Impairment-aided channel state information determination method for communication systems

Abstract

IMPAIRMENT-AIDED CHANNEL STATE INFORMATION DETERMINATION METHOD FOR COMMUNICATION SYSTEMS The invention is a channel state information determination method performed by a communication system (10). The method utilizes adaptive filters (130) to model a transmitter front end (120) of the transmitter device (100) in order to perform a channel state information determination on the receiver device (200) side having hardware impairments of the transmitter front end (120).

Description

[0001] DESCRIPTION

[0002] IMPAIRMENT-AIDED CHANNEL STATE INFORMATION DETERMINATION METHOD FOR COMMUNICATION SYSTEMS

[0003] TECHNICAL FIELD

[0004] Invention relates to a channel state information determination method performed by a communication system and the communication system thereof.

[0005] PRIOR ART

[0006] In classical communication systems, the RF (Radio Frequency) front-end plays a critical role in the transmission and reception of signals. However, RF front-end components are inherently non-ideal and introduce various impairments that negatively impact signal quality and overall system performance. These impairments are broadly categorized into linear and nonlinear distortions.

[0007] Linear Impairments:

[0008] Linear impairments (distortions) cause frequency-dependent changes in the amplitude and phase of a signal. These impairments do not alter the frequency content of the signal, meaning they do not generate new harmonics or intermodulation distortion components. However, they can distort the signal's shape in the time domain and increase out-of-band emissions (OOBE). Some linear impairments are gives as follows:

[0009] An example of a linear impairments is l / Q Imbalance. In RF systems, impairments arising from mismatches in the amplitude and phase relationship between the in-phase (I) and quadraturephase (Q) channels are significant sources of signal degradation. Amplitude mismatches result from gain differences between the channels, while phase mismatches stem from discrepancies in phase alignment. These mismatches lead to signal distortion and a subsequent decline in overall system performance.

[0010] In more detail, amplitude and phase imbalances may occur during the splitting of the local oscillator (LO) signal into the I and Q channels. Instabilities in amplitude or phase can lead to significant mismatches and impairments. Differences in the output amplitudes and timing of DACs in the I and Q channels can result in gain mismatches and phase offsets, leading to l / Q imbalance. Amplitude and phase errors in mixers can contribute to significant impairments, affecting the accuracy and quality of the signal processing. Amplitude and phase response mismatches in filters and amplifiers can lead to unequal signal processing, resulting in distortion and degradation of system performance. Parasitic capacitance and inductance arising from PCB design can introduce unintended signal distortion, impedance mismatches, and degraded system performance.

[0011] Prior art [1] discloses solutions such as Hardware-Level Improvements where The use of precise and matched components can significantly reduce mismatches and minimize impairments, enhancing system performance and reliability; Adaptive Algorithms and Predistortion Techniques where Adaptive algorithms dynamically compensate for impairments by adjusting system parameters in real-time, while predistortion techniques preemptively counteract nonlinearities by applying inverse distortion at the transmitter, effectively mitigating their effects on the received signal and PCB Design Improvements minimises parasitic effects through careful PCB layout optimization, such as reducing stray capacitance and inductance, ensures improved signal integrity and reduces performance degradation.

[0012] Another linear impairment is DC Offset. The presence of an unwanted direct current (DC) component in a signal is referred to as DC offset. DC offset is defined as the addition of a constant value to the signal, which is a linear operation. It introduces a component at zero frequency (DC) in the signal's frequency spectrum but does not generate new frequency components, such as harmonics or intermodulation products. However, DC offset can affect the signal's dynamic range, leading to distortions. This can cause components such as power amplifiers (PA) and analog-to-digital converters (ADC) to operate outside their dynamic range, resulting in clipping and, consequently, the generation of harmonics.

[0013] In more details, the local oscillator (LO) can contribute to DC offset. Inadequate isolation between the LO and RF signals can result in leakage, which leads to the introduction of unwanted DC components in the signal. Offset voltages at the output of DACs introduce DC components into the signal, contributing to DC offset. The nonlinear characteristics of amplifiers can contribute to DC offset by introducing unintended DC components into the signal. Non-ideal mixers can contribute to the generation of DC offset by introducing unintended DC components during the mixing process. Prior art [2] discloses Hardware-Level Solutions which utilized AC Coupling with Capacitors (DC Blocking), Improved Circuit Design and DC Offset Measurement and Removal in order to reduce the effects of DC offset.

