Transmitting device for generating OOK modulated spread spectrum DFT-S-OFDM wake-up signal

Abstract

Embodiments of the present invention relate to a transmitting device (100) for transmitting an OFDM signal (510) in a communication system (500). The OFDM signal (510) is obtained by spreading a sequence of Nbit bits to obtain Nsymb modulation symbols by multiplying each bit of the sequence of Nbit bits by a corresponding spreading sequence of a sequence of Nbit spreading sequences. Each spreading sequence of the sequence of the Nbit spreading sequences is a linear phase sequence having a constant rotational phase angle [phi]. And the Nsymb modulation symbols are multiplied by a discrete Fourier transform precoder, so that Nsymb Fourier coefficients can be obtained. Finally, the OFDM signal (510) is transmitted, the OFDM signal (510) comprising the Nsymb number of Fourier coefficients mapped onto the K number of OFDM subcarriers. The invention further relates to a corresponding method and to a computer program.

Description

Technical Field

[0001] Embodiments of the present invention relate to a transmitting device for a communication system. In addition, embodiments of the present invention also relate to corresponding methods and computer programs. Background Art

[0002] The concept of a wake-up signal (WUS) has been introduced in several communication standards in the wireless industry. Its goal is to help devices significantly reduce their functionality, thereby reducing power consumption, until they receive this specific WUS.

[0003] In previous 3GPP releases, long-term evolution machine-type communication (LTE-M) and narrowband Internet of Things (NB-IoT) specified the machine-type communication (MTC) wake-up signal (MWUS) and narrowband WUS (NWUS), respectively. These are specific types of OFDM signals designed for normal 3GPP New Radio (NR) receivers: an OFDM-modulated Zadoff-Chu (ZC) sequence that encodes the cell identity (ID). As such, it is an OFDM-based WUS, preserving orthogonality with other signals, but also requiring high-level synchronization and high-precision analog-to-digital conversion (ADC) for detection, such as provided by the primary radio of the receiving NR device. MWUS / NWUS enable energy savings at the receive detector because their transmissions are shorter and carry only a small number of bits compared to other data channels, which typically require repetition to extend coverage. However, energy savings are still limited because the NR receiver still needs to be in deep sleep mode, which accounts for a significant portion of the overall energy consumption of user equipment (UE). So far, the NWUS / MWUS functionality does not appear to have been deployed in products by network operators.

[0004] In addition, the current 3GPP RAN1Rel-18 standard is specifically studying low-power WUS (LP-WUS). It envisions that if the main radio of the NR receiver can be completely shut down when no messages arrive, more energy savings can be achieved. To this end, the NR device will be equipped with an additional low-power detection receiver called a low-power wake-up receiver (LP-WUR). The WUR will monitor possible incoming traffic, while the main radio can be completely shut down to maximize power savings, triggering only when necessary. Summary of the Invention

[0005] An object of the embodiments of the present invention is to provide a solution to reduce or solve the disadvantages and problems of conventional solutions.

[0006] Another object of embodiments of the present invention is to provide a low complexity on-offkeying (OOK) signal that can be used, for example, as a WUS.

[0007] The above and other objects are achieved by the subject-matter of the independent claims. Further embodiments of the invention are provided in the dependent claims.

[0008] According to a first aspect of the present invention, the above and other objects are achieved by a transmitting device for a communication system, the transmitting device being configured to:

[0009] By adding N bit Each bit in the sequence of bits is bit The corresponding spreading sequences in the sequence of the N spreading sequences are multiplied to obtain the N bit The sequence of bits is spread to obtain N symb modulation symbols, wherein the N bit Each spreading sequence in the sequence of spreading sequences is a linear phase sequence with a constant rotation phase angle Φ;

[0010] The N symb The modulation symbols are multiplied by the discrete Fourier transform precoder to obtain N symb Fourier coefficients;

[0011] An orthogonal frequency-division multiplexing (OFDM) signal is transmitted, wherein the OFDM signal includes the N subcarriers mapped onto K OFDM subcarriers. symb Fourier coefficients.

[0012] The transmitting device may be a part of any suitable communication device for communicating in a communication system or may be fully integrated in the communication device.In addition, the transmitting device may also have the capability of receiving communication signals in the communication system, rather than just the capability of transmitting communication signals.

[0013] The transmitting device according to the first aspect has the advantage of providing a multi-bit OOK signal with lower complexity compared to conventional solutions. Furthermore, it can provide a flatter ON / OFF modulation state, thereby improving robustness against quantization errors of low-precision ADCs at the receiving device. Compared to conventional solutions, the transmitting device according to the first aspect can also better control the signal spectrum.

[0014] In an implementation of the sending device according to the first aspect, the N bit The bits are spread based on:

[0015] Repeat N bit bits to get N symb A sequence of repeated bits;

[0016] The N symb The repeated bits are multiplied by the concatenated spreading sequence to obtain the N symb modulation symbols, wherein the concatenated spreading sequence is the N bit The cascade of the spreading sequences makes the cascaded spreading sequence a linear phase sequence with a constant rotation phase angle Φ.

[0017] The advantage of this implementation is that the designation and implementation of a single cascaded spreading sequence can be faster than N bit The specification and implementation of the sequence of the spreading sequence is simpler.

[0018] In an implementation of the sending device according to the first aspect, the N bit bits are based on N bit / 2 bits of the Manchester coded sequence.

[0019] The advantage of this implementation is that Manchester coding can achieve a constant energy level of the transmitted signal at the expense of halving the information rate, and furthermore, it is no longer necessary to determine a threshold value for detection at the receiving device.

[0020] In an implementation of the transmitting device according to the first aspect, the spreading sequence r l [m] is given by the following formula:

[0021]

[0022] Where l is the bit index, m is the modulation symbol index, e is the natural exponential function, j is the imaginary unit, Φ l is a constant angle that depends on the bit index l.

[0023] The advantage of this implementation is that only two angles Φ and Φ need to be specified and stored in the sending device. l , to generate the spreading sequence.

[0024] In an implementation manner of the transmitting device according to the first aspect, the constant rotation phase angle Φ is equal to π.

[0025] The advantage of this implementation is that the choice of phase angle minimizes the envelope fluctuation of the OOK state.

[0026] In an implementation of the transmitting device according to the first aspect, the spreading sequence r l [m] is an alternating sequence of values +1 and –1.

[0027] The advantage of this implementation is that its complexity is very low, since the sign change does not require computation, i.e., multiplication.

[0028] In an implementation of the transmitting device according to the first aspect, the spreading sequence r l [m] is an alternating sequence of two binary shift keying symbols.

[0029] The advantage of this implementation is that it reuses constellation symbols that have been specified and implemented in the 3GPP system.

[0030] In an implementation manner of the transmitting device according to the first aspect, the constant rotation phase angle Φ is given by the following formula:

[0031]

[0032] Among them, N seg is the spreading sequence r l length [m], k null are the indices of the Fourier coefficients to be zeroed, and λ is any non-zero integer.

[0033] The advantage of this implementation is that it is able to zero certain Fourier coefficients, ie set the Fourier coefficients to zero, for example the DC subcarrier which can be filtered out by the circuitry of the receiving device.

[0034] In an implementation of the transmitting device according to the first aspect, the size of the discrete Fourier transform precoder is N symb ≤K.

[0035] The advantage of this implementation is that the discrete Fourier transform (DFT) precoder size N can be selected symb , so that it is the number of bits N bit Integer factors of , and therefore, each bit can be spread by the same spreading factor and thus sent with the same energy. In addition, the complexity of the DFT precoder size smaller than the WUS bandwidth K is much lower than the complexity of the typical OFDM inverse fast Fourier transform (IFFT). To further reduce the complexity, the DFT precoder size N symb It may be chosen to be a power of 2, for example.

[0036] In an implementation manner of the sending device according to the first aspect, the sending device is configured to:

[0037] Based on the N symb The periodic repetition of the Fourier coefficients, the N symb Fourier coefficients are expanded into K Fourier coefficients.

[0038] The advantage of this implementation is that it can symb Fourier coefficients are mapped to a greater number of subcarriers K. Using more subcarriers can generate an OOK signal with faster transitions between the ON and OFF states and less fluctuation within the state. Using more subcarriers can also exploit frequency diversity to improve detection at the receiving device.

[0039] In an implementation manner of the sending device according to the first aspect, the sending device is configured to:

[0040] The N symb The K Fourier coefficients are multiplied by the frequency domain spectrum shaping window coefficients to obtain frequency shaping Fourier coefficients.

[0041] The advantage of this implementation is that the frequency domain spectrum shaping will further flatten the OOK state, which will improve the robustness to detection errors.

[0042] In an implementation manner of the transmitting device according to the first aspect, the frequency domain spectrum shaping window coefficients are real-valued symmetric coefficients from a bell-shaped function.

