Common phase error (CPE) correction for simplified PLL hardware design in high-frequency radio systems

Abstract

Provided is a method of common phase error (CPE) correction in a radio system, allowing for e.g. simplified PEL hardware design in high- or higher-frequency radio systems. The radio system includes a plurality of phase-locked-loops (PLLs; 112-1 to 112-M) and / or frequency multipliers for generating a plurality of oscillator signals (rj(t)). The method includes generating a plurality of mixed signals by mixing a reference signal (rref(t)) with each of the oscillator signals; generating a plurality of down-converted signals (sj(t)) by low-pass filtering each of the mixed signals; performing CPE detection, tracking and / or prediction for the oscillator signals based on the plurality of down-converted signals, and performing one or more CPE corrections during operation of the radio system using the outcome of the CPE detection, tracking and / or prediction. A corresponding radio system, as well as a computer program and computer program product are also provided.

Description

COMMON PHASE ERROR (CPE) CORRECTION FOR SIMPLIFIED PLL HARDWARE DESIGN IN HIGH-FREQUENCY RADIO SYSTEMS Technical field[oooi] The present disclosure generally relates to radio systems. In particular, the present disclosure relates to a solution for common phase error (CPE) correction allowing for simplified PLL hardware design, in particular in higher-frequency radio systems.Background

[0002] In many radio systems, there is a need to both provide and distribute multiple oscillator signals within the system, such as for example in multiple antenna radio systems. For example, a radio system with multiple antennas may include a plurality of frequency mixers responsible for converting between e.g. baseband (or intermediate frequency, IF) and radio frequency, wherein each mixer is responsible for mixing an oscillator / carrier signal with the signal received and / or to be sent using its own associated antenna element(s). Examples include multiple-input and multiple-output (MIMO) systems, massive MIMO systems, phased antenna array systems, radar systems, etc., and combinations thereof.

[0003] An oscillator signal may be affected by phase noise (PN), wherein PN is defined as short term (on the order of e.g. seconds or less) variations in phase or frequency of the signal, e.g. of a local oscillator (LO) and / or phase-locked loop (PLL) used to generate the signal. In e.g. multi-carrier orthogonal frequency-division multiplexing (OFDM) systems, within the operating bandwidth (BW), PN is known to degrade signal quality and hence increase bit / symbol error rate (BER / SER) within the beam. PN-effects may be particularly visible in multi-antenna systems operating at higher frequencies, such as from centimeter-wave systems and up (in terms of frequency). High PN may also elevate adjacent-channel interference (ACI) levels and reduce receiver sensitivity. For e.g. beamforming operations, high PN-levels may result in time-varying beam squint and, as a result thereof, in a loss of beam-gain in the intended direction of the beam, thereby negatively impacting user throughput). As identified in 3GPP NR technical report TR38.803, the impact of PN on wide-bandwidth system performance when operating at e.g. millimeter-wave and higher frequencies may be sever.

[0004] The behavior of PN may generally be characterized by its power spectral density (PSD), wherein lower-frequency parts contribute to a slow time-varying common phase error (CPE), that may serve as a dominant impairment source at high frequencies for OFDM-based systems. Meanwhile, higher-frequency parts contribute to inter-carrier interference (ICE). New Radio (NR) systems are typically designed to minimize ICI by appropriate sub-carrier spacing (SCS) selection, and a substantial emphasis from the 3GPP NR standardization activities has been focused on the design of subcarrier spacing in OFDM-based systems in order to deal with ICI impact in high band (HB) operation. In addition, PN-tracking reference signals (PT-RSs) are used for CPE-tracking between OFDM symbols, and, for data demodulation, the effect of PN maybe removed by PT-RS-tracking and channel equalization with respect to the demodulation reference signal. For such operations, there is however an underlying assumption that each LO / carrier frequency generation circuit (e.g. PLL) is designed for low PN conditions along with a suitable choice of distribution architecture for the oscillator signals. When using a single LO / PLL, active radio frequency (RF) components such as frequency multipliers and slow drift in the common reference clock / signal are known to produce slow time-varying CPE in the system, which makes stable clock generation with low PN CPE-characteristics important.

[0005] Focusing on e.g. large MIMO systems (like massive or ultra-MIMO systems operating at centimeter- or millimeter-wave frequency with large / extra-large aperture array), such system designs are moving towards the use of multiple LOs / PLLs to generate multiple LO signals, e.g. LO / PLL-based sub-array designs. Other possibilities include to use sub-array designs using centralized LO generation, wherein e.g. (local) oscillator signals for sub-arrays are supplied using active frequency multipliers. For both cases, the design of the LO signal distribution architecture plays an important role in defining PN impact on e.g. error vector magnitude (EVM) and functions such as beamforming (BF), antenna array calibration (AC), and similar. Systems with less PN are less susceptible to CPE-related degradations. The designers of RF systems often aim at optimizing end-to-end (e2e) system performance, by designing better quality carrier generators (LOs / PLLs) with lower PN levels at low frequency offset from the LO frequency. Designing such high-quality LOs / PLLs and / or e.g. frequency multipliers maybe challenging and both power- and time-consuming, and may lead to higher productresearch and development (R& D) costs and longer time-to-market. The design challenge is known to grow larger with increasing operating frequencies, especially above e.g. 6 GHz. For systems with centralized LO generation, distribution of the common reference clock over the antenna array with low phase drift may be challenging as well.

[0006] A typical large or extra-large antenna array system uses a plurality of PLLs, or a single crystal oscillator (XO) and multiple frequency multipliers, to generate the multiple LO signals (or clock signals, for digital processing). For example, in an extralarge aperture array system, the array may have one or multiple PLLs driving the array, and all the PLLs may be connected to a common reference clock. In all such architectures, effective CPE behavior is slow time-varying and the PN is partially correlated or uncorrelated between the oscillator signals depending on the correlation among the PLLs. Partial correlation in CPE may also result from the common reference clock driving the PLLs, and CPE-estimation / tracking may become an important task to achieve sufficient accuracies in functions like AC, BF, radar, localization and positioning, and similar.

