WO2023179837A1

WO2023179837A1 - Synchronization for d-mimo

Info

Publication number: WO2023179837A1
Application number: PCT/EP2022/057293
Authority: WO
Inventors: Henrik Sjöland; Harsh TATARIA
Original assignee: Telefonaktiebolaget Lm Ericsson (Publ)
Priority date: 2022-03-21
Filing date: 2022-03-21
Publication date: 2023-09-28

Abstract

A method is disclosed for over-the-air frequency synchronization between a plurality of access points of a distributed multiple-input multiple-output (D-MIMO) system. The method comprises determining a group of one or more transmitting access points, and (for each of the transmitting access points) scheduling a plurality of reference signal transmission occasions, wherein a time between transmission occasions is dynamically set based on expected frequency error, and instructing the transmitting access point to transmit a reference signal for frequency synchronization in each of the scheduled transmission occasions. The method may further comprise (for each scheduled transmission occasion) instructing access points, that are not instructed to transmit in the scheduled transmission occasion, to perform phase measurements during the transmission occasion. The method may be performed by a central control node of the D-MIMO system. A corresponding method for performance by an access point of the D-MIMO system is also disclosed, as well as corresponding apparatuses, computer program product, central control node, access point, and D-MIMO system.

Description

SYNCHRONIZATION FOR D-MIMO

TECHNICAL FIELD

The present disclosure relates generally to the field of wireless communication. More particularly, it relates to synchronization between access points for a distributed multiple-input multiple-output (D-MIMO) system.

BACKGROUND

A distributed multiple-input multiple-output (D-MIMO) system typically comprises a central control node and a plurality of access points controlled by the central control node, wherein each access point may have a single antenna, or relatively few (e.g., two) antennas.

In typical D-MIMO system deployments, the access points are spread out within an area to be served by the D-MIMO system. Thereby, the D-MIMO system as a whole enables (massive) MIMO operation, wherein several access points are active for communication with a user device.

For transmission by a D-MIMO system, the signal from each participating access point should preferably arrive - at the receiver of the user device - substantially in phase with the corresponding signals transmitted from the other participating access points (i.e., there should preferably be phase alignment between the received signals). The process of achieving phase alignment may be seen as a form of synchronization between access points of a D-MIMO system.

When phase alignment can be achieved, the received signals will combine constructively at the receiver for phase-coherent reception. When the signals are not properly phase aligned at the receiver, destructive combination may occur, and the received signal might be too weak to enable retrieval of the transmitted information.

In one approach for phase alignment in D-MIMO systems, phase measurements are performed by the access points on reference signaling transmitted by the user device, and the result of the phase measurements is used to determine the phases to be applied for transmission from the access points. However, this approach often provides inferior phase alignment at the receiver of the user device. This can typically lead to one or more other problems related to communication performance (e.g., inferior reception, decreased throughput, increased signaling overhead, increased power consumption, etc.).

Therefore, there is a need for alternative approaches to synchronization between access points of a D-MIMO system.

SUMMARY

It should be emphasized that the term "comprises/comprising" (replaceable by "includes/including") when used in this specification is taken to specify the presence of stated features, integers, steps, or components, but does not preclude the presence or addition of one or more other features, integers, steps, components, or groups thereof. As used herein, the singular forms "a", "an" and "the" are intended to include the plural forms as well, unless the context clearly indicates otherwise.

Generally, when an arrangement is referred to herein, it is to be understood as a physical product; e.g., an apparatus. The physical product may comprise one or more parts, such as controlling circuitry in the form of one or more controllers, one or more processors, or the like.

It is an object of some embodiments to solve or mitigate, alleviate, or eliminate at least some of the above or other disadvantages.

A first aspect is a method for over-the-air frequency synchronization between a plurality of access points of a distributed multiple-input multiple-output (D-MIMO) system. The method comprises determining a group of one or more transmitting access points. The method also comprises (for each of the transmitting access points) scheduling a plurality of reference signal transmission occasions, wherein a time between transmission occasions is dynamically set based on expected frequency error, and instructing the transmitting access point to transmit a reference signal for frequency synchronization in each of the scheduled transmission occasions.

In some embodiments, the method further comprises (for each scheduled transmission occasion) instructing access points that are not instructed to transmit in the scheduled transmission occasion to perform phase measurements during the transmission occasion. In some embodiments, the group of one or more transmitting access points comprises the plurality of access points, or a sub-set of the plurality of access points.

In some embodiments, the method further comprises (for an access point which has an expected frequency error that exceeds a frequency error threshold value) excluding that access point from the group of one or more transmitting access points.

In some embodiments, a number of transmission occasions for a first transmitting access point is higher than a number of transmission occasions for a second transmitting access point when a distance from the first transmitting access point to the excluded access point is shorter than a distance from the second transmitting access point to the excluded access point.

In some embodiments, each transmission occasion is scheduled for one or more of the transmitting access points.

In some embodiments, the method further comprises assigning different reference signals to different transmitting access points with one or more coinciding transmission occasions.

In some embodiments, the reference signal is a tone, or a modulated signal.

In some embodiments, the method further comprises indicating one or more parameters of the reference signal to the transmitting access point and/or to access points configured to perform phase measurements on the reference signal.

In some embodiments, a first time between transmission occasions based on a first expected frequency error is shorter than a second time between transmission occasions based on a second expected frequency error when the first expected frequency error is larger than the second expected frequency error.

In some embodiments, the time between transmission occasions is shorter than half a period of a largest expected frequency error between the transmitting access point and any access point in the plurality of access points.

In some embodiments, the expected frequency error comprises a frequency error determined during previous frequency synchronization event. In some embodiments, the expected frequency error comprises an individual frequency error for the transmitting access point.

In some embodiments, the expected frequency error increases with increasing time since previous frequency synchronization event.

In some embodiments, the method further comprises setting the expected frequency error to a maximum value for an initial frequency synchronization event.

In some embodiments, the scheduling and instructing steps are performed in response to triggering of a current frequency synchronization event.

In some embodiments, the method further comprises triggering the current frequency synchronization event in response to one or more of: an access point of the plurality of access points being initialized, a change of user device location, determining that a predetermined time has elapsed since previous frequency synchronization event, and detecting a communication performance deterioration for one or more of the plurality of access points.

In some embodiments, the plurality of access points comprises one or more of: access points located in a same geographical area, access points within a same radio signal vicinity, and access points that provide for communication with a same user device location.

In some embodiments, the method further comprises receiving (from one or more of the plurality of access points) a respective measurement report, wherein each measurement report is indicative of an estimated frequency error for the corresponding access point.

In some embodiments, the method further comprises determining a respective frequency correction for one or more of the plurality of access points, wherein the frequency correction is based on the estimated frequency error for the corresponding access point.

In some embodiments, determining the respective frequency correction comprises application of bias for maintenance of an average frequency control setting among the plurality of access points.

In some embodiments, the method further comprises conveying an indication of the respective frequency correction to the corresponding access point. In some embodiments, the method of the first aspect is performed by a central control node of the D-MIMO system.

A second aspect is a method for over-the-air frequency synchronization of an access point in relation to a plurality of access points of a distributed multiple-input multiple-output (D-MIMO) system. The method comprises (for each of a first plurality of reference signal transmission occasions, wherein a time between transmission occasions is dynamically set based on expected frequency error) performing phase measurements for frequency synchronization, wherein the phase measurements are relative a transmitting access point during the transmission occasion.

In some embodiments, the method further comprises receiving an instruction to transmit a reference signal for frequency synchronization in each of a second plurality of scheduled transmission occasions, wherein a time between transmission occasions is dynamically set based on expected frequency error, and performing reference signal transmission accordingly.

In some embodiments, the phase measurements are performed for all reference signal transmission occasions except those comprised in the second plurality of scheduled transmission occasions.

In some embodiments, the method further comprises receiving an instruction to perform the phase measurements during each transmission occasion of the first plurality of reference signal transmission occasions.

In some embodiments, the method further comprises determining an estimated frequency error for the access point based on the phase measurements.

In some embodiments, the estimated frequency error for the access point is determined as an average of respective frequency offsets relative transmitting access points of a collection of transmitting access points.

In some embodiments, the method further comprises, for a transmitting access point corresponding to an unreliable frequency offset: excluding that transmitting access point from the collection of transmitting access points, or weighting the frequency offset relative that transmitting access point before determining the average. In some embodiments, determining the respective frequency offset comprises determining a gradient of an affine function fitted to a phase-time representation of the phase measurements.

In some embodiments, determining the respective frequency offset comprises determining an average, or median, of phase differences between time-adjacent phase measurements.

In some embodiments, the method further comprises determining a frequency correction based on the estimated frequency error.

In some embodiments, the method further comprises providing (to a central control node of the D-MIMO system) a measurement report, wherein the measurement report is indicative of the estimated frequency error.

In some embodiments, the method further comprises receiving an indication of a frequency correction from the central control node, wherein the frequency correction is based on the estimated frequency error for the corresponding access point.

In some embodiments, the method further comprises applying the frequency correction.

In some embodiments, the frequency correction is applied for one or more of: a reference oscillator, a frequency synthesizer, and a baseband signal rotation.

In some embodiments, the method of the second aspect is performed by the access point.

A third aspect is a computer program product comprising a non-transitory computer readable medium, having thereon a computer program comprising program instructions. The computer program is loadable into a data processing unit and configured to cause execution of the method according to any of the first and second aspects when the computer program is run by the data processing unit.

A fourth aspect is an apparatus for over-the-air frequency synchronization between a plurality of access points of a distributed multiple-input multiple-output (D-MIMO) system. The apparatus comprises controlling circuitry configured to cause determination of a group of one or more transmitting access points, and (for each of the transmitting access points) scheduling of a plurality of reference signal transmission occasions, wherein a time between transmission occasions is dynamically set based on expected frequency error, and instruction of the transmitting access point to transmit a reference signal for frequency synchronization in each of the scheduled transmission occasions.

A fifth aspect is a central control node for a distributed multiple-input multiple-output (D- MIMO) system, wherein the central control node comprises the apparatus of the fourth aspect.

