Broadcast and multicast transmission in distributed massive mimo networks

Abstract

Systems and methods for broadcast / multicast transmission in a distributed cell-less massive multiple-input multiple-output (MIMO) network are disclosed. In one embodiment, a method for broadcasting / multicasting data to a user equipment (UE) includes obtaining, at each of two or more access points (APs), long-term channel state information (CSI) for the UE, and transmitting, to a central processing system, the long-term CSI for the UE. The method further includes receiving, at the central processing system, the long-term CSI for the UE from each of the APs, computing, based on the long-term CSI, a precoding vector (w) for the UE over the APs, and transmitting, to the APs, the precoding vector (w). The method further includes obtaining, at each of the APs, the precoding vector (w) from the central processing system, precoding, based on the precoding vector (w), data to be broadcast / multicasted to the UE, and broadcasting / multicasting, to the UE, the precoded data.

Description

[0001] Related applications

[0002] This application claims the benefit of U.S. Provisional Patent Application No. 62 / 851,951, filed May 23, 2019, the disclosure of which is incorporated herein by reference in its entirety. Technical Field

[0003] This disclosure relates to distributed massive multiple-input multiple-output (DM-MIMO) systems, and more particularly, to broadcast or multicast transmissions in DM-MIMO systems. Background Technology

[0004] Massive Multiple-Input Multiple-Output (MIMO) (also known as massive antenna systems and ultra-massive MIMO) is a multi-user MIMO technology in which each base station is equipped with a large number (typically more than 50) of antenna elements to serve many terminals that share the same time and frequency band and are spatially separated. A key assumption is that there are far more base station antennas than terminals—at least twice as many, but ideally as many as possible. Massive MIMO offers several advantages over conventional multi-user MIMO. First, conventional multi-user MIMO is not a scalable technology because it is designed to support systems with roughly equal numbers of serving antennas and terminals, and practical implementations of conventional multi-user MIMO typically rely on frequency division duplex (FDD) operation. In contrast, in massive MIMO, the large number of redundant serving antennas on active terminals operating in time division duplex (TDD) results in a significant increase in throughput and radiated energy efficiency. These benefits stem from strong spatial multiplexing achieved through proper shaping of the signals transmitted and received by the base station antennas. By applying precoding to all antennas, the base station can ensure constructive interference between signals at the intended terminal location and destructive interference at almost all other locations. Furthermore, as the number of antennas increases, energy can be concentrated with extreme precision into small areas of space. Other benefits of massive MIMO include the use of simple, low-power components (because it relies on simple signal processing techniques), reduced latency, and robustness against intentional jamming.

[0005] When operating in TDD mode, massive MIMO can exploit the channel reciprocity property, according to which the channel response is the same in both uplink and downlink. Channel reciprocity allows the base station to obtain channel state information (CSI) from the pilot sequences transmitted by the terminals in the uplink, which is then useful for both uplink and downlink. According to the law of large numbers, the effective scalar channel gain seen by each terminal is close to a deterministic constant. This is known as channel hardening. Thanks to channel hardening, terminals can reliably decode downlink data using only long-term statistical CSI, thus making most of the physical layer control signaling redundant, i.e., low-cost CSI acquisition. This makes traditional resource allocation concepts unnecessary and leads to a simplification of the medium access control (MAC) layer. These benefits explain why massive MIMO occupies a central role in the fifth generation (5G).

[0006] However, massive MIMO system performance is affected by several limiting factors. Channel reciprocity requires hardware calibration. Moreover, the so-called pilot contamination effect is a fundamental phenomenon that deeply limits the performance of massive MIMO systems. In theory, each terminal in a massive MIMO system can be assigned a mutually orthogonal uplink pilot sequence. However, the upper limit to the maximum number of orthogonal pilot sequences that can exist is the size of the coherence interval, which is the product of the coherence time and the coherence bandwidth. Therefore, as the number of terminals increases, the adoption of orthogonal pilots leads to inefficient resource allocation, or it is physically impossible to perform when the coherence interval is too short. The result is that pilots must be reused across cells, or even within the home cell (for higher cell density). This inevitably leads to interference between terminals sharing the same pilot. Pilot contamination does not disappear as the number of base station antennas increases, so it is an asymptotically preserved impairment.

[0007] In order to implement massive MIMO in a wireless network, two different architectures can be adopted:

[0008] • Centralized massive MIMO (C-maMIMO), in which all antennas, both at the base station side and at the user side, are co-located in a compact area, as shown in Figure 1 . It represents a traditional massive MIMO system.

[0009] • Distributed massive MIMO (D-maMIMO), in which base station antennas (herein referred to as access points (APs)) are geographically scattered in a large area in a carefully planned or random manner, as shown in Figure 2The antennas are connected together and to a central processing unit (CPU) through a high-capacity backhaul link (e.g., optical cable). It is also referred to as a cell-free massive MIMO system.

[0010] The D-maMIMO architecture is an important enabler for network MIMO in future standards. Network MIMO is a term used for a cell-free wireless network in which all base stations deployed within a coverage area act as a single base station with distributed antennas. From a performance perspective, this can be considered as an ideal network infrastructure as the network has a strong ability to spatially multiplex users and accurately control the interference caused to each person.

[0011] The difference between D-maMIMO and traditional distributed MIMO lies in the number of antennas involved in coherently serving a given user. In D-maMIMO, each antenna serves each user. Compared to C-maMIMO, D-maMIMO has the potential to improve both network coverage and energy efficiency due to the increase in macro-diversity gain. This comes at the cost of higher fronthaul requirements and distributed signal processing needs. In D-maMIMO, information related to payload data and power control coefficients is exchanged via the backhaul network between the APs and the CPU. There is no exchange of instantaneous CSI between the APs or the central unit, i.e., CSI acquisition can be performed locally at each AP.

[0012] Due to the network topology, D-maMIMO suffers from different degrees of path loss caused by different access distances to different distributed antennas, and very different shadowing phenomena that are not necessarily better (e.g., antennas deployed at street level are more likely to be blocked by buildings than antennas deployed at elevated locations). Moreover, since the location of the antennas in D-maMIMO has a significant impact on system performance, optimization of the antenna locations is crucial. Furthermore, D-maMIMO systems can suffer from low degrees of channel hardening. As mentioned previously, the channel hardening property is key in massive MIMO to suppress small-scale fading and stems from the large number of antennas involved in coherent transmission. In D-maMIMO, the APs are distributed over a wide area and many APs are far away from a given user. Therefore, each user is actually served by a small number of APs. As a result, channel hardening can be less pronounced. This will greatly impact system performance.

[0013] The performance of any wireless network obviously depends on whether there is good enough CSI to facilitate phase-coherent processing at multiple antennas. Intuitively, it should be easier to obtain high-quality CSI using C-maMIMO than using D-maMIMO with antennas distributed over a large geographical area. Nonetheless, the macro-diversity gain is of significant importance and leads to improved coverage and energy efficiency.

[0014] One problem with massive MIMO deployments is that the large number of antennas generates a large amount of data. This means that for a traditional radio-to-antenna interface, a very large capacity fiber network is needed to mix this data. Fiber is both expensive and requires skilled personnel for installation, both of which limit the deployment scenarios for massive MIMO. There is also a scalability problem, as different sizes of baseband units are needed to handle different array sizes, e.g. one handling 32 antennas, another for 128 antennas, etc.

[0015] From a practical point of view, the C-maMIMO solution, where all antenna elements (i.e. APs) are placed close together, has many disadvantages compared to the D-maMIMO solution, where the antenna elements are distributed over a larger area. These disadvantages are for example:

[0016] • Very large service variation: User Equipment (UEs) that happen to be located close to the central massive MIMO node will experience very good service quality, while for more distant UEs the service quality will quickly degrade.

[0017] • Sensitive to blockage: Especially at high frequency bands, the signal is easily blocked by obstacles that block the line of sight between the UE and the C-maMIMO node. In D-maMIMO, multiple antenna elements can be blocked, but it takes a much larger obstacle to block all antenna elements.

[0018] • High heat concentration: Due to heat concentration, it is very difficult to make the C-maMIMO node very small. In D-maMIMO, each antenna element (and its associated processing) only generates a small amount of heat, which simplifies miniaturization.

[0019] • Large and visible installation: The C-maMIMO installation can become large, especially at lower frequency bands. The D-maMIMO installation is actually even larger, but can be made so that the visual impact is almost negligible.

[0020] • Installation requires personnel with “radio skills”: Installing complex hardware in a single location requires planning, and it is likely that it also needs to be installed correctly by certified personnel. In a D-maMIMO installation, it is less important to install each of the very many antenna elements in a very good location. It is sufficient that most of the elements are installed in a good enough location. The D-maMIMO deployment can significantly relax the installation requirements.

[0021] • Regulatory-restricted power (e.g., specific absorption rate (SAR)): If antenna elements are close together, there will be an area near the installation where electromagnetic safety regulations apply. This can limit the total radiated radio frequency power in many installations. In a D-maMIMO installation, a user may be close to a small number of antenna elements, but not physically close to many elements distributed over a large area.