[0043] The advantage of this implementation is that it is known that such a FDSS window concentrates the energy of the DFT-s-OFDM pulse well in time, which improves the shape of the OOK signal.

[0044] In an implementation manner of the transmitting device according to the first aspect, the frequency domain spectrum shaping window coefficient is a Kaiser window coefficient with a shaping parameter β=2.

[0045] The advantage of this implementation is that it provides a good least squares approximation of the ideal OOK signal.

[0046] In an implementation manner of the transmitting device according to the first aspect, the frequency domain spectrum shaping window coefficient W0[k] is given by the following formula:

[0047]

[0048] Among them, N fft is the number of samples of the OFDM signal, and sin() is the sine function.

[0049] The advantage of this implementation is that it corresponds to the best least squares approximation of the ideal OOK signal.

[0050] In an implementation manner of the sending device according to the first aspect, the sending device is configured to:

[0051] The frequency-shaped Fourier coefficients are multiplied by the frequency-domain phase shift to obtain the phase-shifted Fourier coefficients, wherein the frequency-domain phase shift is based on the shift parameter T shift .

[0052] The advantage of this implementation is that it can improve the temporal position of the OOK state by maximizing the energy of the OOK state within its target time domain period.

[0053] In an implementation of the transmitting device according to the first aspect, the shift parameter T shift The value depends on the number of samples N of the OFDM signal fft and the N symb modulation symbols.

[0054] The advantage of this implementation is that since the OOK signal spans N fft The OFDM signal of N samples is symb It is constructed from the multiplexing of time-domain pulses and is therefore sufficient to control the temporal position discussed above.

[0055] In an implementation of the transmitting device according to the first aspect, the shift parameter T shift The value of is given by either of the following formulas:

[0056]

[0057] Among them, N fftis the number of samples of the OFDM signal (510), is the ceiling function, is the floor function, and round[] is the rounding function.

[0058] The advantage of this implementation is that it allows close to optimal time positioning, since it corresponds to half the time difference between two consecutive time-domain pulses.

[0059] In an implementation manner of the transmitting device according to the first aspect, the OFDM signal is a wake-up signal.

[0060] According to a second aspect of the present invention, the above and other objects are achieved by a method for sending a device, the method comprising:

[0061] By adding N bit Each bit in the sequence of bits is bit The corresponding spreading sequences in the sequence of the N spreading sequences are multiplied to obtain the N bit The sequence of bits is spread to obtain N symb modulation symbols, wherein the N bit Each spreading sequence in the sequence of spreading sequences is a linear phase sequence with a constant rotation phase angle Φ;

[0062] The N symb The modulation symbols are multiplied by the discrete Fourier transform precoder to obtain N symb Fourier coefficients;

[0063] Send an OFDM signal, the OFDM signal including the N mapped onto K OFDM subcarriers symb Fourier coefficients.

[0064] The method according to the second aspect can be extended to an implementation corresponding to the implementation of the sending device according to the first aspect. Therefore, an implementation of the method includes one or more features of the corresponding implementation of the sending device.

[0065] The advantages of the method according to the second aspect are the same as the advantages of the corresponding implementation of the sending device according to the first aspect.

[0066] An embodiment of the present invention further relates to a computer program, characterized by program code, which, when executed by at least one processor, causes the at least one processor to perform any method provided by an embodiment of the present invention. In addition, an embodiment of the present invention further relates to a computer program product, comprising a computer-readable medium and the computer program, wherein the computer program is included in the computer-readable medium and may include one or more of the following groups: read-only memory (ROM), programmable ROM (PROM), erasable PROM (EPROM), flash memory, electrically erasable PROM (EEPROM), hard disk drive, etc.

[0067] Other applications and advantages of embodiments of the present invention will be apparent from the following detailed description. BRIEF DESCRIPTION OF THE DRAWINGS

[0068] The accompanying drawings are intended to illustrate and explain different embodiments of the present invention, in which:

[0069] Figure 1 The sending device provided by the embodiment of the present invention is shown;

[0070] Figure 2 A flowchart of a method for sending a device provided by an embodiment of the present invention is shown;

[0071] Figure 3 The receiving device provided by the embodiment of the present invention is shown;

[0072] Figure 4 The communication system provided by the embodiment of the present invention is shown;

[0073] Figure 5 Another block diagram of a sending device provided by an embodiment of the present invention is shown;

[0074] Figure 6 The OOK state flattening effect of the selected linear phase ramp is shown for the spreading sequence r[m], where the 2-bit string [1 0] is transmitted with K=72 without FDSS;

[0075] Figure 7 The OOK state flattening effect of the linear phase ramp selected in the spreading sequence r[m] is shown, where the 8-bit string [1 0 0 11 0 1 0] is transmitted in FDSS with K = 72 and γ = 4.

[0076] Figure 8The difference between equations (21) and (22) and their approximation is shown, Φ = π, where the length is N bit0 =4 Manchester coded bit string is sent with K=72 and uses FDSS with β=4. Figure 8 (a) shows the envelope of [1 0 0 1 1 0 10], Figure 8 (b) shows the average power of the DFT coefficients;

[0077] Figure 9 Shows N bit0 =4 and the DC subcarrier of K=72 is set to zero, where FDSS with β=4 is used. Figure 9 (a) shows the envelope of [1 0 01 1 0 1 0], Figure 9 (b) shows the average power of the DFT coefficients;

[0078] Figure 10 The BER is shown as a function of the phase ramp angle Φ and different FDSS coefficients β. Figure 10 (a) corresponds to α BPF =2 and S e =128 shows that Figure 10 (b) Corresponding to α BPF =1 and S e =64 shows;

[0079] Figure 11 shows the BER as a function of SNR, where Figure 11 In (a), S e =128, in Figure 11 In (b), S e =64;

[0080] Figure 12 The BER performance is shown, where the WUR simulation circuit filters out the DC component;

[0081] Figure 13 It shows that the OOK signal is time-shifted corrected by FD phase shift;

[0082] Figure 14 The benefit of FD phase shift (TD cyclic shift) is shown. Figure 14 (a) shows the envelope of [1 0 0 1 1 0 1 0], Figure 14 (b) shows Manchester decoding after downsampling;

[0083] Figure 15 Shown for N bit0 =4 and α BPF = 1, BER performance with and without FD phase shift;

[0084] Figure 16 The PAPR is shown as a function of the phase ramp angle Φ and different FDSS coefficients β. Figure 16 (a) Corresponding to K=24, N e =0 and N bit0 =2 shows that, Figure 16 (b) Corresponding to K=24, N e =8 and N bit0 =8 shows that Figure 16 (c) Corresponding to K=72, N e =0 and N bit0 =4 shows that, Figure 16 (d) Corresponding to K=72, N e =8 and N bit0 =8 shows;

[0085] Figure 17 A representation of TD pulse multiplexing is shown;

[0086] Figure 18 The benefit of spectrum spreading in creating equal spreading factors for each bit is shown. DETAILED DESCRIPTION

[0087] To achieve very low power consumption, WUR can better use a simple non-coherent envelope detector, so WUS using OOK modulation is considered very suitable. OOK modulates the bit by two amplitude values (usually named as the state ON and OFF), see Table 1. In practice, the amplitude value of the signal state fluctuates and depends on the pulse shaping. Ideally, the OFF state will have a constant amplitude value of 0, and the ON state will have a constant amplitude value A≠0, usually assuming A=1 by convention.

[0088] Table 1: Simple OOK

[0089] Information bits state 1 ON 0 OFF

[0090] Since all current NR channels / signals use OFDM modulation, it is expected that a conventional OFDM-based NR transmitter will be able to generate WUS even if it uses a different waveform such as OOK. In addition, it is expected that WUS can be directly orthogonally frequency multiplexed with other concurrent OFDM transmissions without interfering with them. To achieve this, WUS should be generated based on OFDM by filling some dedicated subcarriers. Here, a set of subcarriers (K) used for WUS is multiplexed with subcarriers carrying other data symbols. These subcarriers can be multiplexed with N before adding the cyclic prefix (CP). fft Click IFFT to process them together.

[0091] Formally, the transmitted OFDM signal s[n] is WUS sW [n] and the data signal s generated by a single OFDM modulation D [n]. One with sample index -N CP ≤n≤N fft The CP-OFDM symbol of -1 is calculated as (any normalization coefficients are omitted for simplicity)

[0092]

[0093] in,

[0094]

[0095] Make

[0096]

[0097] in

[0098]

[0099] and

[0100]

[0101] Therefore, WUS can be expressed as

[0102]

[0103] Otherwise, it can be assumed that N fft =2048, where CP length N CP =144, which is a common 3GPP parameter set (numerology), but is not limited thereto.

