Channel acquisition using pilot signals with orthogonal time frequency space modulation

Abstract

Techniques for performing channel estimation in an orthogonal time, frequency, and space (OTFS) communication system include receiving a wireless signal including a data signal portion and a pilot signal portion, where the pilot signal portion includes a plurality of pilot signals multiplexed together in an OTFS domain, performing a two-dimensional channel estimation in a time-frequency domain based on a minimum mean square error (MMSE) optimization criterion, and recovering information bits using channel estimates obtained from the two-dimensional channel estimation.

Description

[0001] Cross-references to related applications

[0002] This patent application claims priority to U.S. Provisional Application No. 62 / 316,437, filed March 31, 2016, entitled "CHANNEL ACQUISITIONUSING ORTHOGONAL TIME FREQUENCY SPACE MODULATED PILOT SIGNALS". The entire contents of the aforementioned patent application are incorporated herein by reference. Technical Field

[0003] This patent document relates to wireless communication, and more specifically, to receiver-side processing of orthogonal time-frequency space (OTFS) modulated signals. Background Technology

[0004] Due to the explosive growth in the number of wireless user devices and the amount of wireless data that these devices can generate or consume, current wireless communication networks are rapidly exhausting their bandwidth to accommodate this high growth in data traffic and to provide users with high-quality service.

[0005] The telecommunications industry is working to develop next-generation wireless technologies that can meet the performance requirements of wireless devices and networks. Summary of the Invention

[0006] This document discloses receiver-side techniques for performing channel estimation or channel acquisition from received signals, including OTFS pilot signals.

[0007] In one exemplary aspect, a wireless communication method for recovering information bits from a received signal is disclosed, the method being implemented by a wireless communication receiver. The method includes: receiving a wireless signal comprising a data signal portion and a pilot signal portion, wherein the pilot signal portion comprises a plurality of pilot signals multiplexed together in an orthogonal time-frequency space (OTFS) domain; performing a two-dimensional channel estimation in the time-frequency domain based on a minimum mean square error (MMSE) optimization criterion; and recovering the information bits using the channel estimation obtained from the two-dimensional channel estimation.

[0008] In another exemplary aspect, a technique for recovering information bits from a received signal is disclosed, implemented by a wireless communication receiver. The technique includes: receiving a wireless signal comprising a data signal portion and a pilot signal portion on a communication channel, wherein the pilot signal portion comprises a plurality of pilot signals multiplexed together in an orthogonal time-frequency space (OTFS) domain; transforming the received wireless signal into a delayed time-domain signal; applying a delayed time-domain mask to separate each of the contributions of the plurality of pilot signals to the delayed time-domain signal into a corresponding individual received pilot signal contribution; using the individual received pilot signal contributions to estimate a communication channel estimate at the delayed time-domain location of the plurality of pilot signals; and interpolating the communication channel estimate to obtain a complete communication channel estimate.

[0009] In another exemplary aspect, a wireless communication method for recovering information bits from a received signal is disclosed, the method being implemented by a wireless communication receiver. The method includes: receiving a wireless signal comprising a data signal portion and a pilot signal portion on a communication channel, wherein the pilot signal portion comprises a plurality of pilot signals multiplexed together in an orthogonal time-frequency space (OTFS) domain; transforming the received wireless signal into a delayed Doppler domain signal by applying a two-dimensional Sin-Fourier transform; applying a delayed Doppler domain mask to separate each of the contributions of the plurality of pilot signals to the time-frequency domain signal into a corresponding individual received pilot signal contribution; using the individual received pilot signal contributions to estimate a communication channel estimate at the delayed Doppler domain location of the plurality of pilot signals; and interpolating the communication channel estimate to obtain a complete communication channel estimate.

[0010] These and other features are described in this patent document. Attached Figure Description

[0011] The accompanying drawings described herein are provided to further understand this application and form part of this application. The exemplary embodiments and their illustrations are used to explain the technology, not to limit its scope.

[0012] Figure 1 An example communication network is shown.

[0013] Figure 2 This is a diagram illustrating an exemplary decomposition of the time-frequency grid into three parts.

[0014] Figure 3 This is a diagram illustrating an example of an interpolation problem. After receiving distorted pilot signals on the downlink channel subgrid, the remote device estimates the channel over the past portion of the data grid.

[0015] Figure 4 This is a diagram illustrating an example of a prediction problem. A remote device estimates channels on a future portion of the data grid.

[0016] Figure 5 This is a diagram illustrating an example of a channel in the time-frequency domain.

[0017] Figure 6 This is a diagram illustrating an example of a channel in the delayed Doppler domain.

[0018] Figure 7 This is a diagram illustrating an example of a channel in the time domain with delay.

[0019] Figure 8 An example of a linear frequency-modulated sequence on a delayed Doppler plane is shown. The span of the plane is 1 / df in the delay dimension and 1 / (Ndt) in the Doppler dimension.

[0020] Figure 9 This is a diagram illustrating an example of a signal received by the antenna of a single remote device in the time-delay domain.

[0021] Figure 10 These are diagrams illustrating an exemplary interpolation problem. Left: Channel response measured on the pilot subgrid. Right: Desired channel response.

[0022] Figure 11 This is a diagram illustrating an example of a one-dimensional interpolation problem. For each delay value τ, the channel on the data grid is estimated using the channel values ​​on the pilot subgrid.

[0023] Figure 12 These are examples of the results of cubic spline interpolation. Top: Showing the complete spline. Bottom: Zoomed in to the region where channel estimation is desired. Splines provide excellent channel estimators.

[0024] Figure 13 An example of a ring with a delayed circumference greater than the Doppler circumference is shown.

[0025] Figure 14 An example of the pilot (red) and data (green) tf planes is shown when k = N-1 and M = 1.

[0026] Figure 15 This is a diagram showing an example of a packet with 8 pilots (4×2) in the delayed Doppler plane.

[0027] Figure 16 An exemplary embodiment is shown with 20 pilots at grid point (5+i50,9+j18) in the delayed Doppler plane, where i = 0 to 49 and j = 0,1.

[0028] Figure 17 yes Figure 16 An enlarged view of an exemplary pilot in the diagram.

