WO2022188981A1

WO2022188981A1 - Apparatus, method and computer program for determining a property related to a quality of one or more alternative configurations of a wireless communication link

Info

Publication number: WO2022188981A1
Application number: PCT/EP2021/056206
Authority: WO
Inventors: Peter Jung; Slawomir Stanczak; Vlerar SHALA; Andreas Pfadler
Original assignee: Volkswagen Aktiengesellschaft; Fraunhofer-Gesellschaft Zur Förderung Der Angewandten Forschungde E.V.
Priority date: 2021-03-11
Filing date: 2021-03-11
Publication date: 2022-09-15
Also published as: EP4305816A1

Abstract

The present disclosure relates to an apparatus, method and computer program for determining a property related to a quality of one or more alternative configurations of a wireless communication link. The method comprises estimating channel coefficients of a wireless channel of the wireless communication link. The method comprises predicting a Signal-to- Interference-and-Noise-Ratio, SINR, of the one or more alternative configurations of the wireless communication link based on a function for predicting the SINR of an alternative configuration based on the estimated channel coefficients, the alternative configuration and a current configuration. The method further comprises determining the property related to the quality of the one or more alternative configurations based on the predicted SINR of the one or more alternative configurations of the wireless communication link.

Description

Apparatus, Method and Computer Program for Determining a Property Related to a Quality of One or More Alternative Configurations of a Wireless Communication Link

The present invention relates to an apparatus, method and computer program for determining a property related to a quality of one or more alternative configurations of a wireless communication link.

New requirements in terms of reliability and efficiency in high mobility environments, such as vehicle-to-vehicle (V2V) communication, are pushing legacy systems to their limits. Orthogonal frequency-division multiplexing (OFDM) is a popular and well-known modulation scheme, but it may suffer from substantial performance degradation and inflexibility in environments with high Doppler spreads. Consequently, novel modulation schemes may be considered and perused which are flexible, efficient and robust in doubly-dispersive channels.

Future vehicular communication systems require high reliability and efficiency under various mobility conditions. Furthermore, they are multilateral as different types of communication links exist. Vehicles are connected to infrastructure, i.e., vehicle-to-infrastructure (V2I), but also using direct vehicle-to-vehicle (V2V) communication. Especially, V2V channels are distinct compared to conventional cellular channels. For communication between high mobility users, large Doppler shifts are expected due to the large relative velocity. Legacy systems, such as OFDM, may experience considerable performance degradation under high Doppler shifts.

New modulation schemes such as orthogonal time frequency and space (OTFS) address the challenges for future communication systems. The key idea behind OTFS is to multiplex a data symbol (e.g., QAM, quadrature amplitude modulation) in the signal representation called the delay-Doppler representation. OTFS was introduced by Hadani et. al as a promising recent combination of classical pulse-shaped Weyl-Heisenberg (or Gabor) multicarrier schemes with a distinct time-frequency (TF) spreading. Data symbols are spread with the symplectic finite Fourier transform (SFFT) over the whole time-frequency grid. This particular linear pre-coding accounts for the doubly-dispersive nature of time-varying multipath channels seen as linear combinations of time-frequency shifts. Several studies show that OTFS outperforms OFDM in such situations. Other research focuses on a performance comparison of OFDM, generalized frequency division multiplexing (GFDM), and OTFS. It reveals significant advantages of OTFS in terms of bit error rate (BER) and frame error rate (FER) in relation to the others. With sufficiently accurate channel information it offers a promising increase in terms of reliability and robustness for high mobility users when using sophisticated equalizers. So far, OTFS was researched with the assumption of perfect grid-matching, often with idealized pulses violating the uncertainty principle and in many cases with ideal channel knowledge (including the cross-talk channel coefficients).

OTFS is a new modulation scheme that addresses the challenges of 5^th Generation mobile communication systems (5G). The key idea behind OTFS is to multiplex a QAM (quadrature amplitude modulation) or QPSK (Quadrature Phase Shift Keying) symbol (data) in the delay- Doppler signal representation. In order to do channel equalization, the wireless channel needs to be estimated at the receiver. This can be done by the insertion of pilots at the transmitter. The a-priory known pilot tones can be used by the receiver to estimate the channel.

Further background can be found in:

[1] T. Wang, J. G. Proakis, E. Masry, and J. R. Zeidler, “Performance degradation of OFDM systems due to Doppler spreading,” IEEE Trans on Wreless Commun., vol. 5, no. 6, pp. 1422- 1432, 2006;

[2] R. Hadani, S. Rakib, M. Tsatsanis, A. Monk, A. J. Goldsmith, A. F. Molisch, and R. Calderbank, “Orthogonal time frequency space modulation,” in 2017 IEEE Wireless Commun. and Netw. Conf. (WCNC), pp. 1-6, IEEE, 2017;

[3] R. Hadani, S. Rakib, A. F. Molisch, C. Ibars, A. Monk, M. Tsatsanis, J. Delfeld, A. Goldsmith, and R. Calderbank, “Orthogonal Time Frequency Space (OTFS) modulation for millimeter- wave communications systems,” in 2017 IEEE MTT-S Int. Microwave Symp. (IMS), pp. 681- 683, June 2017;

[4] M. Kollengode Ramachandran and A. Chockalingam, “MIMO-OTFS in High-Doppler Fading Channels: Signal Detection and Channel Estimation,” in 2018 IEEE Global Commun. Conf. (GLOBECOM), pp. 206- 212, Dec 2018;

[5] P. Raviteja, Y. Hong, E. Viterbo, and E. Biglieri, “Practical PulseShaping Waveforms for Reduced-Cyclic-Prefix OTFS,” IEEE Trans on Vehicular Technol., vol. 68, pp. 957-961, Jan 2019;

[6] A. Nimr, M. Chafii, M. Matthe, and G. Fettweis, “Extended GFDM Framework: OTFS and GFDM Comparison,” in 2018 IEEE Global Commun. Conf. (GLOBECOM), pp. 1-6, Dec 2018.

[7] W. Kozek, “Matched Weyl-Heisenberg expansions of nonstationary environments,”

1996; [8] K. Liu, T. Kadous, and A. M. Sayeed, Orthogonal time-frequency signaling over doubly dispersive channels,” IEEE Trans on Inf. Theory, vol. 50, no. 11, pp. 2583-2603, 2004;

[9] P. Jung and G. Wunder, “WSSUS pulse design problem in multicarrier transmission,” IEEE Trans on Commun., vol. 55, no. 9, pp. 1822-1822, 2007;

[10] W. Kozek and A. F. Molisch, “Nonorthogonal pulseshapes for multicarrier communications in doubly dispersive channels,” IEEE J. on Sel. Areas in Commun., vol. 16, pp. 1579-1589, Oct 1998;

[11] T. Zemen, M. Hofer, D. Loeschenbrand, and C. Pacher, “Iterative detection for orthogonal precoding in doubly selective channels,” in 2018 IEEE 29th Annual Int. Symp. on Pers., Indoor and Mobile Radio Commun. (PIMRC), pp. 1-7, IEEE, 2018;

[12] X. Ma and W. Zhang, “Fundamental limits of linear equalizers: diversity, capacity, and complexity,” IEEE Trans on Inf. Theory, vol. 54, no. 8, pp. 3442-3456, 2008;

[13] T. Zemen, M. Hofer, and D. Loeschenbrand, “Low-complexity equalization for orthogonal time and frequency signaling (OTFS),” arXiv preprint arXiv:1710.09916, 2017;

[14] A. Pfadler, P. Jung, and S. Stanczak, “Pulse-Shaped OTFS for V2X Short-Frame Commun. with Tuned One-Tap Equalization,” in WSA 2020; 24rd Int. ITG Workshop on Smart Antennas, pp. 1-6, VDE, 2020;

[15] Z. Prusa, P. L. Sondergaard, N. Holighaus, C. Wiesmeyr, and P. Balazs, “The Large Time-Frequency Analysis Toolbox 2.0,” in Sound, Music, and Motion, LNCS, pp. 419-442, Springer Int. Publishing, 2014;

[16] S. Jaeckel, L. Raschkowski, K. Borner, and L. Thiele, “QuaDRiGa: A 3D multi-cell channel model with time evolution for enabling virtual field trials,” IEEE Trans on Antennas and Propag., vol. 62, no. 6, pp. 3242- 3256, 2014;

[17] R. Hadani and S. S. Rakib, “OTFS methods of data channel characterization and uses thereof,” Sept. 132016. US Patent 9,444,514;

[18] K. Grochenig, Foundations of time-frequency analysis. Springer Science & Business Media, 2013;

[19] G. Matz, D. Schafhuber, K. Grochenig, M. Hartmann, and F. Hlawatsch, “Analysis, optimization, and implementation of low-interference wireless multicarrier systems,” IEEE Trans on Wireless Commun., vol. 6, no. 5, pp. 1921-1931, 2007;

[20] P. Raviteja, K. T. Phan, and Y. Hong, “Embedded Pilot-Aided Channel Estimation for OTFS in Delay-Doppler Channels,” IEEE Trans on Vehicular Technol., pp. 1-1, 2019;

[21] P. Bello, “Characterization of randomly time-variant linear channels,” IEEE Trans on Commun. Syst., vol. 11, no. 4, pp. 360-393, 1963;

[22] P. Jung, W. Schuele, and G. Wunder, “Robust path detection for the LTE downlink based on compressed sensing,” in 14th Int. OFDM-Workshop, Hamburg, 2009; [23] A. Pfadler, P. Jung and S. Stanczak (2020). Mobility Modes for Pulse-Shaped OTFS with Linear Equalizer. IEEE Globecom 2020, December 7-11, in Taipei, Taiwan; and

[24] Kozek, Werner, and Andreas F. Molisch. "Nonorthogonal pulseshapes for multicarrier communications in doubly dispersive channels." IEEE Journal on selected areas in communications 16.8 (1998): 1579-1589.