[0022] Compared to C-maMIMO, D-maMIMO offers many significant advantages. However, in existing technology solutions, wiring between antenna elements and internal communication within D-maMIMO are not feasible. Connecting a separate cable between each antenna element and the CPU in a D-maMIMO installation (e.g., in a star topology) is economically impractical. Any or optimal AP topology can lead to excessively high costs for backhaul components, as well as installation costs for distributed processing and setup.

[0023] The principle of “radio stripes” was previously described in WO2018 / 103897A1, entitled “Improved Antenna Arrangement for Distributed Massive MIMO”. The actual “base station” in a “radio stripe system” can include circuit-mounted chips within a protective housing of a cable or strip. Receive and transmit processing for each antenna element is performed after that actual antenna element itself. Because it is assumed that the total number of distributed antenna elements is large (e.g., hundreds), the RF transmit power of each antenna element is very low.

[0024] Figure 3 The example depicts a system model and shows a light-emitting diode (LED) lighting strip connected to a box. This diagram is for illustrative purposes only and illustrates how a practical distributed massive MIMO base station can be constructed. The CPU (or strip station) is connected to one or more radio strips (or distributed MIMO active antenna cables).

[0025] A real radio strip may contain tape or adhesive on the back, as in the example of the LED strip, or it may simply contain a very small daily radio processing unit and antenna protected by plastic covering the cable.

[0026] An important observation to be made is that both transmitter and receiver processing can be distributed under certain assumptions, for example, see [link to relevant documentation]. Figure 4 Using low-complexity precoding methods such as conjugate beamforming, each antenna element can be equipped with a control entity (antenna processing unit (APU)) that determines beamforming weights without communicating with all other APUs.

[0027] One area of interest for distributed massive MIMO is on-demand broadcast and / or multicast transmission over a specific area or complete coverage area in 5G New Radio (NR). The existing fourth generation (4G) Long Term Evolution (LTE) system supports broadcast and multicast transmission over a wide area using Single Frequency Network (SFN) or Single Cell Point-to-Multipoint mode of operation (currently named Multimedia Broadcast Multicast Service (MBMS)). Specifically, in Multimedia Broadcast Multicast Service Single Frequency Network (MBSFN), multiple base stations over multiple cells transmit the same data in the same resource block through a special frame dedicated for MBMS service. Alternatively, in Single Cell Point-to-Multipoint (SC-PTM), the same data is transmitted to multiple users in a single cell using Physical Downlink Shared Channel (PDSCH). It is expected that 5G NR access technology will also soon support this operation. However, the radio session setup procedure for LTE MBMS is complex and time consuming. In either of the above modes, MBMS requires separate user plane infrastructure for connecting the Radio Access Network (RAN) with the core network. Furthermore, broadcast transmission in LTE is always an additional functionality with the above-mentioned limitations. It is not trivial to directly adapt LTE MBMS for 5G NR while achieving fast and efficient radio session and joint unicast, broadcast and multicast transmission.

[0028] Depending on the application, low and high rate broadcast transmission with guaranteed reliability can be required. A first example includes broadcast of signals for synchronization, system information and paging during initial access operation, which must be transmitted over a wide area but not necessarily at high rate. In certain cases, such as live events or in a stadium, higher rate data transmission can be required and there is no time or radio resources to obtain instantaneous or small-scale CSI with uplink reference signals before transmitting data to the UEs.

[0029] Therefore, there is a need for systems and methods for efficient use of beamforming (e.g., open-loop transmission used in LTE or NR) in networks with distributed massive MIMO deployment. SUMMARY

[0030] Systems and methods for broadcast or multicast transmission in a distributed cell-free massive multiple-input multiple-output (MIMO) network are disclosed. In one embodiment, a method for broadcasting or multicasting data to user equipments (UEs) in a distributed cell-free massive MIMO network includes, at each of two or more access points (APs): obtaining long-term channel state information (CSI) for at least one UE; and transmitting the long-term CSI for the at least one UE to a central processing system. The method further includes, at the central processing system: for each of the two or more APs, receiving the long-term CSI for the at least one UE from the AP; computing precoding vectors w for the at least one UE over the two or more APs based on the long-term CSI for the at least one UE received from the two or more APs; and transmitting the precoding vectors w to the two or more APs. The method further includes, at each of the two or more APs: obtaining the precoding vectors w from the central processing system; precoding data to be broadcast or multicast to the at least one UE based on the precoding vectors w; and broadcasting or multicasting the precoded data to the at least one UE. In this way, an efficient way of on-demand data broadcasting in a distributed or cell-free massive MIMO network can be provided. Moreover, the reliability of the broadcast / multicast can be greatly improved. Furthermore, the scheme is suitable for operation at microwave and millimeter wave frequencies.

[0031] An embodiment of a method performed at a central processing system is also provided. In one embodiment, a method performed at a central processing system for broadcasting or multicasting data to UEs for a distributed cell-free massive MIMO network includes, for each of two or more APs in the distributed cell-free massive MIMO network, receiving long-term channel state information (CSI) for at least one UE from the AP; computing precoding vectors w for the at least one UE over all of the two or more APs based on the long-term CSI for the at least one UE received from the two or more APs; and transmitting the precoding vectors W to the two or more APs.

[0032] In one embodiment, the at least one UE is two or more UEs.

[0033] In one embodiment, computing the precoding vector (w) comprises: (a) at iteration 0, initializing the precoding vector w to provide a precoding vector w(0) for iteration 0, (b) computing a local average signal-to-interference-and-noise ratio (SINR) for the at least one UE, (c) identifying a weakest UE from among the at least one UE based on the computed local average SINR, (d) at iteration n+1, updating the precoding vector w based on the long-term CSI obtained from the at least one UE for the weakest UE to provide a precoding vector w(n+1) for iteration n+1, (e) normalizing the precoding vector w(n+1) for iteration n+1, and (f) repeating steps (b) through (e) until a stopping criterion is met, such that the normalized precoding vector for the last iteration is provided as the precoding vector w. In one embodiment, the stopping criterion is convergence or a maximum number of iterations has been reached.

[0034] In one embodiment, initializing the precoding vector w comprises initializing the precoding vector w to a value

[0035] In one embodiment, computing the local SINR for the at least one UE comprises computing the local SINR for the at least one UE as k k = Pw * Θ k w, k = 1,..., K. In one embodiment, the weakest UE is identified as the UE whose index k' is

[0036] In one embodiment, updating the precoding vector w comprises updating the precoding vector w at iteration n+1 as w(n+1) = (I + μΘ k′ )w(n).

[0037] In one embodiment, normalizing the precoding vector w(n+1) for iteration n+1 comprises normalizing the precoding vector w(n+1) for iteration n+1 as w(n+1) = w(n+1) / ‖w(n+1)‖.

[0038] In one embodiment, the long-term CSI comprises an estimated path loss.

[0039] In one embodiment, a coherence interval of the long-term CSI spans greater than 1,000 symbols, and the method further comprises scheduling, by the at least one UE, transmission of a dedicated pilot.

[0040] ​​A corresponding embodiment of a central processing system is also disclosed. In one embodiment, a central processing system for a distributed cell-less massive MIMO network for broadcasting or multicasting data to UEs is disclosed, the central processing system adapted to: receive, from each of two or more APs in the distributed cell-less massive MIMO network, long-term CSI for at least one UE; compute, based on the long-term CSI for the at least one UE received from the two or more APs, a precoding vector w for the at least one UE across all of the two or more APs; and transmit the precoding vector w to the two or more APs.

[0041] In one embodiment, a central processing system for a distributed cell-less massive MIMO network for broadcasting or multicasting data to UEs is disclosed, the central processing system comprising a network interface and processing circuitry associated with the network interface. The processing circuitry is configured to cause the central processing system to: receive, from each of two or more APs in the distributed cell-less massive MIMO network, long-term CSI for at least one UE; compute, based on the long-term CSI for the at least one UE received from the two or more APs, a precoding vector w for the at least one UE across all of the two or more APs; and transmit the precoding vector w to the two or more APs.

[0042] An embodiment of a method performed at an AP is also disclosed. In one embodiment, a method performed at an access point, AP, in a distributed cell-less massive MIMO network for broadcasting or multicasting data to UEs is disclosed, wherein the network comprises two or more APs, the method comprising: obtaining long-term CSI for at least one UE; transmitting the long-term CSI for the at least one UE to a central processing system; obtaining a precoding vector w from the central processing system, wherein the precoding vector w is for the at least one UE across all of the two or more APs; precoding data to be broadcast or multicasted to the at least one UE based on the precoding vector w; and broadcasting or multicasting the precoded data to the at least one UE.

[0043] In one embodiment, the at least one UE is two or more UEs.

[0044] In one embodiment, the long-term CSI comprises an estimated path loss.

[0045] In one embodiment, the coherence interval of the long-term CSI spans greater than 1,000 symbols, and obtaining the long-term CSI for the at least one UE comprises estimating the long-term CSI for the at least one UE over a fixed time interval based on transmission of dedicated pilots to the at least one UE.

[0046] A corresponding embodiment of an AP is also disclosed. In one embodiment, an access point (AP) in a distributed cell-less massive MIMO network is provided for broadcasting or multicasting data to UEs, wherein the network comprises two or more APs, wherein the AP is adapted to obtain long-term CSI for at least one UE; transmit the long-term CSI for the at least one UE to a central processing system; obtain a precoding vector w from the central processing system, the precoding vector w being for the at least one UE on all of the two or more APs; precode data to be broadcast or multicasted to the at least one UE based on the precoding vector w; and broadcast or multicast the precoded data to the at least one UE.