[0007] It is to be noted that attempting to reduce PN -induced degradations in EVM and negative effects on AC, BF and time-varying beam squint often produces conflicting requirements for system designers. For example, in a MIMO system with multiple good LOs / PLLs, the optimal number of numbers of PLLs depends on how well designer can balance both sides of the performance requirements, i.e. to provide a design with a number of PLLs large enough to satisfy EVM requirements and small enough to satisfy e.g. AC and BF requirements. However, both the cost and complexity of designing good LOs / PLLs with low PN-levels often grow exponentially with increased operating frequency. In addition, routing of a single oscillator signal to e.g. all beamformers in a large antenna array may be equally challenging. Generally, PN-levels approximately increases by 6 decibels (dB) as carrier frequency doubles, and a system and method that may help to relax PN-requirements in LO / PLL-design (allowing lower-quality LOs / PLLs to be used in such systems) and that scales along with increasing LO frequency may prove particularly beneficial for e.g. future Fifth-Generation plus (5G+), pre-sixth-generation (pre-6G) and 6G system design and product development.[ooo8] The presence of PN causes two types of impairments to e.g. an OFDM system: i) a common random phase error (PN-CPE) that is the same on each subcarrier, and ii) inter-carrier interference (PN-ICI) between OFDM sub-carriers. While 3GPP standards provide mechanisms to handle PN-CPE during the data demodulation process, the same standards do not, to the knowledge of the Applicant behind the present patent application, specify any solutions for handling e.g. timevarying CPE that impacts functions like BF, AC, radar-sensing and localization, and similar.Summary

[0009] In light of the above, the present disclosure aims to at least partially alleviate some of the above-identified shortcomings of contemporary technology. For this purpose, the present disclosure provides a method of CPE correction in a radio system as well as a radio system implementing such functionally, as well as corresponding computer programs and computer products, as defined in and by the accompanying independent claims. Various embodiments of the method, system, computer program and computer program product are defined in and by the accompanying dependent claims.

[0010] According to a first aspect, there is provided a method of CPE correction in a radio system, wherein the system includes a plurality of PLLs and / or frequency multipliers configured for generating a plurality of (higher- frequency) oscillator signals (e.g. from a common lower-frequency signal). The method includes generating, during operation of the radio system, a plurality of down-converted signals by mixing a reference signal with each of the oscillator signals, followed by low-pass filtering each of the mixed signals. The method further includes performing CPE detection, tracking and / or prediction for the plurality of oscillator signals based on the plurality of down-converted signals, and performing one or more CPE corrections during (continued) operation of the radio system using an outcome of the CPE detection, tracking and / or prediction.

[0011] The first aspect improves upon contemporary technology in that it allows CPE detection, tracking and / or prediction outside of the LO / PLL system (for e.g. an OFDM-based system), irrespective of its level, frequency and time-variability. The proposed solution avoids to increase the complexity of the LO / PLL design, butinstead allows to relax PN requirements when designing e.g. the PLLs, thus simplifying PLL hardware design in radio systems. The proposed solution allows generation of oscillator signals and the distribution thereof using simpler-to-design LOs / PLLs that can have poorer PN-performance, and thus reduces both R& D costs and time-to-market. The proposed solution maybe particularly suitable for large and extra-larger array radios operating in centimeter-wave and above frequencies, and may further be applied to any functionality that requires accurate channel estimates (such as AC, BF, radar, localization and sending, and similar). The proposed solution applies to centimeter-wave, millimeter-wave, sub-THz and e.g. THz systems. The proposed solution is frequency-scalable, and allows the use of e.g. lower-quality PLLs for high-mid band (e.g. FR3 above 6 GHz and below 15 GHz) and high band systems (FR2), and similar. The proposed solution may lower PN-impact on AC, allowing lower-complex systemization of AC functionality for massive and ultra-MIMO systems. The proposed solution may ensure beamforming weight stability over time, and may prevent time-varying beam squint and sidelobe effects during e.g. single user MIMO (SU-MIMO), multiple user MIMO (MU-MIMO) scheduling windows in FR3 and / or FR2 hybrid beamforming (HB), and similar. The proposed method may ensure reciprocal phase-coherency over time, enabling high-performance SU-MIMO and MU-MIMO for reciprocity-based BF methods. The proposed solution may further simplify the complexity-level of baseband data-path PN handling algorithms, at least partially contributing to the improved e2e performance and R& D costreduction. If applied to radar systems, such as integrated sensing and communication (ISAC) radar, the proposed solution may improve the radar, sensing and / or localization functionality including impacts on Radio Interface based Base station Synchronization (RIBS) functionality and similar.

[0012] According to a second aspect, there is provided a radio system. The system includes a plurality of PLLs and / or frequency multipliers configured for generating the plurality of (higher-frequency) oscillator signals (e.g. from a common lower-frequency signal). The system further includes a common phase error (i.e. CPE) observer. The CPE observer is configured to generate, during operation of the radio system, the plurality of down-converted signals, by mixing the reference signal with each of the oscillator signals, and by low-pass filtering each of the mixed signals. The system further includes a CPE detector, tracker and / or predictor. The CPE detector, tracker and / or predictor is configured to obtain the plurality of down-convertedsignals generated by the CPE observer (i.e. directly or indirectly from the CPE observer); to perform, based on the plurality of down-converted signals, the CPE detection, tracking and / or prediction for the plurality of oscillator signals, and to output the outcome of the CPE detection, tracking and / or prediction as one or more CPE estimations and / or predictions for the plurality of oscillator signals. The system further includes a CPE corrector. The CPE corrector is configured to receive the one or more CPE estimations and / or predictions from the CPE detector, tracker and / or predictor (i.e. directly or indirectly receiving the output from the CPE detector, tracker and / or predictor), and to use the one or more CPE estimations and / or predictions to perform the one or more CPE corrections during the operation of the radio system. The second aspect thus provides a radio system that implements the functionality of the method of the first aspect. The implementation can be performed using hardware (digital and / or analog circuits), software or a combination of both hardware and software. The second aspect provides the same technical benefits over contemporary technology as those of the method of the first aspect, including the simplified PLL hardware design of the radio system enabled by the envisaged solution.

[0013] According to a third aspect, there is provided a computer program including instructions that, when executed by processing circuitry of a radio system (such as that of the second aspect or any embodiment thereof), cause the radio system to generate, during operation of the radio system, the plurality of oscillator signals; to implement a common phase error, CPE, observing functionality, including to mix the reference signal with each (or copies of each) of the oscillator signals, and to low-pass filter each of the mixed signals, to generate the plurality of down-converted signals. The instructions are further such that they cause the radio system to implement a CPE detection, tracking and / or prediction functionality, including to obtain (i.e. directly or indirectly from the CPE observing functionality) the plurality of down-converted signals generated by the CPE observing functionality; to perform, based on the plurality of down-converted signals, CPE detection, tracking and / or prediction for the plurality of oscillator signals, and to generate an outcome of said CPE detection, tracking and / or prediction as one or more CPE estimations and / or predictions for the plurality of (local) oscillator signals. The instructions are further such that they cause the radio system to implement a CPE correction functionality, including to obtain the one or more CPE estimations and / or predictions (i.e. directlyor indirectly from the CPE detection, tracking and / or prediction functionality) generated by the CPE detection, tracking and / or prediction functionality, and to use the one or more CPE estimations and / or predictions to perform the one or more CPE corrections during operation of the radio system. The computer program of the third aspect is thus such that it causes e.g. the radio system of the second aspect to perform in accordance with the method of the first aspect, with the same technical benefits thereof, including the simplified PLL hardware design.

