A data waveform design method for large-scale internet of things communication

Abstract

The application discloses a data waveform design method for large-scale Internet of Things communication, and is suitable for large-scale Internet of Things communication limited by cost, energy and spectrum. Compared with existing waveforms, the parameters of the general waveform provide more degrees of freedom for the selection of different device conditions and channel conditions. In the parameters of the general waveform, the initial frequency of the waveform is determined, the frequency variation speed of the waveform is controlled, c represents the phase and amplitude of the waveform. The proposed waveform form is a subset of LoRa modulation waveform and FSK waveform, and has more parameter adjustment degrees of freedom, so that higher energy efficiency and spectrum efficiency are provided for large-scale Internet of Things, so that the network life and the number of access users of large-scale Internet of Things are increased.

Description

Technical Field

[0001] This invention belongs to the field of low-energy-consumption technology of Internet of Things (IoT), and in particular relates to a data waveform design method for large-scale IoT communication. Background Technology

[0002] For IoT devices, especially those requiring large-scale deployment due to resource, size, and cost constraints, active power is scarce, necessitating battery power. Typical battery capacities range from 500mAh (one button cell) to 5000mAh (two AA batteries), with a voltage of 3V. Due to the difficulty and expense of battery replacement, it is even impractical in some scenarios (e.g., subcutaneous implantable sensors in medical IoT). Therefore, the battery life of a device fundamentally determines the lifespan of the IoT. Low-energy communication (LOC) is crucial for reducing carbon footprint, paving the way for a "green IoT." While the energy consumption of a single device may seem small, the sheer volume of data generated and exchanged across the vast number of devices in the IoT constitutes a massive overall energy expenditure. Therefore, any reduction in energy consumption by a single device translates into significant overall energy savings, potentially leading to profound environmental impacts. Existing LOC physical layer solutions utilize a proprietary technology called LoRa, whose main advantages lie in its low cost, simple structure, and high power amplifier efficiency (LoRa modulation waveforms exhibit broad band characteristics). However, compared to traditional FSK modulation waveforms, LoRa modulation waveforms perform slightly worse in AWGN channels or frequency-flat fading channels. In terms of spectral efficiency, LoRa modulation has a lower upper limit, far less than QAM-type modulation techniques, which will limit large-scale device access. Summary of the Invention

[0003] The purpose of this invention is to provide a data waveform design method for large-scale Internet of Things (IoT) communication. This general waveform is a subset of LoRa modulation waveform and FSK waveform, and its parameters can provide more degrees of freedom for selection under different device conditions and channel conditions, thereby providing higher energy efficiency and spectral efficiency for large-scale IoT, and thus increasing the network lifetime and the number of connected users of large-scale IoT.

[0004] To solve the above-mentioned technical problems, the specific technical solution of the present invention is as follows:

[0005] A method for designing data waveforms for large-scale Internet of Things (IoT) communication, wherein the waveform duration is T. sym The passband waveform is mathematically represented as

[0006]

[0007] Where f c Represents the carrier frequency, waveform parameter β kBoth parameter α and parameter c carry information bits. This is the baseband waveform;

[0008] Waveform parameter β κ This can be understood as the initial frequency of a waveform, where the initial frequency interval between different waveforms is w. sym When the waveforms are integer multiples of each other, they are orthogonal to each other. Set the allocated channel bandwidth to W, and change β. κ At most, M = WT can be formed sym Each waveform carries m = log2M information bits, similar to multi-level frequency shift keying (MFSK). To ensure that the initial frequency variation range of the baseband section does not exceed [-W / 2, W / 2) and is symmetrically distributed, the waveform parameter β... κ The set is represented as

[0009] β κ ∈{κ·w sym +(w sym / 2-W / 2)|κ=0,1,…,M-1};

[0010] The instantaneous carrier frequency of the baseband waveform at time t is

[0011] Δf α,κ (t)=2αt+β κ ,0≤t<T sym

[0012] If parameter α = 0, the frequency of the waveform is fixed at β. κ The waveform is similar to that of frequency shift keying modulation; if parameter α ≠ 0, the frequency changes linearly at a rate of 2α per unit time; to meet the system bandwidth limit, Δf must be guaranteed. α,κ (t)∈[-W / 2,W / 2), for Δf α,κ (t) Perform a modulo division operation with respect to bandwidth W, i.e.

