paths for respectively transmitting a sensing signal between a transmitter side and a receiver side of the signal transmitter-receiver module . The sensor device further comprises an accelerometer to provide an accelerometer output signal representing an acceleration of the sensor device . The sensor device comprises a controller being configured to select at least one of the sensing signals for further2024PF01209 January 9, 2026

[0017] P2024, 0920 WO N - 3 -

[0018] processing in dependence on a signal comparison of each of the sensing signals with the accelerometer output signal .

[0019] According to the proposed approach of a sensor device, given a multitude of available sensing signals that are received at the receiver side of the signal transmitter-receiver module and negatively affected by motion, an intelligent selection of the sensing signals for further processing is performed by the controller using data of an also present accelerometer . The sensing signal that is least influenced by the movement of the optical sensor device is selected by the controller from the multitude of sensing signals received via various signal paths for further signal processing.

[0020] In order to select a suitable sensing signal for further processing, the sensing signals received at the receiver side of the signal transmitter-receiver module are compared with the accelerometer output signal . The sensing signal that is least similar to the accelerometer output signal is selected by the controller for further signal processing. It is assumed that this signal is also least affected by any disturbing movement of the optical sensor device .

[0021] Accordingly, only those signal sources / signal transmitter devices from which the sensing signal selected for further processing originates are selected or activated on the transmitter side of the signal transmitter-receiver module for generating the sensing signals . The signal transmitter devices which generate sensing signals not selected for further processing can be deactivated which results in energy savings .2024PF01209 January 9, 2026

[0022] P2024, 0920 WO N 4

[0023] Of course, several of the received sensing signals that are least affected by a motion of the sensor device can also be used for further signal processing. Accordingly, only those signal sources which generate the sensing signals selected for further processing are activated on the transmitter side of signal transmitter-receiver module . On the other hand, the proposed configuration of the optical sensor device with sensing signal selecting enables the discontinuation of a signal collection from other unused signal sources or signal transmitter devices, either permanently or intermittently, thereby reducing energy consumption and conserving battery power, especially in wearable devices .

[0024] In conclusion, given a multitude of available signal sources or signal transmitter devices at the transmitter side of the signal transmitter-receiver module and a multitude of available sensing signals at the receiver side of the signal transmitter-receiver module, the proposed intelligent selection of at least one sensing signal which is least affected by the motion of the sensor device using data of an accelerometer of the sensor device leads to optimal signal quality and maximum motion entanglement .

[0025] According to a possible embodiment of the sensor device, the controller is configured to carry out the signal comparison by performing a similarity analysis of each of the sensing signals with the accelerometer output signal . Using similarity / correlation / relationship / dependence analysis between the available accelerometer data and the different available sensing signals or signal sources enables to intelligently select the best available sensing

[0026] signal ( s ) / signal source (s) , showing the smallest relationship to the disturbance f actor / conf ounder motion. The output of2024PF01209 January 9, 2026

[0027] P2024, 0920 WO N 5

[0028] this analysis can be used for adaptively switching between signal sources and selectively turning them on and off which results in reduced energy consumption of the sensor device .

[0029] The controller may be configured to perform the similarity analysis permanently or intermittently. This enables the continuous or intermittent processing of the least motion distorted sensing signal (s) which leads to best performance in subsequent signal processing.

[0030] According to an embodiment of the sensor device, the controller is configured to select at least a first one of the sensing signals for further processing and to exclude at least a second one of the sensing signals from further processing, if a lower similarity between the at least one first sensing signal and the accelerometer output signal than between the at least one second sensing signal and the accelerometer output signal is detected by the controller . The lower similarity between the at least one first sensing signal and the accelerometer output signal indicates that this sensing signal is less affected by a disturbing motion of the sensor device and thus better suited for further signal processing than the at least one second sensing signal .

[0031] According to an embodiment of the sensor device, the signal transmitter-receiver module comprises at least a first and a second signal transmitter device and at least a first and a second signal receiver device . The signal transmitterreceiver module is configured so that a respective one of the signal paths is created from each of the at least one first and second signal transmitter device to each of the at least one first and second signal receiver device .2024PF01209 January 9, 2026

[0032] P2024, 0920 WO N - 6 -

[0033] For example, the sensor device can be arranged on an obj ect to be examined, so that the sensing signals emitted by the signal transmitter devices on the transmitter side of the signal transmitter-receiver module penetrate into the body. The sensing signals are reflected at structures or substances inside the body in the direction of the signal receiver devices on the receiver side of the signal transmitterreceiver module .