[0014] According to a fourth aspect, there is provided a computer program product that includes a computer-readable storage medium storing the computer program of the third aspect (or any embodiment thereof). In some embodiments, the storage medium may be non-transitory. In other embodiments, the storage medium maybe transitory. The computer program product serves the purpose of providing a solution for logistically distributing the computer program of the third aspect, e.g. such that they computer program maybe installed at / on the radio system of the second aspect.

[0015] Other objects and advantages of the present disclosure will be apparent from the following detailed description, the drawings and the claims. Within the scope of the present disclosure, it is envisaged that all features and advantages described with reference to e.g. the method of the first aspect are relevant for, apply to, and maybe used in combination with also any one of the radio system of the second aspect, the computer program of the third aspect, and the computer program product of the fourth aspect, and vice versa.Brief description of the drawings

[0016] Exemplifying embodiments will be described below with reference to the accompanying drawings, in which:Figure 1A schematically illustrates an example CPE observer according to the present disclosure, in combination with a multiple-PLL oscillator signal network for providing a plurality of oscillator signals;Figure 1B schematically illustrates an example of an alternative oscillator signal network using a common (lower-frequency) signal and a plurality of frequency multipliers;Figure 2 schematically illustrates one or more examples of a CPE observer and CPE detector / tracker / predictor arrangement according to the present disclosure;Figure 3 schematically illustrates one or more examples of a radio system including the CPE and CPE detector / tracker / predictor arrangement of Figure 2 together with a CPE correction functionality, for an example radio transceiver system;Figure 4 schematically illustrates an example of a more generic radio system including, or affected by, the envisaged CPE correction chain of the present disclosure;Figure 5 schematically illustrates a flowchart of one or more example methods for CPE correction in a radio system, according to the present disclosure;Figures 6A and 6B schematically illustrate one or more examples of a radio system according to the present disclosure, andFigure 7 schematically illustrates one or more examples of a computer program and computer program product according to the present disclosure.

[0017] In the drawings, like reference numerals will be used for like elements unless stated otherwise. Unless explicitly stated to the contrary, the drawings show only such elements that are necessary to illustrate the example embodiments, while other elements, in the interest of clarity, maybe omitted or merely suggested.Detailed description

[0018] Figure 1A schematically illustrates an example of an arrangement 100 of a multiple-PLL network no and a CPE observer 120 as envisaged herein. The network 110 includes a plurality of PLLs 112-1 to 112-M, where M is an integer indicating a total number of such PLLs. Each PLL 112-j is configured to generate a (higher-frequency) oscillator signal rj(t) (with j e [1, M]), e.g. from a (lower-frequency) signal (not shown) provided to each of the PLLs 112-1 to 112-M. As envisaged herein, each PLL may for example be used to generate a local oscillator (LO) signal, e.g. by controlling a voltage-controlled oscillator (VCO) using a feedback loop and based on a detected, low-pass filtered difference between a desired phase difference between thelower-frequency signal and the output from the respective PLL 112-j, as is common practice. The network 110 includes the circuitry necessary to distribute each of the signals τj(t) to a respective component of a radio system in which the arrangement 100 is to be included. Exactly for what purpose each signal r7(t) is used, and by which component, of course depends on the exact configuration / purpose of the radio system, as long as there is a need for each such component to receive the respective oscillator signal τj(t). Such components may for example be frequency multipliers used for conversion between baseband and passband, between IF band and passband, between baseband and radio frequency band, and similar, each provided for a different path of e.g. a multi-antenna radio system, such as a MIM0 system or derivatives thereof, and similar. In addition to being distributed (not shown) to components of the system that requires such oscillator signals for their operation, such as a plurality of mixers for up- / down-conversion, the signals (e.g. copies of the signals) τj(t) are also provided as input to the CPE observer 120.

[0019] Each of the signals r7(t) may for example be referred to as local oscillator (LO) signals, and the collection of PLLs 112-1 to 112-M may be considered to form a distributed PLL, a distributed LO (generation) system, or similar. The PLLs 112-1 to 112-M may for example be distributed within an antenna array, a MIMO-based system, a radar system, and / or similar.

[0020] As envisaged herein, phase noise (PN) is present among the signals τj(t), and may originate from e.g. the various generators / PLLs 112-1 to 112-M. Each PLL may have noise transfer functions associated with each component in the loop, and the control loop as well as e.g. frequency translation may contribute to such PN. Sources of uncorrelated PN among an array of branches (such as one branch for each antenna element or subgroup of antenna elements) may originate from the PLL PN noise associated with the branch. If using a same (lower-frequency) signal input (not shown) for each PLL 112-1 to 112-4, PN behavior may also be partially correlated among the various branches. PN induced by LOs / PLLs is an important parameter, that may impact e.g. NR high band system performance. Practically achievable PN PSD levels (as measured in e.g. dBc / Hz) for millimeter-wave systems may impact OFDM SCS and achievable signal quality, as elaborated further on in 3GPP TR 38.803.

[0021] Systems with larger number of antennas (such as Massive and Ultra - MIMO systems) that operate at high frequency may include multiple LOs / PLLs in order to avoid high-frequency signal routing challenges through the RF circuit systems, to simplify RF sub-system design. Such systems may require accurate antenna array calibration to meet required beamforming performance levels. As briefly mentioned earlier herein, there is a tradeoff between achievable antenna calibration accuracy and the EVM contribution in the data path (i.e. in the beam). Fewer PLLs in the system may result in better antenna array calibration performance, which may however degrade the EVM metric (as further discussed in e.g. T. Hohne et al., “Phase Noise in Beamforming”, IEEE Trans. On Wireless Communications, Vol.9, No. 12, December 2010). On the other hand, the higher the number of PLLs, the better EVM metric maybe achieved in the beam at the cost of antenna array calibration accuracy. Time-varying nature of PN may result in loss of array calibration over time and may, hence, result in time varying beam squint effects (as further discussed in e.g. M. Umar et al., “Beamforming and Transceiver Optimization in Phase Noise for mmWave and THz Bands”, 2019 16thInternational Symposium on Wireless Communication Systems (ISWCS), Oulu, Finland, 2019). This may lead to losses, such as significant losses, of system performance for key NR functionalities such as achievable peak throughput in SU-MIMO and MU-MIMO within the scheduling interval. High-frequency NR systems are generally to be designed with multiple PLLs to cater to larger antenna arrays and each PLL may have high-performance requirements in order to lower the impact of PN in distributed or semi-distributed LO systems. Designing of high-quality LOs / PLLs with low PN levels maybe challenging and R& D time-cost consuming, and PLL design technology may not be sufficiently scalable with frequency.

[0022] As will now be described in more detail, envisaged herein is to provide a frequency-scalable solution that enables to relax LO / PLL design requirements with respect to target PN levels. The proposed solution may simplify PLL designs, shorten R& D time and may allow PLLs with higher PN levels (in e.g. dBc / Hz) with little or no impact on e2e communication system performance.

