SIETWORK NODE AND METHOD THEREIN FOR FOR DETERMINING A BEAM TO BE TRANSMITTED FOR AT LEAST A FIRST USER EQUIPMENT

ECHNICAL FIELD

Embodiments herein relate to a network node and a method therein. In particular, ley relate to determining a beam to be transmitted for at least a first User Equipment JE).

IACKGROUND

In a typical wireless communication network UEs, also known as wireless ommunication devices, mobile stations, and/or Stations (STAs), communicate via a iadio Access Network (RAN) to one or more core networks (CN). The RAN covers a eographical area, which is divided into service areas, or cell areas, which may also be Bferred to as a beam or a beam group. Each service area or cell area is served by a base tation, which may also be referred to as a radio network node, a radio access node, a Vi-Fi access point, a Radio Base Station (RBS), a NodeB (NB) or eNodeB (eNB). A ervice area or cell area is a geographical area where radio coverage is provided by the adio network node. The base station communicates over an air interface operating on adio frequencies with the wireless device within range of the radio network node.

A Universal Mobile Telecommunications System (UMTS) is a third generation (3G) slecommunication network, which evolved from the second generation (2G) Global !ystem for Mobile Communications (GSM). The UMTS terrestrial radio access network JTRAN) is essentially a RAN using Wideband Code Division Multiple Access (WCDMA) nd/or High Speed Packet Access (HSPA) for UEs. In a forum known as the Third Jeneration Partnership Project (3GPP), telecommunications suppliers propose and agree pon standards for third generation networks, and investigate enhanced data rate and adio capacity. In some RANs, e.g. as in UMTS, several radio network nodes may be onnected, e.g., by landlines or microwave, to a controller node, such as a Radio Network !ontroller (RNC) or a Base Station Controller (BSC), which supervises and coordinates arious activities of the plural radio network nodes connected thereto. This type of onnection is sometimes referred to as a backhaul connection. The RNCs and BSCs arepically connected to one or more core networks. Specifications for the Evolved Packet System (EPS), also called a Fourth

Jeneration (4G) network, have been completed within the 3rd Generation Partnership 'roject (3GPP) and this work continues in the coming 3GPP releases, for example to pecify a Fifth Generation (5G) network. The EPS comprises the Evolved Universal errestrial Radio Access Network (E-UTRAN), also known as the Long Term Evolution _TE) radio access network, and the Evolved Packet Core (EPC), also known as System architecture Evolution (SAE) core network. E-UTRAN/LTE is a variant of a 3GPP radio ccess network wherein the radio network nodes are directly connected to the EPC core etwork rather than to RNCs. In general, in E-UTRAN/LTE the functions of an RNC are istributed between the radio network nodes, e.g. eNodeBs in LTE, and the core network, ^s such, the RAN of an EPS has an essentially "flat" architecture comprising radio etwork nodes connected directly to one or more core networks, i.e. they are not onnected to RNCs. To compensate for that, the E-UTRAN specification defines a direct lterface between the radio network nodes, this interface being denoted the X2 interface.

Multi-antenna techniques can significantly increase the data rates and reliability of a tireless communication network. The performance is in particular improved if both the ansmitter and the receiver are equipped with multiple antennas, which results in a lultiple-lnput Multiple-Output (MIMO) communication channel. Such systems and/or slated techniques are commonly referred to as MIMO.

Future access technologies are expected to support a large amount of transmit ntennas, and especially on the network side. In the context of Massive MIMO, as an xample, the number of antennas is expected to be huge, where a single transmission oint may have in the order of several hundreds or even thousands of antenna elements, v large, albeit much smaller, number of antennas may potentially be expected also in the IE at the high carrier frequencies, since the physical size of the antenna elements at lose frequencies may be made very small.

For the LTE standard, 3GPP has introduced the Channel State Information (CSI)- ieference Signals (RS) for CSI acquisition for up to 8 antenna ports, in see e.g. in !elease 1 1 , 3GPP 36.21 1 and 36.213 for CSI reporting procedures. Further, up to 16 ntenna ports see e.g. in 3GPP Release 13 and in the coming 3GPP Release 14 up to 32 oris 8 antenna ports. This enables UEs to evaluate a channel between the eNB and the IE for up to 8 independent channels on the eNB side. The UE evaluates the channel and sports CSI. This increased number of antenna elements, makes it possible to form more irective antenna patterns as compared to what is possible with the older antenna ystems of today. The more capable system can focus its transmitted and/or received ignal much more efficiently towards the UE being served, whilst suppressing the lterference from/to other UEs. Each transmit direction towards a UE is typically referred D as a beam, whereas the entire process of focussing energy towards a UE being served 5 referred to as beam forming.

Active Antenna Systems (AAS) are an important part of LTE evolution and an ssential part of 5G.

AAS is a generic term that is often used to describe radio base stations that icorporate a large number of separate transmitters that can be used for MIMO and eamforming, and integrate active transmitter components and radiating elements. There re several advantages to AAS implementation. One is that the integrated design reduces )ss factors and can reduce overall power consumption. Secondly, form factor, i.e. size nd weight may be reduced. Related to this, there is some potential for site simplification, or LTE, AAS base stations may offer beamforming functionality, such as cell splitting, ariable down tilt and user specific beamforming. In 3GPP, a number of air interface nhancements have been specified including the possibility of up to 16 CSI-RS ports and ;SI-RS beamforming to facilitate the exploitation of AAS base stations.

It is envisaged that a 5G air interface design, also referred to as NX, and 5G may perate in higher frequency bands than today. For example, 4GHz is discussed for first ystems in Japan, whilst World Radio Conference 2015 (WRC 15) may allocate spectrum p to 6GHz. Further into the future, it is envisaged that International Telecommunications Inion (ITU) and/or regional regulators may allocate millimeter wave spectrum in in the ange 10-100GHz. Radio waves in this band have wavelengths from ten to one millimeter, iving it the name millimeter band or millimeter wave, also referred to as MMW or mmW.

At higher frequencies, propagation losses are much greater than in today's bands, urthermore, it is envisaged that transmissions will take place within higher bandwidths. !ince the transmit power of both base stations and devices is limited by physical onstraints and considerations such as Maximum allowed Electromagnetic Field strength EMF), it is not possible to compensate the increased penetration losses and provide ufficient Signal to Interference and Noise Ratio (SINR) within wider bandwidths simply rith increased transmit power. In order to achieve the link budgets required for high data ates, beamforming will be necessary. It is therefore expected that integrated active arrays nil become a mainstream base station building practice in the 5G era.

In general, to comply with Radio Frequency (RF) EMF exposure limits, AAS base tations will be installed so that access to the compliance boundary is prevented. The ompliance boundary of a base station is described as a volume around the transmitting ntenna aperture outside which the exposure level is below a specified limit. The ompliance boundary of current macro base stations may extend several meters in the lain beam direction. In Figure 1 , the maximum Equivalent Isotropic Radiated Power EIRP) is plotted as function of compliance distance with respect to the International !ommission on Non-Ionizing Radiation Protection (ICNIRP) and the Federal

!ommunications Commission (FCC) limits.

Some methods to reduce transmitting peak power exist already today. The methods lay comprise selection of modulation and coding schemes to limit the amount of peak lipping that needs to be performed, or other smart algorithms to reduce the peak power t the output of the transceiver. However, the solutions available do not consider the patial domain. For an AAS base station, it is very likely that more beamforming apabilities will be part of the design.