[0034] The sensor device can, for example, be configured as an optical sensor device, for example, as a PPG sensor for PPG monitoring of a living organism, which can be used, for example, to determine a heartrate or oxygen saturation in the blood. The optical sensor device comprises optical transmitter devices on the transmitter side and optical receiver devices on the receiver side of an optical transmitter-receiver module . The optical sensing signals received on the receiver side pass through different optical signal paths inside the body, starting from the optical transmitter devices that generate them. On the various optical signal paths, the optical signals are affected to a greater or lesser extent by a motion of the body.

[0035] According to a possible embodiment of the sensor device, the at least one first signal transmitter device is configured for generating the first sensing signal, and the at least one second signal transmitter device is configured for generating the second sensing signal . The signal transmitter-receiver module is configured to operate the at least one first signal transmitter device in an activated state to continue with the generation of the first sensing signal and to operate the at least one second signal transmitter device in a deactivated2024PF01209 January 9, 2026

[0036] P2024, 0920 WO N 7

[0037] state to interrupt the generation of the second sensing signal, if a lower similarity between the at least one first sensing signal and the accelerometer output signal than between the at least one second sensing signal and the accelerometer output signal is detected by the controller .

[0038] After choosing best signal sources or signal transmitter devices to generate the sensing signals which are less affected by motion along their travel on the respective signal paths and used for further signal processing, recording of sensing signals emitted by other signal sources or signal transmitter devices can be stopped, thereby reducing energy consumption especially in mobile / wearable applications .

[0039] According to an embodiment of the sensor device, the controller is configured to perform the similarity analysis by performing a correlation analysis of each of the sensing signals with the accelerometer output signal . Those sensing signals that show the lowest correlation or even no correlation at all with the accelerometer output signal are selected by the controller for further signal processing.

[0040] According to an embodiment of the sensor device, the controller is configured to perform the similarity analysis by calculating a respective variable derived from each of the sensing signals and a variable of the accelerometer output signal . The controller is configured to perform the similarity analysis by comparing the respective variable derived from each of the sensing signals with the variable derived from the accelerometer output signal . The controller may be configured to determine the respective variables by2024PF01209 January 9, 2026

[0041] P2024, 0920 WO N - 8 -

[0042] calculating a transformation or representation of each of the sensing signals and the accelerometer output signal .

[0043] For example, respective first derivatives of the received sensing signals and the accelerometer output signal or low pass filtered variants of the received sensing signals and the accelerometer output signal can be calculated and a similarity between these parameters can be determined to find the signal path showing lowest similarity to accelerometer data .

[0044] According to a possible embodiment of the sensor device, the controller is configured to perform the similarity analysis by calculating respective frequency domain parameters of each of the sensing signals and frequency domain parameters of the accelerometer output signal . The controller is configured to perform the similarity analysis by comparing the respective frequency domain parameters of each of the sensing signals with the frequency domain parameters of the accelerometer output signal .

[0045] According to another possible embodiment of the sensor device, the controller is configured to perform the similarity analysis by calculating a respective FFT spectrum of each of the sensing signals and an FFT spectrum of the accelerometer output signal . The controller is configured to perform the similarity analysis by comparing the respective FFT spectrum of each of the sensing signals with the FFT spectrum of the accelerometer output signal .

[0046] According to another possible embodiment of the sensor device, the controller is configured to perform the similarity analysis by calculating respective time domain2024PF01209 January 9, 2026

[0047] P2024, 0920 WO N 9

[0048] parameters of each of the sensing signals and time domain parameters of the accelerometer output signal . The controller is configured to perform the similarity analysis by comparing the respective time domain parameters of each of the sensing signals with the time domain parameters of the accelerometer output signal .

[0049] According to a possible embodiment of the sensor device, the controller is configured to perform the similarity analysis by calculating respective statistical parameters of each of the sensing signals and statistical parameters of the accelerometer output signal . The controller is configured to perform the similarity analysis by comparing the respective statistical parameters of each of the sensing signals with the statistical parameters of the accelerometer output signal .