[0047] In one embodiment, an access point (AP) in a distributed cell-less massive MIMO network is provided for broadcasting or multicasting data to UEs, wherein the network comprises two or more APs, wherein the AP comprises a network interface; at least one transmitter; at least one receiver; and processing circuitry associated with the network interface, the at least one transmitter, and the at least one receiver. The processing circuitry is configured to cause the AP to obtain long-term CSI for at least one UE; transmit the long-term CSI for the at least one UE to a central processing system; obtain a precoding vector w from the central processing system, wherein the precoding vector w is for the at least one UE on all of the two or more APs; precode data to be broadcast or multicasted to the at least one UE based on the precoding vector w; and broadcast or multicast the precoded data to the at least one UE. BRIEF DESCRIPTION OF DRAWINGS

[0048] The accompanying drawings, which are incorporated in and form a part of the specification, illustrate several aspects of the present disclosure and, together with the description, serve to explain the principles of the present disclosure.

[0049] Figure 1 A centralized massive multiple-input multiple-output (C-maMIMO) is shown;

[0050] Figure 2 A distributed massive multiple-input multiple-output (D-maMIMO) is shown;

[0051] Figure 3An example of a strip architecture for a D-maMIMO system is shown;

[0052] Figure 4 An example in which both transmitter and receiver processing in a D-maMIMO system are distributed is shown;

[0053] Figure 5 An example of a D-maMIMO system in which embodiments of the disclosure can be implemented is shown;

[0054] Figure 6 A comparison of the cumulative distribution of the local average signal to interference and noise ratio (SINR) for a baseline precoder and an iteratively optimized precoder according to an example embodiment of the disclosure is shown;

[0055] Figure 7 Performance advantages of an example embodiment of an iteratively optimized precoder are shown;

[0056] Figure 8 An iterative process for computing precoding weight vectors for broadcasting / multicasting data from multiple access points (APs) in a D-maMIMO system to user equipment (UEs) according to one embodiment of the disclosure is shown;

[0057] Figure 9 Operation of a D-maMIMO system according to some embodiments of the disclosure; Figure 5 in the context of a fifth generation (5G) new radio (NR) access network is shown;

[0058] Figure 10 and 11 is a schematic block diagram of an example embodiment of an AP according to the disclosure;

[0059] Figures 12 to 14 is a schematic block diagram of an example embodiment of a central processing system according to the disclosure;

[0060] Figure 15 An example embodiment of a communication system in which embodiments of the disclosure can be implemented is shown;

[0061] Figure 16 An example embodiment of a host computer, a base station and a UE of Figure 15 is shown; and

[0062] Figures 17 to 20 is a flowchart of an example embodiment of a method implemented in a communication system such as Figure 15 is shown. DETAILED DESCRIPTION

[0063] The examples set forth below represent the necessary information to enable those with ordinary skill in the art to practice the examples and demonstrate the best mode of practicing the examples. Upon reading the following description, one skilled in the art will understand how to implement the concepts to support the examples and applications explained below. The concepts and applications explained herein are not limited to the examples described below, which are intended only to demonstrate the broadness and nature of the concepts and applications.

[0064] Generally, all terms used herein are to be interpreted according to their ordinary meaning in the technical field, unless a different meaning is clearly given and / or is implied from the context in which it is used. All references to a / an / the element, apparatus, component, means, step, etc. are to be interpreted openly as referring to at least one instance of whatever is referred to unless indicated otherwise. The steps of any methods disclosed herein do not have to be performed in the exact order disclosed, unless explicitly stated as such. Any of the embodiments disclosed herein can be applied to any other embodiments, wherever technically feasible. Similarly, any advantages, features, or benefits described in relation to any of the embodiments disclosed herein can be applied to any other embodiment, wherever technically feasible. Further objects, features and advantages of the embodiments disclosed herein will become apparent from the following description.

[0065] Radio node: As used herein, a “radio node” is a radio access node or a wireless device.

[0066] Radio access node: As used herein, a “radio access node” or “radio network node” is any node in a radio access network (RAN) of a cellular communications network for wirelessly transmitting and / or receiving signals. Some examples of radio access nodes include, but are not limited to, base stations (e.g., a NR base station (gNB) in a Third Generation Partnership Project (3GPP) 5G NR network or an enhanced or evolved Node B (eNB) in a 3GPP LTE network), high- power or macro base stations, low-power base stations (e.g., micro, pico, home eNBs, etc.), and relay nodes.

[0067] Core network node: As used herein, a “core network node” is any type of node in a core network. Some examples of core network nodes include, for example, a mobility management entity (MME), a packet data network gateway (P-GW), a service capability exposure function (SCEF), etc.

[0068] Wireless device: As used herein, a “wireless device” is any type of device that accesses a cellular communications network (i.e., is served by it) by wirelessly transmitting and / or receiving signals to a radio access node. Some examples of wireless devices include, but are not limited to, a UE in a 3GPP network and a machine-type communication (MTC) device.

[0069] Network node: As used herein, a "network node" is any node that is part of the radio access network or core network of a cellular communications network / system.

[0070] Note that the description given herein focuses on 3GPP cellular communications systems, and thus, 3GPP terminology or terminology similar to 3GPP terminology is often used. However, the concepts disclosed herein are not limited to 3GPP systems.

[0071] Note that in the description herein, reference can be made to the term "cell"; however, especially with respect to 5G NR concepts, beams can be used instead of cells, and thus, it is important to note that the concepts described herein equally apply to both cells and beams.

[0072] There are certain challenges at present. While existing LTE Multimedia Broadcast Multicast Service (MBMS) supports broadcast transmission with Multimedia Broadcast Multicast Service Single Frequency Network (MBSFN) and Single Cell Point to Multipoint (SC-PTM), there are some limitations. Specifically, in MBSFN, multiple time-synchronized base stations transmit the same content in the same resource blocks over a larger area. The system capacity of MBSFN is limited by the spatial distribution of users, especially by the channel quality of the worst-case users. Alternatively, in SC-PTM, data is broadcast on a per-cell basis using the Physical Downlink Shared Channel (PDSCH); however, performance suffers from inter-cell interference. It can be desirable to further simplify the complex radio session establishment procedure of LTE MBMS and tailor it for 5G NR to enable fast RAN sessions and efficient utilization of resources.

[0073] In traditional cellular systems, antennas are concentrated at specific locations of the base station working at low carrier frequencies. In emerging distributed massive multiple-input multiple-output (D-maMIMO) networks, each user is served by multiple antenna points, and the link quality of distributed MIMO is relatively robust to blockage (i.e., the probability that all links suffer from blockage is small due to macroscopic diversity). Therefore, D-maMIMO will be suitable to work at both microwave and millimeter wave carrier frequencies. However, there is no efficient solution for broadcasting data content in such cell-less / distributed networks.

[0074] One approach is to employ omnidirectional transmission at each antenna point, which can serve as a baseline for the solutions proposed in this disclosure.

[0075] Certain aspects of the present disclosure and embodiments thereof can provide solutions to the above-described or other challenges. In some embodiments, long-term channel state information (CSI) is leveraged at an access point (Ap). Based on this information:

[0076] • users (i.e., wireless devices or UEs) receiving multicast or broadcast (multicast / broadcast) information are associated with the APs; and / or

[0077] • precoders are determined for each AP based on the long-term CSI for users receiving the same multicast / broadcast information.

[0078] In some embodiments, an iterative procedure is provided for computing a plurality of precoder weights at the APs. In each iteration, the plurality of precoder weights at the APs are computed based on the average signal-to-interference-and-noise ratio (SINR) of the worst-case user (i.e., worst-case wireless device or UE).

[0079] In some embodiments, a distributed cell-less massive MIMO network is provided. The distributed cell-less massive MIMO network includes at least two APs, at least one UE, and a central processing unit (CPU). The CPU can also be referred to herein as a central unit (CU) or more generally as a "central processing system" and should be distinguished from a computer architecture "central processing unit" or "CPU" (e.g., an Intel CPU, such as an Intel i3, i5, or i7 CPU), which can be referred to herein as a "computer CPU" or "computer central processing unit." Each AP obtains long-term CSI (e.g., filtering or processing based on uplink measurements), which is communicated to the CPU. At the CPU, an optimization (e.g., iterative) is performed to compute precoding vectors at all APs. Each AP then broadcasts a precoded multicast transmission to UEs that need the same content. Furthermore, in some embodiments, the statistical channel information between the APs and the UEs includes or consists of an estimated path loss. By exploiting signals transmitted and received during initial access (i.e., before radio resource control (RRC) connection setup), the AP estimates the long-term CSI related to UEs within its coverage area. For example, each UE transmits a physical random access channel (PRACH) preamble in the uplink, which can be used to estimate the link average signal-to-noise ratio (SNR), which is equivalent to the long-term channel gain between the AP and the UE. Since the massive channel coherence interval spans thousands of symbols, in some embodiments, the network can schedule a dedicated pilot in the uplink and reliably estimate the long-term CSI in a fixed time interval.