There may be a desire for providing an improved concept for the use of OTFS or OTFS-like modulation in real-world scenarios.

This desire is addressed by the subject-matter of the independent claims.

Embodiments are based on the finding that, in the discussion of OTFS in literature, perfect grid matching was assumed. To fully exploit diversity in OTFS, the 2D-deconvolution implemented by a linear equalizer should approximately invert the doubly-dispersive channel operation, which however is a twisted convolution. In theory this is achieved by choosing the time-frequency grid and the Gabor synthesis and analysis pulses based on the delay and Doppler spread of the channel. However, in practice one always has to balance between supporting high granularity in delay-Doppler spread and multi-user and network aspects.

Fixed, pre-defined configurations of the time-frequency grid, which are called mobility modes, may be used to reduce the complexity. Instead of dealing with hundreds of possible configurations, a limited number of different mobility modes may be used for communication.

The mobility modes may be chosen based on the channel conditions, i.e. , based on the delay spread and the Doppler spread of the channel. Various aspects of the present disclosure provide a mechanism for estimating the performance of different alternative mobility modes based on the performance of a mobility mode currently being used. For example, the performance estimation may be used to select a more appropriate mobility mode.

Various examples of the present disclosure relate to a method for determining a property related to a quality of one or more alternative configurations of a wireless communication link between a first transceiver and a second transceiver. The wireless communication link is based on a current configuration. The method comprises estimating channel coefficients of a wireless channel of the wireless communication link. The method comprises predicting a Signal-to- Interference-and-Noise-Ratio, SINR, of the one or more alternative configurations of the wireless communication link based on a function for predicting the SINR of an alternative configuration based on the estimated channel coefficients, the alternative configuration and the current configuration. The method further comprises determining the property related to the quality of the one or more alternative configurations based on the predicted SINR of the one or more alternative configurations of the wireless communication link. For example, the estimated channel coefficients may represent an instantaneous (as opposed to long-term) state of the wireless channel, and be used to predict how well the alternative configurations are suited for the wireless channel between the two transceivers, while the function for predicting the SINR of an alternative configuration is used to deduce the performance of the alternative configuration based on the performance experienced in the current configuration.

In various examples, the method comprises determining a spreading function based on the estimated channel coefficients. For example, the function for predicting the SINR of an alternative configuration may be based on the spreading function that is based on the estimated channel coefficients, the alternative configuration and the current configuration. In other words, the spreading function, which is based on the estimated channel coefficients, may be used to determine the instantaneous state of the wireless channel.

In some examples, machine-learning may be used as part of the prediction. For example, the SINR of the one or more alternative configurations may be tracked and predicted with the help of a machine-learning model. The machine-learning model may be used to estimate a future spreading function based on a plurality of spreading functions. For example, the SINR of the one or more alternative configurations may be predicted for a point in time in the future based on the predicted spreading function. In some examples, a Long Short-Term Memory or a kernel- based machine-learning algorithm may be used.

For example, the channel coefficients may be estimated based on a spreading function of the wireless channel obtained using a pilot signal. For example, based on a difference between the pilot signal, as transmitted, and the pilot signal, as received, the channel coefficients may be determined.

In some examples, the function for predicting the SINR of an alternative configuration is based on the estimated channel coefficients, a time-frequency-grid-configuration and a pulse configuration of the current configuration and a time-frequency-grid-configuration and a pulse configuration of the alternative configuration. In other words, a relationship between the time- frequency-grid-configuration and the pulse configuration of the current configuration and the time-frequency-grid-configuration and the pulse configuration of the alternative configuration may be used to deduce the performance of the alternative configuration based on the performance experienced in the current configuration.

Alternatively, the function for predicting the SI NR of an alternative configuration is based on the estimated channel coefficients, a time-frequency-grid-configuration and a pulse configuration of the current configuration and a time-frequency-grid-configuration of the alternative configuration. In this case, merely the time-frequency-grid-configuration of the alternative configuration may be considered in relation to the time-frequency-grid-configuration and the pulse configuration of the current configuration and be used to deduce the performance of the alternative configuration based on the performance experienced in the current configuration.

Alternatively, the function for predicting the SINR of an alternative configuration is based on the estimated channel coefficients, a time-frequency-grid-configuration of the current configuration and a time-frequency-grid-configuration of the alternative configuration. In this case, merely the time-frequency-grid-configurations of the current and alternative configuration may be considered and be used to deduce the performance of the alternative configuration based on the performance experienced in the current configuration.

Determining the property related to the quality of the one or more alternative configurations may comprise determining a predicted performance of the one or more alternative configurations based on the predicted SINR of the one or more alternative configurations of the wireless communication link. For example, a Bit-Error Rate (BER), a Frame-Error Rate (FER) or a latency of the wireless communication link may be determined.

For example, the predicted performance of the one or more alternative configurations may be determined for two or more different modulation and coding schemes, for two or more different waveforms, or for two or more different Multiple-Input-Multiple-Output configurations. In other words, not only the pulse shape and/or the time-frequency-grid configuration may be considered, but also other factors that influence the real-world performance of the respective configurations.

In some examples, the method comprises switching the configuration of the wireless communication link based on the determined property of the one or more alternative configurations. For example, if the predicted performance of an alternative configuration is vastly better than the performance of the current configuration, the configuration may be switched. In general, the wireless communication link may be based on a multicarrier transmission-based wireless communication system. For example, the wireless communication link may be based on one of an Orthogonal Frequency Division Multiplexing-based wireless communication system, an Orthogonal Time-Frequency-Space-based wireless communication system a Generalized Frequency Division Multiplexing-based wireless communication system, and a Filter Bank Multi-Carrier-based wireless communication system. Such multicarrier transmission- based wireless communication systems may benefit from the selection of an appropriate mobility mode (i.e., configuration).

In some examples, the one or more alternative configurations are different from the current configuration with respect to a time-frequency-grid-configuration of the wireless communication link. Alternatively, the one or more alternative configurations are different from the current configuration with respect to a time-frequency-grid-configuration of the wireless communication link and/or with respect to a pulse configuration of the wireless communication link. In other words, in some cases, the pulse configuration may be varied in addition to the time-frequency- grid configuration of the respective configurations.

Various examples of the present disclosure further provide a computer program having a program code for performing the above method, when the computer program is executed on a computer, a processor, or a programmable hardware component.

Various examples of the present disclosure further provide an apparatus comprising one or more interfaces for communicating in a mobile communication system. The apparatus further comprises a control module configured to carry out the above method.

Some other features or aspects will be described using the following non-limiting embodiments of apparatuses or methods or computer programs or computer program products by way of example only, and with reference to the accompanying figures, in which:

Figs. 1a and 1b show flow charts of example of a method for determining a property related to a quality of one or more alternative configurations of a wireless communication link;

Fig. 1c shows a block diagram of an example of an apparatus for determining a property related to a quality of one or more alternative configurations of a wireless communication link; Fig. 2 shows a schematic diagram of a machine-learning model used as part of a determination of a Signal-to-lnterference-and-Noise-Ratio; and

Figs. 3a to 3g show various schematic diagrams illustrating the mathematical background behind mobility modes.

Various example embodiments will now be described more fully with reference to the accompanying drawings in which some example embodiments are illustrated. In the figures, the thicknesses of lines, layers or regions may be exaggerated for clarity. Optional components may be illustrated using broken, dashed or dotted lines.

Accordingly, while example embodiments are capable of various modifications and alternative forms, embodiments thereof are shown by way of example in the figures and will herein be described in detail. It should be understood, however, that there is no intent to limit example embodiments to the particular forms disclosed, but on the contrary, example embodiments are to cover all modifications, equivalents, and alternatives falling within the scope of the invention. Like numbers refer to like or similar elements throughout the description of the figures.

As used herein, the term, "or" refers to a non-exclusive or, unless otherwise indicated (e.g., “or else” or “or in the alternative”). Furthermore, as used herein, words used to describe a relationship between elements should be broadly construed to include a direct relationship or the presence of intervening elements unless otherwise indicated. For example, when an element is referred to as being “connected” or “coupled” to another element, the element may be directly connected or coupled to the other element or intervening elements may be present.