In order to achieve beamforming, an AAS base station will possess a number of )gical components as illustrated in Figure 2. Figure 2 illustrates an example of an AAS rchitecture. The base station comprises and Antenna Array (AA), a Radio Distribution letwork (RDN), a Transceiver Unit Array (TXRUA) and a Baseband processing. The aseband processing may be used to perform user specific Transmit (TX) and Receive RX) beamforming. Although a logical part of the base station, the baseband processing lay not be physically co-located with the other components. The TXRUA comprises ctive circuits that may perform actions such as signal conditioning, amplification and Itering in transmit and receive. There may be a different number of transmitters to Bceivers and the transmitters and receivers may be implemented as single modules or eparately. The RDN distributes TX signals between the TXRUA and the transmit antenna lements and RX signals between receive antenna elements and the TXRUA. The RDN lay comprise splitting and combining of signals. The antenna array comprises a group of adiating elements, i.e. single antennas. The radiating elements may be TX only, or RX nly, or both RX and TX. Beamforming is performed by applying amplitude and/or phase variations to the ignals radiated from different antenna elements. The amplitude and phase variations lay be applied at any stage in the architecture of Figure 2. Typically, for very dynamic nd/or user specific beamforming, the amplitude and phase weights are set in the aseband.

A very common type of array, known as a uniform linear array is shown in Figure 3. he depicted array comprises a set of antenna elements arranged in one dimension with uniform spacing.

It is possible for the depicted uniform linear array to transmit beams at different ngles with respect to the antenna plane, as depicted in Figure 4. Figure 4 illustrates an xample of different beams steered from a uniform linear array.

A simple means to direct a beam is to apply a so-called linear phase progression, alculated as follows:

.2πζ _η cos{3 _c )

Where:

Yl is an index into the antennas,

Z _n is the position of the n ^th antenna along the z axis in the coordinates system, W« are the phase weights applied at each of the N antennas, λ is the wavelength, and

3

c is the beam steering direction,

More complex types of array include 2D linear arrays, arrays with non-uniform ntenna spacing in a single plane and conformal arrays, in which antenna elements are rranged in three dimensions.

The 3GPP specifications include a number of methods for generating precoding weights. A commonly used method is codebook based precoding, in which a standardized odebook consists of sets of beamforming weights. Each codebook entry comprises a set f weights that can generate a beam. One of the codebook entries is selected, according D feedback from the UE. Transmissions modes 4, 5, 6, 9 and 10 use codebook based recoding, see 3GPP 36.213.

Alternatively, transmission modes 7, 8 and 9 include a UE specific demodulation Bference signal that is precoded along with the symbols carrying data. This enables the ase station to select precoding weights in any manner without the need for any odebook. An example of a means for selecting weights is reciprocity based precoding in ime Division Duplex (TDD), in which knowledge of the TDD channel based on uplink leasurements is used to derive downlink precoding weights.

An AAS array may typically use methods such as those described above to direct eams in different directions at different times. A beam is directed towards a particular UE r group of UEs when the UE is scheduled. By directing energy using beamforming, 3ceived signal power to the scheduled UE may be increased whilst interference towards ther UEs is decreased.

An interesting parameter relating to array is a spatial profile of directed radiated ower such as the so-called Equivalent Isotropic Radiated Power (EIRP). In radio ommunication systems, EIRP or, alternatively, equivalent isotropically radiated power is ie amount of power that a theoretical isotropic antenna, which evenly distributes power in II directions, would emit to produce the peak power density observed in the direction of laximum antenna gain. EIRP can take into account the losses in transmission line and onnectors and includes the gain of the antenna. The EIRP is often stated in terms of ecibels over a reference power emitted by an isotropic radiator with an equivalent signal trength. The EIRP allows comparisons between different emitters regardless of type, size r form. From the EIRP, and with knowledge of a real antenna's gain, it is possible to alculate real power and field strength values.

EIRP is associated with a direction in relation to the base station. EIRP is the mount of power that would need to be input to an ideal, isotropic antenna in order to xperience the same amount of field strength in the direction under consideration as is xperienced from the actual antenna. If the actual antenna array has directivity, then in sality a lower power will need to be provided to the antenna than needs to be provided to n isotropic antenna.

A particularly important value of EIRP is the EIRP in the direction of the main lobe of ie transmitted signal. Figure 5 depicts a simplified example of a beam EIRP profile with irection in one dimension. Regulators typically specify a maximum amount of power that may be transmitted by base station. The maximum power level may be driven by interference or health onsiderations. The maximum power may be a conducted level, that is, it may be the sum f the power output of each transmitter unit in the transceiver unit array of Figure 2.

lowever, regulators may also place a restriction on peak EIRP from the base station.

The purpose of applying UE specific beamforming is to increase the EIRP in the irection of the scheduled user, in order to increase SINR and achieve higher data rates, lowever, a restriction on the maximum EIRP will limit the amount of beamforming gain lat can be achieved. In the worst case, the EIRP limit may limit the range and data rate upportable by the base station.

E.g., a base station vendor may declare that a certain set of beams can be ansmitted with a declared rated maximum EIRP per beam. For example, it may be that a ase station can transmit a group of cell specific beams for one or more cells with a eclared EIRP level. Furthermore, the specification such as e.g. 3GPP, service provider, lanufacturer or operator, may allow for a tolerance interval around the declared EIRP alue. For example, a tolerance of +-2dB around the declared EIRP may be allowable, or example, the current value in the 3GPP 37.105 specification is now 2.2dB.

However, it may be a problem when the allowed tolerance around the declared !IRP is exceeded. If the power should fall below the lower end of the tolerance range or se above the upper end of the tolerance range then the base station would no longer be ompliance with the specified requirement, and may also breach health and /or lterference limits.

!UMMARY

It is therefore an object of embodiments herein to further improve the performance f a wireless communications network.

According to a first aspect of embodiments herein, the object is achieved by a lethod performed by a network node for determining a beam to be transmitted to at least first User Equipment, UE.

The network node determines a beam to be transmitted to at least a first UE based n an obtained average spatial profile of radiated power in each direction. The average patial profile of radiated power is based on a spatial profile of radiated power averaged ver any one or more out of a frequency interval and a time interval.

According to a second aspect of embodiments herein, the object is achieved by a etwork node for determining a beam to be transmitted to at least a first User Equipment, IE. The network node is configured to determine a beam to be transmitted to at least a rst UE, based on an obtained average spatial profile of radiated power in each direction, Ί which the average spatial profile of radiated power is based on an spatial profile of adiated power averaged over any one or more out of a frequency interval and a time iterval.

Since the average spatial profile of radiated power is taken into account when a earn for a UE is determined, the average power over time e.g. considered by a regulator lay be reduced. This in turn results in that the performance of the wireless

ommunications network is further improved.