[0023] For this purpose, there is provided the CPE observer 120 to which (copies of) the signals rj(t) are provided as input. Provided as input to the CPE observer 120 is also a reference signal Tref( )> in the example of Figure 1A provided by a referencePLL 114 (PLL-REF). Preferably, the reference signal rrey(t) has a same frequency as each of the oscillator signals τj(t). The CPE observer 120 further includes a plurality of frequency mixers 124-1 to 124-M that are configured to mix each of the signals τj(t) with the reference signal τref(t). The output from the mixers 124-1 to 124-M is thus a plurality of mixed signals. As envisaged herein, a reference PLL (such as 114) with similar design quality as for the target PLL(s) 112-1 to 112-M may be used for the signal mixing operation. The mixers 124-1 to 124-M of the CPE observer 120 are configured to generate M parallel mixed PN signals with a common phase-biasing. If required, as envisaged herein, the various mixers 124-1 to 124-M maybe configured to alter e.g. the signal Tref(t in anyway needed to perform down-conversion of the signals τj(t), including e.g. to phase-shift the signalrrey(t) before mixing it with each of the signals τj(t), and / or similar.

[0024] To generate a plurality of down-converted signals sj(t), the mixed signals are then passed through a respective low-pass filter (LPF) 126-1 to 126-M. Herein, s7(t) thus corresponds to the down-converted version of the signal τj(t). If the signals τj(t) are higher-frequency, radio frequency, IF band or passband signals, the plurality of down-converted signals sj(t) are e.g. baseband signals. In particular, the down-converted signals sj(t) correspond to low-frequency parts of the PN with an equal phase biasing for all LOs / PLLs, as the signal τref(t) from the reference PLL 114 is used in the down-conversion process. The low-pass filters 126-1 to 126-M may for example have a predefined passband width (e.g. at or around 100 kHz, and e.g. lower than a bandwidth of the original signal τj(t) to filter only lower-frequency or very low-frequency parts of the PN). The signals sj(t) are thus baseband PN signals that capture (mostly) slow time-varying components of the PN, i.e. the low-pass filters 126-1 to 126-M serve to isolate the slow time-varying part of the PN (or slow drift of the carrier / reference signal(s)).

[0025] Consequently, the envisaged CPE observer 120 provides as output the signals sj(t) that are low-frequency PN observations for the respective PLL 112-j. For example, if assuming that the output signals from the PLL 112-j is the signalT;.(t) =where ω is e.g. a common IF band or passband frequency, t is time and φj(t) is a time-varying phase (i.e. PN), and where the reference signal is the signalTτref=with phase φref(t), the corresponding down-converted, filtered signal sj(t) can be written as the low-frequency PN observationssj(t) = ej(φ(t)−φ(t)),assuming that the reference signal rrey(t) has a same frequency as each of the signals τj(t).

[0026] Figure 1B schematically illustrates an alternative network 111 for generating (and distributing) the signals τj(t). In the network 111, instead of the multiple PLLs 112-1 to 112-M, a single (lower-frequency) PLL (or XO) 113 is used to feed a plurality of frequency multipliers 116-1 to 116-M, where the respective frequency multiplier 116-j generates an output signal τj(t) with a frequency that is a (integer) factor N times the frequency of the signal from the PLL 113. A reference PLL 115 (PLL-REF) is used to generate a (lower-frequency) signal that, after passing through a frequency multiplier 116, is the (higher-frequency) reference signal τref(t). Of course, in some examples, the PLL 115 may be such that it directly generates the (higher-frequency) signal τref(t), eliminating the need for the frequency multiplier 116. In the network 111, sources of uncorrelated PN (CPE) in various array branches may include contributions from PLL / XO frequency drift and frequency errors introduced by the frequency multipliers. In some examples, one or more of the plurality of frequency multipliers 116-1 to 116-M maybe implemented using one or more frequency mixers.

[0027] As envisaged herein, the plurality of oscillator signals τj(t) may thus be provided using a plurality of separate PLLs (as in 110, that may e.g. improve EVM), or e.g. by using a single PLL and / or XO and a plurality of frequency multipliers (as in 111, in which case CPE drift maybe corelated and similar for e.g. all antennas, resulting in improved beam forming performance).

[0028] In summary, the input to the envisaged CPE observer 120 may origin from either the PLL network 110 or the network 111, or e.g. from any other networksuitable for providing a plurality of LO / PLL signals such as the signals rj(t) and reference signal rrey(t). Further, in some examples involving the PLL (distribution) network 110 as envisaged herein, one of the PLLs 112-1 to 112-M, and / or one of the frequency multipliers 116-1 to 116-M, maybe used to generate the reference signal τref(t), while in other examples, an additional PLL and / or frequency multiplier may instead be used to generate the reference signal τref(t). Phrased differently, what is in here referred to as the e.g. plurality of PLLs 112-1 to 112-M and the PLL 114 are not necessarily disjoint sets, as the PLL 114 maybe one of the PLLs 112-1 to 112-M (or e.g. that the frequency multiplier 116 may be any one of the frequency multipliers 116-1 to 116-M). If using a separate PLL / frequency multiplier, such a PLL / frequency multiplier may have similar or same characteristics as those of the PLLs / frequency multipliers used to generate the signals τj(t). The PLLs as described herein maybe analog or digital, and one or more suitable digital-to-analog converters (DACs) and / or analog-to-digital converters (ADCs) maybe included in the processing chain as required to meet the specific architecture used.

[0029] After the CPE observer 120 has generated the plurality of down-converted / filtered signals s7(t), i.e. the plurality of low-frequency PN observations, these are used to ultimately perform CPE estimation / prediction and subsequently corrections in the system, as will now be described in more detail with reference also to Figure 2.

[0030] Figure 2 schematically illustrates various examples of an envisaged processing chain 200 to generate such CPE corrections. The output signals s7(t) from the CPE observer 120 are provided to a CPE detector / tracker / predictor 210. The CPE detector / tracker / predictor 210 performs detection, tracking and / or prediction of CPE for the various branches of the system based on the plurality of signals sj(t), and then generates and outputs one or more CPE estimations and / or predictions, that in Figure 2 are assumed to form part of (or at least be indicated by) CPE estimations and / or predictions signals 220-1 to 220-M (cj(t)), where cj(t) is such a signal for the respective j-th branch / PLL 112-j. In some examples, the signals cj(t) may be referred to as CPE correction signals, that may be used to perform the desired CPE corrections during (continued) use of the system. Each output of the CPE detector / tracker / predictor 210 maybe referred to as e.g. being for a particular PN-CPE correction path, or similar, and the CPE corrections may be applied as e.g. pre-and / or post-corrections as per the functional requirements at e.g. transmitter and / or receiver paths.