[0080] In some embodiments, the long-term CSI available at each AP is transmitted to a central cloud, an iterative algorithm is performed at the central cloud, and then the precoding weights are communicated to the APs to influence the APs prior to data transmission. In other words, the CPU can be implemented "in the cloud."

[0081] Particular embodiments can provide one or more of the following technical advantages. Embodiments of the present disclosure provide an efficient way of on-demand data broadcasting in distributed or cell-less massive MIMO. The reliability of the broadcast can be greatly improved. Embodiments of the present disclosure provide a broadcast / multicast scheme suitable for operation at both microwave and millimeter wave frequencies.

[0082] Figure 5 One example of a D-maMIMO system 500 in which embodiments of the present disclosure can be implemented is shown. As shown, the D-maMIMO system 500 includes a plurality of APs 502-1 through 502-N (collectively referred to herein as APs 502 and individually as AP 502) that are geographically scattered, e.g., in a planned or random manner, over an area (e.g., a large area). The APs 502 are connected to a central processing system 504 (e.g., a CPU) via corresponding backhaul links (e.g., high-capacity backhaul links such as optical cables). The D-maMIMO system 500 is also referred to as a cell-less massive MIMO system. The APs 502 provide radio access to a plurality of UEs 506. In some embodiments, the D-maMIMO system 500 is a cellular communication system, e.g., a 5G NR system, or is part of a cellular communication system, e.g., a 5G NR system.

[0083] A description of some embodiments of the present disclosure will now be provided.

[0084] Consider a distributed massive MIMO system (e.g., the D-maMIMO system 500 of Figure 5 ), in which APs (e.g., the APs 502) and users (e.g., the UEs 506) are characterized by a single or multiple antennas. Each AP serves K strongest users within a certain range (e.g., R c ). Each user is then served by multiple APs that are uniformly distributed within the range R c . For example, in a typical transmission, M denotes the number of APs that serve a considered user (indexed by k). For illustration, embodiments of the present disclosure are first presented for the case of single-antenna APs and single-antenna users, and then generalized to the case of multiple antennas. The corresponding precoding / beamforming vectors at the various APs are assumed to be w = [w1, w2,..., w M ] T where w m is the precoding weight applied at the m-th AP serving the user such that ||w|| = 1.

[0085] The signal observation at the k-th user is

[0086] y k = H kwx + n k

[0087] where x is the transmitted signal with power constraint k n is the aggregate interference and noise with covariance

[0088] The AP only knows the large-scale fading coefficients, but not the small-scale fading coefficients, which are mainly used to obtain the following correlations:

[0089]

[0090] Each of these correlations is estimated by the AP during connection setup. In the context of distributed massive MIMO, there is a relatively large spacing between antennas or APs. For example, if one considers a 2 gigahertz (GHz) carrier frequency or wavelength λ = 0.15 meters, then for an inter-AP distance of [15, 30] meters, the inter-antenna spacing lies in the range [100λ, 200λ]. For such a large antenna spacing, the antennas become independent and the correlation coefficients drop to zero, which thus gives a diagonal correlation matrix whose diagonal entries essentially reflect the path loss and shadowing between the AP and the user. While the channel matrix changes on a fast time scale due to small-scale fading, the correlation Θ k (k = 1,..., K) changes on a relatively slow time scale.

[0091] The local average SINR of the kth user is:

[0092]

[0093] which is the metric used to optimize the precoder w, instead of the instantaneous SINR, which is affected by the fast fading whose knowledge is not available at the AP.

[0094] The baseline precoder is

[0095] Alternatively, the precoding vector is iteratively optimized so that at each iteration w, the precoding vector is manipulated to maximize the average SINR of the worst-case user by exploiting the gradient of the average SINR as Θ k w (k = 1,..., K). The steps of this procedure are shown in Figure 8 and described as follows. Note that these steps are preferably performed by a CPU (e.g., central processing system 504).

[0096] • Step 800: Initialization

[0097] • Step 802: Compute Θ k ​​The local average SINR of a user is computed as γ k = Pw * Θ k w, k = 1,..., K;

[0098] • Step 804: Identify the weakest user, whose index is k', such that

[0099] • Step 806: At iteration n+1, w(n+1) = (I + μΘ k′ )w(n);

[0100] • Step 808: Normalize the precoder as w(n+1) = w(n+1) / ‖w(n+1)‖;

[0101] • Step 810: Repeat steps 802-808 until convergence or a fixed number of iterations is reached, which is set a priori based on convergence criteria.

[0102] Note that the mathematical formulas in the above steps are just examples. Variations will be apparent to those skilled in the art. For example, one variation is that if all antennas are co-located at one base station or access point, then this precoding computation algorithm can be performed at the access point itself without the need to communicate with a CPU. As another example, it is also possible to perform between these two variations. And, it is also possible to exploit different iterative algorithms by formulating different objective criteria, e.g., to maximize the average SINR of high-priority users while guaranteeing minimum SINRs for other users by assigning priorities to them, instead of maximizing the worst-case user average SINR at each iteration.

[0103] The performance of the proposed procedure is evaluated by numerical results given below.

[0104] Evaluation Method: Consider a live event located in the city center with a radius of 500 meters (m), where K users are requesting the same data. The locations of these K users are random. Assume that M = 64 APs are uniformly distributed at random locations in a circular region with a radius of 750 m around the live event. All APs are connected to a CPU via high-speed links. To reflect a realistic interference environment, we extend the network to a radius of 2 kilometers (km) such that interfering APs are distributed with the same density. We simulate 5000 such network implementations to collect statistical data.

[0105] Figure 6 The cumulative distribution of the local average SINR is compared for the baseline precoder and the precoder after iterative optimization, both with K = 4. As shown in Figure 6 , the gain in the local average SINR corresponding to the worst 10% point is about 8.4 decibels (dB), which is significant.

[0106] To verify the performance advantages of the proposed method, we repeated the experiment, where APs were placed in fixed positions according to a hexagonal layout, with the distance between APs set to 200 meters. This yielded approximately M = 55 APs in a circular region of 750 meters. Figure 7 The cumulative probability distribution function (CDF) corresponding to the locally averaged SINR was compared. From Figure 7 As can be seen, the performance advantages of the proposed pre-encoder are significant.

[0107] Table 1 shows the gain of the local average SINR of the iteratively optimized precoder relative to the baseline precoder for M=64 and K varying at 10 percent of CDF.

[0108]

[0109] Table 1: Gain in average SINR achieved by the iteratively optimized precoder relative to the baseline precoder.

[0110] In some embodiments, when the number of multicast users exceeds a predefined threshold, the network falls back to using a “baseline precoder” (or some other precoder that does not use average CSI).

[0111] In some embodiments, when the number of users receiving the multicast service is small (e.g., below a threshold), the network uses a unicast transmission format for the multicast service.

[0112] Figure 9 Some embodiments according to this disclosure are shown. Figure 5 The D-maMIMO system 500 operates in the context of a 5G NR access network. As shown in the figure, the central processing system 504 configures uplink measurements for UEs 506-1 to 506-K (hereinafter referred to as UE-1 to UE-K) near APs 502-1 to 502-M (hereinafter referred to as AP-1 to AP-M) (step 900). UE-1 to UE-K transmits uplink transmissions (step 902), which are received by AP-1 to AP-K.

[0113] Each AP (denoted here as AP-m, where m = 1,...,M) performs uplink measurements on the received uplink transmissions, processes its uplink measurements to obtain long-term CSI estimates for UE-1 to UE-K, and sends the obtained long-term CSI estimates to the central processing system 504 (step 904). Note that Θ k It is calculated based on the long-term CSI estimate, which can be interpreted as the channel estimation covariance.

[0114] The UE-to-AP association decision is performed (step 906). The UE-to-AP association can be autonomous on each AP. For example, each UE is associated with the AP that provides the higher received signal power (RSRP) in the downlink (e.g., above a certain RSRP threshold). This is part of the network's connection setup procedure or initial access. Each AP aims to serve the UEs that are associated with it.

[0115] The central processing system 504 uses the long-term CSI estimates received from the APs (AP-1 to AP-K) to compute the precoding weights w to be used for broadcasting / multicasting data to the UEs (UE-1 to UE-K) (step 908). More specifically, the central processing system 504 uses the above-described procedure (see, e.g., steps 800-810 above) to compute the precoding weights w to be applied on each AP for broadcasting / multicasting data to the UEs (UE-1 to UE-K). Figure 8

[0116] The central processing system 504 provides the computed precoding weights w to the APs. In this way, the APs obtain the precoding weights w from the central processing system 504 (step 910). Note that each AP only obtains the precoder weights for the UEs that are associated with that AP (via the UE-to-AP association decision). The APs then use the precoding weights w to precode the data to be multicast / broadcast to the UEs (UE-1 to UE-K) and multicast / broadcast the resulting precoded data (step 912).

[0117] The procedure can be repeated (step 914) at fixed intervals of, e.g., 100 milliseconds, depending on the traffic variations. Note that the interval can be based on the "time horizon", which is the time over which the conditions for the algorithm will stabilize. For example, it is the time after which the conditions have changed significantly and the algorithm needs to be updated. The interval value 100 ms is a good example of a typical value, but it depends on, e.g., the frequency band (higher frequencies require more frequent updates), the UE speed (high UE speed requires more frequent updates), and the speed of other objects in the environment. In very stable environments, an update period of 1 second can be good enough.