IRIEF DESCRIPTION OF THE DRAWINGS

Examples of embodiments herein are described in more detail with reference to ttached drawings in which: igure 1 is a schematic diagram illustrating below a maximum EIRP plotted as function of compliance distance with respect to ICNIRP and FCC limits.

igure 2 is a schematic block diagram Ilustrating prior art.

igure 3 is a schematic block diagram Ilustrating prior art.

igure 4 is a schematic block diagram Ilustrating prior art.

igure 5 is a schematic block diagram Ilustrating prior art.

igure 6 is a schematic block diagram Ilustrating a declared tolerance range, according to prior art.

igure 7 is a schematic block diagram illustrating embodiments of a communications network.

igure 8a is a schematic diagram of a spatial profile of EIRP averaged over frequency igure 8b is a schematic diagram illustrating an example embodiment related to a

frequency domain.

igure 8c is a schematic diagram of a spatial profile of EIRP averaged over time. gure 9 is a flowchart depicting embodiments of a method in a network node, gure 10 is a schematic diagram illustrating an example embodiment herein, gure 1 1 is a schematic diagram illustrating an example embodiment herein, gure 12 is a schematic diagram illustrating an example embodiment herein, gure 13 is a schematic block diagram illustrating embodiments of a network node, gure 14 is a schematic diagram illustrating an example embodiment herein, gure 15 is a flowchart depicting embodiments of a method in a network node, gure 16 is a schematic diagram illustrating an example embodiment herein, gure 17 is a flowchart depicting embodiments of a method in a network node, gure 18 is a flowchart depicting embodiments of a method in a network node gure 19 is a schematic block diagram illustrating embodiments of a network node.

⁾ ETAILED DESCRIPTION

As part of developing embodiments herein, the inventors realized some problems lat first will be discussed.

As mentioned above, regulators may place a restriction on peak radiated power, uch as EIRP, which will be used as a specific example of radiated power in many of the escribed embodiments below, from the base station due to e.g. health considerations, ypically, such a restriction is a limitation of the radiated power in a certain direction, mich is often calculated as an average of radiated power in a certain direction over a ^• equency interval or a time interval. However, when using beams this limit of average adiated power in a certain direction may decrease the gain of beamforming. Therefore, it 5 desirable to be able to transmit a beam in a certain direction without exceeding the limit f average radiated power in said direction.

A solution to the above stated problem may be achieved by calculating the average adiated power in a certain direction over either frequency or time, and rescheduling the earn at a different frequency or at a different point in time since the average would therwise exceed the limit, wherein the limit may be seen as a threshold. This allows a ase station to achieve high beamforming gains while avoiding exceeding the threshold.

In addition, as mentioned above, a base station manufacturer or service provider lay declare that a certain set of beams may be transmitted with a declared rated laximum radiated power or EIRP per beam. For example, a base station may transmit a roup of cell specific beams for one or more cells with a declared average spatial EIRP rofile level. Furthermore, a specification e.g. related to a service provider, manufacturer r operator may allow for a tolerance interval around the declared EIRP value. For xample, a tolerance of +-2dB around the declared average spatial EIRP profile may be llowable.

However, it may be a problem when the allowed tolerance around the declared verage spatial EIRP profile is exceeded. If the average spatial EIRP profile would rise bove a threshold or would fall below the threshold, the base station would not be ompliant to requirements. Likewise if the declared average EIRP for the beams would ot be kept.

This is since currently, if the average spatial EIRP profile is known and declared ccurately, there is no method to adapt the EIRP in beams dependent on inter beam load nd traffic conditions whilst keeping the average spatial EIRP profile to the declared value nd meeting the tolerance limit. The average special EIRP profile may also be referred to s the average spatial profile of EIRP.

Figure 6 illustrates an example of a declaration of an average spatial EIRP profile Dr a beam in a specific beam pointing direction, together with a tolerance range around ie declared average spatial EIRP profile within which the instantaneous EIRP must 3main. In this figure, as in figures 10,1 1 ,12,15,16,17 and 18, the "average EIRP" in the gure text corresponds to an average spatial profile of EIRP.

!xample embodiments herein provide a method to adapt the EIRP in beams, e.g.

ependent on inter beam load and traffic conditions. This makes it possible to keep the verage spatial EIRP profile below a regulated average spatial EIRP profile limit ccording to some embodiments and to stay within a tolerance range of a declared verage spatial EIRP profile meeting a tolerance limit according to some other mbodiments.

Embodiments of the method relates to determining a beam to be transmitted for at iast a first UE in a wireless communication network, such as for example scheduling earns to meet an EIRP limit.

Some embodiments relate to a network node for performing beamforming to one or lore UEs in which the network node calculates the beam pattern it transmits to a UE and eeps track of the average spatial profile of radiated power,_such as e.g. the average patial EIRP profile, in each direction around the base station. The average being over ^• equency or time. This is such that the network node can determine beams and/or take cheduling decisions to avoid exceeding a threshold for average spatial EIRP profile in ny individual direction, or to maintain average spatial EIRP profile at a specific level in ertain directions.

Figure 7 is a schematic overview depicting a wireless communication network

00 in which embodiments herein may be implemented. The wireless communication etwork 100 comprises one or more RANs and one or more CNs. The wireless ommunication network 100 may use a number of different technologies, such as Wi-Fi, ong Term Evolution (LTE), LTE-Advanced, 5G, Wideband Code Division Multiple Access i/VCDMA), Global System for Mobile communications/enhanced Data rate for GSM !volution (GSM/EDGE), Worldwide Interoperability for Microwave Access (WiMax), or lltra Mobile Broadband (UMB), just to mention a few possible implementations,

!mbodiments herein relate to recent technology trends that are of particular interest in a G context, however, embodiments are also applicable in further development of the xisting wireless communication systems such as e.g. WCDMA and LTE.

A network node 110 operates in the wireless communication network 100. The etwork node 1 10 provides radio coverage over a geographical area, which may also be Bferred to as providing beams or a beam group of beams. The network node 1 10 uses a adio access technology (RAT), such as 5G, LTE, Wi-Fi or similar. The network node 1 10 lay be a transmission and reception point e.g. a radio access network node such as a Vireless Local Area Network (WLAN) access point or an Access Point Station (AP STA), n access controller, a base station, e.g. a radio base station such as a NodeB, an volved Node B (eNB, eNode B), a base transceiver station, a radio remote unit, an access Point Base Station, a base station router, a transmission arrangement of a radio ase station, a stand-alone access point or any other network unit capable of

ommunicating with a UE within the service area served by the radio network node 1 10 epending e.g. on the first radio access technology and terminology used. The network ode 1 10 comprises a multiple antenna system, for example an AAS array and may be Bferred to as an AAS node.

A number of UEs operate in the wireless communication network 100, such as e.g. ne or more first UEs 121 and one or more second UEs 122. The first and second UEs 21 , 122 may each be a mobile station, a non-access point (non-AP) STA, a STA, a tireless terminal, communicate via one or more Access Networks (AN), e.g. RAN, to one r more Core Networks (CN). It should be understood by the skilled in the art that "UE" is non-limiting term which means any terminal, wireless communication terminal, user quipment, Machine Type Communication (MTC) device, Device to Device (D2D) srminal, or node e.g. smart phone, laptop, mobile phone, sensor, relay, mobile tablets or ven a small base station communicating within the wireless communications network 00.

A method for or determining a beam to be transmitted to at least the first user UE 21 , is performed by the network node 1 10. As an alternative, a Distributed Node (DN) 18 and functionality, e.g. comprised in a cloud 119 as shown in Figure 7, may be used Dr performing or partly performing the method.

Embodiments and implementations will be described in terms of radiated power in a pecific direction (θ,φ) such as EIRP, and are exemplified by meeting limits of the adiated power such as a limit of the average spatial EIRP profile in a direction and eclarations of average EIPR. However, generally the embodiments herein may be pplied where there are other metrics of directional field strength relating to interference r exposure limits.

Regarding the terms used herein:

EIRP is a means of expressing transmit power in a specific direction (θ,φ). It may e generically used for any direction. It may be referred to as radiated power in a specific irection (θ,φ).