In contrast, when an element is referred to as being “directly connected” or “directly coupled” to another element, there are no intervening elements present. Similarly, words such as “between”, “adjacent”, and the like should be interpreted in a like fashion.

The terminology used herein is for the purpose of describing particular embodiments only and is not intended to be limiting of example embodiments. As used herein, the singular forms “a,” “an” and “the” are intended to include the plural forms as well, unless the context clearly indicates otherwise. It will be further understood that the terms “comprises,” “comprising,” “includes” or “including,” when used herein, specify the presence of stated features, integers, steps, operations, elements or components, but do not preclude the presence or addition of one or more other features, integers, steps, operations, elements, components or groups thereof. Unless otherwise defined, all terms (including technical and scientific terms) used herein have the same meaning as commonly understood by one of ordinary skill in the art to which example embodiments belong. It will be further understood that terms, e.g., those defined in commonly used dictionaries, should be interpreted as having a meaning that is consistent with their meaning in the context of the relevant art and will not be interpreted in an idealized or overly formal sense unless expressly so defined herein.

Embodiments of the present disclosure relate to wireless communication devices, such as a base station and a mobile transceiver, and to corresponding methods, apparatuses and computer programs. In the following, two wireless communication devices, i.e. wireless transceivers, may be assumed that communicate with each other, e.g., two mobile transceivers or a base station and a mobile transceiver. This communication is usually performed using wireless transmissions that are exchanged between the two wireless communication devices over a (wireless) channel. In at least some embodiments, the channel may be assumed to be a doubly-dispersive channel.

The communication may be performed using so-called data frames, which may be considered to be transmitted in a time frequency plane using one or more time slots and using one or more carrier frequencies, wherein the time slots span across the time dimension of the time-frequency plane, and wherein the carrier frequencies span across the frequency dimension of the time- frequency plane. This time-frequency plane can be used to model a (logical) grid that spans via the time dimension and the frequency dimension. This is a logical construct, which is, during transmission of the data frames, mapped to the time slots and carrier frequencies. In general, this grid in the time-frequency plane is delimited by the bandwidth range being used to transmit the data frame, and by the time that is used to transmit the frame (the time being subdivided into the one or the plurality of time slots). Accordingly, in embodiments, each data frame being transmitted via a wireless communication link may be transmitted based on a two-dimensional grid in a time-frequency plane having a time dimension resolution and a frequency dimension resolution.

Grids (in the time-frequency plane and in the delay-Doppler plane) may be used to represent the signals. In multicarrier transmission-based wireless communication systems, computationally feasible equalizers may suffer from mismatched time-frequency grids. Parity may be achieved with perfect gird matching of the Gabor synthesis and analysis pulses with the delay and Doppler spread of the channel. However, this might not be achieved in practice due to the varying mobility of users, and correspondingly changing channels. This may lead to performance degradation (higher error rates). In many cases, this may be caused by a mismatch of the grid, as perfect grid matching is assumed in theoretical studies on multicarrier transmission-based wireless communication systems, such as OTFS, OFDM and FBMC. Unfortunately, grid mismatch may cause significant performance degradation.

To obtain an improved performance, a time resolution and frequency resolution for the grid in the time-frequency plane that matches the channel that is used for the communication between the wireless communication devices may be chosen. Such a time resolution and frequency resolution for the grid in the time-frequency plane that matches the channel that is used for the communication between the wireless communication devices may be denoted an ideal time- frequency-grid configuration for the communication over the wireless communication link. For example, in different scenarios, signals transmitted via the channel may experience different amounts of delay spread and Doppler spread. To account for such different channels, the grid may be chosen such that the respective properties of the channel are taken into account. For lower relative velocities, less resolution in the time domain may be required, and a higher resolution in the frequency domain may be desired if higher delays occur. For example, at higher relative velocities, a grid having a higher resolution (i.e., more points) in the time dimension may be advantageous (to allow for a higher Doppler spread), while at lower relative velocities, a grid having a higher resolution (i.e., more points) in the frequency dimension may be advantageous.

Various embodiments of the present disclosure relate to a method for instantaneous predicting the best mobility mode based on the channel estimation for multicarrier transmission systems.

Mobility modes are proposed with distinct grid and pulse matching for different doubly dispersive channels (see e.g., the mathematical model shown in connection with Figs. 3a to 3g, and [23]). To account for remaining self-interference, the minimum mean square error (MMSE) linear equalizer may be tuned without the need of estimating channel cross-talk coefficients, as shown in [9], [24] In [23], the results indicate that with an appropriate mobility mode, the potential OTFS gains can be indeed achieved with linear equalizers to significantly outperform OFDM. It should be noted that mobility modes are not only applicable for OTFS, but any other multicarrier schemes, such as OFDM. In [23], the concept of mobility modes was discussed. However, the prediction of the best mode was left to future work.

For example, a vehicle or mobile transceiver may currently be using a certain mobility mode and wants to predict how performance would be with other mobility modes. Therefore, the user wants to predict the radio channel and the best mobility mode in the future to cope with the effect of the channel. The prediction, then can be used to decide either the mode should be switched or not.

Various examples of the present disclosure address the prediction of a better, or the best, mobility mode based on a SINR prediction of the current used mode and the channel estimation and collected time series of the past, for example using a Long Short-Term Memory (LSTM).

Figs. 1a and 1b show flow charts of example of a (computer-implemented) method for determining a property related to a quality of one or more alternative configurations of a wireless communication link between a first transceiver 100 and a second transceiver 200 (as shown in Fig. 1c). The wireless communication link is based on a current configuration. The method comprises estimating 110 channel coefficients of a wireless channel of the wireless communication link. The method comprises predicting 130 a Signal-to-lnterference-and-Noise- Ratio, SINR, of the one or more alternative configurations of the wireless communication link based on a function for predicting the SINR of an alternative configuration based on the estimated channel coefficients, the alternative configuration and the current configuration. The method comprises determining 140 the property related to the quality of the one or more alternative configurations based on the predicted SINR of the one or more alternative configurations of the wireless communication link.

Fig. 1c shows a block diagram of an example of a corresponding apparatus 10 for determining the property related to a quality of one or more alternative configurations of the wireless communication link. The apparatus 10 comprises one or more interfaces 12 for communicating in a mobile communication system. The apparatus 10 comprises a control module 14 configured to carry out the method of Fig. 1a and/or 1b. Fig. 1c further shows the first transceiver 100 comprising the apparatus 10. Alternatively or additionally, the second transceiver 200 may comprise the apparatus 10. For example, one of the first and second transceiver may be a base station (i.e., base station transceiver), and the other may be a mobile transceiver, or both the first and second transceiver may be mobile transceivers. For example, the method may be performed (entirely) by one of the transceivers, e.g., by the base station or by the mobile transceiver. For example, the mobile transceiver or mobile transceivers may be a vehicle or vehicles. Consequently, the wireless communication link may be a wireless communication link according to a wireless vehicular communication protocol. Various embodiments of the present disclosure relate to a method, apparatus and computer program for determining the property related to a quality of the one or more alternative configurations of the wireless communication link between the first transceiver 100 and the second transceiver 200. In other words, embodiments relate to the prediction of a quality of the one or more alternative configurations of the wireless communication link. For example, the property related to the quality of the one or more alternative configurations may relate to a performance of the respective alternative configuration, e.g., with respect to Bit Error Rate (BER), Frame Error Rate (FER) or latency. In other words, the quality of the one or more alternative configurations may be predicted. This prediction may subsequently be used to select one of the one or more alternative configurations as configuration to use for the wireless communication link, e.g., instead of the current configuration. For example, the one or more alternative configurations may be a plurality of alternative configurations. In connection with Figs. 3a to 3g, seven different configurations (i.e. , mobility modes) are shown. For example, one of the configurations may be the current configuration, and the other six configurations may be the one or more alternative configurations.

A configuration within the context of the present disclosure relates to transmissions parameters being used for communicating using the wireless communication link. In the context of the present disclosure, the different configurations are also denoted “mobility modes”, which comprise two aspects - a time-frequency-grid being used, and a pulse configuration being used. In general, the first and second transceiver may support a plurality of pre-defined configurations (mobility modes), e.g., seven mobility modes as introduced above. For example, the first and second transceiver may support at most 20 (or at most 16, or at most 10) different configurations. In other words, the first transceiver and the second transceiver may support a plurality of pre-defined configurations for communicating over the wireless communication link. Accordingly, the plurality of pre-defined configurations, which comprise the current configuration and the one or more alternative configurations, may be fixed, i.e. immutable. For example, the plurality of pre-defined configurations, and therefore the current configuration and the one or more alternative configurations, may be defined by a communication standard specifying the communication via the wireless communication link, e.g., a communication standard defined by the 3^rd-generation partnership project (3GPP). The configuration mode being used to communicate over the wireless communication link may be selected from the plurality of pre- defined configurations.