[0118] Note that the above description is presented in terms of single-antenna APs and single-antenna users, but generalizes to the multi-antenna case. For example, if the AP has multiple antennas, the scalar precoding weight w becomes a precoding vector of dimension equal to the number of antennas at the AP, and similar procedures can be applied as in the single-antenna AP case. m

[0119] Figure 10 ​​​is a schematic block diagram of an AP 502 according to some other embodiments of the disclosure. The AP 502 includes one or more modules 1100, each of which is implemented in software. The modules 1100 provide the functionality of the AP 502 described herein.

[0120] In some embodiments, a computer program including instructions which, when executed by at least one processor, causes the at least one processor to carry out the functionality of the AP 502 according to any of the embodiments described herein is provided. In some embodiments, a carrier containing the aforementioned computer program product is provided. The carrier is one of an electronic signal, an optical signal, a radio signal, or a computer readable storage medium (e.g., a non-transitory computer readable medium such as memory).

[0121] Figure 11 is a schematic block diagram of an AP 502 according to some other embodiments of the disclosure. The AP 502 includes one or more modules 1100, each of which is implemented in software. The modules 1100 provide the functionality of the AP 502 described herein.

[0122] Figure 12is a schematic block diagram of a central processing system 504 according to some embodiments of the present disclosure. As shown, the central processing system 504 includes one or more processors 1204 (e.g., CPUs, ASICs, FPGAs, etc.), memory 1206, and a network interface 1208. The one or more processors 1204 are also referred to herein as processing circuitry. The one or more processors 1204 operate to provide one or more functions of the central processing system 504 as described herein. In some embodiments, the functions are implemented in software stored in, for example, the memory 1206 and executed by the one or more processors 1204.

[0123] Figure 13 is a schematic block diagram illustrating a virtualized embodiment of a central processing system 504 according to some embodiments of the present disclosure. As used herein, a “virtualized” central processing system 504 is an implementation of the central processing system 504 in which at least a portion of the functionality of the central processing system 504 is implemented as a virtual component (e.g., via a virtual machine executing on a physical processing node in a network). As shown, in this example, the central processing system 504 includes one or more processing nodes 1300 coupled to or included as part of a network 1302. Each processing node 1300 includes one or more processors 1304 (e.g., CPUs, ASICs, FPGAs, etc.), memory 1306, and a network interface 1308.

[0124] In this example, the functionality 1310 of the central processing system 504 described herein is implemented at the one or more processing nodes 1300. In some particular embodiments, some or all of the functionality 1310 of the central processing system 504 described herein is implemented as a virtual component executed by one or more virtual machines implemented in a virtual environment hosted by the processing nodes 1300.

[0125] In some embodiments, a computer program including instructions which, when executed by at least one processor, cause the at least one processor to carry out the functionality of the central processing system 504 according to any of the embodiments described herein, or a node (e.g., processing node 1300) implementing one or more functions 1310 of the central processing system 504 in a virtual environment, is provided. In some embodiments, a carrier containing the aforementioned computer program product is provided. The carrier is one of an electrical signal, an optical signal, a radio signal, or a computer readable storage medium (e.g., a non-transitory computer readable medium such as memory).

[0126] Figure 14is a schematic block diagram of a central processing system 504 according to some other embodiments of the present disclosure. The central processing system 504 comprises one or more modules 1400, each implemented in software. The modules 1400 provide the functionality of the central processing system 504 described herein. The discussion is equally applicable to Figure 13 the processing nodes 1300, where the modules 1400 can be implemented at one of the processing nodes 1300 or distributed across multiple processing nodes 1300.

[0127] With reference to Figure 15 According to an embodiment, a communication system includes a telecommunication network 1500, such as a 3GPP-type cellular network, which comprises an access network 1502, such as a RAN, and a core network 1504. The access network 1502 comprises a plurality of base stations 1506A, 1506B, 1506C (e.g. Node Bs, eNBs, gNBs) or other types of wireless APs, each defining a corresponding coverage area 1508A, 1508B, 1508C. Each base station 1506A, 1506B, 1506C is connectable to the core network 1504 over a wired or wireless connection 1510. A first UE 1512 located in coverage area 1508C is configured to wirelessly connect to, or be paged by, the corresponding base station 1506C. A second UE 1514 in coverage area 1508A is wirelessly connectable to the corresponding base station 1506A. While a plurality of UEs 1512, 1514 are illustrated in this example, the disclosed embodiments are equally applicable to a situation where a sole UE is in the coverage area or where a sole UE is connecting to the corresponding base station 1506.

[0128] The telecommunication network 1500 is itself connected to a host computer 1516, which can be embodied in hardware and / or software and can be embodied as a standalone server, a cloud-implemented server, or a distributed server. The host computer 1516 can be controlled by a service provider or be operated by a service provider or on behalf of a service provider. Connections 1518 and 1520 between the telecommunication network 1500 and the host computer 1516 can extend directly from the core network 1504 to the host computer 1516 or can go via an optional intermediate network 1522. The intermediate network 1522 can be one of, or a combination of more than one of, public, private or hosted networks; it can form part of a provider's network, or it can be a wholly separate network. In particular, the intermediate network 1522 can comprise two or more sub-networks (not shown), each of which can have one or more dedicated servers.

[0129] Overall, Figure 15The communication system enables connectivity between the connected UEs 1512, 1514 and the host computer 1516. The connectivity can be described as an over-the-top (OTT) connection 1524. The host computer 1516 and the connected UEs 1512, 1514 are configured to communicate data and / or signaling over the OTT connection 1524 using the access network 1502, the core network 1504, any intermediate network 1522, and possible further infrastructure (not shown) as the medium. The OTT connection 1524 can be transparent in the sense that the participating communication devices through which the OTT connection 1524 passes are unaware of the routing of uplink and downlink communications. For example, a base station 1506 can not or need not be aware that the downlink communication it transmitted reached the connected UE 1512 originating from the host computer 1516 located elsewhere. Similarly, the base station 1506 can not or need not be aware that an uplink communication it received was destined for the host computer 1516 located elsewhere. The OTT connection 1524 can be configured to use a single network entity or entities for some or all of the UEs 1512, 1514 and / or a single network entity or entities for some or all data flows.

[0130] According to an embodiment, example implementations of the UE, base station, and host computer discussed in the previous paragraphs will now be described with reference to Figure 16 In the communication system 1600, host computer 1602 comprises hardware 1604 including communication interface 1606 configured to set up and maintain a wired or wireless connection with an interface of a different communication device of the communication system 1600. The host computer 1602 further comprises processing circuitry 1608, which can have storage and / or processing capabilities. In particular, the processing circuitry 1608 can comprise one or more programmable processors, ASICs, FPGAs, or combinations of these (not shown) adapted to execute instructions. The host computer 1602 further comprises software 1610, which is stored in or accessible by the host computer 1602 and executable by the processing circuitry 1608. The software 1610 includes a host application 1612. The host application 1612 can be operable to provide services to a remote user, such as the UE 1614 connecting using an OTT connection 1616 terminating at the UE 1614 and the host computer 1602. In providing services to the remote user, the host application 1612 can provide user data which is transmitted using the OTT connection 1616.

[0131] The communication system 1600 further includes the base station 1618 provided in a telecommunication system, and the base station 1618 comprises hardware 1620 enabling it to communicate with the host computer 1602 and with the UE 1614. The hardware 1620 can include a communication interface 1622 for setting up and maintaining a wired or wireless connection with an interface of a different communication device of the communication system 1600, as well as a communication management unit 1624 for managing said connection. The base station 1618 further comprises processing circuitry 1626, which can have storage and / or processing capabilities. In particular, the processing circuitry 1626 can comprise one or more programmable processors, ASICs, FPGAs, or combinations of these (not shown) adapted to execute instructions. The base station 1618 further comprises software 1628, which is stored in or accessible by the base station 1618 and executable by the processing circuitry 1626. The software 1628 includes a base station application 1630. The base station application 1630 can be operable to provide services to the UE 1614 using the OTT connection 1616 terminating at the UE 1614 and the base station 1618. In providing services to the UE 1614, the base station application 1630 can provide user data which is transmitted using the OTT connection 1616. Figure 16The UE 1614 (not shown) has at least a radio interface 1624 for a wireless connection 1626. A communication interface 1622 can be configured to facilitate a connection 1628 with a host computer 1602. The connection 1628 can be direct, or it can be connected via the core network of a telecommunications system. Figure 16 (Not shown) and / or via one or more intermediate networks outside the telecommunications system. In the illustrated embodiment, the hardware 1620 of base station 1618 also includes processing circuitry 1630, which may include one or more programmable processors, ASICs, FPGAs, or combinations thereof (not shown) adapted to execute instructions. Base station 1618 also has software 1632 stored internally or accessible via an external connection.