[0029] Figure 18This is a diagram illustrating an example of received pilots after passing through a channel with two delayed Doppler paths (shown as 4 pilots).

[0030] Figure 19 An example of a marker for the received signal level above -40dB of the received pilot peak is shown.

[0031] Figure 20 An example of a TF grid is shown, which indicates the location of the reference signals (red, purple) of two time-frequency packets multiplexed with the data (green).

[0032] Figure 21 An example of receiving pilots after passing through a channel with two delayed Doppler paths (showing two pilots) is shown.

[0033] Figure 22 An example of a marker for the received signal level where the received pilot peak is higher than -40dB is shown.

[0034] Figure 23 Shown in the tf plane Figure 14 Examples of the real parts of pilots P1 and P2.

[0035] Figure 24 A flowchart illustrating an exemplary wireless communication receiving method is shown.

[0036] Figure 25 A flowchart of another exemplary wireless communication receiving method is shown.

[0037] Figure 26 A flowchart of another exemplary wireless communication receiving method is shown.

[0038] Figure 27 An example of a wireless transceiver device is shown.

[0039] Best practice

[0040] To make the objectives, technical solutions, and advantages of the present invention clearer, various embodiments are described in detail below with reference to the accompanying drawings. Unless otherwise stated, the embodiments and features described herein can be combined with each other.

[0041] Current wireless technologies are not expected to meet the growing demands of wireless communication. Many industry organizations have begun working on standardizing next-generation wireless signal interoperability standards. Fifth-generation (5G), developed by the 3rd Generation Partnership Project (3GPP), is one such example, and is used throughout this document for illustrative purposes. However, the technologies disclosed are applicable to other wireless networks and systems.

[0042] This document uses chapter headings to improve the readability of the descriptions and does not in any way limit the discussion to individual chapters.

[0043] Figure 1 An exemplary communication network 100 is shown that can implement the disclosed techniques. Network 100 may include a base station transmitter that transmits wireless signals s(t) (downlink signals) to one or more receivers 102, and received signals are denoted as r(t), which may be located in various locations, including inside or outside buildings and in moving vehicles. Receivers may transmit uplink transmissions to a base station typically located near the wireless transmitter. The techniques described herein may be implemented at receiver 102.

[0044] As explained herein, channel acquisition or channel estimation of wireless signals, including OTFS pilots and data, can be performed in different ways. In some embodiments, the OTFS pilot signal may be in the form of, for example, an OTFS domain pulse.

[0045] 1. Introduction

[0046] This paper discloses different channel acquisition mechanisms using OTFS-based pilots. The theory and implementation scheme of OTFS-based pilots are described in [1]. This paper focuses on introducing a variety of different ways to estimate the channel using these pilots.

[0047] 2. Channel Acquisition

[0048] In the OTFS paradigm, the channel is understood as two domains:

[0049] Time-frequency domain. In this domain, the channel operates through multiplication and appears as a linear combination of two-dimensional waves.

[0050] Delayed Doppler domain. In this domain, the channel operates through convolution and appears as a linear combination of two-dimensional sine waves.

[0051] The two channel domains are correlated via a symptotic Fourier transform.

[0052] The channel acquisition process includes pilot separation, channel interpolation, and / or channel prediction. The following sections will illustrate different ways that channel acquisition can be performed in different domains, including hybrid domains.

[0053] For the purposes of this document, we assume that the time-frequency grid is divided into two or more parts. A fine grid is used for data transmission, and one or more coarse subgrids are used for channel acquisition (pilot transmission). Figure 2 An example of a data grid with two channel acquisition subgrids is shown, one for downlink and one for uplink channel acquisition.

[0054] On the downlink subgrid, the signal received by the antenna of the remote device is a linear combination of pilots transmitted by the hub antenna with channel distortion. The goal is for the remote device to use this received signal to estimate the channel on some portion of the data grid (for the remainder of this paper, we focus on the downlink because the problem is symmetrical with respect to the uplink). Channel acquisition can be divided into two problems: in Figure 3 and Figure 4 The interpolation and prediction methods summarized in the text.

[0055] Channel interpolation is a much easier problem to solve and is sufficient for receiver equalizers. However, for transmitter equalizers, channel prediction is required.

[0056] 3. Time-frequency domain channel acquisition

[0057] The channel acquisition method described in this chapter utilizes the time-frequency domain and can be applied to both pilot separation and channel interpolation and / or prediction.

[0058] One approach is 2D MMSE channel acquisition. This paper describes the theory and mathematical formulas used to implement 2D MMSE channel acquisition. One approach involves performing the following steps:

[0059] (1) Generate the squared autocorrelation matrix R of the pilots of the positive acquisition channel. x The size of this matrix is ​​N·MxN·M, where NΔtxNΔf is the size of the observation field.

[0060] (2) By using R x The restrictions are as follows: R xy =Rx·RES H To calculate the cross-correlation matrix R xy The size of this matrix is ​​N·M x P, where P is the number of pilot samples in the observation window.

[0061] (3) Calculate R y =RES·Rx·RES H +R W =RES·Rxy+R W The size of the matrix is ​​P x P.

[0062] (4) Calculation The size of this matrix is ​​N·M x P

[0063] (5) Calculate the estimated pilot signal vector as C·Y, and reformat the vector into an N×M matrix to form an estimated pilot sample matrix of size N × M on the time-frequency grid Λ.

[0064] (6) Calculate the channel estimate in, Let denote the conjugate of each element of X, and .· denotes element-by-element multiplication. The size of this matrix is ​​N x M.

[0065] (7) Select the part of Ch that represents the time-frequency part that needs to be acquired for channel acquisition.

[0066] A better approach to achieve the same result is to compute the minimum size matrix required in each step to reach the desired portion of Ch. Here is an example of this simplification:

[0067] (1) Only for the portion of the channel that needs to be acquired, the restricted R is directly calculated. x To calculate R using the elements xy For example, if only L < N time samples are needed (for interpolation or prediction beyond the last pilot sample), then the size of the matrix will be L·M x P.

[0068] (2) Only calculate the double-constrained R directly x To calculate R using the elements y .

[0069] (3) Calculation The size of this matrix will now be L·M x P.