A sixth aspect is an apparatus for over-the-air frequency synchronization of an access point in relation to a plurality of access points of a distributed multiple-input multiple-output (D-MIMO) system. The apparatus comprises controlling circuitry configured to cause (for each of a first plurality of reference signal transmission occasions, wherein a time between transmission occasions is dynamically set based on expected frequency error) performance of phase measurements for frequency synchronization, wherein the phase measurements are relative a transmitting access point during the transmission occasion.

A seventh aspect is an access point for a distributed multiple-input multiple-output (D-MIMO) system, wherein the access point comprises the apparatus of any of the sixth aspect.

An eighth aspect is a distributed multiple-input multiple-output (D-MIMO) system comprising a central control node according to the fifth aspect and a plurality of access points according to the seventh aspect.

In some embodiments, any of the above aspects may additionally have features identical with or corresponding to any of the various features as explained above for any of the other aspects.

An advantage of some embodiments is that alternative approaches are provided for synchronization between access points of a D-MIMO system.

An advantage of some embodiments is that frequency synchronization is provided (to enable proper phase alignment) between access points of the D-MIMO system. For example, the frequency offsets between access points of the D-MIMO system may be decreased compared to other approaches.

An advantage of some embodiments is that no wired connections are required between access points of the D-MIMO system (or between access points and the central control node) to achieve frequency synchronization. An advantage of some embodiments is that phase measurements performed by the access points on reference signaling transmitted by the user device, can be reliably used to determine the phases to be applied for transmission from the access points, even when some time has passed between the phase measurements and the transmission from the access points. For example, if access points are not frequency synchronized, time passing between the phase measurements and the transmission from the access points results in that the phase references of the access points drift relative each other, which typically results in inferior phase alignment at the receiver of the user device. Proper frequency synchronization mitigates such problems.

An advantage of some embodiments is that drift of an average frequency control setting of the access points of the D-MIMO system is mitigated (i.e., no additional frequency drift is introduced by the frequency synchronization). For example, a bias for maintenance of an average frequency control setting among the plurality of access points may be applied in the frequency correction determination to mitigate drift of the average frequency control setting.

An advantage of some embodiments is that channel changes (e.g., due to access point movements, and/or due to movements of obstructing and/or reflecting objects) occurring between measurements during frequency synchronization and the communication transmission from the access points can be mitigated. For example, removing deviating measurements from consideration in the frequency synchronization might decrease the impact of channel changes.

An advantage of some embodiments is that the reference signaling for frequency synchronization between access points of the D-MIMO system is dynamically scheduled. For example, an access point having an expected frequency error which is relatively low, may be scheduled for reference signal transmission more seldom than an access point having an expected frequency error which is relatively high. This may enable proper frequency synchronization while keeping the amount of overhead and/or power consumption due to reference signaling and measurements at an acceptable level.

An advantage is that a burst of densely spaced reference signal transmissions from one transmitting access point may be triggered to capture a relatively large expected frequency error (e.g., that exceeds a frequency error threshold value). For example, for an initial frequency synchronization event (e.g., at system startup) it may be assumed that the frequency errors of access points are rather large, and burst reference signal transmissions may be beneficial.

An advantage of some embodiments is that the frequency correction can be determined by the central control node and/or by each access point, which may provide for flexible frequency synchronization. For example, a coarse frequency correction may be determined by the central control node, and a refined frequency correction may be determined by each access point. Alternatively or additionally, the central control node may determine frequency corrections repeatedly (e.g., periodically), and each access point may determine respective frequency corrections in the duration between frequency correction determinations by the central control node. Thus, the amount of overhead due to frequency synchronization may be kept at an acceptable level.

An advantage of some embodiments is that application in relatively large D-MIMO systems is facilitated. For example, when the frequency correction is determined by each access point, there is no delay related to transmission of measurement reports to the central control node and corresponding reception of a frequency correction instruction from the central control node.

An advantage of some embodiments is that there are various possibilities for application of the frequency correction, which may provide for flexible frequency synchronization. For example, a coarse frequency correction may be applied in a radio frequency (RF) domain, and a refined frequency correction may be applied in a baseband domain. Alternatively or additionally, the central control node may determine a frequency correction applied on the radio frequency domain, and each access point may determine respective frequency corrections applied in the baseband domain.

An advantage of some embodiments is that reference signals forfrequency synchronization may be designed to satisfy spectral regulations. For example, reference signals for frequency synchronization may be modulated if using a tone as reference signal would violate spectral regulations (e.g., by having unallowably high spectral density).

An advantage of some embodiments is that two or more reference signals for frequency synchronization may be transmitted simultaneously by corresponding two or more access points. For example, different simultaneously transmitting access points may apply different reference signals (e.g., different tones and/or different modulations) to enable discrimination. Thus, the amount of overhead (in the time domain) due to reference signaling may be kept at an acceptable level.

BRIEF DESCRIPTION OF THE DRAWINGS

Further objects, features and advantages will appear from the following detailed description of embodiments, with reference being made to the accompanying drawings. The drawings are not necessarily to scale, emphasis instead being placed upon illustrating the example embodiments.

Figure 1 is a schematic drawing illustrating an example D-MIMO system according to some embodiments;

Figure 2 is a signaling diagram illustrating example signaling according to some embodiments;

Figure 3 is a flowchart illustrating example method steps according to some embodiments;

Figure 4 is a flowchart illustrating example method steps according to some embodiments;

Figure 5A is a schematic plot illustrating example principles according to some embodiments;

Figure 5B is a schematic plot illustrating example principles according to some embodiments;

Figure 6 is a schematic block diagram illustrating an example apparatus according to some embodiments;

Figure 7 is a schematic block diagram illustrating an example apparatus according to some embodiments; and

Figure 8 is a schematic drawing illustrating an example computer readable medium according to some embodiments.

DETAILED DESCRIPTION

As already mentioned above, it should be emphasized that the term "comprises/comprising" (replaceable by "includes/including") when used in this specification is taken to specify the presence of stated features, integers, steps, or components, but does not preclude the presence or addition of one or more other features, integers, steps, components, or groups thereof. As used herein, the singular forms "a", "an" and "the" are intended to include the plural forms as well, unless the context clearly indicates otherwise.

Embodiments of the present disclosure will be described and exemplified more fully hereinafter with reference to the accompanying drawings. The solutions disclosed herein can, however, be realized in many different forms and should not be construed as being limited to the embodiments set forth herein.

Figure 1 schematically illustrates an example D-MIMO system 100 according to some embodiments. The D-MIMO system 100 comprises a central control node (CCN) 110 and a plurality of access points (API, AP2, AP3, AP4, AP5, AP6) 121, 122, 123, 124, 125, 126. The access points 121, 122, 123, 124, 125, 126 are controlled by the central control node 110.

Control communication between the central control node and the plurality of access points may be based on wireless, or wired, communication. It may be noted that avoiding wired interfaces between the access points, and/or between access points and the central control node, allows for increased flexibility, and/or scalability, of the D-MIMO deployment.

In the example D-MIMO system 100, the control communication between the central control node and the plurality of access points relies on wireless communication between the central control node and the plurality of access points, as illustrated by 140. Furthermore, the access points have no wired interface between them.

Each access point 121, 122, 123, 124, 125, 126 may have a single antenna, or relatively few (e.g., two) antennas, while the D-MIMO system 100 enables (massive) MIMO operation, wherein several access points are active for communication with a user device (e.g., a user equipment, UE) 130. For example, a plurality 120 of access points, comprising the access points 122, 123, 124, 125 that are closest to the user device 130 (e.g., in a geographical sense and/or in a radio propagation sense), may be used for MIMO communication with the user device 130.

As already mentioned, it is important to strive for phase alignment of signals transmitted from access points of a D-MIMO system. The possibility to achieve proper phase alignment may, in some situations, depend on whether or not the access points are frequency synchronized.

For example, phase alignment in D-MIMO systems may comprise performing phase measurements by the access points on reference signaling transmitted by the user device, and using the result of the phase measurements to determine the phases to be applied for transmission from the access points.

This approach to phase alignment is particularly relevant for D-MIMO systems with time division duplex (TDD) operation. Assuming channel reciprocity (i.e., that the uplink, UL, channel used by the user device for transmission of the reference signals corresponds to the downlink, DL, channel to be used for transmission from the access points to the user device), phase alignment is provided in such systems for signals transmitted from access points to a user device. Channel reciprocity can, typically, be assumed when all of the following conditions are fulfilled: the UL and DL channels relate to the same frequency carrier, there is no movement (e.g., of one or more of: the user device, the access point(s), and obstructing/reflecting object(s)) affecting the channel between UL measurements and DL transmission, there is no phase drift in the system (e.g., between access points, and/or between access point and user device), and the transceiver phase differences (i.e., the phase difference between reception path and transmission path of a transceiver) of access points and user device have been characterized and accounted for by reciprocity calibration.

However, there are scenarios where channel reciprocity cannot be assumed to hold. For example, TDD operation implies that there is a time span between UL measurements and DL transmission. This may entail that phases for DL transmission, which are determined based on UL measurements, might not be accurate enough for phase alignment.

For example, when D-MIMO access points are physically separated, which they inherently are, they will typically have slightly different internal reference frequencies, which causes relative phase drift among the access points. Differences in internal reference frequencies may, for example, be due to one or more of: component discrepancies (e.g., caused by manufacturing), component aging, and environmental discrepancies (e.g., temperature, humidity, etc.). Furthermore, internal reference frequencies may vary over time due (e.g., due to aging and/or varying environmental conditions). Exemplifying the severity of this problem for relatively high carrier frequencies, it is noted that two carrier signals at 5 GHz with a 1 ppm frequency difference will drift by 5000 periods per second with respect to each other. Thus, a time span between UL measurements and DL transmission of 100 ps corresponds to a phase drift of a half-period (180 degrees), which leads to that a phase alignment attempt based on the UL measurements results in destructive combining (since the signals are completely out-of-phase instead of in-phase as intended).

Hence, to enable constructive combination at a user device, of signals transmitted by several access points of a D-MIMO system, the transmitting access points typically need to be frequency synchronized relative each other. For example, it may be desirable to keep the total phase error between UL measurements and DL transmission below some threshold value (e.g., at tens-of- degrees). There may furthermore be other sources of phase error than just offset between internal frequencies of access points, which motivates frequency synchronization (e.g., to keep the offset between internal frequencies at a fraction of 1 ppm).