[0014] Yet another linear impairment is Carrier Frequency Offset (CFO). The difference between the carrier frequencies of the transmitter and receiver is referred to as frequency offset. Ideally, the transmitter and receiver should operate at the same carrier frequency. However, due to errors and tolerances in oscillators, a discrepancy between the transmitter and receiver carrier frequencies can occur. Frequency offset causes phase shifts during demodulation, leading to fluctuations in signal amplitude and phase. Additionally, it disrupts the orthogonality between subcarriers, resulting in degradation of system performance.

[0015] In more detail, Errors occur in the frequency stability of oscillators. Deviations from the nominal frequency due to manufacturing tolerances. Thermal effects can alter the oscillator frequency. Noise and errors in PLLs (Phase-Locked Loops) and frequency synthesizers contribute to Carrier Frequency Offset (CFO).

[0016] Prior art [3] discloses Use of Pilot Symbols and Synchronization Signals, Frequency-Locked Loops (FLL), Phase-Locked Loops (PLL) in order to reduce effects of CFO.

[0017] Yet another linear impairment is phase noise. Random fluctuations or deviations in the phase of a signal generated by an oscillator are referred to as phase noise. An ideal oscillator should produce a pure sine wave at a specific frequency with constant amplitude. However, real-world oscillators exhibit random phase and amplitude variations, or jitter, due to factors such as thermal noise, component tolerances, and other imperfections. These phase fluctuations lead to the formation of sidebands around the carrier frequency in the spectrum, degrading the signal's spectral purity. Phase noise causes inter-carrier interference (ICI) among subcarriers, which can affect system performance. Additionally, spectrum broadening due to phase noise can result in interference with adjacent channels. Thermal Noise and Flicker Noise in Oscillator Designs; Fluctuations in the Phase-Locked Loop (PLL) process are considered as phase noise.

[0018] Pho art solutions use Temperature-Controlled High-Quality Oscillators (e.g., TCXO, OCXO); Optimized PLL, VCO designs; Pilot Design for Phase Noise Detection. However, these techniques has [4] disadvantages such as, High-quality oscillators being expensive, showing poor performance at high frequencies and requiring complex algorithm designs and high computational power.

[0019] Yet another linear impairment is group delay. Group delay refers to the frequency-dependent variation in the delay experienced by different frequency components of a signal as it passes through a system. This phenomenon leads to distortion of the signal's shape in the time domain, manifesting as signal spreading, dispersion, or deformation. In wideband signals, it can result in information loss. Additionally, differing delays among subcarriers disrupt orthogonality, degrading system performance.

[0020] In more detail, the amplitude and phase responses of filters vary with frequency. High-order filters or sharp-transition filters can cause rapid variations in group delay across frequencies, leading to signal distortion and performance degradation. Amplifiers with uneven frequency responses amplify and delay different frequency components of a signal unevenly. The frequency-dependent phase responses of phase shifters and mixers contribute to group delay distortion, affecting the signal's time-domain integrity and system performance.

[0021] Prior art [5] discloses solutions such as Group Delay Equalizers, Adaptive Equalizers, Linear Phase Response Filters, Frequency domain processing in order to reduce effects of group delay.

[0022] Nonlinear impairments:

[0023] Nonlinear distortions occur due to the nonlinear relationship between the input and output of system components, leading to the generation of new frequency components in the signal's spectrum. The main types of nonlinear distortions include:

[0024] Power Amplifiers: These distortions arise when power amplifiers exhibit nonlinear behavior due to saturation when processing high-amplitude signals. By exhibiting nonlinear effects on high-amplitude input signals, power amplifiers generate amplitude modulation to amplitude modulation (AM / AM) and amplitude modulation to phase modulation (AM / PM) distortions, leading to various nonlinear impacts on the signal. At the results of this effect; harmonic distortion, generation spectral regrowth and increased OOBE occur. These effects arise from the cases listed below:

[0025] - Nonlinear Transfer Functions of Amplifiers

[0026] - Exceeding the Dynamic Range of Signal Levels - Temperature and Power Supply Variations