Where Θ is the azimuth angle in relation to the base station, and where φ is the levation angle in relation to the base station .

Spatial profile of radiated power may e.g. be EIRP(6,q>,f=F,t=T); The profile of !IRP for each (θ,φ) at time T and frequency F. Other metrics of power may be devised uch as field strength or relative to a different type of antenna.

A spatial power profile is a function of (θ,φ) that provides the radiated power in very (θ,φ) combination, applicable for a specific time and frequency. Thus spatial power rofile = power (θ,φ, t, f). A Spatial EIRP profile is similar to spatial power profile, but where the power is xpressed as EIRP.

Averaged spatial profile of radiated power may also be referred to as the verage spatial power profile, and is an average over a time interval and/or a frequency iterval of the spatial power profile. E.g. when averaged over time, ASPP = power (theta, hi, f). When averaged over both, ASPP = power(theta,phi).

Averaged spatial profile of radiated power may be described as

iean(EIRP(6,(p,f=all F,t=T) or EIRP(6,(p,f=F,t=all T) or EIRP(6,(p,f=all F,t= all T); i.e. !IRP in EIRP(6,q>) averaged over either time, or frequency or both.

In a typical multi-user scenario in the wireless communication network 100, the etwork node 1 10 may schedule different beams in different beam directions in order to 2ach different users such as the one or more first UEs 121 and one or more second UEs 22 at different frequency intervals. The network node 1 10 such as a scheduler in the etwork node 1 10 may decide to transmit different beams with different beam directions n different resource blocks. An LTE channel is divided into resource blocks. This implies lat over frequency, there may not be a single beam pointed in a single direction. If the !IRP is averaged over frequency in any one direction, the EIRP will be different to the !IRP experienced in any single beam. See Figure 8a. Figure 8a depicts frequency ivided into resource blocks (1 ,2,3,4,5). Averaging according to embodiments herein, may ivolve averaging the EIRP in each RB 1 ,2,3,4,5. See also Figure 8b illustrating a spatial rofile of EIRP in direction (θ,φ) averaged over 8 RBs.

A similar consideration may apply in the time domain^ The network node 1 10 may chedule different beams in different beam directions in order to reach different users uch as the one or more first UEs 121 and one or more second UEs 122 at different oints in time. This implies that a beam is not pointed in a single direction continuously. Vhen the radiated power such as e.g. the EIRP is averaged in each beam direction over period of time, the average spatial EIRP profile will be different to the EIRP experienced uring the period of time that a beam is pointed in a particular direction. The advantage is lat a beam may be pointed in a particular direction for a period of time with higher power lan the allowed average power. This would allow more beamforming gain in order to rovide the first UE 121 with a higher data rate. See Figure 8c. Figure 8c illustrates time ivided into intervals (1 ,2,3,4,5). Averaging according to embodiments herein, may involve veraging the EIRP in each interval 1 ,2,3,4,5.

The network node 1 10 such as the scheduler of the network node 1 10 may use a irection dependent metric to keep track of the average spatial EIRP profile transmitted in ach direction around the network node 1 10. The average may be the average over time, r the average over frequency, or both.

The method according to embodiments herein takes average spatial profile of adiated power, also referred to as spatial distribution of the radiated power, into account rhen a beam for a UE is determined, such as e.g. average spatial EIRP profile. This is .g. to reduce transmission power, or to apply less beamforming.

The information of average spatial profile of radiated power such as e.g. the verage spatial EIRP profile may at least be used for two purposes. The first purpose is D ensure that the average radiated power such as e.g. transmitted EIRP, remains below threshold level for meeting e.g. a regulatory requirement specified in one or more irections. The second purpose is to ensure that the average spatial profile of radiated ower, such as the average spatial EIRP profile, remains equal within a tolerance range f, a predetermined profile, at least in certain directions, in order to meet a manufacturer eclaration, whilst allowing the radiated power e.g. EIRP for some individual resource locks or beams to deviate from the profile for certain times or frequencies.

Example embodiments of a method performed by the network node 1 10 for etermining a beam to be transmitted to at least the first UE 121 will be described in a eneral way with reference to a flowchart depicted in Figure 9. After this a more detailed xplanation and exemplification follows.

The method comprises the following actions, which actions may be taken in any uitable order. Actions that are optional are presented in dashed boxes in Figure 9.

According to an example scenario, a beam is to be transmitted to at least the first IE 121 to enable data transmission between the network node 1 10 and the at least first IE 121 .

Action 901 A spatial profile of radiated power in each of different directions from the network ode 1 10 need to be established, e.g. to be able to determine or schedule a beam so that does not exceed a regulated limit of average radiated power, such as a regulated ireshold, in the direction of the beam during the scheduled frequency and/or time of the earn, or to be able to determine or schedule a beam so that it stays within a tolerance ange of a declared average of radiated power in the beam direction. Thus, the network ode 1 10 optionally establishes a spatial profile of radiated power in each of different irections from the network node 1 10, related to beams transmitted in each of the Bspective different directions.

According to an example scenario, extra parameters may be added into a cheduling algorithm in the network node 1 10, such that e.g. a scheduler in the network ode 1 10 may establish, e.g. estimate the radiated EIRP in every direction and direction f maximum EIRP when a beam shall be determined for a UE such as the first UE 121 eing scheduled. In order to make this estimation, the network node 1 10 e.g. the cheduler of the network node 1 10 may use information about the architecture of its ansceiver unit array and RDN and the geometry of the transmitter antenna array.

Information about the architecture of a transceiver unit array and/or RDN of the etwork node 1 10 may be needed for the estimation since the spatial pattern of EIRP is irectly related to the geometry of the antenna array and the mapping of transceivers to ntenna elements. Information about the geometry of the transmitter antenna array may e needed for the estimation since the geometry of the antenna array directly impacts the patial profile of EIRP.

When the network node 1 10 has information about its transceiver unit array the ntenna array geometry and/or RDN then the network node 1 10 will be able to establish, .g. by calculating a spatial profile of radiated power, such as e.g. an EIRP profile, for any earn that is transmitted, whether the beam is based on codebook based precoding or ome other method such as reciprocity.

Action 902

The network node 1 10 may then obtain an average spatial profile of radiated power Ί each direction by averaging the established spatial profile of radiated powers in each of ie respective directions, averaged over any one or more out of: a frequency interval and time interval. A purpose of averaging may be to establish that a requirement on average patial EIRP profile set over a defined time or frequency interval is met. This means that the network node 1 10 obtains one average spatial profile for each irection, wherein each one of said average spatial profiles are based on radiated power Ί said direction during different times or different frequencies.

The instantaneous spatial profile at any time and frequency is created by estimating ie radiated power, such as EIRP in each direction. The average spatial profile is created y averaging instantaneous spatial profiles over time or frequency.

Action 903

When the network node 1 10 shall determine the beam for at least the first UE 121 it nil take the average spatial profile of radiated power in each direction into account. It is n advantage since to take the average spatial profile of radiated power in in each irection into account will allow for the network node 1 10 to achieve high beamforming ain whilst not radiating too little or excessive average power in any direction.

The network node 1 10 thus determines a beam to be transmitted to at least the first IE 121 , based on an obtained average spatial profile of radiated power in in each irection. The average spatial profile of radiated power is based on a spatial profile of adiated power averaged over any one or more out of a frequency interval and a time iterval. The determined beam may be user specific.