Frequency synchronization may preferably be repeatedly performed to account for internal reference frequencies varying over time.

Achieving frequency synchronization between access points in a D-MIMO system may be challenging. When the access points have no wired interface between them, the frequency synchronization is performed using wireless, over-the-air (OTE), signals. Thereby, the frequency synchronization process is affected by the propagation channel between the access points, which makes the frequency synchronization more cumbersome than that of, for example, a centralized massive MIMO system. Furthermore, the signaling for frequency synchronization may increase power consumption and/or decrease throughput.

In the following, various embodiments will be described that provide alternative approaches to synchronization between access points of a D-MIMO system. Particularly, embodiments presented herein enable over-the-air frequency synchronization between a plurality of access points of a D-MIMO system (e.g., the D-MIMO system, 100 of Figure 1).

Generally, it should be noted that - even though some problematic situations, where some embodiments may be beneficial, are described with reference to TDD and channel reciprocity- embodiments may be equally applicable for other scenarios (e.g., frequency division duplex, FDD, and/or non-reciprocal channels).

Figure 2 illustrates example signaling according to some embodiments, for over-the-air frequency synchronization between a plurality of access points of a D-MIMO system (e.g., between access points 121, 122, 123, 124, 125, 126 of the D-MIMO system, 100 of Figure 1). The signaling is shown to involve a central control node (CCN) 210 (compare with the CCN 110 of Figure 1) a transmitting access point (TAP) 240 and a measuring access point (MAP) 250.

According to the example of Figure 2, the central control node 210 instructs the transmitting access point 240 to transmit a reference signal for frequency synchronization in each of a plurality of scheduled transmission occasions. The instruction for the transmitting access point 240 is illustrated by 201. As will be elaborated in later herein, a time between transmission occasions is dynamically set by the central control node 210, based on expected frequency error for the transmitting access point 240.

The central control node 210 may also instruct the measuring access points 250, to perform phase measurements during the plurality of scheduled transmission occasions, wherein the transmitting access point 240 is instructed to perform reference signal transmission for frequency synchronization. The instruction for the measuring access point 250 is illustrated by 202.

In some embodiments, each access point is aware of predetermined transmission occasions and is configured to perform phase measurements during transmission occasions, where it is not instructed to perform reference signal transmission for frequency synchronization. Thus, the explicit instruction 202 from the central control node 210 to perform phase measurements is optional.

When used, the explicit instruction 202 from the central control node 210 to perform phase measurements may be conveyed together with the instruction 201 from the central control node 210 to transmit a reference signal for frequency synchronization (e.g., as a general instruction for all access points). Alternatively, the explicit instruction 202 from the central control node 210 to perform phase measurements may be conveyed separately from the instruction 201 from the central control node 210 to transmit a reference signal for frequency synchronization.

Responsive to the instruction 201, the transmitting access point 240 performs the reference signal transmission, as illustrated by 203, and the measuring access point 250 performs phase measurements for frequency synchronization on the reference signal.

In some embodiments, the measuring access point 250 provides a measurement report to the central control node 210, as illustrated by optional signaling 204, wherein the measurement report is indicative of an estimated frequency error determined from the phase measurements. The central control node 210 can determine a frequency correction for the measuring access point 250 based on the estimated frequency error, and convey an indication of the frequency correction to the measuring access point 250, as illustrated by optional signaling 205.

Alternatively, or additionally, the measuring access point 250 can determine a frequency correction for itself, based on an estimated frequency error determined from the phase measurements.

In either case, the frequency correction is applied by the measuring access point 250, thereby improving frequency synchronization in relation to the transmitting access point 240.

It should be noted that the TAP 240 may represent more than one transmitting access point and/or that the MAP 250 may represent more than one measuring access point. Typically, different access points act as transmitting access points for different transmission occasions, and - consequently - different access points act as measuring access points for different transmission occasions. Thus, the signaling illustrated in Figure 2 is typically performed for each transmitting access point and for each measuring access point.

In some embodiment, provision of the instruction(s) 201, 202 and/or determination of frequency correction (and related signaling 204, 205) is performed once for each transmission occasion.

In some embodiments, the instruction(s) 201, 202 is/are provided once for a collection of transmission occasions (before the earliest transmission occasion of the collection); specifying different transmitting access points for different transmission occasions. Then, the reference signal transmission 203 is performed for each transmission occasion in the collection. The determination of frequency correction and related signaling 204, 205 may, typically, be performed once for the collection of transmission occasions (after the latest transmission occasion of the collection); handling frequency synchronization relative all of the transmitting access points.

Generally, the reference signals may be specific reference signals for frequency synchronization, or may be usable also for other purposes.

Figure 3 illustrates an example method 300 according to some embodiments. The method 300 is for over-the-air frequency synchronization between a plurality of access points of a D-MIMO system.

Typically, the method 300 is performed by a central control node of the D-MIMO system. Alternatively, the method 300 may be performed by another node adapted to control the access points of the D-MIMO system (e.g., by control signaling via the central control node). Yet alternatively, some steps of the method 300 may be performed by the central control node and some (other) steps of the method 300 may be performed by another node.

For example, the method 300 may be performed by the central control node 110 of Figure 1 and/or by the central control node 210 of Figure 2.

The plurality of access points may comprise all access points of the D-MIMO system (e.g., access points 121, 122, 123, 124, 125, 126 of the D-MIMO system 100 of Figure 1) or a subset thereof. For example, at least (e.g., only) all access points that are used for communication with a user device may be comprised in the plurality (e.g., the plurality 120 relating to communication with the user device 130 of Figure 1). Alternatively or additionally, at least (e.g., only) all access points that are located within a specified area or space (e.g., in a geographical sense and/or in a radio propagation sense) may be comprised in the plurality. Thus, the plurality of access points may comprise one or more of: access points located in a same geographical area, access points within a same radio signal vicinity, and access points that provide for communication with a same user device location (e.g., with the same user device, or with several user devices in a same area of location).

The method 300 comprises determining a group of one or more transmitting access points, as illustrated by step 320. The group of one or more transmitting access points may be seen as a specification of which access points are to transmit reference signals during execution of a frequency synchronization event.

For example, the group of one or more transmitting access points may comprise all access points of the plurality of access points, or a subset thereof. When a subset of the plurality of access points are included in the group of one or more transmitting access points, the access points of the subset may be spread out (e.g., in a geographical sense and/or in a radio propagation sense) among the plurality of access points, or may comprise a cluster (e.g., in a geographical sense and/or in a radio propagation sense) within the plurality of access points.

In some embodiments, an access point which has an expected frequency error (e.g., a frequency error determined during previous frequency synchronization event) that exceeds a frequency error threshold value may be excluded from the group of one or more transmitting access points, as illustrated by optional sub-step 325.

Generally, different frequency synchronization events (e.g., executed at different points in time) may apply different groups of one or more transmitting access points. For example, access points with relatively stable internal reference frequencies might be included in the group of one or more transmitting access points more seldom than access points with relatively varying internal reference frequencies.

The method 300 also comprises scheduling a plurality of reference signal transmission occasions for each of the transmitting access points, as illustrated by step 330. Each transmission occasion is typically scheduled for one or more of the transmitting access points. Thus, each transmission occasion may have one or more transmitting access points.

For each transmitting access point, the scheduling of step 330 comprises dynamically setting a time between transmission occasions based on expected frequency error. The time between transmission occasions may be the same for all transmitting access points, or may differ between different transmitting access points. Alternatively or additionally, the time between transmission occasions for a transmitting access point may be the same throughout a frequency synchronization event, or may differ between different transmission occasion intervals throughout the frequency synchronization event. The dynamic setting of time between transmissions provide flexibility. For example, a relatively short time between transmissions enables capturing of relatively large frequency errors, while a relatively long time between transmissions may provide relatively high accuracy for the frequency error determination (e.g., assuming that no phase wrap-around that cannot be dewrapped occurs during the time between transmissions; which may be ensured when the time between transmission occasions is shorter than half a period of the frequency error).

The time between transmissions may decrease when the expected frequency error increases. Thus, a first time between transmission occasions based on a first expected frequency error may be shorter than a second time between transmission occasions based on a second expected frequency error when the first expected frequency error is larger than the second expected frequency error. For example, an expected frequency error that exceeds a frequency error threshold value may trigger a burst of densely spaced reference signal transmissions from one transmitting access point to capture the frequency error.

In some embodiments, the time between transmission occasions is set to be shorter than half a period of a largest expected frequency error between the transmitting access point and any access point in the plurality of access points, to facilitate phase de-wrapping.

The expected frequency error may comprise an individual frequency error for one or more of the transmitting access points.

Generally, embodiments may provide for correction of the individual frequency errors of access points (i.e., reducing the difference/spread in frequencies between access points).

For example, the expected frequency error may comprise a frequency error determined during a previous frequency synchronization event. When there is no previous frequency synchronization event (e.g., for an initial frequency synchronization event), the expected frequency error may be set to an initial expected frequency error value (e.g., a maximum value), as illustrated by optional step 305. The initial expected frequency error value may correspond to a maximum value for the expected frequency error (e.g., a worst case frequency error).

Alternatively, or additionally, data provided by environmental sensors (e.g., temperature sensors, humidity sensors, etc.) that are co-located with the access points may be used to determine the expected frequency error. For example, if data from an environmental sensor indicates relatively high variability, the co-located access point may be expected to have a relatively large frequency error. A relatively high variability may be manifested, for example, in that there is a relatively large change (e.g., above a threshold value) between data values for different frequency synchronization events, and/or in that a data value differs significantly (e.g., more than a threshold value) from an average data value of the sensor (e.g., as initialized).

Typically, a frequency error increases with increasing time since a previous frequency synchronization event. Therefore, the expected frequency error may be determined based on the time since a previous frequency synchronization event (e.g., such that the expected frequency error increases with increasing time since the previous frequency synchronization event).