[0132] The communication system 1600 also includes the previously mentioned UE 1614. The hardware 1634 of UE 1614 may include a radio interface 1636 configured to establish and maintain a radio connection 1626 with a base station serving the coverage area where UE 1614 is currently located. The hardware 1634 of UE 1614 also includes processing circuitry 1638, which may include one or more programmable processors, ASICs, FPGAs, or combinations thereof (not shown) suitable for executing instructions. UE 1614 also includes software 1640, which is stored in or accessible by UE 1614 and executable by processing circuitry 1638. Software 1640 includes a client application 1642. The client application 1642 is operable to provide services to human or non-human users via UE 1614 with the support of host computer 1602. In host computer 1602, the executing host application 1612 can communicate with the executing client application 1642 via an OTT connection 1616 terminated between UE 1614 and host computer 1602. When providing services to a user, client application 1642 can receive request data from host application 1612 and provide user data in response to the request data. OTT connection 1616 can transmit both request data and user data. Client application 1642 can interact with the user to generate user-provided user data.

[0133] Notice, Figure 16 The host computer 1602, base station 1618, and UE 1614 shown can be respectively connected to Figure 15 The host computer 1516, base stations 1506A, 1506B, and 1506C, and UEs 1512 and 1514 are similar to or identical to each other. That is to say, the internal working principles of these entities can be as follows: Figure 16 As shown, and independently, the surrounding network topology can be Figure 15 The surrounding network topology.

[0134] In Figure 16 which OTT connection 1616 has been drawn abstractly to illustrate the communication between host computer 1602 and UE 1614 via base station 1618, without explicit reference to any intermediary devices and the precise routing of messages via these devices. Network infrastructure can determine the routing, which it can configure to hide from UE 1614 or from the service provider operating host computer 1602, or both. While OTT connection 1616 is active, the network infrastructure can further take decisions that cause it to dynamically change the routing (for example, based on load balancing consideration or reconfiguration of the network).

[0135] Wireless connection 1626 between UE 1614 and base station 1618 is in accordance with the teachings of the embodiments described throughout this disclosure. One or more of the various embodiments improve the performance of OTT services provided to UE 1614 using OTT connection 1616, in which wireless connection 1626 forms the last segment.

[0136] A measurement procedure can be implemented for the purpose of monitoring the data rate, delay, and other factors on which the one or more embodiments improve. There can further be an optional network functionality for reconfiguring the OTT connection 1616 between host computer 1602 and UE 1614, in response to a change in the measurement results. The measurement procedure and / or the network functionality for reconfiguring the OTT connection 1616 can be implemented in software 1610 and hardware 1604 of host computer 1602 or in software 1640 and hardware 1634 of UE 1614, or both. In some embodiments, sensors (not shown) can be deployed in or in association with communication devices through which OTT connection 1616 passes; the sensors can participate in the measurement procedure by providing values of the monitored quantities exemplified above, or providing values of other physical quantities from which software 1610, 1640 can compute or estimate the monitored quantities. The reconfiguring of the OTT connection 1616 can include message format, retransmission settings, preferred routing, etc. The reconfiguring does not need to affect base station 1618, and it can be unknown or imperceptible to base station 1618. Such procedures and functionalities can be known and practiced in the art. In certain embodiments, measurements can involve proprietary UE signaling facilitating host computer 1602’s measurements of throughput, propagation times, delays, etc. The measurements can be implemented due to software 1610, 1640 causing messages to be transmitted that would not otherwise be transmitted, in particular empty or “dummy” messages.

[0137] Figure 17is a flowchart illustrating a method implemented in a communication system, in accordance with one embodiment. The communication system includes a host computer, a base station and a UE which can be those described with reference to Figure 15 and 16 to the extent that they are not already included by the references in the previous paragraphs. In order to simplify the present disclosure, only drawing references to Figure 17 will be included in this section. In step 1700, the host computer provides user data. In substep 1702 (which can be optional) of step 1700, the host computer provides the user data by executing a host application. In step 1704, the host computer initiates a transmission carrying the user data to the UE. In step 1706 (which can be optional), the base station transmits to the UE the user data which was carried in the transmission that the host computer initiated, in accordance with the teachings of the embodiments described throughout this disclosure. In step 1708 (which can also be optional), the UE executes a client application associated with the host application executed by the host computer.

[0138] Figure 18 is a flowchart illustrating a method implemented in a communication system, in accordance with one embodiment. The communication system includes a host computer, a base station and a UE which can be those described with reference to Figure 15 and 16 to the extent that they are not already included by the references in the previous paragraphs. In order to simplify the present disclosure, only drawing references to Figure 18 will be included in this section. In step 1800 of the method, the host computer provides user data. In an optional substep (not shown), the host computer provides the user data by executing a host application. In step 1802, the host computer initiates a transmission carrying the user data to the UE. The transmission can pass via the base station, in accordance with the teachings of the embodiments described throughout this disclosure. In step 1804 (which can be optional), the UE receives the user data carried in the transmission.

[0139] Figure 19 is a flowchart illustrating a method implemented in a communication system, in accordance with one embodiment. The communication system includes a host computer, a base station and a UE which can be those described with reference to Figure 15 and 16 to the extent that they are not already included by the references in the previous paragraphs. In order to simplify the present disclosure, only drawing references to Figure 19with reference to the accompanying drawings. In step 1900, which can be optional, the UE receives input provided by the host computer. Additionally or alternatively to step 1900, in step 1902, the UE provides user data. In substep 1904 of step 1900, which can be optional, the UE provides the user data by executing a client application. In substep 1906 of step 1902, which can be optional, the UE executes a client application which provides the user data in reaction to the received input provided by the host computer. In providing the user data, the executed client application can further take into account user input received from the user. Regardless of the specific manner in which the user data is provided, the UE initiates, in substep 1908, transmission of the user data to the host computer. In step 1910 of the method, the host computer receives the user data transmitted from the UE, in accordance with the teachings of the embodiments described throughout this disclosure.

[0140] Figure 20 is a flowchart illustrating a method implemented in a communication system, in accordance with one embodiment. The communication system includes a host computer, a base station and a UE which can be those described with reference to Figure 15 and 16 in the preceding paragraphs. To simplify the present disclosure, in this section only drawing references to Figure 20 will be included. In step 2000, which can be optional, the base station receives user data from a UE, in accordance with the teachings of the embodiments described throughout this disclosure. In step 2002, which can be optional, the base station initiates transmission of the received user data to a host computer. In step 2004, which can be optional, the host computer receives the user data carried in the transmission initiated by the base station.

[0141] Any appropriate steps, methods, features, functions, or benefits disclosed herein can be performed through one or more functional units or modules of one or more virtual apparatuses. Each virtual apparatus can comprise a number of these functional units. These functional units can be implemented via processing circuitry, which can include one or more microprocessor or microcontrollers, as well as other digital hardware, which can include digital signal processors (DSPs), special-purpose computer chips, application- specific integrated circuits (ASICs), field-programmable gate arrays (FPGAs), or other computational logic. The processing circuitry can be configured to execute program code stored in memory, which can include one or several types of memory such as read-only memory (ROM), random-access memory (RAM), cache memory, flash memory devices, optical storage media, etc. Program code stored in memory includes program instructions for executing one or more telecommunications and / or data communications protocols as well as instructions for carrying out one or more of the techniques described herein. In some implementations, the processing circuitry can be used to cause the respective functional unit(s) to perform corresponding functions according one or more embodiments of the present disclosure.

[0142] Although the processes in the figures can show a particular order of executing; it is understood that this order is an example (e.g., alternative embodiments can execute the operations in a different order, combine certain operations, overlap certain operations, etc.).

[0143] Some example embodiments of the present disclosure are as follows.

[0144] Group A embodiments

[0145] Embodiment 1: A method for broadcasting or multicasting data to user equipments, UEs, in a distributed cell-less massive multiple-input multiple-output, MIMO, network, the method comprising any one or more of the following acts:

[0146] • at each access point, AP, (502-m) of the two or more APs (502-1,..., 502-M):

[0147] obtaining (904) long-term channel state information, CSI, for at least one UE (506); and

[0148] transmitting (904) the long-term CSI for the at least one UE (506) to a central processing system (504);

[0149] • at the central processing system (504):

[0150] receiving (904) the long-term CSI for the at least one UE (506) from each AP (502-m) of the two or more APs (502-1,..., 502-M);

[0151] computing (908) a precoding vector w for the at least one UE (506) across the two or more APs (502-1,..., 502-M) based on the long-term CSI for the at least one UE (506) received from the two or more APs (502-1,..., 502-M); and

[0152] transmitting (910) the precoding vector w to the two or more APs (502-1,..., 502-M); and

[0153] • at each AP (502-m) of the two or more APs (502-1,..., 502-M):

[0154] obtaining (910) the precoding vector w from the central processing system (504);

[0155] precoding (912) the data to be broadcast or multicast to the at least one UE (506) based on the precoding vector w; and

[0156] broadcasting or multicasting the precoded (912) data to the at least one UE (506).

[0157] Embodiment 2: The method of embodiment 1, wherein the at least one UE (506) is two or more UEs (506-1,..., 506-K).

[0158] Embodiment 3: The method of embodiment 1 or 2, wherein computing (908) the precoding vector w comprises:

[0159] a) at iteration 0, initializing the precoding vector w to provide a precoding vector w(0) for iteration 0;

[0160] b) computing a local average signal-to-interference-and-noise ratio, SINR, for the at least one UE (506);

[0161] c) identifying a weakest UE from among the at least one UE (506) based on the computed local average SINR;

[0162] d) at iteration n+1, updating the precoding vector w based on long-term CSI obtained from the at least one UE (506) for the weakest UE, thereby providing a precoding vector w(n+1) for iteration n+1;

[0163] e) normalizing the precoding vector w(n+1) for iteration n+1; and

[0164] f) repeating steps (b) through (e) until a stopping criterion is met (e.g., convergence has been reached or a maximum number of iterations has been reached), such that the normalized precoding vector for the last iteration is provided as the precoding vector w.