[0070] (4) Calculate the estimated pilot signal vector as C·Y, and reform the vector into a matrix of size L·M to form an estimated pilot sample matrix of size L x M on the time-frequency grid Λ.

[0071] (5) Calculate the channel estimate in, Let X denote the conjugate of each element of X, and ... denote element-by-element multiplication. The matrix is ​​of size L x M, and it represents the time-frequency part of the channel acquisition.

[0072] 4. Hybrid Delay Time Domain Acquisition

[0073] The channel acquisition method described in this chapter utilizes a hybrid delay time domain. In this domain, the channel is a local convolution with delay (which facilitates pilot separation) and a nonlocal multiplication with time (which facilitates interpolation and prediction). Figure 5 , Figure 6 and Figure 7 An example of this process is shown.

[0074] 4.1 Pilot Separation in the Delay Time Domain

[0075] Assume four pilot signals are transmitted from four hub antenna ports, and assume the pilot signals are Dirac signals located in the delay-Doppler plane, such as... Figure 8 As shown,

[0076] On the time-frequency pilot subgrid, each receiving antenna can receive a linear combination of each pilot distorted by its own channel. To separate the pilots, the received signal can be transformed from the time-frequency domain to the delay-time domain. This can be done via a one-dimensional Fourier transform in the frequency dimension. Knowing that the channel operates in the delay domain through convolution, the implementation can calculate what the remote device antenna will observe. Let the mapping s i : This represents the signal received by the antenna of the i′th remote device on the downlink subgrid. Then, in the delay time domain, the following equation can be written:

[0077]

[0078] Among them, h ij This represents the channel between the j′th hub antenna and the i′th remote device antenna in the time-delay domain.

[0079] Figure 9 This shows their respective channels after the delay time domain as observed by the receiver. Figure 8 The four pilot frequencies.

[0080] Using the equation above, different pilot (i.e., channel) responses can be isolated through windowing. That is, in order to isolate h... ij Some implementations can perform time-delay operations on s in the time domain. i Apply the following square mask:

[0081]

[0082] Here, Δτ represents the channel delay spread.

[0083] 4.2 Windowing

[0084] Windowing the measured channel in the time-frequency domain (e.g., using a squared cosine window) makes the channel response generally better isolated in the delay domain, thus making pilot separation more effective.

[0085] In addition, M ij Other values ​​can be obtained for (τ, kNdt). When the observation window in the frequency dimension is very large, M ij (τ, kNdt) can be reduced to nearly half the delay spread. In other cases, such as when using a square window in the time-frequency domain, for M ij Larger values ​​of (τ, kNdt) can produce better performance.

[0086] 4.3 Channel Interpolation and Prediction in the Delay Time Domain

[0087] The previous chapter described how the signal received by the receiving antenna can be used to extract the channel between each pair of transmitting and receiving antennas constrained by the pilot subgrid.

[0088] Figure 10 This is a diagram illustrating an exemplary interpolation problem. The left diagram shows the channel response measured on the pilot subgrid. The right diagram shows the desired channel response, where intermediate values ​​of the channel response missing from the values ​​measured from the pilot signal must be generated at the receiver.

[0089] Figure 11 This is a diagram illustrating an example of a one-dimensional interpolation problem. For each delay value τ, some implementations can use the channel values ​​(including two or more pilot subgrid points) on the pilot subgrid to estimate the channel on the data grid.

[0090] The implementation scheme should also address the issues of interpolation and prediction. That is, how to estimate the channel for the past or future portions of the data grid?

[0091] One possible solution for some implementations is to transform the problem into a set of one-dimensional problems. That is, for each delay value, the channel can be interpolated or extrapolated over time.

[0092] The fact that the channel is bandwidth-limited relative to time means that the problem is relatively easy to handle. There are many methods available for performing one-dimensional interpolation or prediction, such as:

[0093] Splines: Find a piecewise polynomial that takes the channel values ​​measured on the downlink subgrid. The values ​​of the resulting function on the data grid are used as channel estimates.

[0094] MMSE: Assume the channel is a random band-limited function. Based on this assumption and channel measurements on the downlink subgrid, calculate the most probable channel value on the data grid.

[0095] Fourier Transform: A Discrete Fourier Transform (DFT) is performed on the channel measured on the downlink subgrid. All Fourier coefficients corresponding to frequencies greater than the channel bandwidth are set to zero. The Fourier coefficients are then zero-padded, and an inverse DFT is applied. The value of the resulting function on the data grid is used as an estimate of the channel.

[0096] All three methods can be implemented efficiently using linear algebra. Figure 12 This demonstrates the application of cubic spline interpolation. Figure 11 The results of the channel shown.

[0097] Once the channel has been interpolated / extrapolated in the time-delay domain, an inverse Fourier transform is used across the delay to transform the channel into the time-frequency domain. In some implementations, the receiving device may perform the following four steps to perform channel estimation:

[0098] [1] The signal measured on the downlink subgrid is transformed to the delay time domain by performing a Fourier transform on the signal.

[0099] [2] By applying a square mask, the channel response between different hubs and remote device antenna pairs is separated.

[0100] [3] Estimate the channel over a portion of the data grid by interpolating or extrapolating the channel response over time.

[0101] [4] The channel is transformed to the time-frequency domain by performing an inverse Fourier transform across the delay.

[0102] Instead of applying the inverse Fourier transform described in step 4 over the full delay span of the delay Doppler plane, some implementations may apply the inverse Fourier transform only over the mask portion, or only over the portion of the delay span that is associated with the pilot grid, and then apply the spline or MMSE method described above again in the time-frequency domain.

[0103] 5. Delayed Doppler Domain Channel Acquisition

[0104] The channel acquisition method described in this chapter utilizes the delayed Doppler domain. In this domain, the channel is a local convolution along the delay and Doppler. In this case, the received signal is transformed from the time-frequency domain to the delayed Doppler domain. This is accomplished through a two-dimensional symplectic Fourier transform.

[0105] Similar to that described in Section 4.1, different pilot (i.e., channel) responses can be isolated using windowing. In this case, a two-dimensional delayed Doppler window is applied to isolate the h responses corresponding to the four pilots in Section 4.1. ij In the delayed Doppler domain, for s i Apply the following square mask:

[0106]

[0107] Here, Δτ and Δv represent the channel delay and Doppler spread.