As discussed above, grids (in the time-frequency plane and in the delay-Doppler plane) may be used to represent the signals being transmitted via the wireless communication link. Thus, a first aspect of the configuration may relate to a configuration for a time-frequency-grid being used for the wireless communication link, i.e., a time-frequency-grid configuration of the wireless communication link. For example, the time-frequency configuration may define the parameters of the grid being used ( A := grid parameters), i.e., time and frequency shifts T and F. Consequently, the one or more alternative configurations may be different from the current configuration (and/or with respect to each other) with respect to a time-frequency-grid- configuration of the wireless communication link.

A time-frequency-grid may be implemented using certain types of wireless communication systems that are based on the concurrent use of multiple carrier frequencies. Consequently, the wireless communication link may be based on a multicarrier transmission-based wireless communication system, such as an Orthogonal Frequency Division Multiplexing (OFDM)-based wireless communication system, an Orthogonal Time-Frequency-Space (OTFS)-based wireless communication system a Generalized Frequency Division Multiplexing (GFDM)-based wireless communication system, and a Filter Bank Multi-Carrier (FBMC)-based wireless communication system.

A second aspect of the configuration may relate to pulse parameters being used for the wireless communication link, such as standard deviation of the pulse, pulse coefficients of the pulse, a length of the pulse, and a type (e.g., Gaussian) of the pulse, i.e., a pulse configuration of the wireless communication link. Consequently, the one or more alternative configurations may be different from the current configuration with respect to a time-frequency-grid-configuration of the wireless communication link and/or with respect to a pulse configuration of the wireless communication link.

The wireless communication link is based on the current configuration. In other words, the time- frequency-grid configuration and the pulse configuration of the current configuration may currently be used to communicate via the wireless communication link.

The proposed concept is based on determining the (instantaneous, current) channel coefficients of the wireless channel between the first and the second transceiver (also denoted S), and using the knowledge about the channel coefficients, together with knowledge about the current configuration (denoted c^*) to be used and one of the alternative configurations (denoted c) to predict the quality of the alternative configuration. In other words, the performance of the alternative configurations (mobility modes) c is predicted, for the current time or for a point in time in the future, based on the current mode c ^*:

A first parameter being used is the channel coefficients h, which are represented as spreading function S in the formula. Consequently, the method comprises estimating 110 the channel coefficients of a wireless channel of the wireless communication link. Optionally, the method comprises determining 120 the spreading function based on the estimated channel coefficients. In this case, the function for predicting the SINR of an alternative configuration may be based on the spreading function that is based on the estimated channel coefficients, the alternative configuration and the current configuration.

In general, the channel coefficients characterize the impulse response of the wireless channel. They can be determined by transmitting a known pilot signal, and comparing the pilot signal, as transmitted, to the pilot signal, as received. For example, the channel coefficients may be estimated based on a spreading function of the wireless channel obtained using a pilot signal (see e.g., formula (17)). In this context, the spreading function may define a function with an output comprising multiple delayed, attenuated and Doppler-shifted versions of an input provided to the function. In other words, the spreading function may represent the delay spread and Doppler spread of the wireless channel (in addition to the attenuation experienced by signals on the channel). For example, the channel coefficients may be indicative of a current delay spread and Doppler spread of the wireless channel between the first and second receiver. For example, the spreading function S may be estimated based on a mapping from the estimated channel coefficients h: S = f(h).

In the following, the SINR of the one or more alternative configurations is predicted. In more general terms, the method may comprise providing an interface for predicting/determining the SINR of a configuration of the plurality of pre-defined configurations. Consequently, the method may comprise one or more of providing an interface for predicting/determining the SINR of the current configuration, providing an interface for predicting/determining the SINR of a preferred configuration of the plurality of pre-defined configurations, and providing an interface for predicting/determining the SINR of a configuration to be used for a future communication over the wireless communication link.

The method comprises predicting 130 the Signal-to-lnterference-and-Noise-Ratio, SINR, of the one or more alternative configurations of the wireless communication link based on a function for predicting the SINR of an alternative configuration based on the estimated channel coefficients, the alternative configuration and the current configuration. As outlined above, different input parameters may be used to predict the SINR of the alternative configuration:

For example, c and c^* relate to both the time-frequency-grid configuration and the pulse configuration of the alternative configuration and of the current configuration, respectively, while

relate only to the time-frequency-grid configuration of the alternative configuration and of the current configuration, respectively. Consequently, the function for predicting the SINR of an alternative configuration may be based on the estimated channel coefficients, a time- frequency-grid-configuration and a pulse configuration of the current configuration and a time- frequency-grid-configuration and a pulse configuration of the alternative configuration Alternatively, the function for predicting the SINR of an alternative configuration

may be based on the estimated channel coefficients, the time-frequency-grid-configuration and the pulse configuration of the current configuration and (only) the time-frequency-grid- configuration of the alternative configuration Alternatively, the function for

predicting the SINR of an alternative configuration may be based on the estimated channel coefficients, (only) the time-frequency-grid-configuration of the current configuration and (only) the time-frequency-grid-configuration of the alternative configuration

In the following, an example implementation of a computation of the SINR values is given. In the following, second-order statistics on the channel are denoted by C, and the instantaneous values of a particular realization are denoted å. For example, the SINR

with g being the filter pulse, g being the transmission pulse, and A being the time-frequency grid, may be calculated using

with being the time-variant channel, eing the channel gain of the point m ∈ I on the

time-frequency grid,

being the interference power from (all) other points on the time-frequency-grid, and s² being the noise variance. H is the channel matrix, and A_gY(μ) = is the well-known cross ambiguity function of g and is a

time-frequency-shifted version of the transmit pulse γ, and S_μ with μ = (μ₁,μ₂) is the time- frequency shift operator.

The averaged interference power

are given (for complex schemes) as

The channel gain may be calculated using

Within the context of the present disclosure, the term “SI NR” is being used. However, the SI NR may also encompass values that are related to the SINR, such as the Signal-to-Noise-Ratio or the Signal-to-lnterference ratio. In particular, the SINR may encompass values that are based on a ratio between the channel gain and the interference power.

In various examples of the present disclosure, the predicted SINR may be associated with a confidence level of the prediction. In other words, the method may comprise determining a confidence level of the predicted SINR. For example, the confidence level may be determined according to a pre-defined time-dependent progress of the confidence level. For example, the confidence level of the prediction may be a first pre-defined value for a first time-interval (e.g. a confidence level of 90% that the prediction of the SINR is correct for the next second), a second pre-defined value for a second time-interval (e.g. a confidence level of 80% that the prediction of the SINR is correct for the next five seconds) etc. In this case, “the next second”, or “the next five seconds” may be the prediction horizon of the prediction. For example, the SINR may be predicted within a range of values (e.g. a range of values encompassing the predicted SINR) having a given confidence value. For example, with a confidence of 90%, the SINR may be predicted to be between a first value and a second value. The longer the prediction horizon, the larger the distance between the two values is likely to become. In some examples, one or more further criteria may be evaluated. For example, the pre-defined values may be multiplied by a factor that is dependent on the one or more further criteria. For example, the one or more further criteria may comprise a criterion that is based on a variance of the spreading over multiple samples of the spreading function. For example, the one or more further criteria may comprise a criterion that is based on the number of samples of the spreading being considered in the prediction of the SINR. For example, the one or more further criteria may comprise a criterion that is based on a second-order statistic of the delay spread and Doppler spread of the wireless channel between the first and second transceiver, e.g. based on a scattering function.

The method comprises determining 140 the property related to the quality of the one or more alternative configurations based on the predicted SINR of the one or more alternative configurations of the wireless communication link. As mentioned above, the property related to the quality of the one or more alternative configurations may relate to a performance of the one or more alternative configurations. For example, pre-limitary functions may be used to map the Signal-to-interference-plus-noise ratio (SINR) to the bit error rate (BER) or frame error rate

For examples, seven alternative mobility modes c =

1 ...7 may be considered. Accordingly, determining 140 the property related to the quality of the one or more alternative configurations may comprise determining 145 a predicted performance, such as the BER, FER or latency, of the one or more alternative configurations based on the predicted SINR of the one or more alternative configurations of the wireless communication link. In more general terms, the method may comprise determining a performance (e.g. BER, FER, latency etc.) of the plurality of pre-defined configurations, e.g. by measuring the performance of the current configurations and determining the predicted performance of the one or more alternative configurations. For example, the predicted performance may be determined for each modulation and coding scheme (MCS), used waveform (e.g., OTFS, OFDM etc.), MIMO order etc. In other words, the predicted performance of the one or more alternative configurations may be determined for two or more different modulation and coding schemes, for two or more different waveforms, or for two or more different Multiple-Input-Multiple-Output configurations (orders).