[0165] Embodiment 4: The method of embodiment 3, wherein initializing the precoding vector w comprises initializing the precoding vector w to a value

[0166] Embodiment 5: The method of embodiment 3 or 4, wherein computing the local SINR for the at least one UE (506) comprises computing the local SINR for the at least one UE (506) as * Θ k w, k = 1,..., K, by Θk.

[0167] Embodiment 6: The method of embodiment 5, wherein the weakest UE is identified as the UE whose index k' is ​

[0168] Embodiment 7: The method of any of embodiments 3-6, wherein updating the precoding vector w comprises, at iteration n+1, updating the precoding vector w to w(n+1) = (I + μΘ k′ )w(n).

[0169] Embodiment 8: The method of any of embodiments 3-7, wherein normalizing the precoding vector w(n+1) for iteration n+1 comprises normalizing the precoding vector w(n+1) for iteration n+1 to w(n+1) = w(n+1) / ‖w(n+1)‖.

[0170] Embodiment 9: The method of any of embodiments 1-8, wherein the long-term CSI comprises an estimated path loss.

[0171] Embodiment 10: The method of any of embodiments 1-9, wherein a coherence interval of the long-term CSI spans greater than 1,000 symbols, and the method further comprises: at the central processing system (504), scheduling, by the at least one UE (506), transmission of a dedicated pilot; and at each AP (502-m) of the two or more APs (502-1,..., 502-M), obtaining (904) the long-term CSI for the at least one UE (506) comprises estimating, based on the transmission of the dedicated pilot, the long-term CSI for the at least one UE (506) over a fixed time interval.

[0172] Embodiment 11: A method performed at a central processing system (504) for a distributed cell-less massive MIMO network for broadcasting or multicasting data to user equipment, UEs, the method comprising any one or more of the following acts:

[0173] • for each AP (502-m) of the two or more access points, APs (502-1,..., 502-M) in a distributed cell-less massive multiple-input multiple-output, MIMO, network, receiving (904) long-term channel state information, CSI, for at least one UE (506) from the AP (502-m);

[0174] • based on the long-term CSI for the at least one UE (506) received from the two or more APs (502-1,..., 502-M), computing (908) a precoding vector w for the at least one UE (506) across all of the two or more APs (502-1,..., 502-M); and

[0175] • transmitting (910) the precoding vector w to the two or more APs (502-1,..., 502-M).

[0176] Embodiment 12: The method according to Embodiment 11, wherein the at least one UE (506) is two or more UEs (506-1,..., 506-K).

[0177] Embodiment 13: The method according to Embodiment 11 or 12, wherein computing (908) the precoding vector w comprises:

[0178] a) at iteration 0, initializing the precoding vector w to provide a precoding vector w(0) for iteration 0;

[0179] b) computing a local average Signal to Interference and Noise Ratio, SINR, for the at least one UE (506);

[0180] c) identifying a weakest UE from among the at least one UE (506) based on the computed local average SINR;

[0181] d) at iteration n+1, updating the precoding vector w based on long-term CSI obtained from the at least one UE (506) for the weakest UE, thereby providing a precoding vector w(n+1) for iteration n+1;

[0182] e) normalizing the precoding vector w(n+1) for iteration n+1; and

[0183] f) repeating steps (b) to (e) until a stopping criterion is met (e.g., convergence has been reached or a maximum number of iterations has been reached), such that the normalized precoding vector for the last iteration is provided as the precoding vector w.

[0184] Embodiment 14: The method according to Embodiment 13, wherein initializing the precoding vector w comprises initializing the precoding vector w to a value

[0185] Embodiment 15: The method according to Embodiment 13 or 14, wherein computing the local SINR for the at least one UE (506) comprises computing the local SINR for the at least one UE (506) as k k = Pw * Θ k w, k = 1,..., K.

[0186] Embodiment 16: The method according to Embodiment 15, wherein the weakest UE is identified as the UE whose index k' is

[0187] Embodiment 17: The method according to any one of Embodiments 13 to 16, wherein updating the precoding vector w comprises updating the precoding vector w at iteration n+1 to w(n+1) = (I + μΘ​​k′ ) w(n).

[0188] Embodiment 18: The method of any of embodiments 13-17, wherein normalizing the precoding vector w(n+1) for iteration n+1 comprises normalizing the precoding vector w(n+1) for iteration n+1 as w(n+1) = w(n+1) / ||w(n+1) ||.

[0189] Embodiment 19: The method of any of embodiments 11-18, wherein the long-term CSI comprises an estimated path loss.

[0190] Embodiment 20: The method of any of embodiments 11-19, wherein a coherence interval of the long-term CSI spans greater than 1,000 symbols, and the method further comprises scheduling, by the at least one UE (506), transmission of a dedicated pilot.

[0191] Embodiment 21: A method performed at an access point, AP, in a distributed cell-free massive multiple-input multiple-output, MIMO, network comprising two or more APs, for broadcasting or multicasting data to a user equipment, UE, the method comprising any one or more of the following acts:

[0192] • obtaining (904) long-term channel state information, CSI, for the at least one UE (506);

[0193] • transmitting (904) the long-term CSI for the at least one UE (506) to a central processing system (504);

[0194] • obtaining (910) a precoding vector w from the central processing system (504), the precoding vector w being for the at least one UE (506) across all of the two or more APs (502-1,..., 502-M);

[0195] • precoding (912) data to be broadcast or multicasted to the at least one UE (506) based on the precoding vector w; and

[0196] • broadcasting or multicasting the precoded (912) data to the at least one UE (506).

[0197] Embodiment 22: The method of embodiment 21, wherein the at least one UE (506) is two or more UEs (506-1,..., 506-K).

[0198] Embodiment 23: The method of any of embodiments 21-22, wherein the long-term CSI comprises an estimated path loss.

[0199] Example 24: The method of any of Examples 21-23, wherein the coherence interval of the long-term CSI spans greater than 1,000 symbols, and obtaining (904) the long-term CSI for the at least one UE (506) comprises estimating the long-term CSI for the at least one UE (506) over a fixed time interval based on transmission of dedicated pilots by the at least one UE (506).

[0200] Example 25: A central processing system for a distributed cell-less massive multiple-input multiple-output, MIMO, network for broadcasting or multicasting data to user equipment, UEs, wherein the central processing system is adapted to perform the method according to any of Examples 11-20.

[0201] Example 26: The central processing system of Example 25, comprising: a network interface; and processing circuitry associated with the network interface, the processing circuitry configured to cause the central processing system to perform the method according to any of Examples 11-20.

[0202] Example 27: An access point, AP, for a distributed cell-less massive multiple-input multiple-output, MIMO, network for broadcasting or multicasting data to user equipment, UEs, wherein the distributed cell-less massive MIMO network comprises two or more APs, the AP adapted to perform the method according to any of Examples 21-24.

[0203] Example 28: The AP of Example 27, comprising: a network interface; at least one transmitter; at least one receiver; and processing circuitry associated with the network interface, the at least one transmitter, and the at least one receiver, wherein the processing circuitry is configured to cause the AP to perform the method according to any of Examples 21-24.

[0204] Example 29: A distributed cell-less massive multiple-input multiple-output, MIMO, network comprising: at least two access points, APs, at least one user equipment, UE, and a central processing unit, CPU. Each AP obtains long-term channel state information, CSI, transmitted to the CPU (e.g., based on filtering or processing of uplink measurements), wherein an optimization is performed (e.g., iterated) to compute precoding vectors on all of the at least two APs. Each AP then broadcasts a precoded multicast transmission to at least one UE that needs the same content.

[0205] Example 30: The method of the previous example, wherein the statistical channel information between the access point, AP, and the user equipment, UE, comprises an estimated path loss. The AP estimates long-term channel state information, CSI, related to UEs within the coverage area by exploiting signals transmitted and received during initial access, i.e., before radio resource control, RRC, connection setup. For example, each UE transmits a physical random access channel, PRACH, preamble in the uplink, which can be used to estimate a link average signal-to-noise ratio, SNR, which is equivalent to a long-term channel gain between the AP and the UE.

[0206] Example 31 : Due to the large-scale channel coherence interval spanning thousands of symbols, in some embodiments, the network is able to schedule a dedicated pilot in the uplink and reliably estimate long-term channel state information, CSI, in a fixed time interval.

[0207] Example 32. The method of any of the preceding examples, further comprising: obtaining user data; and forwarding the user data to a host computer or a wireless device.

[0208] Group B embodiments

[0209] Example 33: A central processing system (504) for a distributed cell-less massive multiple-input multiple-output, MIMO, network for broadcasting or multicasting data to user equipment, UEs, comprising: processing circuitry configured to perform any of the steps performed by the central processing system of any Group A embodiment; and power supply circuitry configured to supply power.