The radiated power in each direction is optionally represented by an EIRP in the irection of the beam. The EIRP is the radiated power in an individual direction. The patial profile is the set of EIRP in every direction of the beam, including the main lobe irection of the beam.

The average spatial profile of radiated power may include or not include the stimated radiated power spatial profile of the beam to be transmitted to the at least first IE 121 . For a prospective beam transmission, the power to be radiated if the beam was D be transmitted is estimated, and the estimated radiated power of the prospective beam 5 then optionally included in the average profile of radiated power. It depends on the nplementation. For some implementations the average spatial profile may include the stimate to ensure that the threshold would not be exceeded if the beam where to be ansmitted. In other implementations, it may be that if the threshold is near to being xceeded even without the estimated radiated power, then the beam is not transmitted ut may be rescheduled in a different frequency or at a different point in time.

Action 904

The network node 1 10 may then transmit the determined beam. The following may e.g. relate to the first purpose of ensuring that the average patial profile of radiated power, such as the average spatial EIRP profile, shall remain elow a threshold for meeting e.g. a regulatory requirement.

Thus, in some embodiments, the average spatial profile of radiated power includes ie estimated radiated power of the beam to be transmitted to the at least first UE 121 . he determining of the beam may be performed such that the average spatial profile of adiated power does not exceed a threshold, wherein the threshold is a regulated limit for ie average spatial profile of radiated power in a certain direction at a certain frequency nd point in time. Further, the determining of the beam may be performed by adjusting eamforming weights for the beam such that the average spatial profile of radiated power oes not exceed a threshold in any direction. E.g. check that threshold is not exceeded in ie direction to transmit the beam in and check in any other direction that threshold is not xceeded there.

In some alternative embodiments, the average spatial profile of radiated power does ot include the estimated radiated power of the determined beam to be transmitted for the t least first UE 121

The determining of the beam may be performed by determining the beam to be ansmitted in a different frequency e.g. a frequency or frequency range (resource blocks) ther than for which the obtained spatial profile of radiated power exceeds a threshold in ie direction of the beam, when the average spatial profile of radiated power averaged ver a frequency interval exceeds a threshold. The average spatial profile of radiated ower exceeds the threshold in any direction, in particular the direction of the beam.

As an alternative the determining of the beam is performed by determining the eam to be transmitted at a different point in time, e.g. a time other than that for which the btained profile of radiated power exceeds a threshold in the direction of the beam, when ie average spatial profile of radiated power averaged over a time interval exceeds a ireshold.

The threshold mentioned above may be represented by a limit for average radiated ower in a direction from the network node.

The following may e.g. relate to the second purpose ensuring that the average patial profile of radiated power, such as the average spatial EIRP profile, remains ssentially equal to a predetermined profile and e.g. within a tolerance range of the eclared average spatial EIRP, at least in certain directions, in order to meet a lanufacturer declaration, whilst allowing the radiated power e.g. EIRP for some individual Bsource blocks or beams to deviate from the profile for certain times or frequencies The etermining of the beam may further be performed such that any one or more out of: The verage spatial profile of radiated power is within a tolerance range of a declared average patial profile of radiated power, and e.g. within a tolerable instantaneous deviation from a eclared average spatial profile of radiated power.

The determining of the beam may comprise determining the beam to have a adiated power that is any one out of increased and decreased, compared to a declared ivel of available power, as long as the determined beam stays within a tolerance range of ie declared spatial profile of radiated power. This will be further explained below.

The determining of the beam may comprise temporarily increasing or decreasing ie radiated power in the direction of the beam, such that the average spatial profile of adiated power in the direction of the beam remains within a tolerance range of a declared verage spatial profile of radiated power, and such that the average radiated power profile f all the beams in the direction is equal to the declared average spatial profile of radiated ower. This will be further explained below.

Embodiment's herein will now be further described and explained. The text below is pplicable to and may be combined with any suitable embodiment described above. EIRP 5 used as an example of radiated power.

First purpose, keeping transmitted EIRP below a threshold level

The application behind the first purpose may be to ensure that regulatory EIRP mits are met. In these examples, the radiated power in each direction is represented by n EIRP in each direction of the beam, but may as well be any kind of average spatial rofile of radiated power.

For frequency dependent averaging, the frequency domain is broken into a number f averaging intervals.

Referring to Action 901 in which the network node 1 10 establishes a spatial profile f radiated power such as EIRP in each of different directions from the network node 1 10, slated to beams transmitted in each of the respective different directions, e.g. related to t least one beam transmitted. This establishing may be performed by a different node, rhich then informs the network node 1 10 of the result. The EIRP is established e.g.

alculated in each possible direction. The EIRP in a particular direction is calculated for ach respective beam that is or will be transmitted to each respective one of the UEs such s e.g. the first UE 121 and the second UEs 122. This is performed in each possible irection considering the beam pointing direction of the beam and e.g. the radiation attern for the beam to be determined for the UE 121 , and applied to the each one of the Bsource blocks scheduled to the first UE 121 . Considering herein means that, to alculate the EIRP in each direction, the calculation will need to take the beam pointing irection and the radiation pattern into account. When the beam pointing direction is ligned with the direction in which the EIRP is calculated, then the EIRP will be high; therwise the EIRP may be low. This means that the average spatial profile of EIRP may e increased in the direction where the EIRP is high, but decreased in directions where ie EIRP is low.

Referring to Action 902 in which the network node 1 10 obtains an average spatial rofile of radiated power e.g. an average spatial EIRP profile in each direction by veraging over a frequency interval, the established spatial profile of radiated powers in ach of the respective directions. This obtaining may be performed by a different node, rhich then informs the network node 1 10 of the result. In each direction, the average patial EIRP profile is then calculated in the considered direction across all Resource llocks (RBs) within the frequency interval to be averaged. Figure 8a mentioned above is schematic diagram illustrating a frequency interval such as a frequency domain, broken p in averaging intervals 1 , 2, 3, 4, 5, each of them e.g. a 100kHz averaging interval for ie purpose of EIRP averaging. See also Figure 8b mentioned above illustrating EIRP in irection (θ,φ) averaged over 8 RBs.

Referring to Action 902 in which the network node 1 10 obtains an average spatial rofile of radiated power e.g. an average spatial EIRP profile in each direction, here by veraging over a time interval, the established spatial profile of radiated powers in each of ie respective directions. In each direction, the average spatial EIRP profile is then alculated in the considered direction across past, and possible also present

ransmission Time Intervals (TTIs) or similar time units to be averaged. Figure 8c lentioned above is a schematic diagram illustrating a time interval such as a time omain, broken up into averaging time intervals 1 , 2, 3, 4, 5 for the purpose of EIRP veraging.

For time dependent averaging, when a beam is scheduled in a particular direction, ie average spatial profile of radiated power such as the EIRP is increased. If no beam is cheduled in a particular direction, the average spatial profile of radiated power such as ie EIRP decreases. This means that if the EIRP is high in a direction, it will tend to lcrease the average; if it is low, it will decrease it. The increase or decrease in the verage spatial profile of radiated power such as the EIRP in each direction is dependent pon the transmission (TX) power and beam pattern in that direction. The average spatial rofile of radiated power such as the EIRP may be increased not only in the beam ointing direction, but also in other directions in which significant energy is radiated by the earn or even potentially in every direction according to the calculated spatial power rofile of the beam. The network node 1 10 can then be configured such that it schedules earns in directions in which the average is low, whilst avoiding directions in which the verage is high. In this way, beams may be scheduled with high EIRP whilst the average lay be maintained within a threshold or tolerance range.