[0031] In some examples, depending on the PN-CPE detection, tracking and / or prediction implementation details (e.g. if analog or digital hardware, along with e.g. software, are needed), the processing chain 200 may further include one or more ADCs and / or DACs 230, 232 before and / or after the CPE detector / tracker / predictor 210 as required. Such converters may have lower sampling rates. A result of the CPE detector / tracker / predictor operation may for example be per clock-signal path or PLL-wise CPE estimations for current symbol and / or CPE predictions for future symbols. If phase tracking / prediction is performed in the analog domain, it may in some examples be possible to implement the CPE correction functionality as part of other functions already available, such as in an IQ correction module.

[0032] As envisaged herein, PN-CPE detection, tracking and / or prediction may be performed using established methods, such as based on known estimators and predictors like e.g. linear prediction techniques, Kalman filter-based methods, and similar. For example, the CPE detection, tracking and / or prediction maybe performed by observing, based on the plurality of down-converted signals s;(t), CPE for the plurality of oscillator signals rj(t) during a predefined time window, and by predicting, based on such observations, the CPE for the plurality of oscillator signals rj(t) during a later time window. In e.g. a tracking process, PN sample observations maybe performed over a predefined time window and used in the tracking (and / or prediction) functionality to generate e.g. current and / or future CPE predictions. With a non-zero observation-prediction-correlation loop delay, an envisaged prediction algorithm may generate CPE predictions for future time instances, for example for upcoming OFDM symbols (if relevant) based on previous PN observations. An accuracy of prediction over time maybe tuned via PN sample collection and prediction time window settings. Such a tuning may in some example be done offline based on system performance requirements. A frequency of such CPE tracking (i.e. observation - prediction - correction) may depend on the target application. For example, for illustrative purposes and in e.g. a reciprocity-based beamforming application, an envisaged tracking process may have two phases: i) model training and ii) phase prediction. In the model training phase, model parameters may beestimated from observed PN samples typically taken in advance of the beamforming scheduling interval. During the beamforming scheduling interval, CPE predictions maybe obtained from the tracking model using PN observations from previous slots or mini-slots. Considering e.g. OFDM-based 3GPP NR system design, such model training may include a PN observation window length on the order of a slot duration (e.g. for multiple NR-OFDM symbols duration). The tracking may generate CPE predictions for multiple OFDM symbols ahead of time. As envisaged herein, the described method may be extended trivially to work also for upcoming 6G and future systems.

[0033] The tracked and / or predicted CPE values cj(t) are then used to perform CPE corrections, e.g. pre- and / or post-correction in for example a transmitter (and / or receiver). Corrections maybe for respective carrier signals, PLLs, antennabranches and similar. CPE (pre-) corrections maybe applied e.g. to current and / or future symbols (valid for a time window), channel estimations, radar functions, and e.g. BF weight calculations within the predefined prediction window length. It is envisaged herein to perform the CPE corrections as part of a closed loop, including CPE observation, tracking / prediction and correction. CPE predictions maybe updated based on accumulated PN sample observations from a previous time window of observation. In particular, in case of e.g. OFDM, all LO / PLL frequencies are placed outside the data subcarrier, resulting in improved HW feasibility and observability.

[0034] For example, as envisaged herein, the one or more CPE corrections may be performed at least partially for pre- and / or post-correction of one or more user signals transmitted or received by the system. For example, if the system is OFDMbased, such pre- and / or post-correction maybe performed at baseband / IF / passband to OFDM symbols, at the transmitted and / or received side.

[0035] For example, as envisaged herein, the one or more user signals may be transmitted or received using an OFDM or OFDM-based scheme, and the CPE may be common for a full OFDM symbol. For example, the CPE maybe common for a full OFDM symbol that consists for multiple sub-carriers.

[0036] An example of a radio system in which the processing chain 200 may be included will now be exemplified in more detail with reference also to Figure 3.

[0037] Figure 3 schematically illustrates a radio system 300 in form of a multiantenna RF transceiver system. The system 300 includes a digital front end (DFE) block 310 that generates and / or receives user signals in a digital domain. The system 300 further includes an ADC / DAC block 320, a down- / up-conversion block 330, and an analog beamforming (ABF) block 340. Output from the ABF block 340 is provided to a set of transmit / receive modules 350-1 to 350-M each responsible for receiving and / or providing a signal from a respective one of multiple antenna elements 352-1 to 352-M. In particular, in the down- / up-conversion block 330, the signals are bandpass-filtered and amplified as needed, and the signal of each branch is mixed, using a respective one of a plurality of frequency mixers 332-1 to 332-M, with a corresponding oscillator signal τj(t). The oscillator signals τj(t) are, as described earlier herein, provided by the PLL generation and distribution block no (or e.g. 111).

[0038] As envisaged herein, the radio system 300 is complemented by the processing chain 200 for CPE corrections as described e.g. with reference to Figure 2. The signals τj(t), as well as the reference signal τref(t), are provided as input to the chain 200, and passes through the CPE observer 120 and the CPE tracker / predictor 210, optionally via the one or more ADC blocks 230, 232, to generate the PN-CPE corrections cj(t). The corrections cj(t) are used for pre- and / or post-correction, e.g. by multiplying the user signals from / to the DFE block 310 with the corrections cj(t) using a plurality of CPE correction multipliers 360-1 to 360-M as shown in Figure 3.

[0039] As envisaged herein, in addition to, or instead of, CPE correction of user signals such as shown in Figure 3, corrections such as cj(t) may also be used for functions like antenna array calibration (AC), updating of digital beamforming weights (after CSI-based beamforming weight calculations), e.g. in a periodical fashion to reduce time-varying beam squint effects on system e2e performance. Other envisaged example systems suitable to make use of the proposed solution includes any system requiring multiple LO / PLL signals to be generated and distributed, in particular when there is provided a plurality of local PLLs (e.g. to generate a plurality of local higher-frequency oscillator signals from one or more lower-frequency signals), such as often found in at least one of e.g. phase array antenna systems, radar systems, radio sensing systems for detection and ranging, MIMO systems, Massive or Ultra-MIMO systems, and similar.

[0040] Figure 4 schematically illustrates an example of a more general radio system 400 as envisaged herein. The system 400 implements a processing chain 420 that is configured to convert between a first signal 410 and a second signal 430. If the system 400 is a transmitter system, the signal 410 maybe an input signal (or a plurality of input signals, e.g. one or more data bits that are to be encoded and sent by the system 400). The signal 430 maybe an output signal, that is provided to e.g. an antenna array 440 that may form part of the system 400 and be used to transmit the output 430 wirelessly. The processing chain 420 may include any means necessary for e.g. coding, up-converting, filtering, amplifying, etc., the input 410, resulting in the generation of the output 430. The input 410 may for example correspond to baseband or IF or radio frequency, and the output 430 maybe passband frequency, or similar. One or both of the input 410 and / or output 430 may of course be multichannel signals, and the processing chain 420 may include any means necessary for e.g. splitting and / or combining signals into / from multiple signal paths.