[0050] According to another possible embodiment of the sensor device, the controller is configured to perform the similarity analysis by calculating respective entropy-based parameters of each of the sensing signals and entropy-based parameters of each of the accelerometer output signal . The controller is configured to perform the similarity analysis by comparing the respective entropy-based parameters of each of the sensing signals with the entropy-based parameters of the accelerometer output signal .

[0051] According to another possible embodiment of the sensor device, the controller is configured to perform the similarity analysis by calculating respective non-linear dynamic parameters of each of the sensing signals and nonlinear dynamic parameters of the accelerometer output signal . The controller is configured to perform the similarity2024PF01209 January 9, 2026

[0052] P2024, 0920 WO N 10

[0053] analysis by comparing the respective non-linear dynamic parameters of each of the sensing signals with the non-linear dynamic parameters of the accelerometer output signal .

[0054] Additional features and advantages are set forth in the detailed description that follows . It is to be understood that both the foregoing general description and the following detailed description are merely exemplary, and are intended to provide an overview or framework for understanding the nature and character of the claims .

[0055] Brief Description of the Drawing

[0056] The accompanying figures are included to provide further understanding, and are incorporated in, and constitute a part of the specification. As such, the proposed configuration of a sensor device with sensing signal selection will be more fully understood from the following detailed description, taken in conjunction with the accompanying figures of which

[0057] Figure 1 shows a schematic of an operation of a sensor device with sensing signal selection;

[0058] Figure 2 shows a block diagram of a sensor device with sensing signal selection;

[0059] Figure 3 shows an application of a sensor device with sensing signal selection;

[0060] Figure 4 illustrates signal curves of sensing signals of a signal transmitter-receiver module and signal curves of an accelerometer of a sensor device with sensing signal selection; and2024PF01209 January 9, 2026

[0061] P2024, 0920 WO N - 11 -

[0062] Figure 5 illustrates frequency spectra of sensing signals and frequency spectra of an accelerometer output signal for performing a similarity analysis of the sensing signals with the accelerometer output signal .

[0063] Detailed Description

[0064] A sensor device 1 with intelligent selection of at least one sensing signal from a multitude of available sensing signals, whereby the selected sensing signal is less affected by a motion of the sensor device 1 than other sensing signals, is explained in the following with reference to the schematic of the sensor device 1 shown in Figure 1 and the block diagram of the sensor device 1 shown in Figure 2.

[0065] The sensor device 1 with intelligent selection of at least one sensing signal least affected by motion of the sensor device comprises a signal transmitter-receiver module 10 to provide a plurality of signal paths SP1, SP2, SP3 and SP4 for respectively transmitting a sensing signal SS11, SS22, SS12, SS21 between a transmitter side 11 and a receiver side 12 of the signal transmitter-receiver module 10. The transmitterreceiver module may comprise signal transmitter devices LED1, LED2 at the transmitter side 11 for generating the sensing signals and signal receiver devices PD1, PD2 at the receiver side 12 for receiving the sensing signals . The sensor device 1 further comprises an accelerometer 20 to provide an accelerometer output signal AS which represents an acceleration of the sensor device 1.

[0066] According to the proposed approach, the sensor device 1 comprises a controller 30 which is configured to select at2024PF01209 January 9, 2026

[0067] P2024, 0920 WO N - 12 -

[0068] least one of the sensing signals SS11, SS22, SS12, SS21 for further processing by a data processing circuit 50 in dependence on a signal comparison of each of the sensing signals SS11, SS22, SS12, SS21 of the signal paths SP1, SP2, SP3 and SP4 with the accelerometer output signal AS . The sensing signals SS11, SS22, SS12 and SS21 received at the receiver side 12 may be transferred to the controller 30 for further evaluation via an (analog) front end 40. The controller 30 is configured to select an optimal signal path from the multitude of possible signal paths, particularly at least one of the signal transmitter devices to generate the sensing signal less affected by motion and at least one of the signal receiver devices to receive the less affected sensing signal .

[0069] In the illustrated embodiment of Figures 1 and 2, the sensor device 1 is designed as an optical sensor, for example a PPG (photoplethysmography) sensor device . The signal transmitterreceiver module 10 is configured as an optical transmitterreceiver module which comprises (optical) signal transmitter devices LED1, LED2 at the transmitter side 11 and (optical) signal receiver devices PD1, PD2 at the receiver side 12. The (optical) signal transmitter devices LED1, LED2 at the transmitter side 11 can be configured as light-emitting diodes . The (optical) signal receiver devices PD1, PD2 at the receiver side 12 can be configured as photodiodes .