[0104] Therefore, the purpose of the present invention is to propose a scalable OOK-OFDM WUS waveform that is compatible with 3GPP NR transmitters and reuses existing legacy components of 3GPP signals. Another purpose is to provide a solution that is less complex than traditional solutions.

[0105] Therefore, embodiments of the present invention disclose the use of a bit spreading sequence to control the shape of a signal waveform and / or its spectrum. Applications include, but are not limited to, wireless user space (WUS) transmission in 3GPP NR. Embodiments of the present invention also disclose spreading sequences that make the envelope of the signal's ON and OFF states very flat, thereby providing robustness against detection errors due to noise and fading when a low-precision ADC envelope detector is used at the receiving device.

[0106] therefore, Figure 1 FIG1 shows a sending device 100 provided by an embodiment of the present invention. Figure 1In the illustrated embodiment, the transmitting device 100 includes a processor 102, a transceiver 104, and a memory 106. The processor 102 is coupled to the transceiver 104 and the memory 106 via a communication module 108 as is known in the art. The transmitting device 100 can be used for wireless and / or wired communication in a communication system. Wireless communication capabilities can be provided using an antenna or antenna array 110 coupled to the transceiver 104, while wired communication capabilities can be provided using, for example, a wired communication interface 112 coupled to the transceiver 104.

[0107] The processor 102 may be referred to as one or more general-purpose central processing units (CPUs), one or more digital signal processors (DSPs), one or more application-specific integrated circuits (ASICs), one or more field programmable gate arrays (FPGAs), one or more programmable logic devices, one or more discrete gates, one or more transistor logic devices, one or more discrete hardware components, or one or more chipsets. The memory 106 may be a read-only memory, a random access memory (RAM), or a non-volatile RAM (NVRAM). The transceiver 304 may be a transceiver circuit, a power controller, or an interface that provides the ability to communicate with other communication modules or communication devices (such as network nodes and network servers). The transceiver 104, the memory 106, and / or the processor 102 may be implemented in separate chipsets or in a common chipset. In the present invention, the transmitting device 100 being used to perform certain actions may be understood to mean that the transmitting device 100 includes appropriate modules for performing the actions, such as the processor 102 and the transceiver 104.

[0108] According to an embodiment of the present invention, the sending device 100 is configured to bit Each bit in the sequence of bits is bit Multiply the corresponding spreading sequences in the sequence of N spreading sequences. bit The sequence of bits is spread to obtain N symb modulation symbols, where N bit Each of the spreading sequences in the sequence of spreading sequences is a linear phase sequence with a constant rotation phase angle Φ. The transmitting device 100 is further configured to transmit N symb The modulation symbols are multiplied by the discrete Fourier transform precoder to obtain N symbThe transmitting device 100 is further configured to transmit an orthogonal frequency-division multiplexing (OFDM) signal 510, wherein the OFDM signal includes N signals mapped onto K OFDM subcarriers. symb Fourier coefficients.

[0109] In addition, in the embodiment of the present invention, the transmitting device 100 for the communication system 500 includes a processor, wherein the processor is configured to: bit Each bit in the sequence of bits is bit Multiply the corresponding spreading sequences in the sequence of N spreading sequences. bit The sequence of bits is spread to obtain N symb modulation symbols, where N bit Each spreading sequence in the sequence of spreading sequences is a linear phase sequence with a constant rotation phase angle Φ; symb The modulation symbols are multiplied by the discrete Fourier transform precoder to obtain N symb The transmitting device 100 further comprises a transceiver configured to transmit an orthogonal frequency-division multiplexing (OFDM) signal 510, wherein the OFDM signal comprises N signals mapped onto K OFDM subcarriers. symb Fourier coefficients.

[0110] Furthermore, in yet another embodiment of the present invention, the transmitting device 100 for the communication system 500 includes a processor and a memory, wherein the memory has computer-readable instructions stored therein, and when the computer-readable instructions are executed by the processor, the processor performs the following operations: bit Each bit in the sequence of bits is bit Multiply the corresponding spreading sequences in the sequence of N spreading sequences. bit The sequence of bits is spread to obtain N symb modulation symbols, where N bit Each spreading sequence in the sequence of spreading sequences is a linear phase sequence with a constant rotation phase angle Φ; symb The modulation symbols are multiplied by the discrete Fourier transform precoder to obtain N symb Fourier coefficients; sending an orthogonal frequency-division multiplexing (OFDM) signal 510, the OFDM signal comprising N mapped onto K OFDM subcarriers symb Fourier coefficients.

[0111] Figure 2 It is shown that the sending device 100 (such as Figure 1 The method 200 includes: bit Each bit in the sequence of bits is bit Multiply the corresponding spreading sequences in the sequence of N spreading sequences. bit The sequence of bits is spread (202) to obtain N symb modulation symbols, where N bit Each of the spreading sequences in the sequence of spreading sequences is a linear phase sequence with a constant rotation phase angle Φ. The method 200 includes: symb The modulation symbols are multiplied (204) by the discrete Fourier transform precoder to obtain N symb Fourier coefficients. The method 200 includes transmitting (206) an OFDM signal 510, the OFDM signal including N mapped onto K OFDM subcarriers. symb Fourier coefficients.

[0112] Figure 3 FIG. 3 shows a receiving device 300 provided by an embodiment of the present invention. Figure 3 In the illustrated embodiment, receiving device 300 includes a processor 302, a transceiver 304, and a memory 306. Processor 302 is coupled to transceiver 304 and memory 306 via a communication module 308 as is known in the art. Receiving device 300 also includes an antenna or antenna array 310 coupled to transceiver 304, indicating that the receiving device is configured for wireless communication within a wireless communication system.

[0113] The processor 302 may be one or more general-purpose CPUs, one or more DSPs, one or more ASICs, one or more FPGAs, one or more programmable logic devices, one or more discrete gates, one or more transistor logic devices, one or more discrete hardware components, or one or more chipsets. The memory 306 may be a read-only memory, RAM, or NVRAM. The transceiver 104 may be a transceiver circuit, a power controller, or an interface that provides the ability to communicate with other communication modules or communication devices. The transceiver 304, memory 306, and / or processor 302 may be implemented in separate chipsets or in a common chipset. In the present invention, the receiving device 300 being configured to perform certain actions may be understood to mean that the receiving device 300 includes appropriate modules for performing the actions, such as the processor 302 and the transceiver 304.

[0114] According to an embodiment of the present invention, receiving device 300 is configured to receive an OFDM signal 510 transmitted by transmitting device 100. OFDM signal 510 becomes an OOK signal due to bit spreading according to an embodiment of the present invention. Therefore, receiving device 300 decodes the bits of the OOK signal by non-coherently detecting the envelope fluctuations of the OOK signal. A typical low-power wake-up receiver architecture for OOK signal detection first processes received signal 510 in the analog domain via low-pass filtering for interference suppression and noise reduction, followed by direct envelope detection. This processed signal is then sampled and converted to the digital domain, followed by bit detection. If the detected bit string corresponds to a specific bit string implemented in receiving device 300, then if OFDM signal 510 is a WUS, the receiving unit of receiving device 300 configured for WUS detection triggers the wake-up of other radio units of receiving device 300.

[0115] Figure 4 A communication system 500, such as a 3GPP NR, provided in an embodiment of the present invention is shown. The communication system 500 in the disclosed embodiment includes a transmitting device 100 and a receiving device 300 for communicating and operating in the communication system 500. In a non-limiting example, the transmitting device 100 may be part of a network access node (e.g., a base station), and the receiving device 300 may be part of a client device (e.g., a user equipment terminal). The network access node may be connected to a core network of the communication system via a communication interface.

[0116] Therefore, the network access node and the client device are used to communicate in downlink (DL) and uplink (UL), which means that the network access node can send an OFDM signal 510, which includes N mapped onto K OFDM subcarriers generated according to an embodiment of the present invention. symb Fourier coefficients. Other details related to the embodiments of the present invention will be described in the context of 3GPP 5G NR. Therefore, 3GPP 5G terminology, definitions, expressions, and system architecture will be used. However, it should be noted that the embodiments of the present invention are not limited thereto.

[0117] In general, the embodiments of the present invention can be considered to be based on the inherent time domain multiplexing properties of DFT-precoded OFDM, similar to the DFT-s-OFDM that has been standardized in NR. For WUS applications, the size N of the DFT-precoder is symb ≤K, i.e. not more than the number of subcarriers in the WUS bandwidth allocation, but in applications other than WUS, the DFT-precoder may have other sizes. Before DFT precoding, each bit is spread and mapped to a modulation symbol sequence.

[0118] Figure 5 FIG2 shows a block diagram of a transmitting device 100 integrated into a processing chain of a general communication device according to an embodiment of the present invention. A serial to parallel (S / P) block 130 is connected to an input terminal of the transmitting device 100. The bit string is converted into N bits in the S / P block 130. bit parallel bits. Bit string b[l](l=0,…,N bit -1, where b[l]∈{0,1}) is sent to the transmitting device 100 as input after being converted into parallel bits.