Figure 10 shows an example of a profile of average spatial EIRP with respect to irection in one dimension, wherein the maximal average EIRP threshold is the limit for ie maximal average spatial profile of EIRP that is tolerated in a direction according to sgulations . For frequency domain averaging, Figure 10 depicts an average spatial EIRP rofile for a particular frequency averaging interval. For time domain averaging, the figure hows a snapshot of average spatial EIRP profile at a particular point in time. For implicity purposes, the figure considers the average spatial EIRP profile only in one imension, Θ; however if an array can steer in two dimensions then the average spatial !IRP profile would be tracked in both dimensions.

Referring to Action 903 in which the network node 1 10 determines a beam to be ansmitted to at least the first UE 121 , based on an obtained average spatial profile of adiated power e.g. the average spatial EIRP profile in each direction. When making cheduling decisions such as determining a beam for the first UE 121 , the network node 10, such as e.g. its scheduler, takes into account that the average spatial EIRP profile is ot allowed to exceed a certain threshold. By examining the metrics such as the average patial profile of radiated power e.g. the EIRP in each direction, the network node 1 10 is ble to either (i) avoid scheduling UEs such as the first UE 121 in such a manner that the verage spatial EIRP profile threshold would be exceeded, or (ii) adjust the beamforming weights that it will use when scheduling certain users such as the first UE 121 in order to void exceeding the EIRP threshold. If the EIRP in a particular direction is close to the laximum, then the network node 1 10 may avoid transmitting a beam in the affected irection by means of determining the beam either in a different frequency averaging iterval or at a different point in time in the future. This may be achieved by means of cheduling different UEs including the first UE 121 to those beams for which transmitting ie best beam would lead to exceeding the average spatial EIRP profile limit, e.g. by cheduling different UEs using beams in different directions to the direction in which ansmitting the best beam to the first intended UE would lead to exceeding the average patial EIRP profile limit in any direction. As an alternative, it may be achieved by cheduling the intended users including the first UE 121 but with a less optimal beam irection such that the UEs are still reached but the EIRP threshold is not exceeded.

Second purpose, keeping the average spatial EIRP profile to a prescribed alue in certain directions.

The application behind the second purpose may be one in which e.g. a manufacture as declared a certain level of EIRP to be available for certain beams such as cell specific earns, within a tolerance level. In these examples, the radiated power in each direction is spresented by an EIRP in each direction of the beam, but may as well be any kind of verage spatial profile of radiated power.

Apart from the beams for which the EIRP is and has to be maintained and kept fithin a tolerance range of the declared average EIRP, and for which beams the average !IRP over time or frequency in a certain direction has to be maintained at the declared verage EIRP, other beams are to be determined and scheduled, for example user pecific beams.

From time to time, it may be useful to e.g. temporarily increase the EIRP for the ther beams, for example, to increase data rates or reach cell edge UEs In order to rovide power for these beams, resources such as may be reduced for the beams for rhich EIRP is maintained. Resources that may be reduced include power and ansmitters. The reduction in EIRP as a result of the reduced resources may be any mount such that the EIRP remains within the tolerance range of the declared average, igure 11 is a schematic diagram illustrating an example of deliberately deciding to ansmit with an instantaneous EIRP (at instance X) in a specific direction below the eclared average EIRP profile value but within the tolerance range, such as when etermining the beam to be transmitted to at least the first UE 121 . It may be that the !IRP is instantaneously outside of the tolerance range. However, the average remains fithin the tolerance range. In some embodiments it may be that the average EIRP profile alue is remaining at the declared value and the instantaneous EIRP within a tolerance ange as a further alternative In order to keep the average spatial EIRP profile to the declared level, at other times ie EIRP provided to the beams may be boosted by means of increasing the amount of 3sources available for the beams. See Figure 12, illustrating an example of deliberately eciding to transmit with an EIRP at instance Y in a specific direction above the declared alue but within the tolerance range, such as when determining the beam to be ansmitted to at least the first UE 121. Instantaneously it may be out of the tolerance ange, but the average shall stay within the range.

This means that in some embodiments, the average should remain within a

Dlerance range, and the instantaneous may be anything. However in some alternative mbodiments the average is remaining at the declaration and the instantaneous is within tolerance range.

The embodiments relating to the first purpose, i.e. keeping the transmitted average !IRP profile below a threshold level, allow for the network node 1 10 to achieve high eamforming gain whilst avoiding exceeding average spatial EIRP profile thresholds. This nil allow for user specific beamforming to be exploited such that better coverage and igher data rates are obtained.

The benefits of the embodiments relating to the first purpose, i.e. keeping ansmitted average EIRP below a threshold level, are dependent upon regulatory EIRP mits being defined as an average spatial EIRP profile rather than an instantaneous EIRP. !urrently regulatory limits specify a frequency averaging interval for EIRP, and thus ^• equency averaging is most applicable. At this point in time, regulators typically do not onsider time average spatial EIRP profile, however this is because regulators have ased current requirements on passive antenna systems and not beamforming systems Dr which beam patterns may vary dynamically. When addressing AAS, 5G and eamforming systems, consideration may well be given to defining of a measurement me period for EIRP.

The embodiments relating to the second purpose, i.e. keeping the average spatial !IRP profile to a prescribed value in certain directions provide a means for a

lanufacturer to keep EIRP for certain beams to a declared average level and within a Dlerance range whilst temporarily re-assigning resources to other beams. This enables a etter utilization of power and transmitter resources, which can in turn reduce base station ize, energy consumption and cost. An example of the network node 1 10 for performing the method herein is illustrated Ί the schematic block diagram of Figure 13. The network node 1 10 comprises a cheduler 1301 , a beam intensity calculation unit 1302, a baseband 1303, a

ransceiver unit array 1304, an RDN 1305, an antenna array 1306 and other hardware nd software components responsible for management of the network node 1 10 and ansportation of data to and from the network node 1 10 over backhaul. In this example, ie RDN maps transmitters to TX antenna elements using a 1 : 1 mapping, and there are 2 transmitters and antenna elements. Other configurations of RDN and array are also ossible. A scheduler algorithm may also be included in the network node 1 10. The cheduler algorithm collects typical parameters, such as data buffer size, reports of CQI nd CSI, QoS information etc. from a UE such as the first UE 121 and second UEs122. according to embodiments herein, the scheduler1301 may also receive information on the verage spatial profile of radiated power such as e.g. the average spatial EIRP profile in ach of a set of spatial directions around the network node 1 10 e.g. from the beam itensity calculation unit 1302. Optionally, assistance information from scheduled UEs uch as the first UE 121 and possibly the second UEs122, may also be used for alculating the average spatial profile of radiated power such as e.g. the average EIRP.

The beam intensity calculation unit 1302 estimates the spatial profile of radiated ower such as e.g. the EIRP in each direction in space if a particular beam is scheduled uch as the determined beam for the first UE 121 . An example of a beam intensity pattern 5 provided in Figure 14.

Furthermore, the beam intensity calculation unit 1302 may maintain a record of the verage spatial profile of radiated power such as e.g. the average spatial EIRP profile in ach spatial direction. An example calculation of the average spatial EIRP profile

EIRPav) is shown in the equation below.

EIRPav (9, t) = 0.8* EIRPav (9, t-1) +(1 -0.8) EIRP(9, t)

Where:

EIRPav(0, t) is the average spatial EIRP profile in direction Θ at time t, and

EIRP(9, t) is the EIRP of the scheduled beam in direction Θ at time t.