[0041] In other examples, the system 400 may instead be a receiver system, in which the signal(s) 430 instead serves as the input and the signal(s) 410 as the output, and where the processing chain 420 is responsible for e.g. down-converting the input 430, e.g. from passband to baseband or IF frequency. Likewise, in yet other examples, the system 400 maybe a transceiver system, in which each of the signals 410 and 430 maybe both input and output, and wherein the processing chain 420 maybe configured for e.g. down- and up-conversion as required, e.g. in a time-wise alternating fashion. In any example of the system 400, it is envisaged that the processing chain 420 includes e.g. a plurality of frequency mixers (as used for up-and / or down-conversion) and / or other elements that requires provision of a plurality of (higher-frequency) oscillator signals rj(t) as explained herein. The oscillator signals rj(t) are generated (and distributed) using the (PLL / XO) network no (or e.g.111).

[0042] As described herein, in such a system 400, the proposed solution of CPE correction is provided by including (either as part of the system 400 itself or as an additional component interacting with the system 400) the CPEcorrection / processing chain 200, that receives the oscillator signals rj(t) as well as the reference signal τref(t) (using a plurality of PLLs, or e.g. a single PLL / XO and a plurality of frequency multipliers as in the network 111 instead of the network no). Asdescribed herein, the CPE correction / processing chain 200 generates, by observing CPE in the signals τj(t) by down-converting (i.e. by mixing and low-pass filtering) these signals, and by performing CPE detection, tracking and / or predictions based thereon, CPE corrections 220 (i.e. cj(t)). The CPE corrections 220 are then applied as e.g. pre- and / or post-corrections, for example as part of the processing chain 420 as illustrated by the dashed blocks 424 and 426.

[0043] The system 400 may for example be a MIMO (-based) system, such as a MIMO system, Massive MIMO system, Ultra-MIMO system, and similar. The input 410 maybe for multiple users and / or for multiple subcarriers, and the output 430 maybe for multiple antenna elements or for groups of antenna elements, and similar. The system 400 may for example be any system that has multiple antennas and therefore multiple processing paths, wherein the multiple oscillator signals τj(t) are required as reference / clock signals for e.g. mixers in each path for e.g. down- and / or up-conversion after and / or receiving the signal(s). Other examples include the system 400 being a radar system, such as e.g. a MIMO-based radar system, a radio sensing system for detection and ranging, or any other system (such as e.g. a Large Antenna Array System) wherein there is a need to generate and distribute a plurality of oscillator signals, and where PN is present due e.g. the PLLs and / or XOs and frequency multipliers used to generate and distribute such signals. If a radar system, the inherent phase-variations due to the PLLs maybe corrected for to achieve a better quality beamforming, as the phase error (e.g. CPE) over the array is lowered. In turn, this may serve to provide higher beam quality and lower sidelobes, which may in some situations be even more important for radar systems compared to communication systems.

[0044] For example, if the operation of the system 400 includes radio sensing for detection and ranging, e.g. if the system is a radar system, the one or more CPE corrections maybe for channel estimations used in various sensing and / or radar applications. In other examples, the one or more CPE corrections may instead, or in addition, be at least partially for at least one of antenna array calibration (AC, such as for AC measurements) and beamforming (BF) weight calculation / updating.

[0045] In particular, the envisaged radio system 400 thus includes at least a plurality of PLLs and / or frequency multipliers for generating the plurality of signals τj(t) and rre^(t); the CPE observer 120 (as part of the chain 200); the CPE detector,tracker and / or predictor 210 (also as part of the chain 200), and a CPE corrector (as implemented by e.g. one or more of the blocks 424 and 426) for performing the CPE (pre- and / or post-) corrections as part of the processing chain 420.

[0046] Systems as envisaged herein, such as 300 and / or 400, may thus be performed to perform a method as will now be described in more detail with reference also to Figure 5.

[0047] Figure 5 schematically illustrates an example flowchart of a method 500 as performed in a radio system as envisaged herein, such as in the system 300, 400 or alternative examples thereof.

[0048] An operation S510 of the method 500 includes generating the plurality of mixed signals by mixing the reference signal rτref(t) with each of the oscillator signals τj(t).

[0049] An operation S520 of the method 500 includes generating the plurality of down-converted signals, such as sj(t), by low-pass filtering the plurality of mixed signals resulting from the operation S510.

[0050] An operation S530 of the method 500 includes performing the CPE detection, tracking and / or prediction for the plurality of oscillator signals based on the plurality of down-converted signals, such as sj(t) and resulting from the operation S520.

[0051] An operation S540 of the method 500 includes performing the one or more CPE corrections during operation of the radio system, using the outcome of the CPE detection, tracking and / or correction of operation S530.

[0052] An optional operation S505 of the method 500 may include providing the plurality of oscillator signals, e.g. by using a plurality of separate PLLs (such as in the network 110) or by using a single PLL / XO and a plurality of frequency multipliers (such as in the network 111).

[0053] Figure 6A schematically illustrates, in terms of a number of functional units, the components of a radio system 600 according to the present disclosure (such as the system 300 and / or 400). The radio system 600 may form part of a communications network, such as a telecommunications network, form part of a radar network, or similar, and be configured to perform e.g. one or more of thevarious examples of the method 500 of CPE correction as described earlier herein. The radio system 600 includes processing circuitry 610. The processing circuitry 610 is provided using any combination of one or more of a suitable central processing unit (CPU), multiprocessor, microcontroller, digital signal processor (DSP), etc., capable of executing software instructions stored in a computer program product 710 (see Figure 7 and the description thereof), e.g. in form of a storage medium / memory 620 that may also form part of the radio system 600. The processing circuitry 610 may further be provided as at least one application specific integrated circuit (ASIC), or field-programmable gate array (FPGA).

[0054] Particularly, the processing circuitry 610 is configured to cause the radio system 600 to perform a set of operations, or steps, as disclosed above e.g. when describing the method 500 illustrated in Figure 5. For example, the storage medium 620 may store a set of operations, and the processing circuitry 610 maybe configured to retrieve the set of operations from the storage medium 620 to cause the radio system 600 to perform the set of operations. The set of operations maybe provided as a set of executable instructions. Thus, the processing circuitry 610 is thereby arranged to execute examples of the method 500 as disclosed herein e.g. with reference to Figure 5.

[0055] The storage medium 620 may also include persistent storage, which, for example, can be any single or combination of magnetic memory, optical memory, solid state memory or even remotely mounted memory.

[0056] The radio system 600 may further include a communications interface 630 for communications with other entities, functions, nodes, and devices, such as e.g. those of a telecommunications network or any other network associated with the task to be solved. For example, the communications interface 630 may allow the radio system 600 to communicate with e.g. other entities, such as with other systems, network functions (NFs) or nodes of a telecommunications network, or e.g. with one or more internal operational modules within the same system, etc. As such, the communication interface 630 may include one or more transmitters and receivers, including analogue and / or digital components.