[0070] The (optical) signal transmitter-receiver module 10 is configured so that a respective one of (optical) signal paths SP1, SP2, SP3 and SP4 is created from each of the (optical) signal transmitter devices LD1, LD2 to each of the (optical) signal receiver devices PD1, PD2 . For example, a first (optical) signal path SP1 is created between (optical) signal2024PF01209 January 9, 2026

[0071] P2024, 0920 WO N - 13 -

[0072] transmitter device LED1 and (optical) signal receiver device PD1 for transferring an (optical) sensing signal SS11 which is emitted by the (optical) signal transmitter device LED1 and received by the (optical) receiver device PD1 .

[0073] A second (optical) signal path SP2 is created between (optical) signal transmitter device LD2 and (optical) signal receiver device PD2 for transferring an (optical) sensing signal SS22 generated by (optical) signal transmitter device LD2 and received by (optical) signal receiver device PD2 .

[0074] Furthermore, a third (optical) signal path SP3 is created between (optical) signal transmitter device LD1 and (optical) signal receiver device PD2 for transferring an (optical) sensing signal SS12 emitted by (optical) signal transmitter device LD1 and received by (optical) signal receiver device PD2 .

[0075] A fourth (optical) signal path SP4 is created between (optical) signal transmitter device LD2 and (optical) signal receiver device PD1 for transferring an (optical) sensing signal SS21 that is emitted by (optical) signal transmitter device LD2 and received by (optical) signal receiver device PD1 .

[0076] As illustrated in Figure 1, the (optical) sensor device 1 may be arranged for radiating the skin 2 of a subj ect with optical light signals generated by the (optical) signal transmitter devices LED1, LED2 that are reflected by substances, for example blood cells, or components, for example veins, inside the layers of skin 2 in the direction of the (optical) signal receiver devices PD1, PD2 . The reflected light or the sensing signals SS11, SS22, SS12, SS212024PF01209 January 9, 2026

[0077] P2024, 0920 WO N 14

[0078] received by the (optical) receiver devices PD1, PD2 can be evaluated by the data processing device 50 for monitoring vital signs, for example a heartrate, an oxygen saturation or a blood pressure of a patient .

[0079] The sensor device 1 in the configuration of a PPG sensor is usually worn on the body of a person for vital sign monitoring. If the person is moving, and thus the sensor device 1 is also in motion, the received sensing signals SS11, SS22, SS12, SS21 are negatively affected by the motion which often introduces unwanted artifacts and effects that can significantly degrade the quality and reliability of data obtained by evaluating the received sensing signals, leading to erroneous data / output signals and interpretations of vital sign measurements .

[0080] In the proposed approach of the sensor device 1, the controller 30 is configured to select only said one of the received sensing signals SS11, SS22, SS12, SS21 for further signal processing by the data processing device 50 that is less affected by motion than other received sensing signals . The controller 30 can also be configured to select multiple received sensing signals SS11, SS22, SS12, SS21 for further processing, if these signals are less affected by motion than other received sensing signals .

[0081] In order to identify and select the received sensing signal (s) least affected by motion influences for further signal processing by the data processing device 50, the controller 30 is configured to compare the received sensing signals SS11, SS22, SS12, SS21 with the accelerometer output signal AS .2024PF01209 January 9, 2026

[0082] P2024, 0920 WO N 15

[0083] The controller 30 may be configured to carry out the signal comparison by performing a similarity analysis of each of the sensing signals SS11, SS22, SS12, SS21 with the accelerometer output signal AS . The received sensing signal / signals that shows / show a lower similarity to the accelerometer output signal AS than other received sensing signals is / are selected for further signal processing, while the other received sensing signals are not considered for further signal processing .

[0084] An algorithm for signal selection may be incorporated in firmware or software as a signal selection module that can be used by the controller for carrying out the signal comparison or performing the similarity analysis . The signal selection module could be a f irmware / sof tware module, being one subpart of the controller' s f irmware / sof tware . The

[0085] f irmware / sof tware can comprise a lot of other parts for performing other tasks as well .