0.8 is a factor that controls the time duration of the averaging. Two example embodiments are described below relating to the first purpose, eeping transmitted EIRP below a threshold level. In these examples and all examples elow, the radiated power is represented by an EIRP, but may as well be any kind of verage spatial profile of radiated power.

The first example embodiment is one in which the time domain averaging is used. 1 this example, the average spatial profile of radiated power is the average spatial EIRP rofile. The procedure for determining the beam and e.g. scheduling the first UE 121 aking into account average spatial EIRP profile for the first embodiment is outlined in the owchart of Figure 15.

Action 1501 . As mentioned above in Action 902, the network node 1 10 obtains an verage spatial profile of radiated power in each direction by averaging the established patial profile of radiated powers in each of the respective directions, averaged over any ne or more out of: a frequency interval and a time interval. The network node 1 10 such s e.g. its scheduler 1301 obtains the average spatial EIRP profile by request the beam itensity calculation unit 1302 to provide the current average spatial EIRP profile. Using ie example depicted in Figure 10, the average spatial EIRP is approaching the EIRP limit Dr some beam directions, whereas it is low for other beam directions.

Action 1502. The network node 1 10 such as e.g. its scheduler 1301 determines a eam to be transmitted, such as selects a beam to be transmitted. This may be performed y selecting a subset of UEs such as the first UE 121 for those beam directions in which ie average spatial EIRP profile is substantially below a threshold. The threshold may e.g. slates to regulatory EIRP limits. The threshold may be constant or may itself vary epending on spatial direction. Then based on metrics such as a Proportional Fair (PF) letric, CSI, buffer size etc. the network node 1 10 such as e.g. its scheduler 1301 selects UE such as the first UE 121 to schedule and a beam direction or set of precoding weights for the scheduled UE. Precoding weights may be generated from a codebook, or y some other means such as reciprocity based beamforming for TDD.

Action 1503. The network node 1 10 such as e.g. its scheduler 1301 then sends iformation on the selected beam and precoding weights to the beam intensity calculation nit 1302. The beam intensity calculation unit 1302 estimates a spatial distribution of nergy such as the spatial EIRP profile for the suggested beam and temporarily adds this eam intensity pattern, i.e. the spatial EIRP profile into the average spatial EIRP profile. Action 1504. The network node 1 10 such as e.g. its beam intensity unit 1302 then erifies that the maximum average spatial EIRP profile will not be exceeded in any irection if the planned beam is transmitted.

Action 1505. As mentioned above, the network node 1 10 determines a beam to be ansmitted to at least the first UE 121 , based on the obtained average spatial EIRP profile f radiated power in in each direction, and as long as the average EIRP threshold is not xceeded, the beam intensity unit 1302 indicates to the scheduler 1301 that the beam can e transmitted and permanently updates it's beam intensity profile. The average spatial !IRP profile now includes the estimated radiated power of the beam to be transmitted to ie at least first UE 121 . In case the average EIRP threshold would be exceeded, then the earn intensity calculation unit 1302 would indicate to the scheduler 1301 that the beam is ot suitable and would revert to its previous average spatial EIRP profile and a new beam lay be determined in Action 1502.

Action 1506. When the EIRP threshold is not exceeded, network node 1 10 ansmits the determined beam to the first UE 121 .

A number of variations may be conceived to the above architecture and procedure, or example, the scheduler 1301 may not receive information about the average spatial !IRP profile at all, and may poll the beam intensity calculation unit 1302 with different /pes of beam, receiving in each case an indication of whether the beam may be cheduled or not. Alternatively, the scheduler 1301 and beam intensity calculation unit 302 may be merged such that the scheduler 1301 may consider the average spatial !IRP profile over a longer term whilst planning its beam determining and scheduling ecisions.

In the second example embodiment, the frequency domain is broken up into a umber of frequency averaging intervals. The procedure depicted in the flowchart of igure 17 is carried out for each of the averaging intervals.

The network node 1 10 such as e.g. its scheduler 1301 may select 1702 different earns for different resource blocks in each frequency averaging interval. In case the laximum EIRP threshold in a particular direction would be exceeded in a particular "equency averaging interval, then the scheduler may choose to determining and schedule beam that is directed in an affected direction, i.e. in a direction where the EIRP exceeds ie threshold 1504, over fewer resource blocks within the frequency averaging interval, in rder to reduce the average spatial EIRP profile in the direction, or may select to schedule beam directed in the affected direction in a different frequency averaging interval for rhich the average spatial EIRP profile in the affected direction is lower.

In other embodiments relating to the second purpose, of keeping the average patial EIRP profile to a declared or prescribed value in certain directions, the aim of the etermining of the beam and scheduling may be to maintain the average spatial EIRP rofile for certain resource blocks at a certain threshold level and within a tolerance level, urthermore, the instantaneous EIRP at any individual time and frequency for those Bsource blocks may be restricted to not deviating from the threshold level by more than .g. a second acceptable tolerance.

In these embodiments, once UEs such as the first UE 121 are scheduled and the verage spatial EIRP profile is calculated by the beam intensity calculation unit 1302, the etwork node 1 10 such as e.g. its scheduler 1301 is able to increase and decrease the !IRP between different beams or between different carriers or resource blocks such that ie average spatial EIRP profile is maintained in all beam directions, and the

istantaneous EIRP of a beam s within a tolerance range of the average, here the istantaneous EIRP is within the tolerance range, not the average.

Figure 16 depicts an example of an average spatial EIRP profile 1601 in a irection averaged in either time or frequency, instantaneous EIRP 1602 in a irection. The instantaneous EIRP 1601 exceeds the threshold in a direction 1603, but lis is OK because the average does not exceed the threshold 1604.

The procedure for these embodiments is depicted in Figures 17 and 18.

The scenario behind the Figures 17 and 18 may e.g. be that a manufacturer has eclared an average EIRP that is available on cell wide beams transmitted from the base tation, carrying for example Common Reference Symbols (CRS), Broadcast Channel 3CH) etc. A cell wide beam is a beam that is intended to be able to be received by a UE uch as the first UE 121 that is anywhere within a cell area provided by a network node uch as the network node 1 10. There is some tolerance around the declared average !IRP value. The network node 1 10 is also able to transmit user specific beams for some lEs.

Figure 17 depicts a scheduling method allowing additional resources for user pecific beams and reduced EIRP for cell specific beams. At certain times, for UEs such s e.g. the first UE 121 with poor coverage the network node 1 10 needs to achieve a high !IRP for those UEs. At these times, the network node 1 10 over-allocates (increases) Bsources to the user specific beam and under allocates (decreases) to the cell specific earns using the procedure of Figure 17. The network node 1 10 such as e.g. its scheduler 501 is aware of data in the buffer for a number of users, as indicated in Figure 17. Figure 7 depicts a method for allowing additional resources for user specific beams and 2duced EIRP for cell specific beams. The cell specific beam EIRP is re-calculated with swer resources and the average spatial EIRP profile for the cell wide beam must remain fithin the tolerance limits.

Action 1701. The network node 1 10 calculates resources for user specific beam.

Action 1702. The network node 1 10 reduces resources for other beams with eclared EIRP.

Action 1703. The network node 1 10 then recalculates EIRP for beams with reduced Bsources.

Action 1704. The network node 1 10 checks if the EIRP is within the tolerance ange of the declared average EIRP level.

Action 1705. When the EIRP is within tolerance range, the network node 1 10 hecks if the EIRP average is correct, i.e. below a threshold.