For example, the expected frequency error may be determined as an affine (e.g., linear, or starting from a positive expected frequency error for the time of the previous frequency synchronization) increasing function of the time since the previous frequency synchronization. In some embodiments, the frequency error determined during a previous frequency synchronization event is used as a reference value forthe function, corresponding to a reference time since the previous frequency synchronization (e.g., zero, or a positive value). In some embodiments, data from environmental sensors is used to determine the positive gradient of the function (e.g., increasing the gradient when the data from the environmental sensors is highly variable).

When an access point is excluded from the group of one or more transmitting access points (e.g., due to having an expected frequency error that exceeds a frequency error threshold value), transmitting access points that are close (e.g., in a geographical sense and/or in a radio propagation sense) to the excluded access point may be scheduled to transmit more often than other transmitting access points. This approach may enable capturing of a relatively large frequency error for the excluded access point. Thus, a number of transmission occasions for a first transmitting access point may be higher than a number of transmission occasions for a second transmitting access point when a distance (e.g., in a geographical sense and/or in a radio propagation sense) from the first transmitting access point to the excluded access point is shorter than a distance from the second transmitting access point to the excluded access point. The method 300 also comprises instructing each transmitting access point to transmit a reference signal for frequency synchronization in each of the transmission occasions scheduled for the transmitting access point, as illustrated by step 350 (compare with 201 of Figure 2).

In some embodiments, the method 300 also comprises (for each scheduled transmission occasion) instructing access points, that are not instructed to transmit in the scheduled transmission occasion, to perform phase measurements during the transmission occasion. This is illustrated by optional step 360 (compare with 202 of Figure 2).

The reference signal may be a fixed reference signal (e.g., the same reference signal for all transmitting access points, or different preassigned reference signals for different transmitting access points).

Alternatively, the reference signals may be dynamically assigned to the transmitting access points, as illustrated by optional step 340. Step 340 may comprise assigning the same reference signal for all transmitting access points, or assigning different reference signals for different transmitting access points.

For example, when a reference signal transmission occasion is scheduled for two or more transmitting access points, different reference signals may be assigned to the two or more transmitting access points to enable discrimination between them. Thus, different reference signals may be assigned to different transmitting access points with one or more coinciding transmission occasion.

Generally, the reference signal(s) may be any suitable reference signal(s). For example, each reference signal may be a tone, or a modulated signal, and different reference signals may be achieved by using different tone frequencies, or different modulated signals.

When the reference signals are dynamically assigned to the transmitting access points, step 350 may comprise indicating one or more parameters (e.g., tone frequency, modulation, reference signal identifier, etc.) of the reference signal to the transmitting access point, as illustrated by optional sub-step 355, and/or to access points configured to perform phase measurements on the reference signal.

The method 300 may further comprise receiving a respective measurement report from one or more (typically all) of the plurality of access points, as illustrated by optional step 370 (compare with 204 of Figure 2). Each measurement report may be indicative of an estimated frequency error for the corresponding access point.

The method 300 may further comprise determining a respective frequency correction for one or more of the plurality of access points, as illustrated by optional step 380. The frequency correction for an access point may be based - at least - on the estimated frequency error for the corresponding access point. For example, the frequency correction may comprise removal of the estimated frequency error for an access point from a currently used internal reference frequency of the access point.

In some embodiments, the frequency correction for an access point is also based on the estimated frequency error(s) for one or more (e.g., all) of the other access points of the plurality. For example, determining the respective frequency correction may comprise application of bias for maintenance of an average frequency control setting among the plurality of access points, as illustrated by optional sub-step 385. This may be accomplished by determining intermediate (raw) frequency corrections for each of the access points, and apply a bias that corresponds to a difference between an average of currently used frequency control settings of the access points, and an average of currently used frequency control settings of the access points with the intermediate frequency corrections applied.

In some embodiments, the central control node may execute step 380 by calculating updates for the respective frequency control words for the reference oscillators of the access points, based on the estimated frequency errors. For example, the central control node may have information regarding the currently used frequency control words, and add the calculated updates to the currently used frequency control words.

To mitigate drift of the average frequency control setting and/or maintain an average frequency control setting among the plurality of access points, the central control node may calculate an average of the updated frequency control words, and compare that to an average of the currently used frequency control words. The difference between these averages may be used as a bias for the updated frequency control words (e.g., subtracted, or added, to the updated frequency control words). This approach is beneficial to prevent drifting of the average frequency control word, which could cause out-of-range issues forthe frequency control signals, and/or disturbances to other systems operating at nearby frequencies. The method 300 may further comprise conveying indications of the respective frequency corrections (e.g., the updated frequency control words) to the corresponding access points, as illustrated by optional step 390 (compare with 205 of Figure 2)

In some embodiments, the scheduling and instructing steps 320, 350 (as well as any applicable step among the optional steps 340, 355, 360, 370, 380, 385, 390) are performed in response to triggering of a current frequency synchronization event, as illustrated by optional step 310. Step 320, the determination of the group of transmitting access points, may also be performed in response to triggering of a current frequency synchronization event, as implied in Figure 3, or may be performed more seldom (e.g., responsive to one or more of: an access point being initialized, a user device having moved, and a new user device having appeared).The method 300 remains in step 310 (N-path out of step 310) until a current frequency synchronization event is triggered (Y-path out of step 310). Then, the steps 320, 350 (as well as 320, 325, 340, 355, 360, 370, 380, 385, 390; as applicable) are performed as part of the synchronization event. Thereafter, the method 300 returns to step 310.

In some embodiments, the method 300 comprises triggering the current frequency synchronization event in response to one or more of: data provided by environmental sensors indicating a relatively large change since previous frequency synchronization event, data provided by environmental sensors differing significantly from an average data value of the sensor, an access point of the plurality of access points being initialized, a change of user device location (e.g., a user device having moved, or a new user device having appeared), determining that a predetermined time has elapsed since previous frequency synchronization event, and detecting a communication performance deterioration for one or more of the plurality of access points (e.g., in the form of decreasing throughput, decreasing signal-to-noise ratio, SNR, for reception, or similar).

The triggering of a current frequency synchronization event may further comprise collection of information from the plurality of access points that can be used for frequency synchronization. For example, the information may be indicative of currently used internal reference frequencies (e.g., currently used frequency control settings) of the access points, and/or capabilities of the access points (e.g., regarding which type(s) of frequency correction the access point is configured to apply). Alternatively, collection of information from the plurality of access points may be performed more seldom (e.g., once, at initialization of an access point; or only at triggering of some frequency synchronization events). For example, information regarding capabilities of an access point may be collected only once. Alternatively, or additionally, information regarding the currently used internal reference frequency may be collected only for some frequency synchronization events (there between, the central control node may be configured to keep track of the currently used internal reference frequency via the determined frequency corrections).

When the D-MIMO system and/or one or more of the access points have been inactive for some amount of time, an initial frequency synchronization event may be triggered. In an initial frequency synchronization event, neither the currently used internal reference frequencies of the access points, nor frequency errors determined during previous frequency synchronization events, are typically considered useful. Thus, before the initial frequency synchronization event, the currently used internal reference frequencies (e.g., the currently used frequency control settings) of the access points may be reset (e.g., by control signaling from the central control node, similar to step 390), and the expected frequency error may be set to an initial expected frequency error value, as illustrated by step 305. Typically, the currently used internal reference frequencies of the access points may be reset by letting the reference oscillators of each access point assume their mid-point tuning values.

Figure 4 illustrates an example method 400 according to some embodiments. The method 400 is for over-the-air frequency synchronization between a plurality of access points of a D-MIMO system; more particularly for over-the-air frequency synchronization of an access point in relation to the plurality of access points.

Typically, the method 400 is performed by an access point among the plurality of access points, for over-the-air frequency synchronization of the access point in relation to the plurality of access points of the D-MIMO system.

For example, the method 400 may be performed by one or more of the access points 121, 122, 123, 124, 125, 126 of Figure 1 and/or by one or more of the measuring access point 250 and the transmitting access point 240 of Figure 2. The method 400 may comprise receiving an instruction to transmit a reference signal for frequency synchronization in each of a (second) plurality of transmission occasions scheduled for the access point, as illustrated by optional step 410 (compare with 201 of Figure 2 and step 350 of Figure 3). As elaborated on in relation to Figure 3, a time between transmission occasions for an access point is dynamically set based on expected frequency error.

It should be noted that, according to some embodiments, some access points of the plurality of access points to the frequency synchronized only perform phase measurements since they do not belong to (e.g., are excluded from) the group of transmitting access points. Thus, receiving an instruction to transmit a reference signal for frequency synchronization, as illustrated by step 410, is optional.

Alternatively, or additionally, the method 400 may comprise receiving an instruction to perform phase measurements during each of a (first) plurality of reference signal transmission occasions, as illustrated by optional step 420 (compare with 202 of Figure 2 and step 360 of Figure 3).

In some embodiments, the access point is aware of predetermined transmission occasions and is configured to perform phase measurements during transmission occasions, where it is not instructed to perform reference signal transmission for frequency synchronization. Thus, receiving an explicit instruction to perform phase measurements, as illustrated by step 420, is optional.

The first and second pluralities of transmission occasions are typically disjunct (i.e., during each transmission occasion, the access point either transmits a reference signal or performs phase measurements). For example, the phase measurements may be performed for all reference signal transmission occasions except those comprised in the second plurality of scheduled transmission occasions.

The method 400 comprises performing phase measurements for frequency synchronization, as illustrated by step 438, for each of the first plurality of reference signal transmission occasions, as illustrated by 432. The phase measurements performed during a transmission occasion are relative one or more transmitting access points of the transmission occasion. In some embodiments, the method 400 comprises determining, for each transmission occasion

432, whether the access point is instructed (e.g., in step 410) to transmit a reference signal for frequency synchronization in that transmission occasion, as illustrated by optional step 434.

When the access point is instructed to transmit a reference signal for frequency synchronization in that transmission occasion (Y-path out of step 434), the method 400 may comprise performing reference signal transmission accordingly, as illustrated by optional step 436 (compare with 203 of Figure 2).