[0027] Such nonlinear impairments are corrected using the following techniques:

[0028] The use of digital predistortion (DPD) [6]: This method entails modeling the characteristics of the PA and applying its inverse to the transmitted symbols to achieve predistortion. The objective is to extend the linear operating range of the PA, allowing it to function closer to its saturation point without incurring substantial distortion

[0011] , However, implementing DPD introduces significant computational complexity and demands substantial memory resources. Moreover, the development of DPD algorithms that optimize both low latency and performance adds an additional layer of complexity. Striking a balance between efficiency and accuracy in these algorithms is a considerable challenge, as they must adequately compensate.

[0029] Peak to Average Power Ratio (PAPR) reduction techniques [7], [8]: Multicarrier signals, such as OFDM, are characterized by relatively high peak-to-average power ratio (PAPR) values, which further exacerbate the nonlinear effects of the PA. Various PAPR reduction techniques for OFDM, such as clipping, tone reservation, selective mapping, and active constellation extension, occur. Each of these techniques presents specific trade-offs, including challenges such as increased latency, higher complexity, reduced spectral efficiency, and increased power consumption.

[0030] Another nonlinear impairment is Mixer Nonlinearity. Distortions caused by mixers generating unwanted harmonics and intermodulation products while mixing input signals. It is resulted from nonlinear characteristics of mixers, lack of isolation between LO and RF signals, manufacturing deviations. At the end of this effect; generation of unwanted frequency components like intermodulation distortions and harmonics, interference. Effects of such impairments may be reduced by Use of High-Linearity Mixers: Filtering Techniques and Use of High-Linearity Mixers.

[0031] Another nonlinear impairment is intermodulation distortion (IMD). A multi-carrier signal interacts due to the nonlinear characteristics of the PA, causing intermixing among the carriers. These interactions result in the generation of new frequency components (intermodulation products). Consequently, spectral regrowth and out-of-band emissions occur. The sources of this effect are listed below.

[0032] - Nonlinearities in Amplifiers and Mixers:

[0033] - Interaction of Closely Spaced Signals in Multi-Carrier Systems

[0034] The effects on the signal of these products are: - Generation of new frequency components

[0035] - Intra-system and inter-system interference

[0036] - Signal Distortion and Performance Degradation

[0037] Filtering and channel separation, digital pre-distortion and linearization techniques are utilized in order to reduce the effects.

[0038] Another nonlinear impairment is harmonic generation. Distortion caused by the generation of harmonics (multiples of the fundamental frequency) of the input signal. The loss of energy, unwanted frequency components occur. Nonlinear amplifiers and oscillators, high signal levels are the main sources of this effect.

[0039] When considering both linear and nonlinear impairments, the literature offers numerous techniques for compensation. Generally, techniques for mitigating linear impairments tend to be more successful. However, nonlinear impairments can cause significant performance degradation in the system, and existing techniques in the literature often involve trade-offs.

[0040] Furthermore, matched filter, which is also known as correlator receiver, optimum detector, is a kind of detector used in communication systems to maximize the probability of detection. In communication systems matched filter is utilized in steps such as synchronization, channel estimation, equalization to detect the transmitted symbols from transmitter.

[0041] US2020099566 discloses a new method to detect synchronization signal at Narrow Band IOT (NB IOT) applications at User Equipment (UE). For this detection, several matched filters whose size is fitting to each of Orthogonal frequency-division multiplexing (OFDM) symbol used for synchronization are utilized. These matched filters gather the OFDM symbols having both normal cyclic prefix (NCP) and extended cyclic prefix (ECP) and constitute correlation peaks when the detection is realized. In brief, this method claims that it works in the case when Narrow Band Primary Synchronization Signal (NBPSS) and Narrow Band Secondary Synchronization Signal (NBSS) are used.

[0042] In this context, a solution that compensates both linear and nonlinear impairments is needed.

[0043] BRIEF DESCRIPTION OF THE INVENTION

[0044] The present invention relates to a method to eliminate the above-mentioned disadvantages and bring new advantages to the relevant technical field. An object of the invention is to significantly increase the accuracy of channel state information determination in communication systems.