[0086] This intelligent signal selection from a multitude of possible sensing signals received by the receiver devices on the receiver side 12 of the signal transmitter-receiver module 10 can reduce disturbing influences and artifacts on the further signal processing due to movement of the sensor device, leading to optimal signal quality and a maximum motion entanglement . Moreover, this enables the discontinuation of a signal collection from unused signal sources / signal transmitter devices, either permanently or intermittently, depending on whether the similarity analysis is performed permanently or intermittently by the controller 30, thereby reducing energy consumption and conserving battery power, especially when the sensor device 1 is configured as a wearable device . Adaptively and continuously2024PF01209 January 9, 2026

[0087] P2024, 0920 WO N - 16 -

[0088] choosing and processing the least motion distorted signal source (s) leads to best performance in subsequent signal processing .

[0089] Intelligent signal selection is explained below for a sensor device being embodied as an optical sensor device 1 . The optical sensor device may be designed as a wearable device, for example, a f itness / health monitoring device, for example a smart watch which can be worn on a person' s wrist 3, as illustrated in Figure 3. For the sake of simplicity, only two optical paths SP11 and SP22 that penetrate the layers of a person' s skin 2 are shown. Optical path SP11 for transferring sensing signal SS11, for example, is created between signal transmitter device LED1 and signal receiver device PD1, and optical path SP22 for transferring sensing signal SS22, for example, is created between signal transmitter device LED2 and signal receiver device PD2 .

[0090] Figure 4 shows the sensing signals SP11 and SP22 received at the receiver side 12 of the signal transmitter-receiver module 10, and the accelerometer output signal AS having spatial components AccX, AccY and AccZ .

[0091] In the example shown, the signal comparison between the sensing signals SS11, SS22 and the accelerometer output signal AS is carried out by performing a similarity analysis in the frequency domain of the signals . Figure 5 shows the frequency spectra of the sensing signals SS11, SS22 and the frequency spectra of the spatial components AccX, AccY and AccZ of the accelerometer output signal AS .

[0092] In order to find sensing signals that are particularly affected by the motion, the controller 30 searches for a2024PF01209 January 9, 2026

[0093] P2024, 0920 WO N 17

[0094] frequency match of FFT components in certain frequency bands for a certain time window. Referring to Figure 5, FFT spectrum of sensing signal SS22 shows similarity to accelerometer output signal component AccZ in the spatial Z-direction in the FFT spectra of the accelerometer output signal AS (similarity criterion: match of prominent frequencies) . As a result, controller 30 selects sensing signal SS11 for further processing, as it shares less similarities with the confounder, whereas sensing signal SS22 is excluded from further processing.

[0095] Consequently, the signal transmitter-receiver module 10 may operate signal transmitter device LED1 in an activated state to continue with the generation of sensing signal SS11, whereas signal transmitter device LED2 may be operated by signal transmitter-receiver module 10 in a deactivated state to interrupt the generation of sensing signal SS22.

[0096] The sensor device 1 can be designed as a multifunctional device, for example as a f itness / health monitoring device or a smart watch, as shown in Figure 3. Depending on the various applications of such a sensor device, the controller 30 can be designed to perform other tasks in addition to selecting appropriate signals to be further processed, for example for signal buffering, the intelligent signal selection and for further processing the selected signals, such as for estimation of a heart rate .

[0097] The similarity analysis between the different sensing signals transferred via the various optical paths and the accelerometer output signal may be performed using different similarity parameters, for example frequency domain parameters of the sensing signals and accelerometer output2024PF01209 January 9, 2026

[0098] P2024, 0920 WO N 18

[0099] signal, time domain parameters of the sensing signals and accelerometer output signal, statistical parameters of the sensing signals and accelerometer output signal, entropybased parameters of the sensing signals and accelerometer output signal and / or non-linear dynamic parameters of the sensing signals and accelerometer output signal .

[0100] Further, also variables derived from the sensing signals and accelerometer output signal can be used by the controller 30 for signal comparison, for example similarity of the first derivatives, similarity of low pass filtered variants, etc . Basically, all possible transformations or calculated representations of the sensing signals and the accelerometer output signal can be used to assert similarity.

[0101] According to a possible embodiment, a normalization of the sensing signals and the accelerometer output signal may be performed to allow for a signal comparison. This may be due to the different character of the sensing signals and the accelerometer output signal . For example, the amplitudes and the baseline level (or DC part) of the sensing signals and the accelerometer output signal are different .

[0102] A brief overview of the above-mentioned parameters derived from the sensing signals and the accelerometer output signal which may be used for the similarity analysis is given below.