When the access point is not instructed to transmit a reference signal for frequency synchronization in that transmission occasion (N-path out of step 434) and/or is instructed (e.g., in step 420, or implicitly) to perform phase measurements during the transmission occasion, the method 400 comprises performing phase measurements accordingly, as illustrated by step 438.

The transmission of reference signals or performance of phase measurements, as applicable, is performed for each transmission occasion, as illustrated by combining steps 432, 434, 436, 438 into an execution block 430.

Typically, all access points of the plurality that are not instructed to transmit a reference signal during a transmission occasion performs phase measurements during that transmission occasion.

In some embodiments, a measuring access point is adapted to identify the transmitting access point based on the reference signal (e.g., by listening for tones at the frequencies of possible reference signal tones, or by listening at the frequencies that correspond to center frequencies of possible modulated reference signals and applying correlators corresponding to possible modulation sequences).

Step 438 may comprise determining a phase estimation for each received reference signal, corresponding to the phase difference between the local reference oscillators of the measuring access point and the transmitting access point. By monitoring the phase difference over multiple phase measurements from the same transmitter, the frequency difference (i.e., a frequency offset) between the measuring access point and the transmitting access point can be estimated, as will be exemplified in connection with step 440. Typically, the access point performing the method 400 estimates the frequency offsets relative all transmitting access points during a frequency synchronization event (e.g., during execution block 430).

The method 400 may also comprise determining an estimated frequency error for the access point based on the phase measurements, as illustrated by optional step 440.

In some embodiments, the method 400 may comprise determining a frequency correction based on the estimated frequency error, as illustrated by optional step 450.

Alternatively, or additionally, the method 400 may comprise providing a measurement report to a central control node of the D-MIMO system, as illustrated by optional step 460 (compare with 204 of Figure 2 and step 370 of Figure 3). The measurement report may be indicative of the estimated frequency error determined in step 440.

The method 400 may also comprise receiving an indication of a frequency correction from the central control node, as illustrated by optional step 470 (compare with 205 of Figure 2 and step 390 of Figure 3). The frequency correction may be based on the estimated frequency error for the access point, as elaborated on in connection with Figure 3.

In some embodiments, determination of a frequency correction by the access point may be combined with determination of a frequency correction by the central control node (i.e., provision of the measurement report and reception of the indication of a frequency correction). For example, the central control node may be configured to provide a relatively rough frequency correction value, while the access points fine tunes the frequency correction locally.

In either case, the method 400 may further comprise applying the frequency correction, as illustrated by optional step 480. The frequency correction may, for example, be applied for a reference oscillator (e.g., a crystal oscillator), for a frequency synthesizer (e.g., a fractional-N frequency synthesizer of a phase-locked loop, PLL, that generates the reference frequency), or for a baseband signal rotation.

Combinations of the above alternatives are also possible in relation to frequency correction. For example, a relatively rough frequency correction - possibly determined by the central control node - may be performed at radio frequency (e.g., via the reference oscillator and/or the frequency synthesizer), while the frequency correction is fine-tuned in the baseband (e.g., via baseband signal rotation) - possibly based on frequency correction determination by the access point. Alternatively or additionally, a relatively rough frequency correction may be performed more seldom than frequency correction fine-tuning.

When applied for the reference oscillator, the frequency correction may be implemented by tuning of the reference oscillator.

Application of the frequency correction at the reference oscillator may be beneficial by synchronizing both the carrier frequency and the baseband frequency. When the frequency correction is applied at the reference oscillator, several frequency synchronization events may need to be performed, according to some embodiments, to properly reduce the frequency offset (e.g., if the tuning sensitivity of the reference oscillator is not accurately known).

When applied for the frequency synthesizer, the frequency correction may be implemented by changing the control word for the frequency synthesizer.

Application of the frequency correction at the frequency synthesizer may be beneficial by providing high accuracy. However, the baseband frequency may not be synchronized when the frequency correction is applied at the frequency synthesizer.

When applied for the baseband signal rotation, the frequency correction may be implemented by selection of the angle used in the baseband signal rotation such that the angle corresponds to the frequency correction. Typically, the angle is changed linearly with time to correspond to a frequency offset.

Any frequency correction applied for the baseband signal rotation is preferably turned off during transmission occasions of future frequency synchronization events, to enable proper phase measurements for frequency synchronization (i.e., phase measurements relating to the uncompensated local oscillator frequency offset).

Application of the frequency correction in the baseband may be beneficial for relatively small frequency offsets. With application of the frequency correction in the baseband, it is possible to operate somewhat independently of the central control node (e.g., determining the frequency correction locally), and control traffic to and from the central access node may be reduced. Such reduction may be particularly beneficial for relatively large D-MIMO systems. If the frequency correction is turned off during transmission occasions of future frequency synchronization events, there is typically no risk of average frequency drift (i.e., the average frequency among the plurality of access points is maintained) and no bias compensation is necessary.

The method 400 may be performed whenever there is a frequency synchronization event (compare with step 310 of Figure 3). This is illustrated in that the method 400 returns to step 410 after completion of step 480.

In step 440, the estimated frequency error for the access point may be determined in any suitable way, based on the phase measurements. It should be noted that the phase measurements are typically noisy due to being based on over-the-air signaling (e.g., the phase may be altered by disturbances in the wireless channel; typically due to mobility of the access point(s) and/or obstructing/reflecting objects affecting the propagation channel).

For example, a respective frequency offset may be determined for the access point performing the method 400, relative each transmitting access point of a collection of transmitting access points, and the estimated frequency error for the access point may be determined as an average of the respective frequency offsets.

The average may include a frequency offset relative the access point performing the method (i.e., a zero frequency offset), or the average may be normalized by a factor (M — 1)/M, where M — 1 denotes the number of transmitting access points in the collection. The benefit of these approaches may be illustrated by a situation with only two access points. Ifthe two access points have a specific frequency offset with respect to each other, applying a frequency correction with the same magnitude as the frequency offset for both access points would cause the frequency offset to have the same magnitude as before, with opposite sign. Thus, the frequency corrections should preferably have an average magnitude that corresponds to half of the magnitude of the frequency offset; which is generalized to normalization by a factor (M — 1)/M when there are M access points involved.

Typically, the collection of access points comprises all access points of the group of transmitting access points for which the access point performing the method 400 has performed phase measurements. A frequency offset relative a transmitting access point may be determined by considering the measured phases for the transmitting access point over time.

In one example, the frequency offset may be determined as an average, or median, of phase differences between time-adjacent phase measurements (e.g., using the time interval between time-adjacent phase measurements for conversion from phase domain to frequency domain). In some embodiments, phase differences that substantially deviate from other phase differences for the same transmitting access point (a.k.a., outliers) may be removed before determining the average, or median. Alternatively, or additionally, de-wrapping of phase measurements may be performed before or after determining differences between time- adjacent phase measurements. Determination of an average value may comprise weighting the phase differences differently (e.g., giving relatively low weight to outliers).

In one example, the frequency offset may be determined as a gradient (i.e., the slope) of an affine function fitted to a phase-time representation of the phase measurements (e.g., by application of least mean square, LMS, line fitting). In some embodiments, phase measurements that substantially deviate from a phase-time relation formed by other phase measurements for the same transmitting access point (a.k.a., outliers) may be removed before determining the gradient. Alternatively, or additionally, de-wrapping of phase measurements may be performed before line fitting and determination of the gradient.

Generally, a wrap-around occurs when an angular representation of the phase crosses an angular dimension border (e.g., the border defined by zero degrees and 360 degrees) between two samples (here; phase measurements). De-wrapping refers to the process of counteracting wrap-around; i.e., compensating such that the actual phase is provided. De-wrapping may benefit from setting the time between transmission occasions for each transmitting access point to be shorter than half a period of a largest expected frequency error between the transmitting access point and any access point in the plurality of access points.

In some embodiments, a transmitting access point corresponding to an unreliable frequency offset may be excluded from the collection, or may be given less impact on the estimated frequency error than other transmitting access points of the collection (e.g., by weighting the frequency offset before determining the average; typically using a weighting value between zero and one). This is illustrated by optional sub-step 445. For example, a transmitting access point may be determined as corresponding to an unreliable frequency offset when the phase measurements for the transmitting access point correspond to highly varying frequency offset estimations (e.g., when a variance of the phase differences between time-adjacent phase measurements exceeds a variance threshold value).

In some embodiments, there is no determination of respective frequency offsets relative each transmitting access point of a collection of transmitting access points. Instead, the estimated frequency error for the access point may be determined by collectively considering the measured phases for all of the transmitting access points in the collection (e.g., determining an overall average of phase differences, or an overall median of phase differences, or a gradient of an affine function fitted to a phase-time representation of all of the phase measurements; in analogy with the description above). This approach may be particularly useful in noisy situations. It may be beneficial in this approach if the same, or a very similar, amount of phase measurements are performed for each of the transmitting access points of the collection.

It should be noted that some, or all, of the actions performed in step 440 may be performed by the central control node instead of by the access point according to some embodiments. This may be achieved by suitably adapting the reporting step 460 (e.g., letting step 460 be executed before - or in the midst of - step 440, and letting the measurement report indicate everything needed to execute the remainder of step 440 in the central control node.

Figure 5A schematically illustrates example principles according to some embodiments. More particularly, Figure 5A illustrates principles suitable for determining the frequency offset as an average, or median, of phase differences between time-adjacent phase measurements. For example, the principles exemplified in Figure 5A may be applied in step 440 of Figure 4.

A difference between time-adjacent phase measurements converted to frequency domain is illustrated by "x" in Figure 5A, and the differences for each transmitting access point are circled. Thus, the differences relating to a first transmitting access point are circled as 501, the differences relating to a second transmitting access point are circled as 502, and the differences relating to a third transmitting access point are circled as 503.

The x-axis represents frequency offset relative a reference frequency of the measuring access point; the reference frequency being represented as a zero frequency offset 510. Taking the average, or median, for each transmitting access point results in a first frequency offset 511 for the first transmitting access point, a second frequency offset 512 for the second transmitting access point, and a third frequency offset 513 for the third transmitting access point.

When the phase differences between time-adjacent phase measurements for the same transmitting access point are relatively similar (e.g., as shown for the third transmitting access point, 503), the average, or median, value may be determined based on all the phase differences to provide the frequency offset 513.