[0119] The parallel bits are provided to the spreader block 132, where the number of bits N bit are multiplexed together so that each bit b[l] is seg The spreading sequence r l [n] multiplied to generate a total of N symb =N bit N seg N modulation symbols seg The corresponding modulation symbols. The bit spreading factor N seg To obtain N symb =N bit N seg The sequence of modulation symbols is shown below.

[0120] Each individual bit b[l](l=0,…,N bit -1) and the spreading sequence r l [n](n=0,…,N seg -1) are multiplied so that the modulation symbol is given by

[0121]

[0122] Where m = 0,…,N symb -1 and It should be noted that each individual spreading sequence r l [n] depends on the bit index l. This is important because consecutive symbols send different bits and these bits will affect each other.

[0123] In an embodiment of the present invention, a separate spreading sequence r l [n] can be cascaded into a cascaded spreading sequence r[m], m = 0,…,N symb -1 is shown below:

[0124]

[0125] Therefore, equation (7) can be written as

[0126] d[m]=b[lm ]r[m]. (9)

[0127] Therefore, the modulation symbols can be obtained by spreading the bits by simple repetition to obtain a modulated symbol of length N. symb =N bit N seg Repeating bit string Then repeat the bit string b r [m] is multiplied element by element with the concatenated spreading sequence r[m]. Therefore, in the embodiment of the present invention, the transmitting device 100 is configured to transmit the N bit bits to spread N bit bits to get N symb Thereafter, the transmitting device 100 transmits N symb The repeated bits are multiplied by the concatenated spreading sequence to obtain N symb modulation symbols. The concatenated spreading sequence r[m] will be N bit The concatenation of the spreading sequences r[m] makes the concatenated spreading sequence r[m] a linear phase sequence with a constant rotation phase angle Φ, as described above. In the following disclosure, it is shown in more detail how to construct the concatenated spreading sequence r[m] and thus the individual spreading sequence r[m]. l [n] in order to control the shape of the OFDM signal 510 as an OOK signal and its spectrum.

[0128] The modulation symbols output from the spreader block 132 are provided to Figure 5 The modulation symbols are DFT precoded in the DFT precoding block 134 to provide a sequence of Fourier coefficients.

[0129]

[0130] Before mapping the Fourier coefficients to the K subcarriers, the output of the DFT block 134 may be fed to a signal processing block 136 which spreads, shapes, and shifts the phase of the Fourier coefficients.

[0131] Before mapping the Fourier coefficients to K OFDM subcarriers, the Fourier coefficients of the DFT precoder 134 can be provided to an optional signal processing block 136, where the Fourier coefficients are expanded by spectrum extension (SE) to meet the subcarrier allocation K. Frequency-domain spectral shaping (FDSS) window and phase shift can also be considered to achieve further shaping of the OFDM signal to improve performance. This structure is much less complex to implement than traditional solutions because the DFT precoder size is at most equal to the number of subcarriers in the WUS bandwidth allocation, which is usually much smaller than the IFFT size of OFDM.

[0132] SE

[0133] DFT precoder size N symb It is necessary to make the spreading factor N seg =N symb / N bit To achieve this, the discrete Fourier transform precoder size is N symb ≤K, rather than making N symb = K. In one example, based on N symb The periodic repetition of Fourier coefficients, N symb Fourier coefficients are expanded into K Fourier coefficients.

[0134] Given a subcarrier allocation of K subcarriers, choose the size N of the DFT precoder symb , so that N symb ≤K and N symb / N bit is an integer, e.g., the largest possible value, and if necessary, K subcarriers are padded by SE as follows:

[0135] D (se) [k]=D[k+L (mod N symb )], for k=0,…,K-1. (11)

[0136] Where L is an integer shift. Typically, N symb =K, without SE and D (se) [k] = D[k]. Otherwise, a related case is to choose N symb , so that N e =(KN symb ) is an even number, then choose L=-N e / 2(modN symb), since this shift has the benefit of creating a symmetrical spectrum. This can be written as:

[0137]

[0138] SE also has the ability to modify the number of resulting DFT-s-OFDM time-multiplexed pulses that make up the OFDM signal. Fewer pulses mean they are wider, so it provides an additional degree of freedom for modifying the overall signal shaping. In traditional schemes, SE is used as a way to reduce the peak-to-average power ratio (PAPR) at the expense of breaking the orthogonality between pulses, thereby increasing inter-pulse interference. The motivation for using SE in this scheme is different: it is used to control the spreading factor. Furthermore, to create an OOK signal, there is no benefit in maintaining orthogonality between pulses.

[0139] FDSS

[0140] After SE, FDSS can be applied to the Fourier coefficients as follows:

[0141] X′[k]=W[k] D ( se)[k] (13)

[0142] Where {W[0],…,W[K-1]} is the FDSS window coefficient. FDSS can further shape the OOK waveform. Therefore, N symb The Fourier coefficients or K Fourier coefficients are multiplied by the FDSS window coefficients to obtain frequency shaping Fourier coefficients.

[0143] A related embodiment of the FDSS window is a low PAPR window, which is usually real and symmetric, and whose coefficients are derived from Bell-shaped functions. This window further mitigates the fluctuations of the signal envelope, thereby flattening the OOK state.

[0144] In an embodiment of the present invention, due to its convenient parameterization, the FDSS window coefficients are Kaiser window coefficients with a shaping parameter β. This window coefficient has also been shown to effectively focus the DFT-s-OFDM energy in the time domain (TD), which is relevant for OOK signal design. The shaping parameter can be equal to 2, i.e., β = 2. It should be noted that the case β = 0 gives a rectangular window and is therefore equivalent to no FDSS at all.

[0145] There may be other types of FDSS windows, such as a truncated root-raised-cosine (RRC) filter with parameters (0.5,–0.65) or (0.5, 0.1667); a 2-tap filter with coefficients [1–0.28], for example; a 3-tap filter with coefficients [–0.335 1–0.335] or [–0.28 1–0.28]. Therefore, in an embodiment, the FDSS window coefficients are given by:

[0146]

[0147] Among them, N fft is the number of samples of the OFDM signal 510, and sin() is the sine function.

[0148] FD phase shift

[0149] A frequency-domain (FD) phase shift can also be applied to the Fourier coefficients to further shape the signal. Thus, the frequency-shaped Fourier coefficients can be multiplied by the FD phase shift to obtain the phase-shifted Fourier coefficients, where the FD phase shift is based on the shift parameter T shift .

[0150] In the example, an FD phase shift can be applied to the Fourier coefficients as follows:

[0151]

[0152] In order to calculate WUS s in equation (6) W [n] Create TD cyclic shift. This step is used to cyclically shift the OOK signal in order to improve the time position of the OOK state by maximizing the energy of the OOK state in its target time domain period. As discussed in Appendix A, WUS s W [n] is equivalent to the TD pulse multiplexing in equation (31). Without the shift in equation (14), the first pulse g0[n] carrying the first modulation symbol d[0] has a peak at time index 0, and its energy is evenly distributed in a cycle between the start and the end of the OFDM symbol. This operation shifts all pulses so that the energy of the first pulse is mainly at the beginning of the OFDM symbol. However, since only the WUS signal should be shifted and not the concurrent data, an FD implementation of this TD cyclic shift may be required. In the absence of other multiplexed data, neither FDSS nor SE, this operation can be implemented by TD cyclic shift before CP addition. This means that the shift parameter T shift The value of depends on the number of samples N of the OFDM signal 510 fft and the number of modulation symbols Nsymb .

[0153] A related embodiment corresponds to cyclically shifting the pulses by half the time difference between two consecutive pulses, i.e. approximately Therefore, the first and last samples of the OFDM signal are located between the first and last pulses. Small changes (such as ) and integer approximation (such as or ) also provides a similar effect. Therefore, the shift parameter T shift The value of can be given by any of the following formulas:

[0154]

[0155] Among them, N fft is the number of samples of the OFDM signal 510, is the ceiling function, is the floor function, and round[] is the rounding function. For example, T shift The larger value is considered In order to further reduce the energy leakage of the first time multiplexed pulse to the end of the OFDM symbol. In order to avoid the last time multiplexed pulse to have its energy leak at the beginning of the OFDM symbol, the last pulse can be used as a guard pulse by systematically setting the corresponding last input of the DFT precoder to zero.

[0156] By combining equations (11), (13) and (14), the signal processing block 136 converts the output of the DFT block 134 (i.e., the Fourier coefficients D[k], k = 0, ..., N symb , as shown in equation (10)) is transformed into the following WUS Fourier coefficients:

[0157] For k=0,…,K.