[0108] In some implementations, the measured channel can be windowed in the time-frequency domain. Windowing the measured channel in the time-frequency domain (e.g., using a squared cosine window) allows for better isolation of the channel response in the delayed Doppler domain, thereby making pilot separation more effective.

[0109] M ijOther values ​​can be obtained for (τ, v). When the observation window in the frequency dimension is very large, M ij The size of (τ, v) can be reduced to nearly half the delay spread. In other cases, such as when using a square window in the time-frequency domain, a larger size M... ij (τ, v) can produce better performance.

[0110] In various implementations, the pilots can be placed in two dimensions; for example, eight pilots can be Dirac delta functions located on the delayed Doppler plane. In this case, pilots also need to be separated in the Doppler dimension, where windowing can be applied for pilot separation.

[0111] After applying the window, the channel on the desired portion of the time-frequency plane can be estimated by applying a 2D symmetric DFT of the delayed Doppler plane with a mask, or alternatively, by applying a 2D symmetric DFT on a portion of the plane and then applying 2D splines or MMSE in the time-frequency plane.

[0112] 6. Introduction to OTFS-based reference signals

[0113] The following sections 7-9 provide a description of implementation methods for pilot signal generation, transmission, and reception in wireless networks.

[0114] 7. Reference signal based on OTFS

[0115] 7.1 Brief Summary

[0116] Typically, OTFS-based reference signals can be transmitted separately from the data. The idea is that reference signals can be transmitted on a time-frequency grid that is coarser than the data grid.

[0117] Suppose that the time-frequency (tf) grid of the data is defined by the following discrete points:

[0118]

[0119] The TF grid of the reference signal (pilot) will be a subset of the data grid:

[0120]

[0121] The delayed Doppler (τ, v) loop associated with the two grids is:

[0122] A data ring with the following perimeter

[0123] ο

[0124] ο

[0125] Pilot loop with the following perimeter

[0126] ο

[0127] ο

[0128] Figure 13 An example of a ring is shown.

[0129] It can be seen that the function X[K, L] defined on the tf grid in (1) can be transformed into a 2D (continuous periodic) function on the correlation ring using the discrete Fourier transform defined as follows:

[0130]

[0131] The inverse transform from x(τ,ν) to X[K,L] is defined as the following inverse symplectic Fourier transform:

[0132]

[0133] The data and pilot grid are constrained to 2D discrete intervals kxl and nxm, respectively:

[0134]

[0135]

[0136] The transition involves uniformly sampling the corresponding rings into kxl and nxm. These sampled rings create a mesh in the tf plane that interacts with their associated mesh, and are defined as follows:

[0137]

[0138]

[0139] When k = N-1, M = 1, and l = m, the tf plane will appear as Figure 14 As shown in the diagram, the horizontal axis corresponds to time, and the vertical axis corresponds to frequency. Filled dots correspond to pilot frequencies, while hollow circles correspond to data.

[0140] A reference signal can be viewed as the symplectic exponent of a subset of points in a data grid. If the subset is regular (i.e. forms a subgrid), the reference signal is considered structured; otherwise, it is considered unstructured.

[0141] 7.2.2D Structured Reference Signal

[0142] The 2D structured reference signal is a pilot signal generated on the delayed Doppler plane, which is associated with a subgrid of the data grid (N>1, M≥1). These pilot signals are multiplexed with the data in the tf domain.

[0143] Several schemes can be used to efficiently generate 2D structured OTFS based on a reference signal for a given delay Doppler spread and allowed channel consumption. Three schemes are described in the following subsections.

[0144] Some implementations may choose a relaxed version of any of these schemes and / or a combination of schemes, depending on the use case and / or limitations (such as backward compatibility, ease of implementation, etc.).

[0145] 7.2.1 Delayed Doppler Packet Reference Signal

[0146] The delayed Doppler packet reference signal can be generated as follows:

[0147] (1) Select the finest TF pilot grid that meets the allowed consumption (minimum N and M).

[0148] (2) In the continuous delay Doppler loop associated with the selected pilot grid, include as many pilots as possible (spaced as sparsely as possible) while taking into account the receiver’s ability to achieve sufficiently good pilot separation and channel estimation of the expected delay, as well as the Doppler propagation of the channel.

[0149] (3) Transform the delayed Doppler plane into the tf plane (e.g., symmetric Fourier transform).

[0150] (4) Apply an n×m window to create an n×m discrete interval.

[0151] Some parameters that may affect how many pilots can be supported by a given pilot loop are: the size of the pilot observation window (the choice of n and m), the channel delay and Doppler spread, and the receiver implementation (e.g., window size and shape, interpolation, and pilot separation algorithms). A sufficiently large TF window (large n and m) will allow up to The best packet for a pilot:

[0152]

[0153] in, It is the largest integer less than x, Δ τ It is the channel delay spread, and Δ v It is the Doppler spread of the channel.

[0154] Figure 15 An example of packets of eight reference signals in the delayed Doppler plane is shown (τ axis represents time, and v axis represents Doppler).

[0155] 7.2.1.1 Example

[0156] The following example illustrates how an implementation can use delayed Doppler packets to generate a large number of reference signals with low power consumption in an enumeration similar to LTE enumeration.

[0157] Assume the following:

[0158] Channel:

[0159] • Bandwidth: 10 MHz

[0160] • Delayed propagation: Δτ = 5us

[0161] • Doppler propagation: Δ ν =300Hz ptp (150Hz peak)

[0162] Data grid (see equation (1)):

[0163] ·dt=66.67us

[0164] ·df=15kHz

[0165] Permissible consumption: <7%

[0166] Delayed Doppler Packet Solution:

[0167] To meet the allowed consumption requirements, one solution is to assign every 15th grid point in the time domain and all grid points in the frequency domain to the reference signal, thereby generating the following delayed Doppler packet:

[0168] Pilot grid (see equation (2)):

[0169] ·N=15

[0170] M = 1

[0171] Associated pilot loop:

[0172] • Delayed perimeter:

[0173] • Doppler perimeter:

[0174] The best package with a large TF window will support the following number of pilots:

[0175]

[0176] For the purposes of this example, assume that the implementation uses raised cosine (RC) filters in both the transmitter and receiver, with the following Tx / Rx windows:

[0177] • Number of time samples: 50 (n=36, with an additional 14 for RC filters)

[0178] • Number of frequency samples: 626 (m=500, plus 126 for RC filters)

[0179] Pilot Packets: For this example, we will limit the number of pilots to 20 and stack them as follows:

[0180] Number of pilots in the delay domain: 10

[0181] Number of pilots in the Doppler domain: 2

[0182] The pilots will also be located with an offset of 4 grid points in the middle of the delay dimension and its assigned Doppler interval. For the first pilot, this is shifted to grid point (5,9).