Action 1706. When the EIRP average is correct, the network node 1 10 transmits earns.

Action 1707. When the EIRP is not within tolerance, or when the when the EIRP verage is not correct the network node 1 10 selects new beams to reduce resources for.

Action 1708. The network node 1 10 checks if the new beams can be selected.

When the new beams can be selected, the network node 1 10 acts according to action 1703.

Action 1709. When the new beams cannot be selected, the network node 1 10 3schedules a different user specific beam, and then acts according to Action 1701.

At other times, UEs in good coverage which e.g. may be the first UE 121 are cheduled with user specific beams. At these times, the network node 1 10 over-allocates Bsources to the cell specific beams. The reason for over-allocating resources is to keep ie average spatial EIRP profile for the cell specific beams at the declared level and within ie tolerance limits. After over-allocating resources, the remaining resources are used for ie user specific beams. Figure 18 illustrates a procedure for boosting the EIRP in cell pecific beams in order to maintain the average. Action 1801 . The network node 1 10 over-allocates resources, such as e.g.

Ilocates more resources than needed to some cell specific beams.

Action 1802. The network node 1 10 recalculates the EIRP for said some cell pecific beams.

Action 1803. The network node 1 10 then checks if EIRP is within tolerance. Action 1804. When the EIRP is within tolerance, the network node 1 10 update !IRP average.

Action 1805. The network node 1 10 then uses remaining resources to schedule ser specific beams.

Action 1806. When the EIRP is not within tolerance, the network node 1 10 re- llocates resources to cell specific beams and then acts according to Action 1802.

To perform the method actions for determining a beam to be transmitted to at least ie first UE, the network node 1 10 the may comprise the following arrangement depicted Ί Figure 19.

The network node 1 10 is configured to, e.g. by means of a determining module 910 configured to, determine a beam to be transmitted to at least a first UE 121 , based n an obtained average spatial profile of radiated power in each direction, in which the verage spatial profile of radiated power is based on an spatial profile of radiated power veraged over any one or more out of a frequency interval and a time interval.

The radiated power in any direction may be represented by an EIRP.

In some embodiments the average spatial profile of radiated power includes the stimated radiated power of the beam to be transmitted to the at least first UE 121 . In ome of these embodiments the network node 1 10 may further be configured to, e.g. by leans of the determining module 1910 configured to, determine the beam such that the verage spatial profile of radiated power does not exceed a threshold. In some other of lese embodiments, the network node 1 10 further is configured to, e.g. by means of the etermining module 1910 configured to, determine the beam by adjusting beamforming weights for the beam such that the average spatial profile of radiated power does not xceed a threshold in any direction. In some alternative embodiments, the average spatial profile of radiated power does ot include the estimated radiated power of the determined beam to be transmitted for the t least first UE 121 .

The network node 1 10 may further be configured to, e.g. by means of the etermining module 1910 configured to, determine the beam by:

determining the beam to be transmitted in a different frequency, when the average patial profile of radiated power averaged over a frequency interval exceeds a threshold.

The network node 1 10 may further be configured to, e.g. by means of the etermining module 1910 configured to, determine the beam to be transmitted at a liferent point in time, when the average spatial profile of radiated power averaged over a me interval exceeds a threshold.

The network node 1 10 may further be configured to, e.g. by means of the etermining module 1910 configured to, determine the beam such that any one or more ut of: the average spatial profile of radiated power is within a tolerance range of a eclared average spatial profile of radiated power.

The network node 1 10 may yet further be configured to, e.g. by means of the etermining module 1910 configured to, determine the beam to have a radiated power lat is any one out of increased and decreased compared to a declared level of available ower, as long as the determined beam stays within a tolerance range of the declared patial profile of radiated power.

The network node 1 10 may further be configured to, e.g. by means of the etermining module 1910 configured to, determine the beam by increasing or decreasing ie radiated power in the direction of the beam, such that the average spatial profile of adiated power in the direction of the beam remains within a tolerance range of a declared verage spatial profile of radiated power.

In some embodiments, the determined beam is user specific.

The threshold may be represented by a limit for average radiated power.

The network node 1 10 may further be configured to, e.g. by means of an

stablishing module 1920 configured to, establish a spatial profile of radiated power in ach of different directions from the network node 1 10, related to beams to be transmitted Ί each of the respective different directions.

The network node 1 10 may further be configured to, e.g. by means of an obtaining lodule 1930 configured to, obtain an average spatial profile of radiated power in each irection by averaging the established spatial profile of radiated power in each of the Bspective directions, averaged over any one or more out of: a frequency interval and a me interval.

The network node 1 10 may further be configured to, e.g. by means of a

ransmitting module 1940 configured to, transmit the determined beam.

The embodiments herein may be implemented through one or more processors, uch as a processor 1950 of a processing circuitry in the network node 1 10 depicted in igure 19, together with computer program code for performing the functions and actions f the embodiments herein. The program code mentioned above may also be provided as computer program product, for instance in the form of a data carrier carrying computer rogram code for performing the embodiments herein when being loaded into the network ode 1 10. One such carrier may be in the form of a CD ROM disc. It is however feasible rith other data carriers such as a memory stick. The computer program code may jrthermore be provided as pure program code on a server and downloaded to the etwork node 1 10.

The network node 1 10 may further comprise a memory 1960 comprising one or lore memory units. The memory 460 comprises instructions executable by the processor 950.

The memory 1960 is arranged to be used to store e.g. information about CSI of a Drward link channel, a first quality value, precoders, data, configurations, and applications D perform the methods herein when being executed in the network node 1 10.

In some embodiments, a computer program 1970 comprises instructions, which men executed by the at least one processor 1950, cause the at least one processor 1950 D perform actions according to any of the Actions 901 -904.

In some embodiments, a carrier 1980 comprises the computer program 1970, merein the carrier is one of an electronic signal, an optical signal, an electromagnetic ignal, a magnetic signal, an electric signal, a radio signal, a microwave signal, or a omputer-readable storage medium. Those skilled in the art will also appreciate that the modules in the first radio node 10, described above may refer to a combination of analog and digital circuits, and/or one r more processors configured with software and/or firmware, e.g. stored in the memory 60, that when executed by the one or more processors such as the processor 450 as escribed above. One or more of these processors, as well as the other digital hardware, lay be included in a single Application-Specific Integrated Circuitry (ASIC), or several rocessors and various digital hardware may be distributed among several separate omponents, whether individually packaged or assembled into a system-on-a-chip (SoC).

ABBREVIATIONS

Abbreviation Explanation

2D Two Dimensional

5G Fifth Generation

AA Antenna Array

AAS Active Antenna System

AE Antenna Element

CQI Channel Quality Information

CSI Channel State Information

CSI-RS Channel State Information (related Reference Symbols

EIRP Equivalent Isotropic Radiated Power

EMF Electromagnetic Field

FCC Federal Communications Commission

GHz Giga Hertz

ITU International Telecommunications Union

LTE Long Term Evolution

MIMO Multiple Input Multiple Output

ICNIRP International Commission on Non-Ionizing Radiation Protection

NX The name of a potential 5G air interface design

PF Proportional Fair

QoS Quality of Service

RDN Radio Distribution Network

RX Receive

RXU Receive Unit SINR Signal to Interference and Noise Ratio

TDD Time Division Duplex

TX Transmit

TXRUA Transceiver Unit Array

TXU Transmit Unit

WRC World Radio Conference