[0045] To achieve all the objects mentioned above and that will emerge from the following detailed description, the present invention relates to a channel state information determination method performed by a communication system comprising a transmitter device and a receiver device. Accordingly, it comprises following steps performed by the transmitter device:

[0046] - modelling, by an adaptive filter unit which is connected to a transmitter RF front end, the transmitter RF front end and generating model parameters;

[0047] - generating, by a model generating unit, a front end model of Transmitter RF front end representing hardware impairments using model parameters;

[0048] - generating, by a baseband signal generator, a communication signal;

[0049] - processing, by the transmitter RF front end, communication signal which introduces hardware impairments of the transmitter RF front end to the communication signal;

[0050] - transmitting, by the transmitter antenna, processed signal to a communication channel; and following steps performed by the receiver device;

[0051] - receiving, by the receiver antenna, communication signal from the communication channel where received communication signal comprises hardware impairments, communication channel’s channel response and noise;

[0052] - processing, by the receiver RF front end, received communication signal converting it back to baseband;

[0053] - performing, by a baseband processing unit, channel estimation on the communication signal obtaining channel response of the communication channel;

[0054] - performing, by the baseband processing unit, channel equalization on the communication using channel response signal where channel equalized communication signal represents noisy symbols, which include both data and pilot indices received over the channel at the receiver;

[0055] - re-modulating and demodulating, by the baseband processing unit, channel equalized communication signal;

[0056] - applying, by the baseband processing unit, Inverse discrete Fourier transform to demodulated channel equalized communication signal;

[0057] - applying, by the baseband processing unit, front end model to the communication signal; - applying, by the baseband processing unit, discrete Fourier transform to the communication signal; - performing, by the baseband processing unit, channel estimation using communication signal that is transformed to frequency domain and obtaining an updated channel response of the communication channel comprising hardware impairments of the receiver RF front end. Thus, CSI is determined in a significantly more accurate way since RF front end impairments are considered in determination phase.

[0058] A possible embodiment of the invention is characterized in that wherein channel estimation is performed using at least one of Least Squares (LS), Discrete Fourier Transform (DFT)-based methods, or Minimum Mean Squared Error (MMSE) based method.

[0059] Another possible embodiment of the invention is characterized in that wherein adaptive filters (130) uses at least one of a LMS (Least Mean Square) based method, RLS (Recursive Least Square) based method, or GAL (Gradient Adaptive Lattice) based method.

[0060] Another possible embodiment of the invention is characterized in that wherein channel equalization is performed based on one of the Zero-Forcing (ZF) method or Minimum Mean Square Error (MMSE) based method.

[0061] BRIEF DESCRIPTION OF THE DRAWINGS

[0062] Figure 1 is a drawing illustrating schematic view of the system.

[0063] Figure 2 is a drawing illustrating schematic view of the transmitter front end.

[0064] Figure 3 is a drawing illustrating schematic view of the model generator and front end model.

[0065] REFERENCE NUMBERS GIVEN IN THE FIGURE

[0066] 10 Communication system

[0067] 100 Transmitter device

[0068] 110 Baseband signal generator

[0069] 120 Transmitter RF front end

[0070] 121 Converter

[0071] 122 Mixer

[0072] 123 Clock

[0073] 124 Power amplifier

[0074] 125 Local oscillator

[0075] 130 Adaptive filter unit 131 First adaptive filter

[0076] 132 Second adaptive filter

[0077] 133 Third adaptive filter

[0078] 134 Fourth adaptive filter

[0079] 140 Transmitter antenna

[0080] 150 Front end model

[0081] 151 First model parameter

[0082] 152 Second model parameter

[0083] 153 Third model parameter

[0084] 154 Fourth model parameter

[0085] 160 Model generator

[0086] 200 Receiver device

[0087] 210 Baseband processing unit

[0088] 220 Receiver RF front end

[0089] 240 Receiver antenna

[0090] 300 Communication channel

[0091] DETAILED DESCRIPTION OF THE INVENTION

[0092] In this detailed description, the subject matter is explained with references to examples without forming any restrictive effect only in order to make the subject more understandable.

[0093] Invention is a communication system (10) that increases accuracy of communication using hardware impairments while channel state information (CSI) determination in order to increase performance on a receiver device (200).