When some phase differences between time-adjacent phase measurements substantially deviate from other phase differences for the same transmitting access point (e.g., as shown for the first transmitting access point, 501), the average, or median, value may be determined after removing the outliers 591, 592, to provide the frequency offset 511. This may be particularly beneficial when an average value is determined, while determination of a median value implicitly counteracts the impact of outliers.

When the phase differences between time-adjacent phase measurements for the same transmitting access point are relatively varying (e.g., as shown for the second transmitting access point, 502), the corresponding frequency offset 512 may be regarded as unreliable. Consequently, the frequency offset 512 may be weighted before averaging over the transmitting access points to determine the frequency error, or the frequency offset 512 may be discarded.

For example, Figure 5A may be seen as illustrating the frequency offset estimation in a situation where the D-MIMO system comprises four access points to be frequency synchronized (i.e., the plurality of access points comprises four access points). The four access points may be denoted APO (the measuring access point), API (the first transmitting access point), AP2 (the second transmitting access point), and AP3 (the third transmitting access point). In this situation, the four access points have frequency offsets between them that satisfy f3 > f 2 > fl > fO, and f3 — f2 = f 2 — fl = f 2 — fO = Af where f0,fl, f2, f3 are the internal reference frequencies of APO, API, AP2, AP3, respectively. Thus, the frequency offset 511 corresponds to Af, the frequency offset 512 corresponds to 2Af, and the frequency offset 513 corresponds to 3Af. Figure 5B schematically illustrates example principles according to some embodiments. More particularly, Figure 5B illustrates principles suitable for determining the frequency offset as a gradient of an affine function fitted to a phase-time representation of the phase measurements. For example, the principles exemplified in Figure 5B may be applied in step 440 of Figure 4.

The x-axis represents the time for the phase measurements, and the y-axis represents dewrapped phase measurements. The phase measurements for a first transmitting access point are illustrated by "x", the phase measurements for a second transmitting access point are illustrated by "o", and the phase measurements for a third transmitting access point are illustrated by "+".

An affine (here; linear) function 521, 522, 523 is fitted to the phase-time representation of the phase measurements for each of the first, second, and third transmitting access points, and the gradients (i.e., slopes) of the functions may be used as respective (first, second, and third) frequency offsets for the first transmitting access points.

When some phase measurements substantially deviate from a phase-time relation formed by other phase differences for the same transmitting access point (e.g., as shown for the first transmitting access point, 521), the function fitting and/or the gradient determination may be performed after removing the outliers 595, 596.

Thus, alternative approaches are provided for synchronization between access points of a D- MIMO system. According to some embodiments, frequency synchronization is provided (to enable proper phase alignment) between access points of the D-MIMO system. For example, application of some embodiment may entail that the frequency offsets between access points of the D-MIMO system may be decreased compared to other approaches. By application of proper frequency synchronization, phase measurements performed by the access points on reference signaling transmitted by the user device, can be reliably used to determine the phases to be applied for transmission from the access points (i.e., phase alignment), even when some time has passed between the phase measurements and the transmission from the access points. Some embodiments employ a bias in the frequency correction determination (compare with step 385 of Figure 3) to mitigate drift of an average frequency control setting, wherein the bias is for maintenance of an average frequency control setting among the plurality of access points.

According to some embodiments, deviating measurements (e.g., outliers of a transmitting access point, and/or all measurements for a transmitting access point when unreliable) are removed from consideration (compare with step 440 of Figure 4), to make the frequency error estimation more robust to channel changes (e.g., due to access point movements, and/or due to movements of obstructing and/or reflecting objects).

When the reference signaling between access points of the D-MIMO system is dynamically scheduled (compare with step 330 of Figure 3), proper frequency synchronization is enabled (e.g., capturing large frequency errors as well as providing high accuracy), as well as flexibility. Further, the amount of overhead and/or power consumption due to reference signaling and measurements may be kept at an acceptable level when the reference signaling between access points of the D-MIMO system is dynamically scheduled.

The frequency correction can be determined by the central control node (compare with step 380 of Figure 3) and/or locally by each access point (compare with step 450 of Figure 4), which may provide for flexible frequency synchronization. For example, a coarse frequency correction may be determined by the central control node, and a refined frequency correction may be determined by each access point. Alternatively or additionally, the central control node may determine frequency corrections repeatedly, and each access point may determine respective frequency corrections in the duration between frequency correction determinations by the central control node.

There are various possibilities for application of the frequency correction, which may provide for flexible frequency synchronization. For example, a coarse frequency correction may be applied in a radio frequency domain, and a refined frequency correction may be applied in a baseband domain.

Using different reference signals to enable discrimination, two or more reference signals for frequency synchronization may be transmitted simultaneously by corresponding two or more access points, which may provide for flexible frequency synchronization and/or reduced signaling overhead.

Figure 6 schematically illustrates an example apparatus 600 according to some embodiments. The apparatus 600 is for over-the-air frequency synchronization between a plurality of access points of a D-MIMO system.

For example, the apparatus 600 may be comprisable (e.g., comprised) in a central control node (CCN) 610 for a D-MIMO system (compare with the central control node 110 of Figure 1 and the central control node 210 of Figure 2). Alternatively, or additionally, the apparatus 600 may be configured to perform (or cause performance of) one or more method steps as described in connection with the method 300 of Figure 3.

The apparatus 600 comprises a controller (CNTR; e.g., controlling circuitry or a control module) 620.

The controller 620 is configured to cause determination of a group of one or more transmitting access points (compare with step 320 of Figure 3).

To this end, the controller 620 may comprise, or be otherwise associated with (e.g., connected, or connectable, to) a group determiner (GDET; e.g., determining circuitry or a determination module) 621. The group determiner 621 may be configured to determine group of one or more transmitting access points.

The controller 620 is also configured to cause, for each of the transmitting access points, scheduling of a plurality of reference signal transmission occasions, wherein a time between transmission occasions is dynamically set based on expected frequency error (compare with step 330 of Figure 3).

To this end, the controller 620 may comprise, or be otherwise associated with (e.g., connected, or connectable, to) a scheduler (SCH; e.g., scheduling circuitry or a scheduling module) 622. The scheduler 622 may be configured to schedule the plurality of reference signal transmission occasions for each of the transmitting access points.

The controller 620 is also configured to cause, for each of the transmitting access points, instruction of the transmitting access point to transmit a reference signal for frequency synchronization in each of the scheduled transmission occasions (compare with step 350 of Figure 3). In some embodiments, the controller 620 may also be configured to cause, for each scheduled transmission occasion, instruction of access points that are not instructed to transmit in the scheduled transmission occasion, to perform phase measurements during the transmission occasion (compare with step 360 of Figure 3).

For example, the instruction of the transmitting access point and/or the instruction of the access points that are not instructed to transmit may comprise transmission of control signaling by a transceiver (TX/RX; e.g., transceiving circuitry) 630 associated with the controller 620. For example, the transceiver 630 may be comprised in the central control node 610, and/or the transceiver 630 may be connectable (e.g., connected) to the controller 620.

To this end, the controller 620 may comprise, or be otherwise associated with (e.g., connected, or connectable, to) an instructor (INS; e.g., instructing circuitry or an instruction module) 623. The instructor 623 may be configured to instruct the transmitting access point(s) and/or the access points that are not instructed to transmit (e.g., by causing transmission of control signaling by the transceiver 630).

The controller 620 may also be configured to cause assignment of different reference signals to different transmitting access points with one or more coinciding transmission occasions (compare with step 340 of Figure 3).

To this end, the controller 620 may comprise, or be otherwise associated with (e.g., connected, or connectable, to) an assigner (ASN; e.g., assigning circuitry or an assignment module) 624. The assigner 624 may be configured to assign reference signals to transmitting access points.

The controller 620 may also be configured to cause triggering of a (the current) frequency synchronization event (compare with step 310 of Figure 3).

To this end, the controller 620 may comprise, or be otherwise associated with (e.g., connected, or connectable, to) a triggerer (TRIG; e.g., triggering circuitry or a trigger module) 625. The triggerer 625 may be configured to trigger the current frequency synchronization event.

The controller 620 may also be configured to cause reception (e.g., via the transceiver 630), from one or more of the plurality of access points, of a respective measurement report indicative of an estimated frequency error for the corresponding access point (compare with step 370 of Figure 3).

The controller 620 may also be configured to cause determination of a respective frequency correction for one or more of the plurality of access points, wherein the frequency correction is based on the estimated frequency error for the corresponding access point (compare with step 380 of Figure 3).

To this end, the controller 620 may comprise, or be otherwise associated with (e.g., connected, or connectable, to) a frequency correction determiner (FDET; e.g., determining circuitry or a determination module) 626. The frequency correction determiner 626 may be configured to determine the respective frequency correction(s).

The controller 620 may also be configured to cause conveyance (e.g., by transmission via the transceiver 630), of an indication of the respective frequency correction to the corresponding access point (compare with step 390 of Figure 3).

Figure 7 schematically illustrates an example apparatus 700 according to some embodiments. The apparatus 700 is for over-the-air frequency synchronization between a plurality of access points of a D-MIMO system; more particularly for over-the-air frequency synchronization of an access point in relation to the plurality of access points.

For example, the apparatus 700 may be comprisable (e.g., comprised) in an access point (AP) 710 for a D-MIMO system (compare with the access points 121, 122, 123, 124, 125, 126 of Figure 1 and the access points 240, 250 of Figure 2). Alternatively, or additionally, the apparatus 700 may be configured to perform (or cause performance of) one or more method steps as described in connection with the method 400 of Figure 4.

The apparatus 700 comprises a controller (CNTR; e.g., controlling circuitry or a control module) 720.

The controller 720 may be configured to cause reception of an instruction to perform phase measurements during each transmission occasion of a (first) plurality of reference signal transmission occasions (compare with step 420 of Figure 4), and/or an instruction to transmit a reference signal for frequency synchronization in each of a (second) plurality of scheduled transmission occasions (compare with step 420 of Figure 4). For example, the instruction to transmit and/or the instruction to perform phase measurements may be received by a transceiver (TX/RX; e.g., transceiving circuitry) 730 associated with the controller 720. For example, the transceiver 730 may be comprised in the access point 710, and/or the transceiver 730 may be connectable (e.g., connected) to the controller 720.