[0057] The processing circuitry 610 controls the general operation of the radio system 600 e.g. by sending data and control signals to the communications interface 630 and the storage medium 620, by receiving data and reports from thecommunications interface 630, and by retrieving data and instructions from the storage medium 620. Other components, as well as their related functionality, of the radio system 600 are omitted in order not to obscure the concepts presented herein. As generally used herein, a “radio system” may operate e.g. at centimeter-wave, millimeter-wave, sub-THz and / or THz frequencies.

[0058] Figure 6B schematically illustrates, in terms of a number of functional modules 610a, 610b, 610c and 610d, the components of a radio system 600 according to one or more examples of the present disclosure. The radio system 600 includes at least a first module 610a configured to perform operation S510 of the method 500 described with reference to Figure 5. The module 610a maybe referred to as a “mixing module”, “a mixer”, or similar. The radio system 600 also includes a second module 610b configured to perform operation S620. The module 610b maybe referred to as a “filtering module”, a “filter module”, or similar. The modules 610a and 610b may jointly be referred to as a “down-converting module”, a “downconverter”, or similar. The radio system 600 further includes a third module 610c configured to perform operation S530 of the method 500. The module 610c maybe referred to as a “CPE detection, tracking and / or prediction module”, a “CPE detector, tracker and / or predictor”, and similar. The radio system 600 further includes a fourth module 6iod configured to perform operation S540 of the method 500. The module 6iod maybe referred to as a “CPE correction module”, a “CPE corrector”, or similar. In other examples, two or more of the modules 610a to 6iod may instead be provided as part of a single module. For example, modules 610a and 610b maybe combined into a “CPE observing module”, a ”CPE observer module”, or similar, and implement a “CPE observing functionality”. The module 610c may implement a “CPE detection, tracking and / or prediction functionality”, and the module 6iod may implement a “CPE correction functionality”, or similar. The radio system 600 may also include one or more optional functional modules (illustrated by the dashed box 6ioe), such as for performing e.g. the optional operation S505 of the method, and / or for performing any other functionality of the radio system 600 not described herein, such as for beamforming, AC, and similar.

[0059] In general terms, each functional module 610a-e may be implemented in hardware or in software. Preferably, one or more or all functional modules 610a-e maybe implemented by the processing circuitry 610, possibly in cooperation with thecommunications interface 630 and / or the storage medium 620. The processing circuitry 610 may thus be arranged to from the storage medium 620 fetch instructions as provided by a functional module 610a-e, and to execute these instructions and thereby perform any operations of any example of the method 500 performed by / in the radio system 600 as disclosed herein.

[0060] Envisaged herein is of course also to provide e.g. only the functionality required to implement the CPE correction chain 200 on its own, e.g. not as part of a full radio system such as 300 or 400, but as a separate entity that may interact with such a radio system to still perform the desired CPE corrections. For example, it is envisaged herein to provide a CPE correction generator including at least the CPE observer 120 and the CPE detector / tracker / predictor 220 described with reference to e.g. Figures 1A, 1B and 2, configured to generate the corrections cj(t) based on the input r7(t) and rrey(t). In other examples, such a CPE correction generator may also include e.g. the (PLL / XO) network no (or 111), and be able to provide the oscillator signals r7(t) to the system.

[0061] Figure 7 schematically illustrates an example computer program product 710, including computer readable means 730. On the computer readable means 730, a computer program 720 can be stored, which computer program 720 can cause the processing circuitry 610 and thereto operatively coupled entities and devices, such as the communication interface 630 and the storage medium 620, of the radio system 600 to execute one or more of the examples of a method 500 as described with reference to e.g. Figure 5. The computer program 720 and / or computer program product 710 may thus provide means for performing any operations of any method 500 performed by the radio system 600 as disclosed herein.

[0062] In the examples of Figure 7, the computer program product 710 and computer readable means 730 are illustrated as an optical disc, such as a CD (compact disc) or a DVD (digital versatile disc) or a Blu-Ray disc. The computer program product 710 and computer readable means 730 could also be embodied as a memory, such as a random-access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM), or an electrically erasable programmable read-only memory (EEPROM) and more particularly as a non-volatile storage medium of a device in an external memory such as a USB (Universal Serial Bus) memory or a Flash memory, such as a compact Flash memory. Thus, while thecomputer program 720 and data are here schematically shown as a track on the depicted optical disk, the computer program 720 and / or data maybe stored in any way which is suitable for the computer program product 710 and computer readable means 730. As envisaged herein, a computer-readable storage medium may also be transitory and e.g. correspond to a signal (electrical, optical, mechanical, or similar) present on e.g. a communication link, wire, or similar means of signal transferring.

[0063] Also envisaged herein is of course to also provide a computer program (product) that causes the envisaged CPE correction generator to perform as desired, i.e. such that this standalone entity is capable of implementing the CPE observing and CPE detection / tracking / prediction functionality as described herein, e.g. corresponding to at least operations S510, S520 and S530 (an optionally also operation S505) of the method 500.

[0064] The solutions as envisaged and described herein may also be applicable in for example an Open Radio Access Network (O-RAN) solution / implementation. For example, the PN-CPE tracking and / or prediction maybe done in an O-RAN Radio Unit (O-RU) alone, in which case no impact on the O-RAN Fronthaul (O-RAN FH) is necessitated, making such an implementation a preferred choice in some situations. In other examples, joint O-RU and O-RAN Distribution Unit (O-DU) implementations are also envisaged. For example, PN-CPE tracking and / or prediction functionality may be implemented in O-DU, in case which the O-FH specification may be enhanced to carry down-converted PN samples from O-RU to O-DU as part of a Control / User ( / Synchronization) (C / U or C / U / S)-plane enhanced frame structure. In the downlink, PN-CPE predictions maybe carried as part of e.g. enhanced C-plane information.

[0065] In summary of all of the above, it is in the present disclosure provided a solution for how to correct for CPE in systems having a plurality of PLLs and / or frequency multipliers for generation of a plurality of oscillator signals (such as carriers or clock signals), as used for e.g. up- and / or down-conversion in the system, in a way that relaxes the requirements on individual PLLs in terms of PN as well as for signal / LO distribution architectures, especially for centimeter-wave, millimeterwave, sub-THz and THz implementations. The envisaged solution thus provides benefits especially in e.g. 5G NR, 6G and 6G+ systems, wherein such frequencies are to be used and where there is a need to e.g. provide a separate PLL for eachchannel / antenna element, and where PN and CPE become an issue if left unattended. The proposed solution helps to lower R& D cost-time and hence product cost and time-to-market, and is frequency-scalable and may ensure e.g. beamforming weightstability over time, by preventing time-varying beam squint and sidelobe effects during e.g. SU-MIMO and MU-MIMO scheduling windows in e.g. FR3 and FR2 HB systems. The proposed solution ensures reciprocal phase-coherency over time, enabling high-performance SU-MIMO and MU-MIMO for reciprocity-based beamforming methods, and simplifies e.g. BB data path PN handling algorithm complexity levels and results in improved e2e performance. Further, ISAC-radar, sensing and localization functions (including impacts on RIBS functionality) may also be improved in relevant systems implementing the proposed solution.