[0158] In addition, a mapper block 138 maps the WUS Fourier coefficients from the signal processing block 136 to the K allocated subcarriers of the WUS. Other data formulated in equation (2) (e.g., other WUS for other receiving devices or other types of data) can be input to the mapper block 138 so as to be frequency multiplexed together within the same OFDM symbol. The output of the mapper block 138 is fed to the OFDM IFFT block 140 to generate a time-domain OFDM symbol. Finally, the CP block 142 adds a cyclic prefix to the OFDM signal before transmission in the communication system.

[0159] Manchester Code

[0160] OOK modulation is usually performed after optional Manchester encoding of the information bits. The Manchester encoding scheme is shown in Table 2.

[0161] Table 2: Manchester-encoded OOK

[0162] Information bits Coded bits state 1 1 0 ON OFF 0 0 1 OFF ON

[0163] Manchester encoding creates a bit string with a constant average value of 1 / 2. The advantage of this is that the envelope of the modulated OOK signal will have a constant DC component that does not carry any information. Therefore, by estimating this DC component, the optimal threshold for detection can be found. Ideally, the DC level of the envelope can be estimated and subtracted so that the decision boundary of WUS is zero. However, due to fading in wireless transmission, this threshold selection usually does not perform well because ambiguous decoding states such as '0 0' or '1 1' may appear. A better way to utilize the principle of Manchester encoding is to compare the amplitude of the first signal state with the amplitude of the second signal state to obtain the information bit. It should be noted that Manchester-coded OOK is a form of pulse-position modulation (PPM), and the scheme disclosed in this article can be directly applied to the transmission of any OFDM-based pulse position modulation.

[0164] Therefore, in the embodiment of the present invention, the bit string to be spread is of length N bit0 Manchester-encoded version of the original bit string, such that N bit =2N bit0 Or equivalently, N bit bits are based on N bit The Manchester coded bits of a sequence of / 2 bits. Then (before encoding) the number of modulation symbols per information bit is 2N seg , used to create ON and OFF states for each bit. symb =K, the number of modulation symbols in each OOK state, that is, the spreading factor may not be an integer. However, using Manchester coding, the number of modulation symbols per information bit is May be an integer, even if N seg Therefore, we can consider assigning two different spreading factors to the "0" and "1" bits, namely N segOFF and N segON , thus maintaining N segOFF +N segON =2N seg and N symb =N bit0 (N segOFF +N segON), instead of using SE to obtain N seg In Appendix D, it is shown that using SE provides better performance, however, the disclosed scheme can be directly generalized to the case of two different spreading factors for the ON and OFF states.

[0165] As mentioned above, the spreading sequence r[m] used in this paper will be able to control the signal shape and spectrum. To this end, a constant envelope sequence is used as the concatenated spreading sequence r[m] according to the following expression:

[0166]

[0167] Then, the modulated symbol becomes

[0168]

[0169] Equivalently, each individual spreading sequence r in equation (7) l [n] will be equal to

[0170]

[0171] Therefore, the individual spreading sequence r l [m] is given by the formula r l [m]=e jΦm Given, where l is the bit index, m is the modulation symbol index, e is the natural exponential function, and j is the imaginary unit.

[0172] Furthermore, the phase of the concatenated spreading sequence r[m] can be constrained to follow a linear phase with a rotation angle Φ, i.e.,

[0173] φ m =Φm+Φ0, for m=0,…,N symb -1 (18)

[0174] where Φ and Φ0 are constant angles. The constant Φ0 only affects the global phase of the signal and may be irrelevant from the perspective of a receiver with a non-coherent detector, so Φ0 = 0 is considered. Therefore, each individual spreading sequence in equation (17) becomes a linear phase sequence, i.e., a sequence with a rotation angle Φ. Among them, Φ l =ΦlN seg +Φ0 is a constant angle that depends on the bit index l but is independent of the modulation symbol index m.

[0175] In another alternative, a low-complexity embodiment would select the same spreading sequence for each bit, which corresponds to shifting the constant angle Φ l Set to zero, resulting in a linear phase sequence r that is independent of the bit index l[n]=e jΦn While this scheme generally preserves most of the benefits, it is observed from simulations that it is not optimal. It should be noted that this enables the coherent combining of TD multiplexed pulses within each OOK state, rather than between consecutive OOK states.

[0176] Using equation (18) (Φ0=0), the modulation symbol of equation (16) becomes

[0177]

[0178] In Appendix A, it is shown that the optimal constant rotation phase angle Φ that provides a flat ON / OFF state is

[0179] Φ=π. (20)

[0180] This is equivalent to bit spreading by alternating sequences of +1 and -1. Explicitly, equation (14) simplifies to r[m] = (-1) m Therefore, in this case, the individual spreading sequence r l [m] is an alternating sequence of values +1 and –1. It is also equivalent to equation (19) that, as in equation (16), Each bit is spread individually, that is, the same spreading sequence is used for each bit Until the sign changes, such as

[0181] The alternating sequence of +1 and –1 can be interpreted as an alternating sequence of two binary phase-shift keying (BPSK) constellation symbols. It should be noted that in the 3GPP standard, the BPSK constellation is specified as That is, it is a constellation {+1, -1} rotated by π / 4. Therefore, it may be necessary to choose Φ0 = π / 4 in order to convert the sequence of +1 and -1 discussed above into a spreading sequence of two alternating BPSK symbols specified in 3GPP, for example

[0182] These embodiments offer very low implementation complexity and follow the principle of minimizing the phase difference between the overlapping lobes of two adjacent pulses, as explained in Appendix A. The analysis in Appendix A is approximate only in the sense that it considers only two adjacent pulses, while other adjacent pulses also contribute to the fluctuations of the state envelope. The analysis becomes more relevant when shaping is added from the FDSS window, as the other side lobes are increasingly attenuated. Figure 6 and Figure 7The signal shapes obtained depending on the angle used in the linear phase ramp are shown. Without FDSS, several angles (such as Φ = π / 3, π / 2, or π) provide similar fluctuations in the ON state, but Φ = π provides slightly less energy leakage in the OFF state. With FDSS, both the ON and OFF states become more constant as Φ approaches π.

[0183] More precisely, it is shown in Appendix A that the phase difference between the overlapping lobes of two adjacent pulses can be minimized by choosing the phase slope of the concatenated spreading sequence r[m] as follows:

[0184]

[0185] In practice, this equation is well approximated by the value π. It should be noted that in the absence of SE, N symb =K and L=0, and

[0186]

[0187] Numerical evaluation using Φ = π instead of equations (21) and (22) provides almost imperceptible differences in waveform shape, but still corresponds to different Fourier coefficient values, such as Figure 8 As shown. However, it can be observed that the Fourier coefficients often have some null values when choosing Φ = π / M, where M is an integer related to the number of bits per OFDM symbol and the number of symbols modulating ON and OFF states. If this is not ideal, it can be easily circumvented by using small deviations from these angles, as in equations (21) and (22), which do not affect the waveform shape compared to Φ = π, as shown Figure 8 shown.

[0188] Zeroing subcarriers

[0189] Another interesting embodiment is to select the linear phase of the concatenated spreading sequence r[m] to set certain subcarriers to zero. For example, in the design of Wifi WUS, the direct current (DC) subcarrier is selected to be 0 to prevent it from being filtered out by the WUR circuit. In Appendix B, it is shown that if N seg >1, then the constant rotation phase angle can be chosen to be equal to the following formula null ∈{0,…,N symb The output of the DFT precoding of equation (10) with respect to the value of -1} can be set to zero, that is, D[k null ]=0:

[0190]

[0191] Where λ is any non-zero integer, for example, λ = 1 or λ = -1. For the DC subcarrier, the middle index can be selected

[0192] exist Figure 9 In the paper, it is verified that when K=N symb =72, N bit0 = 4 and consider setting index k to zero null The subcarrier zeroing effect is obtained when N = 36. seg =9, resulting in a rotation phase angle from Figure 9 As can be seen in Figure 2, the required middle subcarrier has been eliminated, while the OOK waveform shape is still very good, since Φ≈0.78π, which is quite close to π. It should be noted that in order to obtain a symmetrical distribution of average power among the subcarriers, the phase ramp is chosen to have a random sign between different transmissions, i.e., +Φ or –Φ.

[0193] The disclosed scheme allows a carefully selected set of parameters to produce Fourier coefficients for generating OFDM signals, thereby providing a least square (LS) approximation (also denoted as the LS method) of the ideal OOK signal under a given bandwidth allocation constraint. In traditional schemes, providing LS approximations of Fourier coefficients is a high-complexity approach because it requires the introduction of a second DFT / FFT of the same size as the OFDM modulation. Even if only K FFT outputs are required, only limited complexity reduction can be achieved over the full FFT by using a pruned FFT algorithm. In fact, for K < 100,0 ... <N fft outputs, the performance gain of pruned FFT is usually quite limited, with an order of O(N fft log2K), rather than O(N fft log2N fft ).