The controller 720 is configured to cause, for each of the first plurality of reference signal transmission occasions, performance of phase measurements for frequency synchronization relative a transmitting access point during the transmission occasion (compare with step 438 of Figure 4). Knowledge regarding which are the first plurality of reference signal transmission occasions may be acquired explicitly or implicitly as elaborated on earlier.

To this end, the controller 720 may comprise, or be otherwise associated with (e.g., connected, or connectable, to) a phase measurer (MEAS; e.g., measuring circuitry or a measurement module) 721. The phase measurer 721 may be configured to perform the phase measurements

The controller 720 may also be configured to cause performance of reference signal transmission (e.g., by causing transmission of reference signaling by the transceiver 730) in accordance with an instruction to do so (compare with step 436 of Figure 4).

The controller 720 may also be configured to cause determination of an estimated frequency error for the access point based on the phase measurements (compare with step 440 of Figure 4).

To this end, the controller 720 may comprise, or be otherwise associated with (e.g., connected, or connectable, to) a frequency error estimator (EST; e.g., estimating circuitry or an estimation module) 722. The frequency error estimator 722 may be configured to determine the estimated frequency error.

The controller 720 may also be configured to cause determination of a frequency correction based on the estimated frequency error (compare with step 450 of Figure 4).

To this end, the controller 720 may comprise, or be otherwise associated with (e.g., connected, or connectable, to) a frequency correction determiner (FDET; e.g., determining circuitry or a determination module) 723. The frequency correction determiner 723 may be configured to determine the frequency correction. Alternatively, or additionally, the controller 720 may be configured to cause provision of a measurement report to a central control node of the D-MIMO system (e.g., by causing transmission of report signaling by the transceiver 730), wherein the measurement report is indicative of the estimated frequency error (compare with step 460 of Figure 4).

The controller 720 may also be configured to cause reception (e.g., via the transceiver 730) of an indication of a frequency correction from the central control node (compare with step 470 of Figure 4).

The controller 720 may also be configured to cause application of the frequency correction; as determined by the access point and/or as indicated by the central control node (compare with step 480 of Figure 4).

To this end, the controller 720 may comprise, or be otherwise associated with (e.g., connected, or connectable, to) a frequency corrector (FC; e.g., correcting circuitry or a correction module) 724. The frequency corrector 724 may be configured to apply the frequency correction.

According to some embodiments, a D-MIMO system may comprise a central control node 610 as described in connection with Figure 6 and a plurality of access points 710 as described in connection with Figure 7.

Generally, it should be noted that any feature described herein in relation to one Figure may be equally applicable (mutatis mutandis) to one or more of the other Figures, even if not explicitly mentioned in connection thereto.

The described embodiments and their equivalents may be realized in software or hardware or a combination thereof. The embodiments may be performed by general purpose circuitry. Examples of general purpose circuitry include digital signal processors (DSP), central processing units (CPU), co-processor units, field programmable gate arrays (FPGA) and other programmable hardware. Alternatively or additionally, the embodiments may be performed by specialized circuitry, such as application specific integrated circuits (ASIC). The general purpose circuitry and/or the specialized circuitry may, for example, be associated with or comprised in an apparatus such as a central control node or an access point for a D-MIMO system.

Embodiments may appear within an electronic apparatus (such as a central control node or an access point for a D-MIMO system) comprising arrangements, circuitry, and/or logic according to any of the embodiments described herein. Alternatively or additionally, an electronic apparatus (such as a central control node or an access point for a D-MIMO system) may be configured to perform method steps according to any of the embodiments described herein.

According to some embodiments, a computer program product comprises a non-transitory computer readable medium such as, for example, a universal serial bus (USB) memory, a plugin card, an embedded drive, or a read only memory (ROM). Figure 8 illustrates an example computer readable medium in the form of a compact disc (CD) ROM 800. The computer readable medium has stored thereon a computer program comprising program instructions. The computer program is loadable into a data processor (PROC; e.g., a data processing unit) 820, which may, for example, be comprised in an apparatus 810 such as a central control node or an access point for a D-MIMO system. When loaded into the data processor, the computer program may be stored in a memory (MEM) 830 associated with, or comprised in, the data processor. According to some embodiments, the computer program may, when loaded into, and run by, the data processor, cause execution of method steps according to, for example, any of the methods illustrated in Figures 3 and 4, or otherwise described herein.

Generally, all terms used herein are to be interpreted according to their ordinary meaning in the relevant technical field, unless a different meaning is clearly given and/or is implied from the context in which it is used.

Reference has been made herein to various embodiments. However, a person skilled in the art would recognize numerous variations to the described embodiments that would still fall within the scope of the claims.

For example, the method embodiments described herein discloses example methods through steps being performed in a certain order. However, it is recognized that these sequences of events may take place in another order without departing from the scope of the claims. Furthermore, some method steps may be performed in parallel even though they have been described as being performed in sequence. Thus, the steps of any methods disclosed herein do not have to be performed in the exact order disclosed, unless a step is explicitly described as following or preceding another step and/or where it is implicit that a step must follow or precede another step. In the same manner, it should be noted that in the description of embodiments, the partition of functional blocks into particular units is by no means intended as limiting. Contrarily, these partitions are merely examples. Functional blocks described herein as one unit may be split into two or more units. Furthermore, functional blocks described herein as being implemented as two or more units may be merged into fewer (e.g. a single) unit.

Any feature of any of the embodiments disclosed herein may be applied to any other embodiment, wherever suitable. Likewise, any advantage of any of the embodiments may apply to any other embodiments, and vice versa.

Hence, it should be understood that the details of the described embodiments are merely examples brought forward for illustrative purposes, and that all variations that fall within the scope of the claims are intended to be embraced therein.

Claims

1. A method for over-the-air frequency synchronization between a plurality of access points

(122, 123, 124, 125) of a distributed multiple-input multiple-output, D-MIMO, system (100), the method comprising: determining (320) a group of one or more transmitting access points; and for each of the transmitting access points: scheduling (330) a plurality of reference signal transmission occasions, wherein a time between transmission occasions is dynamically set based on expected frequency error; and instructing (201, 350) the transmitting access point to transmit a reference signal for frequency synchronization in each of the scheduled transmission occasions.

2. The method of claim 1 further comprising - for each scheduled transmission occasion - instructing (202, 360) access points, that are not instructed to transmit in the scheduled transmission occasion, to perform phase measurements during the transmission occasion.

3. The method of any of claims 1 through 2, wherein the group of one or more transmitting access points comprises the plurality of access points, ora sub-set of the plurality of access points.

4. The method of any of claims 1 through 3 further comprising - for an access point which has an expected frequency error that exceeds a frequency error threshold value - excluding (325) that access point from the group of one or more transmitting access points.

5. The method of claim 4, wherein a number of transmission occasions for a first transmitting access point is higher than a number of transmission occasions for a second transmitting access point when a distance from the first transmitting access point to the excluded access point is shorter than a distance from the second transmitting access point to the excluded access point.

6. The method of any of claims 1 through 5, wherein each transmission occasion is scheduled for one or more of the transmitting access points.

7. The method of any of claims 1 through 6 further comprising assigning (340) different reference signals to different transmitting access points with one or more coinciding transmission occasions.

8. The method of any of claims 1 through 7, wherein the reference signal is a tone, or a modulated signal.

9. The method of any of claims 1 through 8 further comprising indicating (355) one or more parameters of the reference signal to the transmitting access point and/orto access points configured to perform phase measurements on the reference signal.

10. The method of any of claims 1 through 9, wherein a first time between transmission occasions based on a first expected frequency error is shorter than a second time between transmission occasions based on a second expected frequency error when the first expected frequency error is larger than the second expected frequency error.

11. The method of any of claims 1 through 10, wherein the time between transmission occasions is shorter than half a period of a largest expected frequency error between the transmitting access point and any access point in the plurality of access points.

12. The method of any of claims 1 through 11, wherein the expected frequency error comprises a frequency error determined during previous frequency synchronization event.

13. The method of any of claims 1 through 12, wherein the expected frequency error comprises an individual frequency error for the transmitting access point.

14. The method of any of claims 1 through 13, wherein the expected frequency error increases with increasing time since previous frequency synchronization event.

15. The method of any of claims 1 through 14 further comprising setting (305) the expected frequency error to a maximum value for an initial frequency synchronization event.

16. The method of any of claims 1 through 15, wherein the scheduling and instructing steps are performed in response to triggering (310) of a current frequency synchronization event.

17. The method of claim 16 further comprising triggering (310) the current frequency synchronization event in response to one or more of: an access point of the plurality of access points being initialized, a change of user device location, determining that a predetermined time has elapsed since previous frequency synchronization event, and detecting a communication performance deterioration for one or more of the plurality of access points.

18. The method of any of claims 1 through 17, wherein the plurality of access points comprises one or more of: access points located in a same geographical area, access points within a same radio signal vicinity, and access points that provide for communication with a same user device location.

19. The method of any of claims 1 through 18 further comprising receiving (204, 370) - from one or more of the plurality of access points - a respective measurement report, wherein each measurement report is indicative of an estimated frequency error for the corresponding access point.

20. The method of claim 19 further comprising determining (380) a respective frequency correction for one or more of the plurality of access points, wherein the frequency correction is based on the estimated frequency error for the corresponding access point.

21. The method of claim 20, wherein determining the respective frequency correction comprises application (385) of bias for maintenance of an average frequency control setting among the plurality of access points.

22. The method of any of claims 20 through 21 further comprising conveying (205, 390) an indication of the respective frequency correction to the corresponding access point.

23. The method of any of claims 1 through 22, performed by a central control node (110, 210,

610) of the D-MIMO system.

24. A method for over-the-air frequency synchronization of an access point in relation to a plurality of access points (122, 123, 124, 125) of a distributed multiple-input multipleoutput, D-MIMO, system (100), the method comprising: for each (432) of a first plurality of reference signal transmission occasions, wherein a time between transmission occasions is dynamically set based on expected frequency error, performing (438) phase measurements for frequency synchronization, wherein the phase measurements are relative a transmitting access point during the transmission occasion.