[0183] Figure 16 The 20 pilots in the delayed Doppler plane are shown.

[0184] Figure 17 The pilot signals of the first 100 grid points in the delay dimension are shown in magnification.

[0185] To illustrate the worst-case effect of windowing on channel time delay spread Figure 18 The diagram shows the received pilot signal (no direct path) when the channel is represented by the following two paths:

[0186] Path 1: Delay = 5µs, Doppler offset = 152.8Hz

[0187] Path 2: Delay = 0.067µs, Doppler offset = -152.8Hz

[0188] These two paths result in the received pilot being precisely positioned in the middle between grid points, leading to the most severe leakage from the receiving window to adjacent grid points, which represents the worst-case scenario of pilot separation.

[0189] exist Figure 18 The received pilot signals in the delayed Doppler plane are shown (magnified to show only the pilot signals in the first 100 grid points in the delay dimension).

[0190] Figure 19 The diagram shows grid points in the delayed Doppler plane where the received signal is -40 dB higher than the highest level of the received pilot. From Figure 19 As can be seen, the leakage of the received pilot to surrounding pilots is 40 dB lower than that of the pilot itself. Other receiver implementations can result in even lower leakage, which would allow for the packetization of more pilots.

[0191] 7.2.2 Time-Frequency Packet Reference Signal

[0192] The time-frequency packet reference signal is generated as follows:

[0193] Choose the coarsest TF pilot grid (maximum N and M) that can support single-channel estimation (one pilot).

[0194] As many as possible These pilot grids can be used at equal intervals between grid points without violating the allowed power consumption.

[0195] A single pilot is placed in a delayed Doppler loop associated with the selected TF pilot grid.

[0196] Transform the delayed Doppler plane into the tf plane (Symphonic Fourier transform).

[0197] As explained in step 2, the tf plane is created. Each copy is separate from the others.

[0198] Apply an nxm window to each of the tf planes (different windows can be applied to each plane).

[0199] The size of the pilot window will affect the quality of channel estimation (a larger grid will improve the quality of channel estimation).

[0200] Figure 20 Examples of time-frequency packets for two reference signals are shown, represented by filled rows of grid points 2002 and 2004, respectively.

[0201] 7.2.3 Potentially Sensitive Packet Reference Signal

[0202] The potential sensitive packet reference signal can be generated as follows:

[0203] (1) Select the finest TF pilot grid that meets the allowable consumption (minimum N and M).

[0204] (2) Select the smallest size pilot observation window in the time domain (minimum n) that supports a pilot in the Doppler ring (delayed Doppler plane).

[0205] (3) In the delayed Doppler loop, while taking into account the receiver’s ability to achieve sufficiently good pilot separation and channel estimation of the expected delay, as well as the Doppler propagation capability of the channel, include as many pilots as possible (spaced as sparsely as possible).

[0206] (4) Transform the delayed Doppler plane into the tf plane (Symphonic Fourier transform).

[0207] (5) Apply the nxm window to the tf plane.

[0208] 7.2.3.1 Example

[0209] Using the same assumptions as in the previous example, the implementation can shorten the window time interval to encapsulate only one pilot in the Doppler dimension. Again assuming an RC filter, the solution can be as follows:

[0210] Number of time samples: 15 (n=9, with an additional 6 for RC filters)

[0211] Number of frequency samples: 626 (m=500, plus 126 for RC filters)

[0212] Pilot Packets: For this example, we limit the number of pilots to 10 and stack them in the delay domain only.

[0213] The pilots will be located with an offset of 4 grid points in the delay dimension and in the middle of the Doppler interval. For the first pilot, this is shifted to grid point (5,5).

[0214] To illustrate the worst-case effect of windowing on channel time delay spread, the received pilot (without a direct path) is shown when the channel is represented by the following two paths:

[0215] Path 1: Delay = 5µs, Doppler offset = 166.7Hz

[0216] Path 2: Delay = 0.067µs, Doppler offset = -166.7Hz

[0217] These two paths result in the received pilot being precisely positioned in the middle between grid points, leading to the most severe leakage from the receiving window to adjacent grid points, which represents the worst-case scenario of pilot separation.

[0218] exist Figure 21 The received pilot signals in the delayed Doppler plane are shown (magnified to show only the pilot signals in the first 100 grid points in the delay dimension).

[0219] Figure 22 The grid points in the delayed Doppler plane are shown, where the received signal is -40 dB higher than the highest level of the received pilot.

[0220] 7.3. Unstructured Reference Signal

[0221] The unstructured reference signal RE pilot is generated on a delayed Doppler ring with the same perimeter as the data ring, transformed to the tf plane, and then constrained by a finite number of grid points on the tf grid. The number of grid points assigned to the reference signal and their positions will determine how many pilots this assignment can support.

[0222] 8. Example of a downlink reference signal based on OTFS

[0223] 8.1 Cell-specific reference signal

[0224] Cell-specific reference signals are transmitted in all downlink (DL) subframes and are available to all UEs in the cell.

[0225] 8.1.1FDD

[0226] In FDD, because DL transmission is regular, some implementations can use delayed Doppler packets with minimal overhead, allowing the number of packets to be equal to the number of antenna ports used by a transmitting device such as an eNodeB. Some implementations can transmit pilots continuously (an infinite grid in the time domain). Each pilot is separately transformed from the delayed Doppler plane to the tf plane and transmitted to its antenna port (see [link to FDD implementation]). Figure 23 The receiver can apply an appropriate window to achieve separation between different pilots and sufficiently good interpolation from the pilot grid to the data grid in the TF plane. Continuous pilot transmission should not affect data delay because the receiver can implement a sliding TF window from power-on, continuously collecting pilot information and preparing to estimate the channel when control information or data is needed. The number of pilots transmitted and their positions in the delay Doppler plane need to be communicated to the UE. The size and shape of the TF window at the receiver are implementation-specific and do not need to be specified.