[0103] Frequency domain parameters concern features derived from the frequency content of a signal, usually obtained via Fourier transform or similar techniques . Examples are power spectral density (PSD) which describes the power distribution of a signal over frequencies, band power which described power within specific frequency bands ( for example subdivided2024PF01209 January 9, 2026

[0104] P2024, 0920 WO N 19

[0105] frequency bands in 0.25 Hz steps) . A similarity analysis based on frequency domain parameters may be performed by searching for matching frequency peaks in the FFT spectra of the sensing signals and the accelerometer output signal (comparing relative peak powers) , as explained above, for example, with reference to Figure 5.

[0106] Time domain parameters concern features that describe signal behaviour directly in the time domain. Examples are zerocrossing rate (ZCR) defined as the rate at which a signal changes sign, often used in audio and vibration analysis, and root mean square (RMS) which is a parameter that measures the magnitude of a varying signal .

[0107] Statistical parameters concern summary statistics that provide insight into the distribution, variability, and structure of a signal . Examples are standard deviation (SD) which represents the signal' s variability, coefficient of variation (CV) defined as standard deviation relative to the mean which is useful for assessing variability in signals with different mean levels, and Pearson correlation (PC) which quantifies the linear relationship between two variables or signals, describing their association.

[0108] Entropy-based and complexity-based parameters are measures that assess the randomness, complexity or predictability of a signal . Examples are approximate entropy (ApEn) which quantifies the regularity of patterns in a signal, sample entropy (SampEn) which is a refinement of ApEn (often more robust and less dependent on signal length) , Shannon entropy (SE) which measures the uncertainty or randomness of data, permutation entropy (PE) which is a measure used to quantify the complexity of time series data, Renyi entropy, which is a2024PF01209 January 9, 2026

[0109] P2024, 0920 WO N 20

[0110] generalization of Shannon entropy, and is sensitive to changes in signal probability distributions, and mutual information (MI ) which measures the amount of shared information between two signals, quantifying how much knowing one signal reduces uncertainty about the other .

[0111] A similarity analysis based on non-linear dynamics and chaos analysis uses parameters that characterize chaotic or nonlinear behaviour in signals, often using fractal or dynamic signal approaches . Examples are Lyapunov exponent which indicates the rate of divergence of nearby points, useful for detecting chaotic behaviour; fractal dimension ( for example, Higuchi' s or Katz' s) which quantifies the complexity and "roughness" of a signal; correlation dimension which estimates the fractal dimension, giving insight into the signal' s dynamic properties; Hurst exponent which reflects the long-term memory of a signal, with values greater than 0.5 suggesting persistent behavior; and Poincare plot analysis which is a graphical representation of signal' s dynamic properties, often used in heartrate variability analysis . Poincare plot analysis is used to detect patterns and irregularities in the time series, revealing information about the stability of dynamical systems, providing insights into periodic orbits, chaotic motions, and bifurcations . From the plot, the standard deviations SD1 and SD2 are derived (along specified axis in the plot) .

[0112] The principle of the proposed approach of a sensor device with intelligent sensing signal selection has been explained above with reference to Figures 1 to 5 using the example of an optical sensor device, for example PPG sensor device with evaluation of (optical) PPG signals . However, an optical sensor device using PPG signals for signal processing is just2024PF01209 January 9, 2026

[0113] P2024, 0920 WO N - 21 -

[0114] one example of the proposed configuration for a sensor device .

[0115] The proposed approach is basically applicable to every signal that can be negatively affected by motion, given a multitude of signal sources are available for selection ( for example two or more impedance measurement signals, two or more ECG signals, two or more audio signals, an ultra sound device having two or more ultra sound paths, a sound-based application having two or more sound paths, etc . ) .

[0116] Especially, the proposed approach of a sensor device with intelligent sensing signal selection may be used for a plurality of wearable applications, where motion is considered a confounding variable and one is interested in working with the least affected signal available, in order to obtain the most accurate output after subsequent processing of the signal, for example PPG-based heartrate, PPG-based oxygen saturation SpO2, PPG-based blood pressure estimation, PPG-based respiration rate estimation, ECG-based heartrate, etc .

[0117] The embodiments of the proposed sensor device with sensing signal selection disclosed herein have been discussed for the purpose of familiarizing the reader with novel aspects of the sensor device . Although preferred embodiments have been shown and described, many changes, modifications, equivalents and substitutions of the disclosed concept may be made by one having skill in the art without unnecessarily departing from the scope of the claims .