[0094] Referring to figure 1, communication system (10) comprises a transmitter device (100) and a receiver device (200). Transmitter device and receiver device may both have transmitting and receiving capabilities. Transmitter device (100) may be a base station and receiving device (200) may be a user equipment.

[0095] Transmitter device (100) comprises a baseband signal generator (111) for generating a baseband signal. Depending on the application and SNR requirements, information and pilot bits, with known sequences like Zadoff-Chu sequences, which can be designed are subjected to channel coding techniques such as convolutional coding. The bits are then mapped according to modulation schemes like QAM. Following this, the baseband signal is generated in the time domain.

[0096] In a possible embodiment, communication system (10) is a Bistatic / Multistatic or ISAC communication system.

[0097] Transmitter device (100) comprises a transmitter RF front end (120) connected to baseband signal generator (111) for processing generated baseband signal. Transmitter RF front end (120) is responsible for conditioning the first signal before it is transmitted. T ransmitter RF front end (120) may comprise processing components for converting first signal to analogue; components for mixing signals; components for amplifying or filtering signals. Transmitter RF front end (120) may comprise more components that are well-known and not disclosed herein. Transmitter RF front end (120) is well known in the art used in communication systems such as monostatic, bi-static or any other suitable system.

[0098] Transmitter device (100) comprises an adaptive filter unit (130). Adaptive filter unit (130) comprises at least an adaptive filter connected between input and output of the transmitter RF front end (120) and configured to model said transmitter RF front end (120) or it comprises plurality of adaptive filters each connected between input and output of at least some of the processing components of transmitter RF front end (120), each configured to model respective processing component of transmitter RF front end (120) and generate model parameters. In a possible embodiment, adaptive filter is connected to input of Transmitter RF front end (120) and output of Transmitter RF front end (120). Adaptive filter receives input signal and outputs an output signal. Output of Transmitter RF front end (120) and output of adaptive filter is subtracted from each other and end result is compared with an error tolerance signal. Adaptive filter is configured to reconfigure its internal model in order to generate an output which satisfies error tolerance on each iteration. Thus, a model parameter for T ransmitter RF front end (120) is generated. In another possible embodiment, plurality of adaptive filters are connected each component of Transmitter RF front end (120) individually. Thus, model parameters for each component is generated.

[0099] Transmitter device (100) comprises a model generator (160). Model generating unit (160) receives model parameter / s from adaptive filter / s and generates a front end model (150) for Transmitter RF front end (120). This model is a mathematical model that is generated using model parameters. For instance, model parameters of serially connected components may be multiplied in order to generate a model in an embodiment. Mathematical model is a model that takes the first signal as input and outputs a signal (figure 3).

[0100] Transmitter device (100) comprises a transmitting antenna (140) for transmitting processed baseband signal.

[0101] Referring to figure 2, in a possible embodiment of the invention, an adaptive filter for each processing component of transmitter RF front end (120) is provided. Each adaptive filter is connected between inputs and outputs of components. Each adaptive filter may be provided with error tolerance signal. Error tolerance signals may be different than each other. In more detail, transmitter RF front end (120) comprises a converter (121) for converting first signal to an analogue. A local oscillator (125) is provided. A mixer (122) is provided for mixing local oscillator’s (125) signal with the signal. A power amplifier (124) is provided for amplifying the signal. A first adaptive filter (131) is connected to converter (121). A second adaptive filter (132) is connected to the mixer (122). A third adaptive filter (133) is connected to power amplifier (124) and a fourth adaptive filter (134) is connected to local oscillator (125). Each adaptive filter models respective component and generates model parameters.

[0102] Adaptive filter techniques such as LMS (Least Mean Square), RLS (Recursive Least Square), and GAL (Gradient Adaptive Lattice) may be used in the invention. In a possible embodiment, the hardware characteristics of RF front-end components can be obtained by using a machine learning-based algorithm to know the baseband transmitted signal.

[0103] The receiver device (200) comprises a receiver antenna (140) that receives transmitted signal reflected back from a communication channel (300).