In some examples, machine-learning may be used to support the prediction of the SINR. In other words, the SINR of the one or more alternative configurations may be tracked and predicted with the help of a machine-learning model. Fig. 2 shows a schematic diagram of a machine-learning model used as part of a determination of a Signal-to-lnterference-and-Noise- Ratio. In Fig. 2, a Long-Short-Term-Memory (LSTM) model is used, i.e. , the machine-learning model may be an LSTM model. However, other types of machine-learning models or algorithms may be used as well, e.g., a kernel-based machine-learning approaches, such as a support vector machine. In particular, machine-learning may be used the following task:

For example, the machine-learning model may be trained to extend the time series S_₂ ...S₀ of the collected and estimated spreading function of the past, based on the time-series S_₂ ... S₀ and based on the current configuration c^*. In other words, the machine-learning model may be trained to perform time-series prediction based on the time series S_₂ ...S₀ of the collected and estimated spreading function of the past and based on the current configuration c^*. In other words, the machine-learning model may be used to, and trained to, estimate a future spreading function based on a plurality of spreading functions, e.g., based on a time-series comprising a plurality of spreading functions. For example, the machine-learning model, e.g., the LSTM model, may perform a historical weighting of the spreading functions, with recent results being more important. A result of this estimation is a prediction of a future spreading function, which may then be used in the determination of the SINR. Accordingly, the SINR of the one or more alternative configurations may be predicted for a point in time in the future based on the predicted spreading function. In other words, the predicted SINR of the one or more alternative configuration may relate to a point in time of the future, e.g., five seconds, ten seconds, or fifteen seconds in the future.

For example, with the help of the machine-learning model, the performance of (e.g., all of) the considered mobility modes c may be estimated based on the current used mode c*.

As mentioned above, the determined property of the one or more alternative configurations may be used to decide on whether the current configuration should be switched to one of the alternative configurations. In other words, the method may comprise switching 150 the configuration of the wireless communication link based on the determined property of the one or more alternative configurations, e.g., based on the performance of the one or more alternative configurations. For example, the method may comprise comparing the performance of the current configuration to the predicted performance of the one or more alternative configurations. If one of the alternatives is significantly better (e.g., if the BER, FER or latency is at least a pre- defined threshold value, e.g., at least 10% or at least 20% better), the configuration may be switched 150. However, the threshold value might not be chosen to be too low, as switching the configuration may carry have an overhead penalty.

The wireless communication link is used by the first transceiver to transmit wireless messages to the second transceiver, and by the second transceiver to transmit wireless messages to the first transceiver. At both ends, the communication over the wireless communication link may be based on the knowledge which configuration (or the plurality of pre-defined configurations) is being used by the other end for transmitting and receiving wireless messages over the wireless communication link (with the configuration being used for transmitting and receiving being the same, or with different configurations being used for transmitting and receiving). Therefore, if one of the transceivers decides to switch to one of the alternative configurations, the other transceiver may be notified of the switch to the other configuration. In other words, the method may comprise notifying, before switching the configuration, the other of the first and second transceiver, of the impending switch of the configuration. For example, a notification message may be transmitted over the wireless communication link to the other transceiver to notify the other transceiver. Accordingly, the method may comprise transmitting the notification message to the other transceiver, or receiving the notification message from the respective other transceiver. For example, the notification message may comprise information on the alternative configuration selected for the switch, e.g. an identifier of the respective configuration. For example, the notification message may comprise information on a tinning of the switch, e.g., a frame number which first uses the new configuration. For example, the notification message may comprise information on a duration of the switch, e.g., how many frames the new configuration is being used. In some examples, the other transceiver may acknowledge the switch. Accordingly, the method may comprise receiving an acknowledgement in response to the notification message from the other transceiver, or transmitting the acknowledgement in response to the received the notification message to the respective other transceiver. In some examples, the switch of the configuration may be negotiated between the first transceiver and the second transceiver.

At least some examples are based on using a machine-learning model or machine-learning algorithm. Machine learning refers to algorithms and statistical models that computer systems may use to perform a specific task without using explicit instructions, instead relying on models and inference. For example, in machine-learning, instead of a rule-based transformation of data, a transformation of data may be used, that is inferred from an analysis of historical and/or training data. For example, the content of images may be analyzed using a machine-learning model or using a machine-learning algorithm. In order for the machine-learning model to analyze the content of an image, the machine-learning model may be trained using training images as input and training content information as output. By training the machine-learning model with a large number of training images and associated training content information, the machine-learning model “learns” to recognize the content of the images, so the content of images that are not included of the training images can be recognized using the machine- learning model. The same principle may be used for other kinds of sensor data as well: By training a machine-learning model using training sensor data and a desired output, the machine-learning model “learns” a transformation between the sensor data and the output, which can be used to provide an output based on non-training sensor data provided to the machine-learning model.

Machine-learning models are trained using training input data. The examples specified above use a training method called “supervised learning”. In supervised learning, the machine-learning model is trained using a plurality of training samples, wherein each sample may comprise a plurality of input data values, and a plurality of desired output values, i.e., each training sample is associated with a desired output value. By specifying both training samples and desired output values, the machine-learning model “learns” which output value to provide based on an input sample that is similar to the samples provided during the training. Apart from supervised learning, semi-supervised learning may be used. In semi-supervised learning, some of the training samples lack a corresponding desired output value. Supervised learning may be based on a supervised learning algorithm, e.g., a classification algorithm, a regression algorithm or a similarity learning algorithm. Classification algorithms may be used when the outputs are restricted to a limited set of values, i.e., the input is classified to one of the limited set of values. Regression algorithms may be used when the outputs may have any numerical value (within a range). Similarity learning algorithms are similar to both classification and regression algorithms, but are based on learning from examples using a similarity function that measures how similar or related two objects are.

Apart from supervised or semi-supervised learning, unsupervised learning may be used to train the machine-learning model. In unsupervised learning, (only) input data might be supplied, and an unsupervised learning algorithm may be used to find structure in the input data, e.g., by grouping or clustering the input data, finding commonalities in the data. Clustering is the assignment of input data comprising a plurality of input values into subsets (clusters) so that input values within the same cluster are similar according to one or more (pre-defined) similarity criteria, while being dissimilar to input values that are included in other clusters.

Reinforcement learning is a third group of machine-learning algorithms. In other words, reinforcement learning may be used to train the machine-learning model. In reinforcement learning, one or more software actors (called “software agents”) are trained to take actions in an environment. Based on the taken actions, a reward is calculated. Reinforcement learning is based on training the one or more software agents to choose the actions such, that the cumulative reward is increased, leading to software agents that become better at the task they are given (as evidenced by increasing rewards).

In general, an LSTM, which is mentioned above, may be trained using a supervised learning algorithm, as the LSTM learns by specifying a training sample and a desired output, using techniques like gradient descent to find a combination of weights within the LSTM that is most suitable for generating the desired transformation. In the proposed concept, spreading functions may be provided at the input of the LSTM, and a desired weighting of the spreading functions may be provided as desired output. Alternatively or additionally, the training may be embedded in a reinforcement learning-based approach, where the weighting is changed using reinforcement learning based on a reward function that is based on a divergence between the predicted SINR and the actual SINR (e.g., as measured or as simulated). In various examples, the LSTM may be trained for time-series prediction, e.g., by using historic time-series data (of the spreading function), providing a window of samples of the time-series data (i.e., a sequence of spreading functions) as training samples and a subsequent sample (i.e. , a subsequent spreading function) as desired output.

Machine-learning algorithms are usually based on a machine-learning model. In other words, the term “machine-learning algorithm” may denote a set of instructions that may be used to create, train or use a machine-learning model. The term “machine-learning model” may denote a data structure and/or set of rules that represents the learned knowledge, e.g., based on the training performed by the machine-learning algorithm. In embodiments, the usage of a machine- learning algorithm may imply the usage of an underlying machine-learning model (or of a plurality of underlying machine-learning models). The usage of a machine-learning model may imply that the machine-learning model and/or the data structure/set of rules that is the machine- learning model is trained by a machine-learning algorithm.

For example, the machine-learning model may be an artificial neural network (ANN). ANNs are systems that are inspired by biological neural networks, such as can be found in a brain. ANNs comprise a plurality of interconnected nodes and a plurality of connections, so-called edges, between the nodes. There are usually three types of nodes, input nodes that receiving input values, hidden nodes that are (only) connected to other nodes, and output nodes that provide output values. Each node may represent an artificial neuron. Each edge may transmit information, from one node to another. The output of a node may be defined as a (non-linear) function of the sum of its inputs. The inputs of a node may be used in the function based on a “weight” of the edge or of the node that provides the input. The weight of nodes and/or of edges may be adjusted in the learning process. In other words, the training of an artificial neural network may comprise adjusting the weights of the nodes and/or edges of the artificial neural network, i.e., to achieve a desired output for a given input. In at least some embodiments, the machine-learning model may be deep neural network, e.g., a neural network comprising one or more layers of hidden nodes (i.e., hidden layers), prefer-ably a plurality of layers of hidden nodes.