[0118] In particular, the design of the proposed sensor device is not limited to the disclosed embodiments, and gives examples2024PF01209 January 9, 2026

[0119] P2024, 0920 WO N 22

[0120] of many alternatives as possible for the features included in the embodiments discussed.

[0121] Furthermore, features recited in separate dependent claims may be advantageously combined. Moreover, reference signs used in the claims are not limited to be construed as limiting the scope of the claims .

[0122] Moreover, as used herein, the term "comprising" does not exclude other elements . In addition, as used herein, the article "a" is intended to include one or more than one component or element, and is not limited to be construed as meaning only one .

[0123] This patent application claims the priority of German patent application with application No . 102025102227.5, the disclosure content of which is hereby incorporated by reference .2024PF01209 January 9, 2026

[0124] P2024, 0920 WO N - 23 -

[0125] References

[0126] 1 sensor device

[0127] 2 skin

[0128] 3 watch

[0129] 4 wrist

[0130] 10 signal transmitter-receiver module 20 accelerometer

[0131] 30 controller

[0132] 40 frontend module

[0133] 50 data processing device

[0134] 11 transmitter side

[0135] 12 receiver side

[0136] SP1, SP4 signal path

[0137] SS11, SS22, SS12, SS21 sensing signal

[0138] AS accelerometer output signal

Claims

2024PF01209 January 9, 2026P2024, 0920 WO N - 24 -Claims1. A sensor device with sensing signal selection, comprising: - a signal transmitter-receiver module ( 10) to provide a plurality of signal paths (SP1, SP2, SP3, SP4 ) for respectively transmitting a sensing signal (SS11, SS22, SS12, SS21 ) between a transmitter side ( 11 ) and a receiver side ( 12 ) of the signal transmitter-receiver module ( 10) ,- an accelerometer (20) to provide an accelerometer output signal (AS) representing an acceleration of the sensor device ( 1 ) ,- a controller (30) being configured to select at least one of the sensing signals (SS11, SS22, SS12, SS21 ) for further processing in dependence on a signal comparison of each of the sensing signals (SS11, SS22, SS12, SS21 ) with the accelerometer output signal (AS) .

2. The sensor device of claim 1,wherein the controller (30) is configured to carry out the signal comparison by performing a similarity analysis of each of the sensing signals (SS11, SS22, SS12, SS21 ) with the accelerometer output signal (AS) .

3. The sensor device of claim 2,wherein the controller (30) is configured to perform the similarity analysis by performing a correlation analysis of each of the sensing signals (SS11, SS22, SS12, SS21 ) with the accelerometer output signal (AS) .

4. The sensor device of claim 2 or 3,- wherein the controller (30) is configured to perform the similarity analysis by calculating a respective variable derived from each of the sensing signals (SS11, SS22, SS12,2024PF01209 January 9, 2026P2024, 0920 WO N 25SS21 ) and a variable derived from the accelerometer output signal (AS ) ,- wherein the controller (30) is configured to perform the similarity analysis by comparing the respective variable derived from each of the sensing signals (SS11, SS22, SS12, SS21 ) with the variable of the accelerometer output signal (AS) .

5. The sensor device of claim 4,wherein the controller (30) is configured to determine the respective variables by calculating a transformation or representation of each of the sensing signals (SS11, SS22, SS12, SS21 ) and the accelerometer output signal (AS) .

6. The sensor device of any of the claims 2 to 5,- wherein the controller (30) is configured to perform the similarity analysis by calculating respective frequency domain parameters of each of the sensing signals (SS11, SS22, SS12, SS21 ) and frequency domain parameters of the accelerometer output signal (AS) ,- wherein the controller (30) is configured to perform the similarity analysis by comparing the respective frequency domain parameters of each of the sensing signals (SS11, SS22, SS12, SS21 ) with the frequency domain parameters of the accelerometer output signal (AS) .

7. The sensor device of any of the claims 2 to 6,- wherein the controller (30) is configured to perform the similarity analysis by calculating a respective FFT spectrum of each of the sensing signals (SS11, SS22, SS12, SS21 ) and an FFT spectrum of the accelerometer output signal (AS) , - wherein the controller (30) is configured to perform the similarity analysis by comparing the respective FFT spectrum2024PF01209 January 9, 2026P2024, 0920 WO N 26of each of the sensing signals (SS11, SS22, SS12, SS21 ) with the FFT spectrum of the accelerometer output signal (AS) .