[0013] Δf α,κ (t)=(2αt+β κ +W / 2)modW-W / 2;

[0014] For the purposes of generality and symbol simplification, hereby... The baseband waveform is then corrected to

[0015]

[0016]

[0017] in The sign(·) distribution represents the sign function, and the u(·) distribution represents the step function; ρ(ι) is the pulse-shaped waveform; where

[0018]

[0019] as well as

[0020]

[0021] The parameter α is variable, and its magnitude is the frequency diversity size of the control waveform, |α|, to cope with different channel conditions; once the magnitude of α is determined, it becomes... Afterwards, the waveform can also simultaneously enable the set. The element in the middle is used as the α parameter, meaning that α can carry an additional 1 bit of information;

[0022] The parameter c is a complex number. If the number of elements in the optional set of parameter c is L, it is similar to a QAM-type modulation symbol with L constellation points, which can carry l = log2L information bits. In the case of M = 16, 8, 4, 2, c takes the constellation point set of PSK to increase spectral efficiency.

[0023] In summary, the signal rate r using this general waveform modulation b The ratio of the number of information bits carried by each waveform to the waveform length, i.e.

[0024] r b = (log2M+1+log2L)·w sym

[0025] Where parameter β κ It can carry log2M bits of information, parameter α can carry 1 bit of information, and parameter c can carry log2L information.

[0026] Furthermore, when c = 1, if Keying is used to carry log2M information, and this general waveform is equivalent to MFSK;

[0027] like Or -M / 2, where M takes values ​​from 128 to 4096. Keying is used to carry log2M information, which is equivalent to a LoRa modulated waveform. Used to maximize the frequency diversity of LoRa symbols.

[0028] The data waveform design method for large-scale Internet of Things (IoT) communication according to the present invention has the following advantages:

[0029] The technical solution of this invention forms a universal waveform design for large-scale Internet of Things (IoT). This waveform can guarantee robust performance under various channel conditions and has higher energy efficiency and spectral efficiency than existing technologies, thus better meeting the requirements of large-scale IoT. Attached Figure Description

[0030] Figure 1 For the present invention, under the AWGN channel, the configuration in Table 1 is adopted. And the energy efficiency and spectral efficiency performance under the c parameter configuration;

[0031] Figure 1 (a) For the present invention, under the AWGN channel, configurations 1-5 in Table 1 are used. And the energy efficiency and spectral efficiency performance under the c parameter configuration;

[0032] Figure 1 (b) For the present invention, under the AWGN channel, the configurations 6-8 in Table 1 are used. And the energy efficiency and spectral efficiency performance under the c parameter configuration;

[0033] Figure 2 To enable this invention to operate in a frequency-flat channel, the configuration shown in Table 1 is used. And the energy efficiency and spectral efficiency performance under the c parameter configuration;

[0034] Figure 2 (a) In the frequency-flat channel, the present invention uses configurations 1-5 in Table 1. And the energy efficiency and spectral efficiency performance under the c parameter configuration;

[0035] Figure 2 (b) In the frequency-flat channel, the present invention employs configurations 6-8 in Table 1. And the energy efficiency and spectral efficiency performance under the c parameter configuration;

[0036] Figure 3 In order to implement this invention in a frequency-selective channel, the configuration in Table 1 is used. And the energy efficiency and spectral efficiency performance under the c parameter configuration;

[0037] Figure 3 (a) In the frequency-selective channel of this invention, configurations 1-5 in Table 1 are used. And the energy efficiency and spectral efficiency performance under the c parameter configuration;

[0038] Figure 3 (b) In the frequency-selective channel, the present invention employs configurations 6-8 in Table 1. And the energy efficiency and spectral efficiency performance under the c parameter configuration; Detailed Implementation

[0039] To better understand the purpose, structure, and function of this invention, the following detailed description of a data waveform design method for large-scale Internet of Things (IoT) communication, in conjunction with the accompanying drawings, is provided.