25. The method of claim 24 further comprising: receiving (201, 410) an instruction to transmit a reference signal for frequency synchronization in each of a second plurality of scheduled transmission occasions, wherein a time between transmission occasions is dynamically set based on expected frequency error; and performing reference signal transmission (203, 436) accordingly.

26. The method of claim 25, wherein the phase measurements are performed for all reference signal transmission occasions except those comprised in the second plurality of scheduled transmission occasions.

27. The method of any of claims 24 through 26 further comprising receiving (202, 420) an instruction to perform the phase measurements during each transmission occasion of the first plurality of reference signal transmission occasions.

28. The method of any of claims 24 through 27 further comprising determining (440) an estimated frequency error for the access point based on the phase measurements.

29. The method of claim 28, wherein the estimated frequency error for the access point is determined as an average of respective frequency offsets relative transmitting access points of a collection of transmitting access points.

30. The method of claim 29 further comprises, for a transmitting access point corresponding to an unreliable frequency offset: excluding (445) that transmitting access point from the collection of transmitting access points, or weighting (445) the frequency offset relative that transmitting access point before determining the average.

31. The method of any of claims 29 through 30, wherein determining the respective frequency offset comprises determining a gradient of an affine function fitted to a phase-time representation of the phase measurements. e method of any of claims 29 through 30, wherein determining the respective frequency offset comprises determining an average, or median, of phase differences between time- adjacent phase measurements. he method of any of claims 28 through 32 further comprising determining (450) a frequency correction based on the estimated frequency error. e method of any of claims 28 through 32 further comprising providing (204, 460) - to a central control node of the D-MIMO system - a measurement report, wherein the measurement report is indicative of the estimated frequency error. e method of claim 34 further comprising receiving (205, 470) an indication of a frequency correction from the central control node, wherein the frequency correction is based on the estimated frequency error for the corresponding access point. he method of claim 33 and/or 35 further comprising applying (480) the frequency correction. e method of claim 36, wherein the frequency correction is applied for one or more of: a reference oscillator, a frequency synthesizer, and a baseband signal rotation. e method of any of claims 24 through 37, performed by the access point (121, 122, 123,

124, 125, 126, 240, 250, 710). computer program product comprising a non-transitory computer readable medium

(800), having thereon a computer program comprising program instructions, the computer program being loadable into a data processing unit and configured to cause execution of the method according to any of claims 1 through 38 when the computer program is run by the data processing unit. apparatus for over-the-air frequency synchronization between a plurality of access points of a distributed multiple-input multiple-output, D-MIMO, system, the apparatus comprising controlling circuitry (620) configured to cause: determination of a group of one or more transmitting access points; and for each of the transmitting access points: scheduling of a plurality of reference signal transmission occasions, wherein a time between transmission occasions is dynamically set based on expected frequency error; and instruction of the transmitting access point to transmit a reference signal for frequency synchronization in each of the scheduled transmission occasions.

41. The apparatus of claim 40, wherein the controlling circuitry is further configured to cause - for each scheduled transmission occasion - instruction of access points, that are not instructed to transmit in the scheduled transmission occasion, to perform phase measurements during the transmission occasion.

42. The apparatus of any of claims 40 through 41, wherein the group of one or more transmitting access points comprises the plurality of access points, ora sub-set of the plurality of access points.

43. The apparatus of any of claims 40 through 42, wherein the controlling circuitry is further configured to cause - for an access point which has an expected frequency error that exceeds a frequency error threshold value - excluding that access point from the group of one or more transmitting access points.

44. The apparatus of claim 43, wherein a number of transmission occasions for a first transmitting access point is higher than a number of transmission occasions for a second transmitting access point when a distance from the first transmitting access point to the excluded access point is shorterthan a distance from the second transmitting access point to the excluded access point.

45. The apparatus of any of claims 40 through 44, wherein each transmission occasion is scheduled for one or more of the transmitting access points.

46. The apparatus of any of claims 40 through 45, wherein the controlling circuitry is further configured to cause assignment of different reference signals to different transmitting access points with one or more coinciding transmission occasions.

47. The apparatus of any of claims 40 through 46, wherein the reference signal is a tone, or a modulated signal.

48. The apparatus of any of claims 40 through 47, wherein the controlling circuitry is further configured to cause indication of one or more parameters of the reference signal to the transmitting access point and/or to access points configured to perform phase measurements on the reference signal.

49. The apparatus of any of claims 40 through 48, wherein a first time between transmission occasions based on a first expected frequency error is shorter than a second time between transmission occasions based on a second expected frequency error when the first expected frequency error is larger than the second expected frequency error.

50. The apparatus of any of claims 40 through 49, wherein the time between transmission occasions is shorter than half a period of a largest expected frequency error between the transmitting access point and any access point in the plurality of access points.

51. The apparatus of any of claims 40 through 50, wherein the expected frequency error comprises a frequency error determined during previous frequency synchronization event.

52. The apparatus of any of claims 40 through 51, wherein the expected frequency error comprises an individual frequency error for the transmitting access point.

53. The apparatus of any of claims 40 through 52, wherein the expected frequency error increases with increasing time since previous frequency synchronization event.

54. The apparatus of any of claims 40 through 53, wherein the controlling circuitry is further configured to cause setting of the expected frequency error to a maximum value for an initial frequency synchronization event.

55. The apparatus of any of claims 40 through 54, wherein the controlling circuitry is configured to cause the scheduling and instruction to be performed in response to triggering of a current frequency synchronization event.

56. The apparatus of claim 55, wherein the controlling circuitry is further configured to cause triggering of the current frequency synchronization event in response to one or more of: initialization of an access point of the plurality of access points, a change of user device location, determination that a predetermined time has elapsed since previous frequency synchronization event, and detection of a communication performance deterioration for one or more of the plurality of access points. e apparatus of any of claims 40 through 56, wherein the plurality of access points comprises one or more of: access points located in a same geographical area, access points within a same radio signal vicinity, and access points that provide for communication with a same user device location. e apparatus of any of claims 40 through 57, wherein the controlling circuitry is further configured to cause reception - from one or more of the plurality of access points - of a respective measurement report, wherein each measurement report is indicative of an estimated frequency error for the corresponding access point. he apparatus of claim 58, wherein the controlling circuitry is configured to cause determination of a respective frequency correction for one or more of the plurality of access points, wherein the frequency correction is based on the estimated frequency error for the corresponding access point. e apparatus of claim 59, wherein the controlling circuitry is configured to cause the determination of the respective frequency correction by causing application of bias for maintenance of an average frequency control setting among the plurality of access points. e apparatus of any of claims 59 through 60, wherein the controlling circuitry is further configured to cause conveyance of an indication of the respective frequency correction to the corresponding access point. central control node for a distributed multiple-input multiple-output, D-MIMO, system, wherein the central control node comprises the apparatus (600) of any of claims 40 through 61. apparatus for over-the-air frequency synchronization of an access point in relation to a plurality of access points of a distributed multiple-input multiple-output, D-MIMO, system, the apparatus comprising controlling circuitry (720) configured to cause: for each of a first plurality of reference signal transmission occasions, wherein a time between transmission occasions is dynamically set based on expected frequency error, performance of phase measurements for frequency synchronization, wherein the phase measurements are relative a transmitting access point during the transmission occasion. e apparatus of claim 63, wherein the controlling circuitry is further configured to cause: reception of an instruction to transmit a reference signal for frequency synchronization in each of a second plurality of scheduled transmission occasions, wherein a time between transmission occasions is dynamically set based on expected frequency error; and performance of reference signal transmission accordingly. e apparatus of claim 64, wherein the controlling circuitry is configured to cause the phase measurements to be performed for all reference signal transmission occasions except those comprised in the second plurality of scheduled transmission occasions. e apparatus of any of claims 63 through 65, wherein the controlling circuitry is further configured to cause reception of an instruction to perform the phase measurements during each transmission occasion of the first plurality of reference signal transmission occasions. e apparatus of any of claims 63 through 66, wherein the controlling circuitry is further configured to cause determination of an estimated frequency error for the access point based on the phase measurements. e apparatus of claim 67, wherein the estimated frequency error for the access point is determined as an average of respective frequency offsets relative transmitting access points of a collection of transmitting access points. e apparatus of claim 68, wherein the controlling circuitry is further configured to cause, for a transmitting access point corresponding to an unreliable frequency offset: exclusion of that transmitting access point from the collection of transmitting access points, or weighting of the frequency offset relative that transmitting access point before determining the average. e apparatus of any of claims 68 through 69, wherein the controlling circuitry is configured to cause the determination of the respective frequency offset by causing determination of a gradient of an affine function fitted to a phase-time representation of the phase measurements.

71. The apparatus of any of claims 68 through 69, wherein the controlling circuitry is configured to cause the determination of the respective frequency offset by causing determination of an average, or median, of phase differences between time-adjacent phase measurements.

72. The apparatus of any of claims 67 through 71, wherein the controlling circuitry is further configured to cause determination of a frequency correction based on the estimated frequency error.

73. The apparatus of any of claims 67 through 71, wherein the controlling circuitry is further configured to cause provision - to a central control node of the D-MIMO system - of a measurement report, wherein the measurement report is indicative of the estimated frequency error.

74. The apparatus of claim 73, wherein the controlling circuitry is further configured to cause reception of an indication of a frequency correction from the central control node, wherein the frequency correction is based on the estimated frequency error for the corresponding access point.

75. The apparatus of claim 72 and/or 74, wherein the controlling circuitry is further configured to cause application of the frequency correction.

76. The apparatus of claim 75, wherein the frequency correction is applied for one or more of: a reference oscillator, a frequency synthesizer, and a baseband signal rotation.

77. An access point for a distributed multiple-input multiple-output, D-MIMO, system, wherein the access point comprises the apparatus (700) of any of claims 63 through 76.

78. A distributed multiple-input multiple-output, D-MIMO, system comprising a central control node (110, 210, 610) according to claim 62 and a plurality of access points (121, 122, 123, 124, 125, 126, 240, 250, 710) according to claim 77.