[0104] The receiver device (200) comprises a Receiver RF front end (220) for preprocessing received signal. Receiver RF front end (220) may comprise amplifiers and filtering elements, or other elements that are known to be utilized in receiver RF front ends (220) in communication devices. Receiver RF front end (220) such as in monostatic, bi-static or other type of communication devices are well known in the art. Receiver RF front end (220) converts received signal back to the baseband. Received second signal comprises characteristics of the communication channel (300) that it is reflected from and hardware impairments of transmitter RF front end (120). The receiver device (200) comprises a baseband processing unit (210) that performs CSI determination and detects objects from the received signal using detector / locator algorithms. Baseband processing unit (210) utilizes front end model (150) for significantly increased CSI determination considering hardware impairments of transmitter RF front end (120).

[0105] According to the above-mentioned details, communication system (10) operates as follows: The baseband signal generator (110) generates a communication signal in time domain. Communication signal may be depicted as [n]. Communication signal can be either a single-carrier signal or a multi-carrier signal, such as OFDM. Time-domain communication signal passes through the transmitter RF front end (120) and exhibits impairments of the transmitter RF front end. The signal comprising the impairments may be denoted as s[t]. Frequency response of the communication channel (300) may be denoted as H[f]. In such a scenario, the frequency domain signal taking by Fourier Transform of the time domain signal obtained at the receiver is denoted as Y[f] and is represented by following:

[0106] Y[f] = S[f] * H[f] + W[f]

[0107] Where s[f] and w[f] denote frequency domain of impairments s[t] and noise sample, respectively.

[0108] The channel estimation is performed using pilot symbols. In the bi-static systems since all the symbols are known at the receiver, the channel estimation is realized over the whole symbols. On the other hand, in multistatic scenarios channel estimator is done using pilot symbols. The channel estimator can utilize well-known techniques from the literature, such as Least Squares (LS), Discrete Fourier Transform (DFT)-based methods, or Minimum Mean Squared Error (MMSE). For instance, it is presented the channel estimator equation for the LS estimator.

[0109] H(f)kp = Y(f)kp / X(f)kp

[0110] Where kp is the pilot index. Then it can be interpolated to obtain the needed sample size.

[0111] By the baseband processing unit (210) channel equalization is performed. At this step, a channel equalizing based on either Zero-Forcing (ZF) or Minimum Mean Square Error (MMSE) techniques can be employed. Following, a simple channel equalizer is presented.

[0112] Y(f) = Y(f) / H(f)

[0113]

[0114] Where represents the noisy symbols, which include both data and pilot indices received over the channel at the receiver.

[0115] The baseband processing unit (210) demodulates and re-modulates

[0116]

[0117] and obtaining x f)- The baseband processing unit performs IDFT (Inverse Discrete Fourier Transform) process thus obtaining ^7) ■ The baseband processing unit adds RF-front end models into the thus obtainingS(.s~t)= x^t) *I(t)isobtained where / (t) is RF front-end model. Transmitting portion of the system transmits / (t) to receiving portion beforehand. Then Discrete Fourier Transform (DFT) process is applied into the

[0118]

[0119] and is obtained. The updated channel estimation is obtained including RF front-end impairments. In this step any kind of channel estimator method like LS, MMSE can be implemented. To illustrate LS type channel estimator is used and

[0120]

[0121] is obtained whereisimpairment added channel estimator.

[0122]

[0123] Accordingly, the update incorporates CSI impairments, resulting in improved estimation accuracy."

[0124] After obtaining CSI (Channel State Information), communication can be performed based on CSI. Additionally, once the CSI is acquired.

[0125] The scope of protection of the invention is specified in the attached claims and cannot be limited to those explained for sampling purposes in this detailed description. It is evident that a person skilled in the art may exhibit similar embodiments in light of the above-mentioned facts without drifting apart from the main theme of the invention. REFERENCES

[0126] [1] Valkama, M. (2010). RF impairment compensation for future radio systems. Multi-Mode / Multi-Band RF Transceivers for Wireless Communications: Advanced Techniques, Architectures, and Trends, 451-496.

[0127] [2] Yih, C. H. (2009). Analysis and compensation of DC offset in OFDM systems over frequency-selective rayleigh fading channels. IEEE transactions on vehicular technology, 58(7), 3436-3446.

[0128] [3] Sourour, E., El-Ghoroury, H., & McNeill, D. (2004, September). Frequency offset estimation and correction in the IEEE 802.11 a WLAN. In IEEE 60th Vehicular Technology Conference, 2004. VTC2004-Fall. 2004 (Vol. 7, pp. 4923-4927). IEEE.