[0040] The waveform's duration is T sym The passband waveform is mathematically represented as

[0041]

[0042] Where f c Represents the carrier frequency, waveform parameter β k Both parameter α and parameter c carry information bits. This is the baseband waveform;

[0043] A data waveform design method for large-scale Internet of Things (IoT) communication includes the following:

[0044] Waveform parameter β κ This can be understood as the initial frequency of a waveform, where the initial frequency interval between different waveforms is w. sym When the waveforms are integer multiples of each other, they are orthogonal to each other. Set the allocated channel bandwidth to W, and change β. κ At most, M = WT can be formed sym Each waveform carries m = log2M information bits, similar to multi-level frequency shift keying (MFSK). To ensure that the initial frequency variation range of the baseband section does not exceed [-W / 2, W / 2) and is symmetrically distributed, the waveform parameter β... κ The set is represented as

[0045] β κ ∈{κ·w sym +(w sym / 2-W / 2)|κ=0,1,…,M-1};

[0046] The instantaneous carrier frequency of the baseband waveform at time t is

[0047] Δf α,κ (t)=2αt+β κ ,0≤t<T sym

[0048] If parameter α = 0, the frequency of the waveform is fixed at β. κ The waveform is similar to that of frequency shift keying modulation; if parameter α ≠ 0, the frequency changes linearly at a rate of 2α per unit time; to meet the system bandwidth limit, Δf must be guaranteed. α,κ (t)∈[-W / 2,W / 2), for Δf α,κ (t) Perform a modulo division operation with respect to bandwidth W, i.e.

[0049] Δf α,κ (t)=(2αt+β κ +W / 2)modW-W / 2;

[0050] For the purposes of generality and symbol simplification, hereby... The baseband waveform is then corrected to

[0051]

[0052]

[0053] in The sign(·) distribution represents the sign function, and the u(·) distribution represents the step function; ρ(ι) is the pulse-shaped waveform; where

[0054]

[0055] as well as

[0056]

[0057] The parameter α is variable, and its magnitude is the frequency diversity size of the control waveform, |α|, to cope with different channel conditions; once the magnitude of α is determined, it becomes... Afterwards, the waveform can also simultaneously enable the set. The element in the middle is used as the α parameter, meaning that α can carry an additional 1 bit of information;

[0058] The parameter c is a complex number. If the number of elements in the optional set of parameter c is L, it is similar to a QAM-type modulation symbol with L constellation points, which can carry l = log2L information bits. In the case of M = 16, 8, 4, 2, c takes the constellation point set of PSK to increase spectral efficiency.

[0059] In summary, the signal rate r using this general waveform modulation b The ratio of the number of information bits carried by each waveform to the waveform length, i.e.

[0060] r b = (log2M+1+log2L)·w sym

[0061] Where parameter β κ It can carry log2M bits of information, parameter α can carry 1 bit of information, and parameter c can carry log2L information.

[0062] When c = 1, if Keying is used to carry log2M information, and this general waveform is equivalent to MFSK;

[0063] like Or -M / 2, where M takes values ​​from 128 to 4096. Keying is used to carry log2M information, which is equivalent to a LoRa modulated waveform. Used to maximize the frequency diversity of LoRa symbols.

[0064] Example 1

[0065] This embodiment provides a data waveform design method for large-scale Internet of Things (IoT) communication, the method comprising:

[0066] In actual systems, the signal bandwidth is infinite and cannot be simply limited by the bandwidth value W. Generally speaking, the signal bandwidth needs to meet a specific adjacent channel leakage ratio (ACLR) or spectrum emission mask. In this embodiment, the signal bandwidth requirements of the NB-IoT system are used to constrain the parameters of the designed waveform [1].

[0067] [1]Jingjing Zhang, Michael Mao Wang, and Tingting Xia, "Practicalsynchronization waveform for massive machine-type communications," IEEETransactions on Communications, vol.67, no.2, pp.1467-1479, February 2019.

[0068] The baseband waveform is generated using the following mathematical form.

[0069]

[0070] in The sign(·) and u(·) distributions represent the sign function and the step function, respectively.

[0071]

[0072] as well as

[0073]

[0074] Assuming normalized frequency shift parameters Set B has M = 128 elements, that is...