[0227] 8.1.2 TDD

[0228] In TDD, the regularity of DL transmission depends on the frame size. In LTE, the shortest DL transmission period is 5ms (half the frame size), which, if used as a time-domain grid point in the tf plane, results in a Doppler perimeter of the pilot loop of 200Hz (1 / 5ms). This translates to the ability to estimate a channel with a Doppler spread not exceeding 200Hz. If shorter DL periodicity is supported, a 2D structured reference signal can be used to support a larger Doppler spread (or more pilots).

[0229] If the regularity of DL transmissions is insufficient to support channel delay spread, one (or both) of the following other options can be used:

[0230] (1) If the grid point allocation of the reference signal supports a structure with dimensions higher than 2D, then the reference signal is used.

[0231] (2) Use 2D unstructured reference signals and limit them to all tf grid points that can be used as reference signals.

[0232] 8.2. UE-specific reference signals

[0233] When an eNodeB transmits data to a UE, UE-specific reference signals may be transmitted only to that specific UE. These reference signals may be transmitted when the eNodeB uses an antenna port different from the cell-specific antenna port (e.g., when the eNodeB uses UE-specific beamforming when transmitting data to the UE). In this case, the reference signals can only be transmitted within the frequency band and time allocated to the specific transmission for the UE.

[0234] Any reference signal described in Section 7 can be used as a UE-specific reference signal. The type chosen will depend on which TF grid points are available for the reference signal, as well as the frequency width and length of the transmission. As an example, a wider bandwidth transmission will be able to send multiple pilots, which can be used to support multiple beams for the same UE.

[0235] 2D unstructured reference signals can be used to transmit one or more pilots for any number of antenna ports on a currently defined LTE UE-specific RS arrangement.

[0236] 9. Example of an uplink reference signal based on OTFS

[0237] 9.1 Demodulation reference signal

[0238] Any reference signal described in Chapter 7 can be used to demodulate a reference signal. The type chosen will depend on which TF trough points are available for the reference signal, as well as the frequency width and length of the transmission. As an example, a wider bandwidth transmission will be able to send multiple pilots, which can be used to support multiple beams.

[0239] 2D unstructured reference signals can be used to transmit one or more pilots on the currently defined LTE demodulation RS arrangement.

[0240] 9.2 Detection reference signal

[0241] The probe reference signal can be regular, and therefore a delayed Doppler packet scheme is recommended to maximize the number of UEs simultaneously transmitting pilots for a given delay spread. To minimize the time length used to estimate the channel, the maximum-size grid spacing can be used in the frequency dimension to select potentially sensitive packets.

[0242] Figure 24 This is a flowchart of an exemplary method 200 for wireless communication. Method 200 may be provided by a wireless receiver (e.g., Figure 1 The receiver 102 described in the document is implemented.

[0243] Method 200 includes receiving, at 202, a wireless signal comprising a data signal portion and a pilot signal portion, wherein the pilot signal portion comprises multiple pilot signals multiplexed together in an orthogonal time-frequency space (OTFS) domain. The data portion may include information bits, such as user data and / or control data.

[0244] Method 200 includes performing a two-dimensional channel estimation in the time-frequency domain at point 204 based on a minimum mean square error (MMSE) optimization criterion. In some implementations, such as those described in Section 3, the two-dimensional channel estimation may be performed by identifying a subset of pilot signal portions from which a preliminary channel estimate in the time-frequency domain is obtained, estimating the preliminary channel estimate using samples of the radio signal received within a two-dimensional observation window, and obtaining a channel estimate by interpolating and / or predicting the preliminary channel estimate.

[0245] As further explained in sections 3 through 6, channel estimation can be performed using all time-domain signal samples, or alternatively, it can be performed only at a subset of, for example, the received pilot signals, to obtain an initial channel estimate, and the entire channel can then be obtained using the estimate from the initial channel estimate through channel interpolation and / or channel prediction.

[0246] In some implementations, the FIR Wiener filter can be used for channel estimation based on the received pilot signal.

[0247] Method 200 includes recovering the information bits at 206 using a channel estimate obtained from a two-dimensional channel estimate.

[0248] Figure 25 This is a flowchart of an exemplary method 300 for wireless communication. Method 300 may be provided by a wireless receiver (e.g., Figure 1 The receiver 102 described in the document is implemented.

[0249] Method 300 includes receiving, at 302, a wireless signal comprising a data signal portion and a pilot signal portion on a communication channel, wherein the pilot signal portion comprises multiple pilot signals multiplexed together in an orthogonal time-frequency space (OTFS) domain.

[0250] Method 300 includes transforming the received wireless signal into a delayed time-domain signal at 304. In some embodiments, a Fourier transform is used to perform the transform.

[0251] Method 300 includes applying a delayed time-domain mask at 306 to separate each of the contributions of a plurality of pilot signals to the delayed time-domain signal into a corresponding individual received pilot signal contribution. In some embodiments, a square mask and zeros may be used, wherein the square mask includes a value "1" for delay and a time value less than its corresponding threshold. Optionally, a smoother window function may be used for the mask.

[0252] Method 300 includes estimating the communication channel at the time-domain location of the delay of multiple pilot signals using the individual received pilot signal contributions at 308. The estimation can be performed using one of several different techniques, such as piecewise polynomial approximation, MMSE, or frequency-limited matched approximation.

[0253] Method 300 includes interpolating the communication channel estimate at 310 to obtain the overall communication channel estimate. For multiple delay values, interpolation can be performed across the time domain using a preliminary channel estimate obtained using the techniques described herein.

[0254] Figure 26 This is a flowchart of an exemplary method 400 for wireless communication. Method 300 may be provided by a wireless receiver (e.g., Figure 1 The receiver 102 described in the document is implemented.