Alternatively, the machine-learning model may be a support vector machine. Support vector machines (i.e., support vector networks) are supervised learning models with associated learning algorithms that may be used to analyze data, e.g., in classification or regression analysis. Support vector machines may be trained by providing an input with a plurality of training input values that belong to one of two categories. The support vector machine may be trained to assign a new input value to one of the two categories. Alternatively, the machine- learning model may be a Bayesian network, which is a probabilistic directed acyclic graphical model. A Bayesian network may represent a set of random variables and their conditional dependencies using a directed acyclic graph. Alternatively, the machine-learning model may be based on a genetic algorithm, which is a search algorithm and heuristic technique that mimics the process of natural selection.

The one or more interfaces 12 may correspond to one or more inputs and/or outputs for receiving and/or transmitting information, which may be in digital (bit) values according to a specified code, within a module, between modules or between modules of different entities. For example, the one or more interfaces 12 may comprise interface circuitry configured to receive and/or transmit information. The control module 14 may be implemented using one or more processing units, one or more processing devices, any means for processing, such as a processor, a computer or a programmable hardware component being operable with accordingly adapted software. In other words, the described function of the control module 14 may as well be implemented in software, which is then executed on one or more programmable hardware components. Such hardware components may comprise a general-purpose processor, a Digital Signal Processor (DSP), a micro-controller, etc.

A base station or base station transceiver can be operable to communicate with one or more active mobile transceivers and a base station transceiver can be located in or adjacent to a coverage area of another base station transceiver, e.g., a macro cell base station transceiver or small cell base station transceiver. Hence, embodiments may provide a mobile communication system comprising one or more mobile transceivers and one or more base station transceivers, wherein the base station transceivers may establish macro cells or small cells, as e.g., pico-, metro-, or femto cells. A mobile transceiver may correspond to a smartphone, a cell phone, user equipment, radio equipment, a mobile, a mobile station, a laptop, a notebook, a personal computer, a Personal Digital Assistant (PDA), a Universal Serial Bus (USB) -stick, a car, a mobile relay transceiver for D2D communication, etc. A mobile transceiver may also be referred to as User Equipment (UE) or mobile in line with the 3GPP terminology.

A base station transceiver can be located in the fixed or stationary part of the network or system. A base station transceiver may correspond to a remote radio head, a transmission point, an access point, radio equipment, a macro cell, a small cell, a micro cell, a femto cell, a metro cell etc. A base station transceiver may correspond to a base station understood as a logical concept of a node/entity terminating a radio bearer or connectivity over the air interface between a terminal/mobile transceiver and a radio access network. A base station transceiver can be a wireless interface of a wired network, which enables transmission of radio signals to a UE or mobile transceiver. Such a radio signal may comply with radio signals as, for example, standardized by 3GPP or, generally, in line with one or more of the above listed systems. Thus, a base station transceiver may correspond to a NodeB, an eNodeB, a Base Transceiver Station (BTS), an access point, a remote radio head, a transmission point, a relay transceiver etc., which may be further subdivided in a remote unit and a central unit.

A mobile transceiver can be associated, camped on, or registered with a base station transceiver or cell. The term cell refers to a coverage area of radio services provided by a base station transceiver, e.g., a NodeB (NB), an eNodeB (eNB), a remote radio head, a transmission point, etc. A base station transceiver may operate one or more cells on one or more frequency layers, in some embodiments a cell may correspond to a sector. For example, sectors can be achieved using sector antennas, which provide a characteristic for covering an angular section around a remote unit or base station transceiver. In some embodiments, a base station transceiver may, for example, operate three or six cells covering sectors of 120° (in case of three cells), 60° (in case of six cells) respectively. A base station transceiver may operate multiple sectorized antennas. In the following a cell may represent an according base station transceiver generating the cell or, likewise, a base station transceiver may represent a cell the base station transceiver generates.

In the following, a more detailed instruction to mobility modes according to an example is given.

Mobility modes are proposed with distinct grid and pulse matching for different doubly dispersive channels. To account for remaining self-interference, the minimum mean square error (MMSE) linear equalizer may be tuned without the need of estimating channel cross-talk coefficients.

The proposed approach was evaluated with the QuaDRiGa channel simulator and with OTFS transceiver architecture based on a polyphase implementation for orthogonalized Gaussian pulses. In addition, OTFS is compared to a IEEE 802.11 p compliant design of cyclic prefix (CP) based orthogonal frequency-division multiplexing (OFDM). The results indicate that with an appropriate mobility mode, the potential OTFS gains can be indeed achieved with linear equalizers to significantly outperform OFDM.

Strict requirements on reliability and efficiency in high mobility communication scenarios, such as vehicle-to-everything (V2X) communication, are pushing legacy systems to their limits. Orthogonal frequency-division multiplexing (OFDM) is a widely-used modulation scheme which however suffers substantial performance degradation and inflexibility in scenarios with high Doppler spreads [1], Consequently, there is a need for the development of novel modulation schemes that are flexible, efficient and robust in doubly dispersive channels.

An orthogonal time frequency and space (OTFS) waveform is introduced by Hadani et. al [2] as a promising combination of classical pulse-shaped Weyl-Heisenberg (or Gabor) multicarrier schemes with a distinct time-frequency (TF) spreading. Data symbols are spread with the symplectic finite Fourier transform (SFFT) over the whole TF grid. This particular linear pre- coding accounts for the doubly dispersive nature of time-varying multipath channels seen as linear combinations of TF shifts. Several studies show that OTFS outperforms OFDM in such situations [3], [4], [5], Another research work focuses on a performance comparison of OFDM, generalized frequency division multiplexing (GFDM), and OTFS [6], It reveals significant advantages of OTFS in terms of bit error rate (BER) and frame error rate (FER) in relation to the others. However, so far research has mainly focused on OTFS with the assumption of perfect grid matching and often with idealized pulses, violating the uncertainty principle. In many cases, ideal channel knowledge is assumed, including the cross-talk channel coefficients.

Different doubly dispersive communication channels provide distinct delay-Doppler (DD) spread and diversity characteristics. Particular single dispersive cases therein are time or frequency- invariant channels, which boil down to simple frequency or time division communication schemes, respectively. For some high mobility scenarios, the channel becomes dispersive in both time and frequency domain. Especially, V2X channels differ in their dissipation in both domains. Depending on the communication scenario, a distinct spreading region is spanned:

where B, L, v, and t are the bandwidth, signal length, Doppler, and delay spread, respectively.

In order to cope with doubly dispersive channels, the synthesis pulse used at the transmitter, the analysis pulse used at the receiver, and their TF grid may match U [7], [8], [9], A common way is to design the ratio of time and frequency shifts T and F as well as TF spreads σ_t and σ_f of the Gabor pulses with respect to the channel scattering function under the wide-sense stationary uncorrelated scattering (WSSUS) assumption:

where is the ratio between the maxima of the delay and the Doppler spread of the

channel. This approach is referred to as pulse and grid matching [7], [10], [8], [9]. With the goal of satisfying the condition of pulse and grid matching in (1), distinct mobility modes are proposed and investigated.

For coherent communication, the doubly dispersive channel operation may be estimated and inverted at the receiver. In general, linear equalizer are favored for channel equalization, since they have a lower complexity compared to e.g., maximum-likelihood equalizer (MLE) or iterative techniques such as interference cancellation [11] Although MLE enjoys the maximum diversity, in some cases linear equalizer can achieve the same diversity gain as MLE [12], for example in the case of non-singular convolutions. In [13], it has been observed that in most cases full OTFS diversity is not achieved when using a common minimum mean square error (MMSE) equalization. On the contrary, MLE or interference cancellation techniques for OTFS are complex and also require accurate estimation of the cross-talk channel coefficients. Indeed, the remaining self-interference caused by suboptimal pulse and grid matching may be estimated and taken into account at the equalizer. A linear equalizer which accounts for self-interference on a frame base was introduced in [14] This approach is used in the presented work to account for the remaining self-interference.

In this section, mobility modes are proposed that control the self-interference on a coarse level and to instantaneously tune the linear MMSE equalizer by estimating from pilot and guard symbols the remaining self-interference power. The main focus of this section can be summarized as follows:

• OTFS is studied from the perspective of the pulse-shaped Gabor signaling with additional TF spreading, implemented using the MATLAB toolbox LTFAT [15],

• Doubly dispersive vehicular channels are considered in a concrete geometry-based scenario generated by the QuaDRiGa channel simulator [16] using pilot-based channel estimation as in [17],

• Mobility modes are proposed with distinct pulse and grid matching, and

• The impact of the remaining self-interference in the equalizer due to imperfect 2D-deconvolution of the twisted convolution affected by grid and pulse mismatch [14] is taken into account.