[0075]

[0076] Table 1

[0077]

[0078] 1,2 M = 128 symbol rate and bit rate

[0079] Table 1 (continued)

[0080]

[0081] To satisfy the spectrum transmission template in [1], different values ​​are required when M = 128. Under the parameter settings, the number of information bits carried by the symbol, the symbol rate, and the bit rate are shown in Table 1:

[0082] In parameter configuration scheme 1 only This is responsible for carrying information bits, i.e., the traditional FSK scheme. (Configuration 2 and Configuration 3) Carrying 1 bit of information, the possible parameter combinations are {-1 / 2, +1 / 2} and {-M / 4, +M / 4}. Parameter configuration scheme 4... Used to increase the frequency diversity of a waveform, also only It undertakes the task of carrying information bits. If the value of M ranges from 128 to 4096, that is, the existing LoRa modulation scheme. Configuration 5 further utilizes parameters based on Configuration 4. It undertakes the task of transmitting one information bit, increasing waveform spectral efficiency. Larger... This implies greater frequency domain dynamics, which will lead to a decrease in the frequency domain cohesion of its spectrum. Given a given spectral requirement (such as [1]), Increasing the value of M results in a decrease in the supported symbol rate (as shown in the table above). Different values ​​of M will also result in different symbol rates; only the rate for M=128 is listed in this table. Configurations 6, 7, and 8, based on configurations 2, 3, and 5 respectively, further utilize parameter c to carry information when the value of M is small (e.g., M=8). Parameter c can be a constellation set of QAM and PSK.

[0083] To adapt to the needs of different devices in large-scale IoT, waveform configurations with different energy efficiencies and spectral efficiencies are obtained through parameter adjustments. In the diagram, orange empty circles represent waveforms using configuration 1, semi-solid circles represent waveforms using configuration 2, solid squares represent waveforms using configuration 3, hollow triangles represent configuration 4 (overlapping with the LoRa modulation portion), and solid triangles represent waveforms using configuration 5. Configurations 5, 6, and 7 are represented by solid spheres, pentagons, and pentagrams, respectively.

[0084] Energy efficiency E here b / N0 represents the bit energy E required for the waveform to achieve a specific bit error probability (e.g., one in ten thousand) at the receiver. b The ratio of the noise power spectral density N0, E bA lower / N0 ratio indicates less energy required per bit, thus higher energy efficiency. Considering that actual power consumption is directly related to the power amplifier efficiency, i.e., the peak-to-average power ratio of the signal, the E here... b / N0 is factored into the power amplifier efficiency at the radiating end.

[0085] Figure 1 In an additive white Gaussian noise channel environment, the waveform using configuration 2 in the low spectral efficiency region and configuration 6 in the high spectral efficiency region exhibits optimal performance. This means that energy efficiency is highest for the same spectral efficiency, and vice versa. This provides higher energy efficiency and spectral efficiency for large-scale IoT, thereby increasing network lifetime and the number of connected users. The solid sphere in configuration 6, the pentagon in configuration 7, and the pentagram in configuration 8 represent cases where parameter c is not constant at 1, meaning parameter c carries information bits. The smaller M is, the lower the waveform's energy efficiency and the higher the spectral efficiency. However, in configuration 2, after M=8, further reducing M to M=4 (the hollow square dashed line part) is less effective at improving spectral efficiency than using configuration 6. That is, keeping M=8 constant and using parameter c to carry information bits. In the diagram, parameter c uses QPSK and 8PSK constellations.

[0086] Figure 2 Using a frequency-flat channel, the waveforms of configuration 3 in the low spectral efficiency region and configuration 7 in the high spectral efficiency region exhibit optimal performance.

[0087] Figure 3 Using a frequency-selective channel, the waveforms of configuration 5 in the low spectral efficiency region and configuration 8 in the high spectral efficiency region exhibit optimal performance. Configuration 1 cannot function in this channel environment and is therefore not shown in the figure.

[0088] In conclusion, Figure 1 , Figure 2 and Figure 3 These represent different conditions under AWGN channels, frequency-flat channels, and frequency-selective channels. And the energy efficiency and spectral efficiency performance under parameter c configuration. In AWGN channels and frequency-flat channels, The optimal performance is achieved under the configuration (i.e., configuration 2 combined with configuration 5); under frequency-selective channels, The performance is better under the configuration (i.e., configuration 5 combined with configuration 8).