[0129] [4] Zhang, W., Xu, Z., Lours, M., Boudot, R., Kersale, Y., Luiten, A. N.,... & Santarelli, G. (2011). Advanced noise reduction techniques for ultra-low phase noise optical-to-microwave division with femtosecond fiber combs. IEEE transactions on ultrasonics, ferroelectrics, and frequency control, 58(5), 900-908.

[0130] [5] Eskandari, A. R., & Mohammadi, L. (2011). Group delay variations in wideband transmission lines: Analysis and improvement. International Journal of Soft Computing Engineering (IJSCE), 1(4), 122-128.

[0131] [6] D. R. Morgan, Z. Ma, J. Kim, M. G. Zierdt, and J. Pastalan, “A generalized memory polynomial model for digital predistortion of rf power amplifiers,” IEEE Trans. Signal Process., vol. 54, no. 10, pp.

[0132] 3852-3860, 2006.

[0133] [7] G. K. Carvajal, M. F. Keskin, C. Aydogdu, O. Eriksson, H. Herbertsson, H. Hellsten, E. Nilsson, M. Rydstr om, K. Vanas, and H. Wymeersch, “Comparison of automotive FMCW and OFDM radar under interference,” IEEE Radar Conf., pp. 1-6, 2020.

[0134] [8] F. H. Raab, P. Asbeck, S. Cripps, P. B. Kenington, Z. B. Popovic, N. Pothecary, J. F. Sevic, and N. O. Sokal, “Power amplifiers and transmitters for rf and microwave,” IEEE T rans. Microw. Theory Tech, vol. 50, no. 3, pp. 814-826, 2002.

Claims

CLAIMS1. A channel state information determination method performed by a communication system (10) comprising a transmitter device (100) and a receiver device (200) characterized in that comprising following steps performed by the transmitter device (100):- modelling, by an adaptive filter unit (130) which is connected to a transmitter RF front end (120), the transmitter RF front end (120) and generating model parameters;- generating, by a model generating unit (120), a front end model (150) of Transmitter RF front end (120) representing hardware impairments using model parameters;- generating, by a baseband signal generator (110), a communication signal;- processing, by the transmitter RF front end (120), communication signal which introduces hardware impairments of the transmitter RF front end to the communication signal;- transmitting, by the transmitter antenna (140), processed signal to a communication channel (300);and following steps performed by the receiver device (200);- receiving, by the receiver antenna (240), communication signal from the communication channel (300) where received communication signal comprises hardware impairments, communication channel’s (300) channel response and noise;- processing, by the receiver RF front end (220), received communication signal converting it back to baseband;- performing, by a baseband processing unit (210), channel estimation on the communication signal obtaining channel response of the communication channel (300);- performing, by the baseband processing unit (210), channel equalization on the communication using channel response signal where channel equalized communication signal represents noisy symbols, which include both data and pilot indices received over the channel at the receiver;- re-modulating and demodulating, by the baseband processing unit (210), channel equalized communication signal;- applying, by the baseband processing unit (210), Inverse discrete Fourier transform to demodulated channel equalized communication signal;- applying, by the baseband processing unit (210), front end model (150) to the communication signal;- applying, by the baseband processing unit (210), discrete Fourier transform to the communication signal;- performing, by the baseband processing unit (210), channel estimation using communication signal that is transformed to frequency domain and obtaining an updatedchannel response of the communication channel comprising hardware impairments of the receiver RF front end (220).

2. The method according to claim 1, characterized in that wherein channel estimation is performed using at least one of Least Squares (LS), Discrete Fourier Transform (DFT)- based methods, or Minimum Mean Squared Error (MMSE) based method.

3. The method according to claim 1 or 2, characterized in that wherein adaptive filters (130) uses at least one of a LMS (Least Mean Square) based method, RLS (Recursive Least Square) based method, or GAL (Gradient Adaptive Lattice) based method.

4. The method according to one of the preceding claims, characterized in that wherein channel equalization is performed based on one of the Zero-Forcing (ZF) method or Minimum Mean Square Error (MMSE) based method.

Images