II. OTFS SYSTEM MODEL. In this section, the system model and the OTFS transceiver structure is introduced. OTFS is a combination of classical pulse shaped multicarrier transmission with Gabor structure, i.e. , TF translations on a regular grid in the TF plane, and additional TF spreading using the SFFT. A. Time-Frequency Grid and Pulse Shaping. The frequency resolution is where B is the

overall bandwidth and M the number of subcarriers. The time resolution is with D being

the frame duration and N the number of time symbols. The TF grid is sampled with T and F period in the time and frequency domain, respectively. The interbank length also depends on the dimensioning of the used synthesis and analysis pulse and the so-called time frequency product T · F. The Gabor filterbanks at the transmitter and at the receiver are configured with pulses y for the synthesis and g for the analysis of the signals, respectively.

Three cases are distinguished: TF > 1, TF = 1, and TF < 1 - sometimes referred to as undersampling, critical sampling, and oversampling of the TF plane, respectively [18], Here, . is assumed, as it is a typical compromise between maximizing the signal to interference ratio (SIR) and the loss in degrees of freedom [19], To guarantee perfect reconstruction in the non- dispersive and noiseless case, the pulses y and g may be required to be biorthogonal:

where it is defined (same for γ_α,β(t), with δ(0) = 1 and zero otherwise.

Here,

is used as inner product on L₂(R), the Hilbert space of signals with finite energy. To ensure uncorrelated noise contributions, the synthesis and analysis pulses are assumed to be equal, resulting in an orthogonal pulse. Given a preliminary prototype pulse, the well-known S^-1/2-trick is used to perform the orthogonalization, i.e., constructing a tight Gabor frame on an adjoint lattice [9], However, exact orthogonality at the output of doubly dispersive channels is usually destroyed resulting in self-interference. By choosing different pulses for the transmitter and receiver, it may even be possible to further reduce the self-interference for classes of doubly dispersive channels.

B. TF-Spreading and De-Spreading. The transceiver structure is essentially the same as in many pulse shaped multicarrier schemes, like pulse-shaped OFDM, biorthogonal frequency division multiplexing (BFDM) or filter bank multicarrier (FBMC). A distinct feature of OTFS is the spreading. All symbols

are pre-coded with the inverse SFFT denoted as . The SFFT differs from the ordinary 2D Fourier transformation by its sign switching in the exponent and coordinates swapping. One can interpret this by mapping an array of discrete DD positions ( l , k ) to an array of grid points (m, n) in the TF plane, since time shifts lead to oscillations in frequency and frequency shifts result in oscillations in time. More precisely, at the transmitter, the pre-coding is given by

where

The received and equalized symbols in the TF plane are de-spreaded again as

such that

C. Structure of the OTFS Frames. A pilot-based channel estimation is used, where a pilot is inserted in the DD domain as proposed by [17], The pilot is sent by the transmitter in the same frame as the data. In doing so, the channel can be easily estimated at the receiver in the DD domain. The symbols to be placed in the DD domain are threefold. The data symbols, usually coming from a particular mod

ulation

alphabet, are placed on positions indexed by the set

Positions used for channel estimation are defined by the set

with

which w contain a single pilot symbol; the other positions are unused and can be seen as guard sy In this context, it is assumed that

where W and Q define the guard region in delay and Doppler domain, respectively. An arbitrary location

is used for the non-zero pilot symbol. Note that W and Q are defined with respect to the expected DD shift [20], Fig. 3a depicts an example of an OTFS frame with data, pilot, and guard symbols. Q and W are chosen with an appropriate dimension for each OTFS mode. A constant product of Q · W, i.e., 1024 symbols is assumed, to compare different configurations with the same pilot overhead (same data rate). For simplicity, the non-zero pilot

with the normalized power of

and all the other symbols in P are zero-valued guard symbols are set.

D. Gabor Synthesis Filterbank. The OTFS frame in the TF plane is then used to synthesize a transmit signal s(t). This is implemented with a Gabor synthesis filterbank configured with a transmit pulse γ [7], This can be formally written as

E. The Doubly Dispersive Channel. For a doubly dispersive channel, the noiseless time- continuous channel output consists of an unknown linear combination of TF translates of the input signal s(t). This operation can be formally expressed as

where the pth discrete propagation path has the delay

· The index set is defined by

, For

is then given by

where can be seen as the discrete DD spreading function [21], In particular, this

simplified model implies that each path has the same range of frequency shifts but with

possibly different coefficients. The set of TF shifts is assumed to be usually

in a box

, which is also known as the underspread assumption. Putting (6) in (7) with (8) yields

F. Gabor Analysis Filterbank. The received signal is down-converted and passed through an analysis filterbank. The output of the noiseless Gabor analysis filterbank in TF slot is

then

III. CHANNEL ESTIMATION AND SELF-INTERFERENCE. In this section, the channel estimation, the equalization and the amount of self-interference which remains in the OTFS transceiver structure is explained in more detail. In particular, the link between the equalization as a 2D-deconvolution and the true channel mapping, given as a twisted convolution, is shown.

A. Impact of the Self-Interference. To reveal the impact of pulse and grid mismatch on self- interference, the inner product in (10) is rewritten and computed separately:

Where A(α,β) = is the cross-ambiguity function. The goal is to design the pulses γ and

g such that

for all values

Roughly speaking, this implies that (taken over data

symbols and channel realizations) of the self-interference

defined to be

becomes negligibly small. Note that since pulses g and y such that A_g

_y(α, β) =

for all the (α,β ) do not exist. Therefore, the goal of matched pulse shaping is

instead to minimize the expected self-interference power.

By considering self-interference in the system model,

is obtained. Applying to (14) shows that in the first order (up to inference) the channel acts as

2D-convolution since

As point-wise multiplication in the TF plane is (circular) 2D-convolution in the DD plane,

where

is the channel transfer function. The magnitude of

is depending on the matching given in (12), i.e., the higher the mismatch the larger the self-interference.

B. Delay-Doppler Channel Estimation. The channel is estimated with the pilot sent by the transmitter in the DD domain. The

is applied to quarter of the guard area, where the channel impulse response (CIR) is obtained by [20]:

for all . Fig. 3a highlights the symbols used for channel estimation in a black dashed

frame. The remaining guard symbols (outside the black dashed frame) are used to avoid interference between the pilot and data symbols. C. Time-Frequency Equalization. It is proposed to use mobility modes to achieve sufficient performance at moderate complexity. The appropriate mobility mode controls the self- interference on a coarse level. In addition, the MMSE equalizer is tuned to account for the remaining self-interference power. The received frame (14) is equalized with the estimated channel (17) by MMSE equalization:

where σ2 is the noise variance. Therefore, it is the mean self-interference power / is estimated, which contains the averaged power of the self-interference and the error of the channel estimation at the receiver. This may be approachd by estimating I as the empirical mean (over

from pilot and guard symbols for each frame to tune the MMSE equalizer instantaneously to the corresponding channel realization. For a given /, the equalized symbols in the DD domain are given by:

An intuitive approach is then to minimize a given error metric d(·,·) between the transmitted (assumed to be known at receiver) and equalized pilot and guard symbols, and

respectively, as proposed in [14]:

As error metric

the

is used on a finite grid as in [14], Finally, each frame is then equalized with its individual I_opt.

IV. MOBILITY MODES. In this section, the mobility modes are introduced to reduce the self- interference caused by grid and pulse mismatches. Coping with different channel conditions, i.e. , distinct delay and Doppler spreads, seven different mobility modes are investigated. The mobility mode may be defined by the long-term expectation of the channel. The proposed mobility modes are aiming to yield a small deviation from equality in (12) and hence to reduce the impact of self-interference. The remaining self-interference power is then estimated in (20) and used for the linear equalization. Table I of Fig. 3b presents the mobility modes I to XII. The higher the resolution in time (N symbols), the fewer resolution in frequency domain (M subcarrier) and vice versa. Mode I represents the case for equal time and frequency resolution. Each mobility mode therefore has its own pulse shape which is achieved by squeezing and orthogonalization according to the procedure explained in the introduction. It is assumed that the transmitter and the receiver use the same mode. The appropriate mode can be selected depending on the second order statistic of the channel. The selection of an appropriate mode is left for future work.

V. NUMERICAL RESULTS. In this section, the approach of using distinct mobility modes for grid and pulse matching is numerically analyzed. Table II of Fig. 3c summarizes the parameters used to obtain the numerical results. In the case of cyclic prefix (CP) based OFDM, the regularized least-squares approach is followed for channel estimation and zero-forcing equalization [22] is used. One OFDM configuration is studied, with the same TF grid as OTFS Mode I (see Table I). The OFDM configuration is close to the 802.11p standard where the rectangular pulses include the CP. The coded BER curves are presented for different communication scenarios for all modes, where convolution codes with a code rate of r = 0.5 are used. Table III of Fig. 3g lists all modes and the corresponding signal to noise ratio (SNR) needed to reach the target BER of 10^-2and 10^-3. The lowest BER reached for each mode is listed. Figs 3d to 3f show the BER for distinct V2X scenarios and different mobility modes. The QuaDRIGa channel simulator [16] with the 3GPP 38.901 and QuaDRIGa UD2D channel model for V2I and V2V scenarios is used, respectively. The convolution coding is using a code rate of r = 0.5. Fig. 3d and 3e depict the BER for the vehicle-to-infrastructure (V2I) scenario under line- of-sight (LOS) and strict non line-of-sight (NLOS) condition, respectively. Each V2X scenario is characterized by a distinct DD spread. Therefore, for each case, a different moblity mode is appropriate, i.e., Mode I or II in LOS and Mode VI or IV in NLOS. In Fig. 3f, a vehicle-to-vehicle (V2V) scenario is presented with a relative speed of Δν = 160 km/h. Here Mode I out-performs the others. In general, it can be observed that OTFS outperforms OFDM with an appropriate mobility mode in all scenarios.