[0255] Method 400 includes receiving, at 402, a wireless signal comprising a data signal portion and a pilot signal portion on a communication channel, wherein the pilot signal portion comprises multiple pilot signals multiplexed together in an orthogonal time-frequency space (OTFS) domain.

[0256] Method 400 includes transforming the received wireless signal into a delayed Doppler domain signal by applying a two-dimensional Sin Fourier transform at 404.

[0257] Method 400 includes applying a delayed Doppler domain mask at 406 to separate each of the contributions of multiple pilot signals to the time-frequency domain signal into a corresponding individual received pilot signal contribution.

[0258] Method 400 includes using individual received pilot signal contributions at 408 to estimate the communication channel at the delayed Doppler domain location of multiple pilot signals.

[0259] Method 400 includes interpolating the communication channel estimate at 410 to obtain the overall communication channel estimate.

[0260] In some embodiments, the wireless communication device may include a memory and a processor, wherein the memory stores instructions that, when implemented, cause the processor to implement method 200, 300, or 400.

[0261] Additional details of methods 200, 300 and 400 are also described in sections 3, 4 and 5, respectively (e.g., time-frequency domain channel acquisition, or hybrid delay time-domain channel acquisition, or delay Doppler domain channel acquisition).

[0262] Figure 27An example of a wireless transceiver device 500 is shown. Device 500 can be used to implement methods 200, 300, or 400. Device 500 includes a processor 502 and a memory 504, wherein the memory 504 stores instructions and data executable by the processor during computation. Device 500 includes receiving and / or transmitting circuitry 506, for example, including radio frequency operations for receiving or transmitting signals.

[0263] It should be understood that a technique for receiving wireless data by performing channel acquisition through OTFS pilot signals has been disclosed.

[0264] The embodiments, modules, and other embodiments and functional operations disclosed and described herein can be implemented as digital electronic circuits or computer software, firmware, or hardware (including the structures disclosed herein and their structural equivalents), or a combination thereof. The disclosed embodiments and other embodiments can be implemented as one or more computer program products, i.e., one or more modules of computer program instructions encoded on a computer-readable medium for execution by a data processing device or for controlling the operation of a data processing device. The computer-readable medium can be a combination of machine-readable storage devices, machine-readable storage substrates, memory devices, substances influencing machine-readable propagation signals, or a combination thereof. The term "data processing device" encompasses all devices, apparatuses, and machines for processing data, including, for example, programmable processors, computers, or multiple processors or computers. In addition to hardware, the device may include code that creates an execution environment for the computer program in question, such as code constituting processor firmware, a protocol stack, a database management system, an operating system, or a combination thereof. Propagation signals are artificially generated signals, such as machine-generated electrical, optical, or electromagnetic signals, generated to encode information for transmission to a suitable receiver device.

[0265] Computer programs (also known as programs, software, software applications, scripts, or code) can be written in any form of programming language, including compiled or interpreted languages, and can be deployed in any form, including as standalone programs or as modules, components, subroutines, or other units suitable for use in a computing environment. A computer program does not necessarily correspond to a document in a documenting system. A program may be stored as part of a document that holds other programs or data (e.g., one or more scripts stored in a markup language document), in a single document dedicated to the program in question, or in multiple coordinating documents (e.g., a document storing one or more modules, subroutines, or portions of code). Computer programs can be deployed to execute on a single computer, located at a single site, or distributed across multiple sites and interconnected via a communications network.

[0266] The processes and logic described herein can be executed by one or more programmable processors, which execute one or more computer programs to perform functions by manipulating input data and generating outputs. These processes and logic can also be executed by special-purpose logic circuitry, such as FPGAs (Field-Programmable Gate Arrays) or ASICs (Application-Specific Integrated Circuits), and the device can also be implemented as said special-purpose logic circuitry.

[0267] For example, processors suitable for executing computer programs include general-purpose and special-purpose microprocessors, as well as any one or more processors in any type of digital computer. Generally, a processor receives instructions and data from read-only memory or random access memory, or both. The fundamental elements of a computer are a processor for executing instructions and one or more memory devices for storing instructions and data. Generally, a computer will also include one or more mass storage devices (e.g., magnetic disks, magneto-optical disks, or optical disks) for storing data, or be operatively connected to receive data from or transfer data to said one or more mass storage devices, or both. However, a computer does not need to have such devices. Computer-readable media suitable for storing computer program instructions and data include all forms of non-volatile memory, media, and memory devices, including, for example: semiconductor memory devices, such as EPROM, EEPROM, and flash memory devices; magnetic disks, such as internal hard disks or removable disks; magneto-optical disks; and CD-ROM and DVD-ROM disks. The processor and memory may be supplemented or incorporated therein by dedicated logic circuitry.

[0268] While this patent document contains numerous details, these should not be construed as limiting the scope of the claimed invention or what may be claimed, but rather as descriptions of features specific to particular embodiments. Certain features described herein in the context of individual embodiments may also be implemented in combination in a single embodiment. Conversely, various features described in the context of a single embodiment may also be implemented individually or in any suitable sub-combination in multiple embodiments. Furthermore, although features may be described above as functioning in certain combinations, and even initially claimed to be so, one or more features from a claimed combination may be removed from said combination in some cases, and the claimed combination may be for sub-combinations or variations thereof. Similarly, although operations are depicted in a specific order in the drawings, this should not be construed as requiring that such operations must be performed in the specific order shown or in a sequential order, or that all shown operations must be performed to achieve the desired result.

[0269] Only a few examples and implementations are disclosed. Changes, modifications, and enhancements can be made to the described examples and implementations, as well as other implementations, based on the disclosed content.

Claims

1. A wireless communication method for recovering information bits from a received signal, the method being implemented by a wireless communication receiver, the method comprising the following steps: Receive a wireless signal comprising a data signal portion and a pilot signal portion on a communication channel, wherein the pilot signal portion comprises multiple pilot signals multiplexed together in the orthogonal time-frequency space (OTFS) domain; The wireless signal is converted into a delayed time-domain signal; A time-delay mask is applied to separate the contribution of each of the plurality of pilot signals for the time-delay signal into the corresponding individual received pilot signal contribution. The individual received pilot signal contributions are used to estimate the communication channel at the time-domain delay location of the plurality of pilot signals; and The communication channel estimate is interpolated to obtain the overall communication channel estimate.