[0194] The disclosed scheme can generate the same minimum LS approximation signal, but the implementation complexity is much lower than the traditional LS scheme. fft / N bit and N seg =N symb / N bit are integers, Φ=π and Then use size N fftThe Fourier coefficients of the direct LS approximation of the DFT precoder and the disclosed scheme are derived in closed form and shown in Appendix C to differ only in the amplitude coefficient and the global phase factor, both of which are independent of the bit data. Therefore, using the appropriate FDSS window W0[k] specified in equation (59), the two methods can produce the same Fourier coefficients, up to the uncorrelated global phase factor. As a direct byproduct, if we consider the combination of the FDSS window W LS A more complex LS method for [k] uses the proposed scheme but uses the FDSS window W[k]=W LS [k]W0[k] can also obtain the same Fourier coefficients.

[0195] Compared to the Kaiser window discussed previously, it can be observed that the expression W0[k] can be adequately approximated by a Kaiser window with shaping parameter β≈ 2. Similarly, if the more complex LS method is combined with a Kaiser window with shaping parameter β LS The disclosed scheme will be used together with the FDSS Kaiser window with approximately The Kaiser window with a larger shaping parameter is a sufficient approximation.

[0196] Compared to the naive approach for LS approximation, the benefit of the disclosed scheme is that the same OOK signal can be obtained with much lower implementation complexity, since the complexity of the two approaches is determined by the size of their corresponding DFT precoders. This is illustrated in Table 3 with two numerical examples, where it can be seen that the complexity can be reduced by 2 to 3 orders of magnitude. This is because the FFT size in OFDM modulation is usually large, while the subcarrier allocation of WUS signals is usually considered small. In addition, here, it is assumed that the naive LS method is implemented with an order of N. fft log2K optimized pruned FFT algorithm, otherwise this may require an order of N fft log2N fft In addition, the FFT size considered here is N fft = 2048, the same as in LTE, but the reference FFT size in the NR specification is twice as high: N fft =4096, which will only double the complexity of the LS method.

[0197] Table 3: Complexity comparison

[0198] Complexity order <![CDATA[K=N symb =72]]> <![CDATA[K=24,N symb =16]]> Naive LS method <![CDATA[N fft log2K]]> 100％ 100％ The disclosed solution <![CDATA[N symb log2N symb ]]> 3.5％ 0.68％

[0199] The disclosed scheme targets good bit error rate (BER) performance with low power consumption. BER evaluation confirms that the embodiment with a spreading sequence r[m] rotated by a phase angle Φ=π provides the best performance, while the impact of the FDSS window is less significant. In the case of a receiver with a DC blocker, it is shown that the embodiment with a corresponding zeroed DC subcarrier maintains good performance.

[0200] BER is calculated as a function of the WUS signal-to-noise ratio (SNR), i.e., the WUS component s of the transmitted signal s[n] W [n] divided by the total noise power. Consider a very simple and low power receiver where a 0.15 BER is considered sufficient for WUS. Further assume that the OFDM transmitter uses N fft = 2048, and a total of 600 modulated subcarriers and 15kHz subcarrier spacing to transmit N bit0 =4 Manchester-coded signal. Assume that the WUS signal consists of K = 72 subcarriers transmitted in the center of the band, while the other subcarriers on either side are modulated with random BPSK symbols. The signal reaches the receiver via a multi-tap wireless channel. Assume the time domain line C (TDL-C) channel model with Rayleigh fading as specified in 3GPP, an expected delay spread of 100 ns, and a speed of 3 km / h.

[0201] The received analog signal first passes through a bandpass filter (BPF) centered on the WUS signal band to eliminate inter-channel interference; then enters an envelope detector to smooth the signal. The envelope detector consists of a norm operator followed by a low-pass filter. Assume that the bandpass filter (BPF) and the low-pass filter (LPF) are third-order Butterworth filters with cutoff bandwidths of coefficients α and α, respectively. BPF and α LPF The signal is then passed through the ADC, after which the bits are decoded. The ADC is considered to have low precision and operates at a minimum sampling rate of one sample per OOK state and 2-bit amplitude quantization. The ADC's sampling goal is to be in the middle of the OOK state at a given reference time, which was previously acquired by synchronization with the aid of a preamble or by blind synchronization (e.g., based on CP redundancy). It is assumed that there are random synchronization errors uniformly distributed over the symmetrical sample interval [-S e ,S e ], the maximum error is equal to S eSince a Manchester-coded signal is assumed, detection is performed by direct amplitude comparison of two consecutive samples.

[0202] exist Figure 10 In Figure 1, the BER at a fixed SNR is shown as a function of the phase ramp angle Φ and different FDSS coefficients β. Here, it is assumed that the bandwidth scaling factor of the LPF matches the WUS bandwidth, i.e., α LPF =1, and consider two values of BPF: α BPF = 2 and 1, for which –2dB and 0dB SNR are selected respectively. Two ranges of synchronization error are selected, samples, covering the entire segment of the OOK state, half of which is S e = 64. It can be observed that, in general, BER decreases as Φ increases until Φ = π. When the synchronization error is less than one OOK state, FDSS shaping β is has a slightly positive effect because FDSS concentrates more energy in the middle range of the state. When the synchronization error is as large as that of the OOK state, the FDSS shaping β is There is always a greater negative impact on There is also a smaller negative impact, as the FDSS window attenuates the edges of the states.

[0203] Figure 11 The BER as a function of SNR is compared between the disclosed scheme and the conventional scheme. Φ = π is selected, where for S e =128, β=0; for S e =64, β=5. Figure 11 It can be seen that only when S e = 64, i.e., when the synchronization error is not too large, the BER of the proposed scheme can be slightly better than that of the LS approximation (referred to as the LS method in the figure) by using more FDSS shaping. This improvement is small, and in general, by using Φ = π and β = 2, the disclosed scheme provides the same BER performance as the LS method, as explained above. Nevertheless, it should be remembered that the complexity of the disclosed scheme is much lower than that of the LS method, while providing more degrees of freedom for optimization.

[0204] For further comparison, the disclosed scheme is shown to provide a significant improvement over a similar but naive scheme, in which bits are spread by mapping them to random symbols of a BPSK or π / 2-BPSK constellation before feeding DFT-s-OFDM modulation. Note that π / 2-BPSK incorporates the signal by constructing a linear phase ramp with an angle of π / 2 between consecutive symbols. Using FDSS with π / 2-BPSK further improves its performance, where the optimal shaping is found to be β = 3. The BER for all curves can be improved by narrowing the bandwidth of the BPF or LPF. However, for lower power consumption, a larger filter bandwidth may be required.

[0205] Finally, consider an embodiment in which the WUR blocks the DC component of the received signal in its analog front end. The BER performance is evaluated as described above, where the parameter K = N symb =72, N bit0 =4, N seg =9, the DC subcarrier is located at index k null =36, so that according to equation (23), the phase slope angle is from Figure 12 (a) is foreseeable and Figure 12 (b) verifies that the angular variation of the phase ramp still provides good performance. This is compared to the LS approximation, where the spectrum is uncontrollable and most of its energy is on the DC subcarrier, where it is filtered out by the receiver.

[0206] An FD phase shift has been considered, which corresponds to cyclically shifting the main lobe of the TD multiplexed pulse by half its period, i.e. As explained in Appendix A, this is because the energy of the first pulse is centered at time zero, so the time position of the OOK waveform is systematically offset from the expected position. An alternative to using CP is to apply T shift This time delay of the samples, i.e. the reference time of the OFDM signal is selected as -T shift This shift is more important when the number of pulses is small because the pulse lobes are larger and thus the energy of the pulse leaks more into the adjacent OOK states.

[0207] Figure 13 The positioning correction obtained on a 2-bit OOK signal is shown, where the ON and OFF states are expected to span each half of the OFDM symbol duration. The signal is obtained using K=24 subcarriers. symb =16 pulses, where N e = 8 for spectrum spreading. Using a linear phase ramp of Φ = π, the Fourier coefficients are shaped by an FDSS window with parameter β = 4. Figure 13As can be seen in , the OFF state has its envelope to reach an amplitude close to 1.

[0208] Correcting this time offset can improve robustness to synchronization errors. Figure 14 Figure 15(b) shows how this FD phase shift helps reduce decoding errors in the case of large synchronization errors. Here, the signal is downsampled at a minimum rate of one sample per state, with downsampling starting at the 220th sample in the middle of the waveform instead of the 128th sample. The 8-bit string considered corresponds to the Manchester encoding of the bits [1 0 1 1]. Therefore, the bits can be decoded by comparing the amplitude (or energy) of two consecutive samples, as shown in Figure 15(b). It can be seen that considering T shift The signal with time offset correction decodes the bits correctly as [10 1 1], while the signal without time offset compensation outputs two decoded bit errors, the string of [1 0 0 0].