VI. CONCLUSIONS. Mobility modes were introduced for pulse-shaped OTFS modulation to enable linear equalization. By selecting an appropriate mobility mode for pulse and grid matching the self-interference level, immanent in doubly dispersive channels, reduces and hence, also the BER. It can be concluded that through the introduction of mobility modes, one can improve the system performance for low-complexity equalizers implementing tuned 2D- deconvolutions instead of dealing with the full twisted convolution. It is pointed out that the tuning of the equalizer for the remaining interference levels provides further gains of the mobility modes. For each V2X scenario a distinct mobility mode outperforms the others and the effect improves with more accurate channel knowledge. In all scenarios at least one OTFS mode outperforms the CP-based OFDM. It is shown the importance of the selection of an appropriate mobility mode. As already mentioned, in embodiments the respective methods may be implemented as computer programs or codes, which can be executed on a respective hardware. Hence, another embodiment is a computer program having a program code for performing at least one of the above methods, when the computer program is executed on a computer, a processor, or a programmable hardware component. A further embodiment is a computer readable storage medium storing instructions which, when executed by a computer, processor, or programmable hardware component, cause the computer to implement one of the methods described herein.

A person of skill in the art would readily recognize that steps of various above-described methods can be performed by programmed computers, for example, positions of slots may be determined or calculated. Herein, some embodiments are also intended to cover program storage devices, e.g., digital data storage media, which are machine or computer readable and encode machine-executable or computer-executable programs of instructions where said instructions perform some or all of the steps of methods described herein. The program storage devices may be, e.g., digital memories, magnetic storage media such as magnetic disks and magnetic tapes, hard drives, or optically readable digital data storage media. The embodiments are also intended to cover computers programmed to perform said steps of methods described herein or (field) programmable logic arrays ((F)PLAs) or (field) programmable gate arrays ((F)PGAs), programmed to perform said steps of the above-described methods.

The description and drawings merely illustrate the principles of the invention. It will thus be appreciated that those skilled in the art will be able to devise various arrangements that, although not explicitly described or shown herein, embody the principles of the invention and are included within its spirit and scope. Furthermore, all examples recited herein are principally intended expressly to be only for pedagogical purposes to aid the reader in understanding the principles of the invention and the concepts contributed by the inventor(s) to furthering the art, and are to be construed as being without limitation to such specifically recited examples and conditions. Moreover, all statements herein reciting principles, aspects, and embodiments of the invention, as well as specific examples thereof, are intended to encompass equivalents thereof. When provided by a processor, the functions may be provided by a single dedicated processor, by a single shared processor, or by a plurality of individual processors, some of which may be shared. Moreover, explicit use of the term “processor” or “controller” should not be construed to refer exclusively to hardware capable of executing software, and may implicitly include, without limitation, Digital Signal Processor (DSP) hardware, network processor, application specific integrated circuit (ASIC), field programmable gate array (FPGA), read only memory (ROM) for storing software, random access memory (RAM), and non-volatile storage. Other hardware, conventional or custom, may also be included. Their function may be carried out through the operation of program logic, through dedicated logic, through the interaction of program control and dedicated logic, or even manually, the particular technique being selectable by the implementer as more specifically understood from the context.

It should be appreciated by those skilled in the art that any block diagrams herein represent conceptual views of illustrative circuitry embodying the principles of the invention. Similarly, it will be appreciated that any flow charts, flow diagrams, state transition diagrams, pseudo code, and the like represent various processes which may be substantially represented in computer readable medium and so executed by a computer or processor, whether or not such computer or processor is explicitly shown.

Furthermore, the following claims are hereby incorporated into the detailed description, where each claim may stand on its own as a separate embodiment. While each claim may stand on its own as a separate embodiment, it is to be noted that - although a dependent claim may refer in the claims to a specific combination with one or more other claims - other embodiments may also include a combination of the dependent claim with the subject matter of each other dependent claim. Such combinations are proposed herein unless it is stated that a specific combination is not intended. Furthermore, it is intended to include also features of a claim to any other independent claim even if this claim is not directly made dependent to the independent claim.

It is further to be noted that methods disclosed in the specification or in the claims may be implemented by a device having means for performing each of the respective steps of these methods.

List of reference signs

10 Apparatus

12 One or more interfaces

14 Control module

100 First transceiver

110 Estimating channel coefficients

120 Determining a spreading function

130 Predicting a Signal-to-Noise-and-lnterference-Ratio

140 Determining a property of one or more alternative configurations

145 Determining a performance of the one or more alternative configurations

150 Switching the configuration

200 Second transceiver

Claims

1. A method for determining a property related to a quality of one or more alternative configurations of a wireless communication link between a first transceiver (100) and a second transceiver (200), the wireless communication link being based on a current configuration, the method comprising:

Estimating (110) channel coefficients of a wireless channel of the wireless communication link;

Predicting (130) a Signal-to-lnterference-and-Noise-Ratio, SI NR, of the one or more alternative configurations of the wireless communication link based on a function for predicting the SI NR of an alternative configuration based on the estimated channel coefficients, the alternative configuration and the current configuration; and

Determining (140) the property related to the quality of the one or more alternative configurations based on the predicted SI NR of the one or more alternative configurations of the wireless communication link.

2. The method according to claim 1, wherein the method comprises determining (120) a spreading function based on the estimated channel coefficients, wherein the function for predicting the SI NR of an alternative configuration is based on the spreading function that is based on the estimated channel coefficients, the alternative configuration and the current configuration.

3. The method according to one of the claims 1 or 2, wherein the SINR of the one or more alternative configurations is tracked and predicted with the help of a machine-learning model wherein machine-learning model is used to predict a future spreading function based on a plurality of previously determined spreading functions, wherein the SINR of the one or more alternative configurations is predicted for a point in time in the future based on the predicted spreading function.

4. The method according to one of the claims 1 to 3, wherein the channel coefficients are estimated based on a spreading function of the wireless channel obtained using a pilot signal.

5. The method according to one of the claims 1 to 4, wherein the function for predicting the SI NR of an alternative configuration is based on the estimated channel coefficients, a time-frequency-grid-configuration and a pulse configuration of the current configuration and a time-frequency-grid-configuration and a pulse configuration of the alternative configuration.

6. The method according to one of the claims 1 to 4, wherein the function for predicting the SI NR of an alternative configuration is based on the estimated channel coefficients, a time-frequency-grid-configuration and a pulse configuration of the current configuration and a time-frequency-grid-configuration of the alternative configuration.

7. The method according to one of the claims 1 to 4, wherein the function for predicting the SINR of an alternative configuration is based on the estimated channel coefficients, a time-frequency-grid-configuration of the current configuration and a time-frequency-grid- configuration of the alternative configuration.

8. The method according to one of the claims 1 to 7, wherein determining (140) the property related to the quality of the one or more alternative configurations comprises determining (145) a predicted performance of the one or more alternative configurations based on the predicted SINR of the one or more alternative configurations of the wireless communication link.

9. The method according to claim 8, wherein the predicted bit performance of the one or more alternative configurations is determined for two or more different modulation and coding schemes, for two or more different waveforms, or for two or more different Multiple-Input-Multiple-Output configurations.

10. The method according to one of the claims 1 to 9, wherein the method comprises switching (150) the configuration of the wireless communication link based on the determined property of the one or more alternative configurations.

11. The method according to one of the claims 1 to 10, wherein the wireless communication link is based on a multicarrier transmission-based wireless communication system.

12. The method according to one of the claims 1 to 11, wherein the wireless communication link is based on one of an Orthogonal Frequency Division Multiplexing-based wireless communication system, an Orthogonal Time-Frequency-Space-based wireless communication system, a Generalized Frequency Division Multiplexing-based wireless communication system, and a Filter Bank Multi-Carrier-based wireless communication system.

13. The method according to one of the claims 1 to 12, wherein the one or more alternative configurations are different from the current configuration with respect to a time- frequency-grid-configuration of the wireless communication link, or wherein the one or more alternative configurations are different from the current configuration with respect to a time-frequency-grid-configuration of the wireless communication link and/or with respect to a pulse configuration of the wireless communication link.

14. A computer program having a program code for performing the method according to one of the previous claims, when the computer program is executed on a computer, a processor, or a programmable hardware component.

15. An apparatus (10) comprising: one or more interfaces (12) for communicating in a mobile communication system; and a control module (14) configured to carry out the method according to one of the claims 1 to 13.