2. The method as described in claim 1, wherein, The transformation is performed using a Fourier transform.

3. The method as described in claim 1, wherein, The time-domain delay mask includes a square mask.

4. The method of claim 1, wherein, The interpolation is performed across time for multiple delay values ​​in the delay time domain.

5. The method of claim 1, wherein, The estimation includes: The communication channel estimate is estimated using spline approximation techniques based on the contribution of the individually received pilot signal, wherein a piecewise continuous polynomial is calculated in the spline approximation techniques.

6. The method of claim 1, wherein, The estimation includes: The communication channel estimate is estimated using the minimum mean square error criterion based on the contribution of the individually received pilot signal.

7. The method of claim 1, wherein, The estimation includes: The communication channel estimate is estimated by applying bandwidth constraints by zeroing the high-frequency coefficients obtained by applying frequency transformation to the contribution of the individually received pilot signal.

8. The method of claim 1, wherein, The interpolation includes: Interpolation is performed using a polynomial approximation.

9. The method of claim 1, further comprising: The information bits are recovered using the channel estimate obtained from the entire communication channel estimate.

10. The method of claim 9, wherein, Recovering the information bits includes: Perform symbol demodulation to recover the information bits.

11. The method of claim 1, further comprising: The communication channel estimation is windowed in the time-frequency domain.

12. The method of claim 11, wherein, The windowing includes windowing using a square cosine window.

13. The method of claim 1, further comprising: In the time-frequency domain, the channel estimation is performed by an inverse Fourier transform across the delay dimension.

14. The method of claim 13, wherein, The application of the inverse Fourier transform includes: The inverse Fourier transform is applied only on the mask portion or only on the portion corresponding to the delay span around the received pilot signal.

15. Wireless communication equipment, including: Memory, used to store data and instructions; A processor is configured to execute the instructions and implement a data receiving method, the data receiving method comprising: Receive a wireless signal comprising a data signal portion and a pilot signal portion on a communication channel, wherein the pilot signal portion comprises multiple pilot signals multiplexed together in the orthogonal time-frequency space (OTFS) domain; The received wireless signal is converted into a delayed time-domain signal; A time-delay mask is applied to separate the contribution of each of the plurality of pilot signals for the time-delay signal into the corresponding individual received pilot signal contribution. The individual received pilot signal contributions are used to estimate the communication channel at the time-domain delay location of the plurality of pilot signals; and The communication channel estimate is interpolated to obtain the overall communication channel estimate.

16. The device as claimed in claim 15, wherein, The transformation is performed using a Fourier transform.

17. The device as claimed in claim 15, wherein, The time-domain delay mask includes a square mask.

18. The device as claimed in claim 15, wherein, The interpolation is performed across time for multiple delay values ​​in the delay time domain.

19. The device as claimed in claim 15, wherein, The estimation includes: The communication channel estimate is estimated using spline approximation techniques based on the contribution of the individually received pilot signal, wherein a piecewise continuous polynomial is calculated in the spline approximation techniques.

20. The device as claimed in claim 15, wherein, The estimation includes: The communication channel estimate is estimated using the minimum mean square error criterion based on the contribution of the individually received pilot signal.

21. The device as claimed in claim 15, wherein, The estimation includes: The communication channel estimate is estimated by applying bandwidth constraints by zeroing the high-frequency coefficients obtained by applying frequency transformation to the contribution of the individually received pilot signal.

22. The device as claimed in claim 15, wherein, The interpolation includes: Interpolation is performed using a polynomial approximation.

23. The apparatus of claim 15, further comprising: The information bits are recovered using the channel estimate obtained from the entire communication channel estimate.

24. The device as claimed in claim 23, wherein, Recovering the information bits includes: Perform symbol demodulation to recover the information bits.

25. A wireless communication method for recovering information bits from a received signal, the method being implemented by a wireless communication receiver, the method comprising the steps of: Receive a wireless signal comprising a data signal portion and a pilot signal portion on a communication channel, wherein the pilot signal portion comprises multiple pilot signals multiplexed together in the orthogonal time-frequency space (OTFS) domain; The wireless signal is transformed into a delayed Doppler domain signal by applying a two-dimensional Sin Fourier transform; Delayed Doppler domain masking is applied to separate the contribution of each of the plurality of pilot signals into the corresponding individual received pilot signal contribution; The individual received pilot signal contributions are used to estimate the communication channel at the delayed Doppler domain location of the plurality of pilot signals; and The communication channel estimate is interpolated to obtain the overall communication channel estimate.

26. The method of claim 25, wherein, The delayed Doppler domain mask includes a square mask.

27. The method of claim 26, wherein, The interpolation is performed across time for multiple delay values ​​in the delayed Doppler domain.

28. The method of claim 25, wherein, The estimation includes: The communication channel estimate is estimated using the minimum mean square error criterion based on the contribution of the individually received pilot signal.

29. The method of claim 25, further comprising: The information bits are recovered using the channel estimate obtained from the entire communication channel estimate.

30. The method of claim 25, further comprising: The multiple pilot signals are isolated by setting the delay and Doppler values ​​outside the entire channel estimate to zero; as well as The multiple pilot signals are transformed back to the time-frequency domain using a two-dimensional inverse symptotic Fourier transform.

31. The method of claim 25, further comprising: The plurality of pilot signals are transformed back to the time-frequency domain by using a two-dimensional inverse sine Fourier transform, and the estimation of the communication channel estimate includes: using an optimization algorithm to estimate the communication channel estimate in the time-frequency domain.

32. The method of claim 31, wherein, The optimization algorithm includes either the minimum mean square error algorithm or the spline optimization algorithm.

33. The method of claim 31, further comprising: The communication channel estimation is windowed in the time-frequency domain.

34. The method of claim 33, wherein, The windowing includes windowing using a square cosine window.

35. The method of claim 25, wherein, The multiple pilot signals are delta functions located at discrete points in the delayed Doppler domain.

36. A computer-readable storage medium on which a computer program is stored, wherein, When the computer program is executed by a processor, the method described in any one of claims 1 to 14 or claims 25 to 35 is performed.

Images