[0209] The BER performance difference between using and not using FD phase shift is as follows: Figure 15 Here, the synchronization error range is chosen to cover a complete OOK state, i.e., S e = 128, and the OOK waveform is generated using Φ = π and β = 2. Two cases of bandwidth allocation are considered, namely K = 24 and K = 72, where it can be verified that the performance gap is larger under smaller bandwidth allocation because the pulse lobe is larger in this case and more energy leaks into other OOK states without the proper TD shift achieved by the proposed FD phase rotation.

[0210] An alternative to FD phase shifting that can be considered is the use of guard symbols, where some symbols at the input of the DFT precoder are systematically set to zero. Guard symbols can also be used between different states to avoid energy leakage between the ON and OFF states. However, the use of guard symbols is generally not optimal because it reduces the width of the ON state, making the signal more sensitive to synchronization errors. It should be noted that at the input of OFDM modulation, guard symbols are different from guard subcarriers. Guard subcarriers are also beneficial for the disclosed scheme to reduce interference from concurrent data transmission.

[0211] For independent WUS transmission, i.e., without other concurrent data transmission, the PAPR performance of the proposed scheme is also considered. bit0 The maximum PAPR as a function of the rotation phase angle Φ is shown in the following table for four different combinations of Figure 16 As shown. bit0 =8, use N e = 8 spectrum expansion. Mainly in the range A good PAPR of about 4dB can be obtained within 100mA, which is consistent with the perspective of providing good BER performance. Figure 16 In this angle range for all cases except case (b), increasing the FDSS shaping β improves the PAPR. In Figure 20(b), for π / 2≤Φ≤π, increasing the FDDS shaping β improves the PAPR to β=3. In addition, it can be found that in this case (b), for each curve with fixed β, there is a clear optimal result at Φ=π / 2. In addition, numerical evaluation shows that when N seg = 1, that is, when there is no bit spreading and there is only one pulse per bit, Φ = π / 2 provides the best PAPR. As the bit spreading increases, the PAPR variation in the range π / 2 ≤ Φ ≤ π is flat. When the bit spreading is not so large, such as in N seg = 4, we can see that the best PAPR for a given β is in the angle between π / 2 and π. Therefore, the angle Φ = 3π / 4 can provide a good balance to cover several scenarios.

[0212] It can be noted that the special case of the Fourier coefficients in equation (58) can be implemented as DFT precoding of bits without explicit spreading as follows:

[0213]

[0214] Among them, only N bit is DFT precoded as:

[0215]

[0216] Then repeat it by symmetrical spectrum expansion

[0217]

[0218] Then, apply FDSS to it as follows:

[0219] X′[k]=W″[k]D ( se)[k] (29)

[0220] Among them, the FDSS window is

[0221]

[0222] You can use the aforementioned FD shift.

[0223] As mentioned above, the transmitting device 100 disclosed herein may be any type of suitable communication device. Non-limiting examples are network access nodes and client devices.

[0224] A network access node herein may also be denoted as a wireless network access node, an access network access node, an access point (AP) or a base station (BS) (e.g., a radio base station (RBS)). In some networks, the base station may be referred to as a transmitter, "gNB," "gNodeB," "eNB," "eNodeB," "NodeB," or "B-node," depending on the standards, technologies, and terminology used. Wireless network access nodes may be of different categories or types, such as macro eNodeB, home eNodeB, or mini base station, based on transmission power and cell size. A wireless network access node may also be a station, which is any device that includes a media access control (MAC) and physical layer (PHY) interface to a wireless medium (WM) that complies with IEEE 802.11. The wireless network access node can be used to communicate in the following: 3GPP-related long term evolution (LTE), advanced LTE, fifth generation (5G) wireless systems, such as new radio (NR) and its evolutions, and IEEE-related Wi-Fi, worldwide interoperability for microwave access (WiMAX) and its evolutions.

[0225] The client device herein may be represented as a user device, user equipment (UE), mobile station, internet of things (IoT) device, sensor device, wireless terminal and / or mobile terminal, capable of wireless communication in a wireless communication system (sometimes also referred to as a cellular wireless system). UE may also be referred to as a mobile phone, cellular phone, computer tablet or laptop with wireless capabilities. For example, the UE in this context may be a portable, pocket storage, handheld, computer-based or vehicle-mounted mobile device that can communicate voice and / or data with another communication entity (such as another receiver or server) over a radio access network (RAN). UE may also be a station, which is any device that includes an IEEE 802.11-compliant MAC and PHY interface to a WM. UE may be used to communicate in 3GPP-related LTE, Advanced LTE, 5G wireless systems (such as NR) and their evolutions, as well as in IEEE-related Wi-Fi, WiMAX and their evolutions.

[0226] Furthermore, any method provided by the embodiments of the present invention can be implemented in a computer program having a code module, which, when executed by a processing module, causes the processing module to perform the steps of the method. The computer program is included in a computer-readable medium of a computer program product. The computer-readable medium can include essentially any memory, such as the aforementioned ROM, PROM, EPROM, flash memory, EEPROM, or hard drive.

[0227] Furthermore, it should be appreciated that the transmitting device 100 includes necessary communication capabilities in the form of, for example, functions, modules, units, elements, etc., for performing or implementing embodiments of the present invention. Examples of other such modules, units, elements, and functions are: a processor, a memory, a buffer, a control logic, an encoder, a decoder, a rate matcher, a derate matcher, a mapping unit, a multiplier, a decision unit, a selection unit, a switch, an interleaver, a deinterleaver, a modulator, a demodulator, an input terminal, an output terminal, an antenna, an amplifier, a receiving unit, a transmitting unit, a DSP, a TCM encoder, a TCM decoder, a power supply unit, a power supply line, a communication interface, a communication protocol, etc., which are appropriately arranged together to perform the scheme.

[0228] Thus, the one or more processors of the transmitting device 1000 may include, for example, one or more instances of a CPU, a processing unit, a processing circuit, a processor, an ASIC, a microprocessor, or other processing logic that can interpret and execute instructions. The term "processor" may therefore refer to a processing circuit that includes multiple processing circuits, such as any, some, or all of the items listed above. The processing circuit may also perform data processing functions for inputting, outputting, and processing data, including data buffering and device control functions, such as call processing control, user interface control, etc.

[0229] Finally, it should be understood that the present invention is not limited to the embodiments described above, but also relates to and incorporates all embodiments within the scope of the appended independent claims.

[0230] Appendix A

[0231] Without loss of generality, we assume that the starting WUS subcarrier is K0 = 0. After inserting equations (9) to (15) into equation (6), and assuming that T shift is an integer, WUS becomes equivalent to

[0232]

[0233] That is, it is N symb The multiplexing of pulses to symbols d[m] is given by:

[0234]

[0235] They are all time-shifted versions of the same pulse shaping filter

[0236]

[0237] The filter is the inverse discrete Fourier transform of the FDSS window. In the absence of FDSS windowing, W[0] = ... = W[N sc -1]=1, which further simplifies to a DFT-s-OFDM pulse in the form of a Dirichlet kernel with N sc modulated subcarriers:

[0238]

[0239] Using the typical windowing function W[k], the pulse|g m [n]|Basically maintains the sinc shape, but with more or less attenuated sidelobes.

[0240] Figure 17 The figure shows the effect of TD pulse multiplexing produced by DFT precoding on OFDM. Here, K = 24 subcarriers are considered, but only N symb = 16 pulses, because the size of N is used e = SE of 8. Pulse shift FDDS is not used here.

[0241] Now by inserting equation (19) into equation (31), the signal is equivalent to

[0242]

[0243] in,

[0244]

[0245] is the OOK waveform of bit b[p].

[0246] Each pulse g m [n] In the sample interval The period has most of its energy, with an energy peak in the middle. This therefore corresponds to time multiplexing of bits, where each bit b[p] is transmitted via waveform O p [n] sent, where most of the potential combined energy lies in the sample interval

[0247]

[0248] In N symb =N bit N seg In this case, this simplifies to

[0249]

[0250] Based on this structure, the main design goal becomes to find relevant pulse phase rotations that significantly guarantee coherent combining of pulses in the same segment.

[0251] The phase difference between two adjacent pulses with index m and (m+1) is

[0252]

[0253] Then, if the FDSS window is real-valued and symmetric, then

[0254]

[0255] The function θ[n]={0 or π} corresponds to the sign difference between the real parts of the pulses and varies as a function of n. However, it can be verified that in the absence of FDSS and without spectrum spreading, this constant is equal to θ[n]=0 for all samples between two adjacent pulses.

[0256] So, assuming θ[n]=0, L=0 and N e =0, then

[0257]

[0258] therefore,

[0259]

[0260] Therefore, by choosing

[0261]

[0262] Can obtain Since the number of pulses is never small in practice, this can be well approximated by choosing Φ = π.