[0066] Although features and elements may be described above in particular combinations, each feature or element may be used alone without the other features and elements or in various combinations with or without other features and elements. Additionally, variations to the disclosed embodiments may be understood and effected by the skilled person in practicing the claimed invention as defined by the appended patent claims, from a study of the drawings, the disclosure, and the appended claims themselves. In the claims, the words “comprising” and “including” does not exclude other elements, and the indefinite article “a” or “an” does not exclude a plurality. The mere fact that certain features are recited in mutually different dependent claims does not indicate that a combination of these features cannot be used to advantage.

Claims

CLAIMS1. A method (500) of common phase error, CPE correction in a radio system, wherein the system comprises a plurality of phase-locked loops, PLLs and / or frequency multipliers configured for generating a plurality of oscillator signals (τj(t)), and wherein the method comprises:- generating (S510, S520), during operation of the radio system, a plurality of down-converted signals (s7(t)) by mixing (S510) a reference signal (rre^(t)) with each of the oscillator signals followed by low-pass filtering (S520) each of the mixed signals;- performing (S530) common phase error, CPE detection, tracking and / or prediction for the plurality of oscillator signals based on the plurality of down-converted signals, and- performing (S540) one or more CPE corrections during operation of the radio system using an outcome of said CPE detection, tracking and / or prediction.

2. The method of claim 1, wherein the method comprises providing (S505) the plurality of oscillator signals by using a plurality of separate PLLs.

3. The method of claim 1, wherein the method comprises providing (S505) each of the plurality of oscillator signals by using a single PLL or crystal oscillator, XO, and a plurality of frequency multipliers.

4. The method of claim 3, wherein the plurality of frequency multipliers are implemented using a plurality of frequency mixers.

5. The method of any one of claims 1 to 4, wherein the method comprises providing the reference signal for the down-conversion by i) one of the plurality of PLLs and / or frequency multipliers, or by ii) an additional PLL or frequency multiplier separate from the plurality of PLLs and / or frequency multipliers.

6. The method of any one of the preceding claims, wherein the reference signal has a same frequency as each of the plurality of oscillator signals.

7. The method of any one of the preceding claims, wherein the one or more CPE corrections are at least partially for pre- and / or post-correction of one or more user signals transmitted or received by the system.

8. The method of claim 7, wherein the one or more user signals are transmitted or received using an orthogonal frequency-division multiplexing, ODFM, scheme, and wherein the CPE is common for a full OFDM symbol.

9. The method of any one of the preceding claims, wherein said operation of the system comprises radio sensing for detection and ranging.

10. The method of any one of the preceding claims, wherein the one or more CPE corrections are at least partially for at least one of antenna array calibration and beamforming weight calculation / updating.

11. The method of any one of the preceding claims, wherein said CPE tracking and / or prediction is performed by observing, based on the plurality of down-converted signals, CPE for the plurality of oscillator signals during a predefined time window, and predicting, based on said observations, CPE for the plurality of oscillator signals during a later time window.

12. The method of any one of the preceding claims, wherein as part of said operation, each of the plurality of oscillator signals are used as input to a plurality of frequency mixers for down- and / or up-converting between any two of a passband, baseband, intermediate frequency, IF, band and RF band.

13. The method of any one of the preceding claims, wherein the radio system is at least one of a multiple-input multiple-output, MIMO, system, Massive MIMO system, Ultra-MIMO system, Large Antenna Array System, and a radar system.

14. A radio system (300, 400, 600), comprising:- a plurality of phase-locked loops, PLLs, (112-1 to 112-M) and / or frequency multipliers (116-1 to 116-M) configured for generating a plurality of oscillator signals (τj(t));- a common phase error, CPE, observer (120), configured to:—generate, during operation of the radio system, a plurality of down- converted signals (sj(t)), by mixing a reference signal (rre^(t)) with each of the oscillator signals followed by low-pass filtering (126-1 to 126-M) each of the mixed signals;- a CPE detector, tracker and / or predictor (210), configured to:— obtain the plurality of down-converted signals generated by the CPE observer;— perform, based on the plurality of down-converted signals, common phase error, CPE, detection, tracking and / or prediction for the plurality of oscillator signals, and— generate an outcome of said CPE detection, tracking and / or prediction as one or more CPE estimations and / or predictions (c;(t)) for the plurality of oscillator signals,and- a CPE corrector (424, 426), configured to obtain the one or more CPE estimations and / or predictions generated by the CPE detector, tracker and / or predictor, and to use the one or more CPE estimations and / or predictions to perform one or more CPE corrections during operation of the radio system.

15. The system of claim 14, further configured to perform the method (500) of any one of claims 2 to 13.

16. The system of claim 14 or 15, wherein the system is at least one of a multipleinput multiple-output, MIMO, system, a Massive MIMO system, an Ultra-MIMO system, a Large Antenna Array System, and a radar system.

17. The system of any one of claims 14 to 16, wherein the radio system comprises processing circuitry (610) and a memory (620) storing instructions (720), wherein the instructions are such that they, when executed by the processing circuitry, cause the radio system to perform at least some functionality of at least one of the CPE observer, the CPE detector, tracker and / or predictor, and the CPE corrector.

18. A computer program (720) comprising instructions that, when executed by processing circuitry (610) of a radio system (600), cause the radio system to:- implement a common phase error, CPE, observing functionality, comprising to mix a reference signal (rre^(t)) with each of a plurality of oscillator signals (τj(t)), followed by low-pass filtering each of the mixed signals, to generate a plurality of down-converted signals (s7(t));- implement a CPE detection, tracking and / or prediction functionality, comprising to obtain the plurality of down-converted signals generated by the CPE observing functionality; to perform, based on the plurality of down-converted signals, CPE detection, tracking and / or prediction for the plurality of oscillator signals, and to generate an outcome of said CPE detection, tracking and / or prediction as one or more CPE estimations and / or predictions (c7(t)) for the plurality of oscillator signals, and- implement a CPE correction functionality, comprising to obtain the one or more CPE estimations and / or predictions generated by the CPE detection, tracking and / or prediction functionality, and to use the one or more CPE estimations and / or predictions to perform one or more CPE corrections during operation of the radio system.

19. The computer program of claim 18, wherein the instructions are further such that they, when executed by the processing circuitry, cause the radio system to perform the method (500) of any one of claims 2 to 13.

20. A computer program product (710) comprising a computer-readable storage medium (730) storing the computer program (720) of claim 18 or 19.

Images