[0089] This invention provides a universal data waveform design method. Compared to existing waveforms, the parameters of this universal waveform offer more degrees of freedom for selection under different device conditions and channel scenarios. The parameters of this universal waveform... middle, Determine the initial frequency of the waveform. The frequency change rate of the controlled waveform is determined, where c represents the phase and amplitude of the waveform. The proposed waveform forms are subsets of LoRa modulated waveforms and FSK waveforms, and offer more degrees of freedom in parameter adjustment. subsets and The subsets exhibit superior spectral and energy efficiency in scenarios with direct-line channels and frequency-selective multipath channels, respectively, making them suitable for large-scale IoT communications where cost, energy, and spectrum are limited.

[0090] Any aspects of this invention not described in detail are well-known to those skilled in the art.

[0091] It is understood that the present invention has been described through some embodiments, and those skilled in the art will recognize that various changes or equivalent substitutions can be made to these features and embodiments without departing from the spirit and scope of the invention. Furthermore, under the teachings of the present invention, these features and embodiments can be modified to adapt to specific situations and materials without departing from the spirit and scope of the invention. Therefore, the present invention is not limited to the specific embodiments disclosed herein, and all embodiments falling within the scope of the claims of this application are within the protection scope of the present invention.

Claims

1. A data waveform design method for massive Internet of Things communication, characterized by, The time length of the waveform is T sym The mathematical form of the passband waveform is where f c represents the carrier frequency, the waveform parameter β k , the parameter α and the parameter c all carry information bits, is the baseband waveform; Waveform parameter β κ The initial frequency of the waveform is understood, when the initial frequency interval between different waveforms is w sym is an integer multiple, the waveforms are orthogonal to each other, wherein The allocated channel bandwidth is set to W, by changing β κ At most M = WT sym orthogonal waveforms are formed, i.e. carrying m = log2M information bits, similar to multi-ary frequency shift keying MFSK; to ensure that the initial frequency variation range of the baseband part does not exceed [-W / 2, W / 2), and is symmetrically distributed, the set of waveform parameters β κ is represented as β κ ∈{κ·w sym +(w sym / 2-W / 2)|κ=0,1,…,M-1} The instantaneous carrier frequency of the baseband waveform at time t is Δf α,κ (t) = 2at + β κ , 0≤t sym If parameter a = 0, the frequency of the waveform is fixed as β κ , the waveform is similar to a frequency shift keying modulation waveform; if parameter a ≠ 0, the frequency linearly changes at a speed of 2a per unit time; to satisfy the system bandwidth limitation, it is necessary to ensure that Δf α,κ (t) ∈ [-W / 2, W / 2), and a modulus operation is performed on Δf α,κ (t) with respect to the bandwidth W, that is, Δf α,κ (t) = (2αt + β κ + W / 2) mod W - W / 2; For the purpose of generality and notational simplicity, we will The baseband waveform is then modified to wherein sign(•) distribution represents a sign function, u(•) distribution represents a step function; p(t) is a pulse shaping waveform; wherein and The parameter a is variable, and its size, i.e., |a| controls the frequency diversity size of the waveform to cope with different channel conditions; when the size of a is determined as After that, the waveform can also enable the set The element in the middle as the a parameter, i.e., a can additionally carry an amount of 1 bit of information; The parameter c is a complex number, and if the number of elements of the optional set of parameters c is L, similar to a QAM-like modulation symbol with L constellation points, l = log2 L information bits can be carried; in the case of M = 16, 8, 4, 2, c takes the constellation point set of PSK to increase the spectral efficiency. In summary, the signal rate r b The ratio of the number of information bits carried by each waveform to the length of the waveform, i.e. r b = (log2M + 1 + log2L) · w sym where parameter β κ The parameter α can carry 1 bit of information, and the parameter c can carry log2L information.

2. The method for data waveform design for massive IoT communications according to claim 1, wherein, When c = 1, if Keying is used to carry log2M information, and the passband waveform is equivalent to MFSK; If or -M / 2, M takes values from 128 to 4096, Keying is used to carry log2 M information, the waveform is equivalent to LoRa modulation waveform, here To maximize the frequency diversity of LoRa symbols.

Images