[0263] In the case of SE, it can also be verified that θ[n] = 0 for samples where the main lobes of adjacent pulses cross. So, we can again assume that θ[n] = 0, and

[0264]

[0265] In practice, it is usually also well approximated by the value π.

[0266] Appendix B

[0267] Recall that the modulation symbol is chosen to be d[m] = b r [m]e jΦm , where br It is a sequence consisting of a segment of 1 and a segment of 0, where the length of each segment of 1 is N seg Assume that there is N ones 1 segment, so that the 1 segment starts from index Let’s start. Consider a set of symb The output of DFT precoding equation (10) for the 1-1} is given by:

[0268]

[0269] Therefore, if N seg >1 and Where λ is any non-zero integer, then And the inner sum is zero, that is, the Fourier coefficient is D[k null ]=0.

[0270] In the absence of SE, D[k null ] is directly mapped to the subcarrier index k null ∈{0,…,K-1}. Otherwise, in the case of SE with shift L, the Fourier coefficients D[k null ] is mapped to k ′ null =L+k null (modN pulse ) Index coefficient D (se) [k ′ null ], where k ′ null ∈{0,…,K-1}. Assume that the DC subcarrier is located at index At this point, you need to index The DFT coefficients D[k null ] is set to zero.

[0271] Appendix C

[0272] As mentioned before, the Fourier coefficients of the LS approximation are generated from N fft Get it from the FFT, as shown below:

[0273]

[0274] Among them, the length is N fft The ideal target OOK signal is of length N bit Repeat each bit in the string b[m] Therefore, for For each segment of samples, are all constants, so

[0275]

[0276] If n=0, then Otherwise, if n≠0, we can use the exponential summation formula to get

[0277]

[0278] Finally, if K is even, the Fourier coefficients of the WUS with index k=0,…,K-1 are

[0279]

[0280] For the middle subcarrier, we get

[0281]

[0282] Otherwise, for other indices k≠K / 2, the closed-form expression is obtained

[0283]

[0284] in, is the uncorrelated global phase.

[0285] For the scheme disclosed herein, bits are multiplied by a factor of N. seg =N symb / N bit Spread spectrum, and the pulse phase varies linearly with the angle Φ, so that for k = 0, ..., N symb -1:

[0286]

[0287] if but and

[0288] Otherwise, any index k such that get

[0289]

[0290] Now, by choosing the pulse phase ramp angle to be Φ = π, this simplifies to

[0291]

[0292] If the Fourier coefficients are shifted For spectrum expansion, for index k=0,…,K-1:

[0293]

[0294] Finally, the Fourier coefficients of WUS are obtained by applying the FDSS window W[k] and FD shift as follows:

[0295]

[0296] For shift values This concludes

[0297]

[0298] Now comparing equations (52) and (58), we find that the Fourier coefficients differ only in magnitude and global phase, both of which are independent of the data bits. Specifically, the FDSS window is defined in the disclosed scheme as

[0299]

[0300] get in, is a global phase factor that is independent of k. Therefore, the disclosed scheme and the conventional LS method provide the same Fourier coefficients up to the global phase.

[0301] Appendix D

[0302] Given a subcarrier allocation K, the spreading factor N seg In the case where it is not directly an integer, it is shown that using SE provides better performance than using two spreading factors in the ON and OFF states when possible. The evaluation scenario uses values of Φ = π, β = 2 and α BPF = 1. Here, K = 72 and N bit0 =8, so Not an integer, but By using N e =8 spectrum expansion, we can get N seg =4 integer spreading factor, so each OOK state has the same spreading factor N segON =N segOFF =4. Alternatively, you can use (N segON =5, N segOFF =4), or (N segON =4, N segOFF =5). Figure 18 As shown, due to SE, each bit has a constant spreading factor, which provides better performance compared to the alternative where the ON and OFF states have different spreading factors.

Claims

1. A transmitting device (100) for a communication system (500), characterized in that The sending device (100) is used for: By adding N bit Each bit in the sequence of bits is bit The corresponding spreading sequences in the sequence of the N spreading sequences are multiplied to obtain the N bit The sequence of bits is spread to obtain N symb modulation symbols, wherein the N bit Each spreading sequence in the sequence of spreading sequences is a linear phase sequence with a constant rotation phase angle Φ; The N symb The modulation symbols are multiplied by the discrete Fourier transform precoder to obtain N symb Fourier coefficients; An orthogonal frequency division multiplexing OFDM signal (510) is transmitted, wherein the OFDM signal includes the N subcarriers mapped onto K OFDM subcarriers. symb Fourier coefficients.

2. The transmitting device (100) according to claim 1, characterized in that For the N bit The bits are spread based on: Repeat N bit bits to get N symb A sequence of repeated bits; The N symb The repeated bits are multiplied by the concatenated spreading sequence to obtain the N symb modulation symbols, wherein the concatenated spreading sequence is the N bit The cascade of the spreading sequences makes the cascaded spreading sequence a linear phase sequence with a constant rotation phase angle Φ.

3. The transmitting device (100) according to claim 1 or 2, characterized in that The N bit bits are based on N bit / 2 bits of the Manchester coded sequence.

4. The transmitting device (100) according to any one of the preceding claims, characterized in that The spreading sequence r l [m] is given by the following formula: Where l is the bit index, m is the modulation symbol index, e is the natural exponential function, j is the imaginary unit, Φ l is a constant angle that depends on the bit index l.

5. The transmitting device (100) according to claim 4, characterized in that The constant rotation phase angle Φ is equal to π.

6. The transmitting device (100) according to claim 4 or 5, characterized in that The spreading sequence r l [m] is an alternating sequence of values +1 and –1.

7. The transmitting device (100) according to claim 4 or 5, characterized in that The spreading sequence r l [m] is an alternating sequence of two binary shift keying symbols.

8. The transmitting device (100) according to claim 4, characterized in that The constant rotation phase angle Φ is given by the following formula: Among them, N seg is the spreading sequence r l length [m], k null are the indices of the Fourier coefficients to be zeroed, and λ is any non-zero integer.

9. The transmitting device (100) according to any one of the preceding claims, characterized in that The size of the discrete Fourier transform precoder is N symb ≤K.

10. The transmitting device (100) according to any one of the preceding claims, characterized in that Used for: Based on the N symb The periodic repetition of the Fourier coefficients, the N symb Fourier coefficients are expanded into K Fourier coefficients.

11. The transmitting device (100) according to claim 9 or 10, characterized in that Used for: The N symb The K Fourier coefficients are multiplied by the frequency domain spectrum shaping window coefficients to obtain frequency shaping Fourier coefficients.

12. The transmitting device (100) according to claim 11, characterized in that The frequency domain spectrum shaping window coefficients are real-valued symmetric coefficients from a bell-shaped function.

13. The transmitting device (100) according to claim 12, characterized in that The frequency domain spectrum shaping window coefficients are Kaiser window coefficients with a shaping parameter β=2.

14. The transmitting device (100) according to claim 11, characterized in that The frequency domain spectrum shaping window coefficient W0[k] is given by the following formula: Among them, N fft is the number of samples of the OFDM signal (510), and sin() is a sine function.

15. The transmitting device (100) according to any one of claims 11 to 14, characterized in that Used for: The frequency-shaped Fourier coefficients are multiplied by the frequency-domain phase shift to obtain the phase-shifted Fourier coefficients, wherein the frequency-domain phase shift is based on the shift parameter T shift .

16. The transmitting device (100) according to claim 15, characterized in that The shift parameter T shift The value of depends on the number of samples N of the OFDM signal (510) fft and the number of modulation symbols N symb .

17. The transmitting device (100) according to claim 16, characterized in that The shift parameter T shift The value of is given by either of the following formulas: Among them, N fft is the number of samples of the OFDM signal (510), is the ceiling function, is the floor function, is the rounding function.

18. The transmitting device (100) according to any one of the preceding claims, characterized in that The OFDM signal (510) is a wake-up signal.

19. A method (200) for a sending device (100), characterized in that The method (200) comprises: By adding N bit Each bit in the sequence of bits is bit The corresponding spreading sequences in the sequence of the N spreading sequences are multiplied to obtain the N bit The sequence of bits is spread (202) to obtain N symb modulation symbols, wherein the N bit Each spreading sequence in the sequence of spreading sequences is a linear phase sequence with a constant rotation phase angle Φ; The N symb The modulation symbols are multiplied (204) by the discrete Fourier transform precoder to obtain N symb Fourier coefficients; An OFDM signal (510) is transmitted (206), the OFDM signal comprising the N mapped onto K OFDM subcarriers symb Fourier coefficients.

20. A computer program having a program code, characterized in that When the computer program is run on a computer, the program code is used to perform the method according to claim 19 .

Images