8. The sensor device of any of the claims 2 to 7,- wherein the controller (30) is configured to perform the similarity analysis by calculating respective time domain parameters of each of the sensing signals (SS11, SS22, SS12, SS21 ) and time domain parameters of the accelerometer output signal (AS ) ,- wherein the controller (30) is configured to perform the similarity analysis by comparing the respective time domain parameters of each of the sensing signals (SS11, SS22, SS12, SS21 ) with the time domain parameters of the accelerometer output signal (AS) .

9. The sensor device of any of the claims 2 to 8,- wherein the controller (30) is configured to perform the similarity analysis by calculating respective statistical parameters of each of the sensing signals (SS11, SS22, SS12, SS21 ) and statistical parameters of the accelerometer output signal (AS ) ,- wherein the controller (30) is configured to perform the similarity analysis by comparing the respective statistical parameters of each of the sensing signals (SS11, SS22, SS12, SS21 ) with the statistical parameters of the accelerometer output signal (AS) .

10. The sensor device of any of the claims 2 to 9,- wherein the controller (30) is configured to perform the similarity analysis by calculating respective entropy-based parameters of each of the sensing signals (SS11, SS22, SS12, SS21 ) and entropy-based parameters of the accelerometer output signal (AS) ,2024PF01209 January 9, 2026P2024, 0920 WO N 27- wherein the controller (30) is configured to perform the similarity analysis by comparing the respective entropy-based parameters of each of the sensing signals (SS11, SS22, SS12, SS21 ) with the entropy-based parameters of the accelerometer output signal (AS) .

11. The sensor device of any of the claims 2 to 10,- wherein the controller (30) is configured to perform the similarity analysis by calculating respective nonlinear dynamic parameters of each of the sensing signals (SS11, SS22, SS12, SS21 ) and nonlinear dynamic parameters of the accelerometer output signal (AS) ,- wherein the controller (30) is configured to perform the similarity analysis by comparing the respective nonlinear dynamic parameters of each of the sensing signals (SS11, SS22, SS12, SS21 ) with the nonlinear dynamic parameters of the accelerometer output signal (AS) .

12. The sensor device of any of the claims 1 to 11, wherein the controller (30) is configured to perform the similarity analysis permanently or intermittently.

13. The optical sensor device of any of the claims 1 to 12, wherein the controller (30) is configured to select at least a first one of the sensing signals (SS11 ) for further processing and to exclude at least a second one of the sensing signals (SS22 ) from further processing, if a lower similarity between the at least one first sensing signal (SS11 ) and the accelerometer output signal (AS) than between the at least one second sensing signal (SS22 ) and the accelerometer output signal (AS) is detected by the controller (30) .2024PF01209 January 9, 2026P2024, 0920 WO N - 28 -14. The sensor device of claim 13,- wherein the signal transmitter-receiver module ( 10) comprises at least a first and a second signal transmitter device (LED1, LED2 ) at the transmitter side ( 11 ) and at least a first and a second signal receiver device (PD1, PD2 ) at the receiver side ( 12 ) ,- wherein the signal transmitter-receiver module ( 10) is configured so that a respective one of the signal paths (SP1, SP2, SP3, SP4 ) is created from each of the at least one first and second signal transmitter device (LED1, LED2 ) to each of the at least one first and second signal receiver device (PD1, PD2 ) .

15. The sensor device of claim 14,- wherein the at least one first signal transmitter device (LED1 ) is configured for generating the first sensing signal (SS11 ) , and the at least one second signal transmitter device (LED2 ) is configured for generating the second sensing signal (SS22 ) ,- wherein the signal transmitter-receiver module ( 10) is configured to operate the at least one first signal transmitter device (LED1 ) in an activated state to continue with the generation of the first sensing signal (SS11 ) and to operate the at least one second signal transmitter device (LED2 ) in a deactivated state to interrupt the generation of the second sensing signal (SS22 ) , if a lower similarity between the at least one first sensing signal (SS11 ) and the accelerometer output signal (AS) than between the at least one second sensing signal (SS22 ) and the accelerometer output signal (AS) is detected by the controller (30) .

Images