[0210] Example 34: A communication system including a host computer comprising: processing circuitry configured to provide user data; and communication interface configured to forward the user data, for transmission to a user equipment, UE, in a cellular network; wherein the communication system comprises a central processing system (504) for a distributed cell-less massive multiple-input multiple-output, MIMO, network for broadcasting or multicasting data to the UE, the central processing system having a network interface and processing circuitry configured to perform any of the steps performed by the central processing system of any Group A embodiment.

[0211] Example 35: The communication system of the previous example, further including: the central processing system.

[0212] Example 36: The communication system of the previous two examples, further including: the UE, wherein the UE is configured to communicate with at least one access point, AP, of the distributed cell-less massive MIMO network.

[0213] Embodiment 37: A communication system according to the three preceding embodiments, wherein: the processing circuitry of the host computer is configured to execute a host application, thereby providing the user data; and the UE comprises processing circuitry configured to execute a client application associated with the host application.

[0214] Embodiment 38: A method implemented in a communication system including a host computer, a base station and a User Equipment, UE, the method comprising: at the host computer, providing user data; and at the host computer, initiating a transmission carrying the user data to the UE via a cellular network comprising a central processing system (504) for a distributed cell-less massive Multiple-Input Multiple-Output, MIMO, network, the central processing system (504) being configured to broadcast or multicast data to the UE, wherein the central processing system (504) performs any of the steps performed by the central processing system in any of the Group A embodiments.

[0215] Embodiment 39: The method according to the preceding embodiment, further comprising: at one or more Access Points, APs, of the distributed cell-less massive MIMO network, transmitting the user data.

[0216] Embodiment 40. The method according to the two preceding embodiments, wherein the user data is provided at the host computer by execution of a host application, the method further comprising: at the UE, executing a client application associated with the host application.

[0217] Embodiment 41. A User Equipment, UE, configured to communicate with a base station, the UE comprising a radio interface and processing circuitry configured to perform the method according to the three preceding embodiments.

[0218] In this disclosure, at least some of the following abbreviations can be used. If there is an inconsistency in abbreviations between the above and the below, the above usage should take precedence. If listed multiple times below, the first listing should take precedence over any subsequent listing.

[0219] • 3GPP Third Generation Partnership Project

[0220] • 4G Fourth Generation

[0221] • 5G Fifth Generation

[0222] • AP Access Point

[0223] • APU Antenna Processing Unit

[0224] • ASIC Application-Specific Integrated Circuit

[0225] • BS Base Station

[0226] • CDF Cumulative Probability Distribution Function

[0227] • C-maMIMO Centralized massive multiple-input multiple-output

[0228] • CPU Central processing unit

[0229] • CSI Channel state information

[0230] • dB Decibel

[0231] • D-maMIMO Distributed massive multiple-input multiple-output

[0232] • DSP Digital signal processor

[0233] • eNB Enhanced or evolved Node B

[0234] • FDD Frequency division duplex

[0235] • FPGA Field-programmable gate array

[0236] • GHz Gigahertz

[0237] • gNB New radio base station

[0238] • km Kilometer

[0239] • LED Light-emitting diode

[0240] • LTE Long term evolution

[0241] • m Meter

[0242] • MAC Medium access control

[0243] • MBMS Multimedia broadcast multicast service

[0244] • MBSFN Multimedia broadcast multicast service single frequency network

[0245] • MIMO Multiple-input multiple-output

[0246] • MME Mobility management entity

[0247] • MTC Machine type communication

[0248] • NR New radio

[0249] • OTT Over-the-top

[0250] • PDSCH Physical downlink shared channel

[0251] • P-GW Packet data network gateway

[0252] • PRACH Physical random access channel

[0253] • RAM Random Access Memory

[0254] • RAN Radio Access Network

[0255] • ROM Read Only Memory

[0256] • RRC Radio Resource Control

[0257] • SAR Specific Absorption Rate

[0258] • SCEF Service Capability Exposure Function

[0259] • SC-PTM Single Cell Point To Multipoint

[0260] • SFN Single Frequency Network

[0261] • SINR Signal to Interference and Noise Ratio

[0262] • SNR Signal to Noise Ratio

[0263] • TDD Time Division Duplex

[0264] • UE User Equipment

[0265] Those skilled in the art will recognize improvements and modifications to the embodiments of the present disclosure. All such improvements and modifications are considered within the scope of the concepts disclosed herein.

Claims

1. A method for broadcasting or multicasting data to a user equipment (UE) in a distributed, cell-free, massive MIMO network, comprising: At each of two or more access points (APs) (502-m): obtain (904) long-term channel state information (CSI) for at least one UE (506); and transmit (904) the long-term CSI for the at least one UE (506) to the central processing system (504); At the central processing system (504): for each of the two or more APs (502-1, ..., 502-M) (502-m), the long-term CSI for the at least one UE (506) is received (904) from the AP (502-m); Based on the long-term CSI received from the two or more APs (502-1,...,502-M) for the at least one UE (506), calculate (908) a precoding vector w for the at least one UE (506) on the two or more APs (502-1,...,502-M); and transmit (910) the precoding vector w to the two or more APs (502-1,...,502-M); as well as At each of the two or more APs (502-m): the precoding vector w is obtained (910) from the central processing system (504); based on the precoding vector w, the data to be broadcast or multicast to the at least one UE (506) is precoded (912); and the precoded (912) data is broadcast or multicast to the at least one UE (506). The calculation of the precoding vector w in (908) includes performing the following steps: (a) In iteration 0, the precoding vector w is initialized to provide the precoding vector w(0) for iteration 0; (b) Calculate the local average signal-to-interference-plus-noise ratio (SINR) of the at least one UE (506); (c) Identify the weakest UE from the at least one UE (506) based on the calculated local average SINR; (d) In iteration n+1, based on the long-term CSI obtained from the at least one UE (506) for the weakest UE, the precoding vector w is updated to provide a precoding vector w(n+1) for iteration n+1. (e) Normalize the precoded vector w(n+1) used for iteration n+1; and (f) Repeat steps (b) to (e) until the stopping criterion is met, so that the normalized precoded vector for the last iteration is provided as the precoded vector w.

2. A method for broadcasting or multicasting data to a user equipment (UE) and executing it at a central processing system (504) for a distributed, cell-free, massive MIMO network, comprising: For each of two or more access points (APs) (502-m) in the distributed cellless massive multiple input multiple output MIMO network, receive (904) long-term channel state information (CSI) for at least one UE (506) from the AP (502-m); Based on the long-term CSI received from the two or more APs (502-1, ..., 502-M) for the at least one UE (506), calculate (908) the precoding vector w for the at least one UE (506) on all APs in the two or more APs (502-1, ..., 502-M); and Transmit (910) the precoding vector w to the two or more APs (502-1, ..., 502-M). The calculation of the precoding vector w in (908) includes performing the following steps: (a) In iteration 0, the precoding vector w is initialized to provide the precoding vector w(0) for iteration 0; (b) Calculate the local average signal-to-interference-plus-noise ratio (SINR) of the at least one UE (506); (c) Identify the weakest UE from the at least one UE (506) based on the calculated local average SINR; (d) In iteration n+1, based on the long-term CSI obtained from the at least one UE (506) for the weakest UE, the precoding vector w is updated to provide a precoding vector w(n+1) for iteration n+1. (e) Normalize the precoded vector w(n+1) used for iteration n+1; and (f) Repeat steps (b) to (e) until the stopping criterion is met, so that the normalized precoded vector for the last iteration is provided as the precoded vector w.

3. The method according to claim 2, wherein, The at least one UE (506) is two or more UEs (506-1, ..., 506-K).

4. The method according to claim 2 or 3, wherein, The stopping criterion is convergence or reaching the maximum number of iterations.

5. The method according to claim 2 or 3, wherein, Initializing the precoding vector w includes: initializing the precoding vector w to a value. Where M is the number of APs.

6. The method according to any one of claims 2 to 3, wherein, Calculating the local average SINR of the at least one UE (506) includes: via Θ k The local average SINR of the at least one UE (506) is calculated as γ. k =Pw*Θ k w, k = 1, ..., K, where γ k It is the local average SINR of the k-th UE, P is the power constraint of the transmitted signal, and Θ k This is the correlation used for the k-th UE, where K is the number of UEs.

7. The method according to claim 6, wherein, The weakest UE is identified as its index k'. UE.

8. The method according to any one of claims 2 to 3, wherein, Normalizing the precoding vector w(n+1) used for iteration n+1 includes: normalizing the precoding vector w(n+1) used for iteration n+1 to w(n+1) = w(n+1) / ||w(n+1)||.

9. The method according to any one of claims 2 to 3, wherein, The long-term CSI includes estimated path loss.

10. The method according to any one of claims 2 to 3, wherein, The long-term CSI has a coherence interval spanning more than 1,000 symbols, and the method further includes: The transmission of the dedicated pilot is scheduled by the at least one UE (506).

11. A central processing system (504) for a distributed, cell-free, massively multi-input multiple-output (MIMO) network (500), for broadcasting or multicasting data to a user equipment (UE) (506), wherein, The central processing system (504) is adapted to perform the method according to any one of claims 2 to 10.

12. A central processing system (504) for a distributed, cell-free, massively multi-input multiple-output (MIMO) network (500), for broadcasting or multicasting data to a user equipment (UE) (506), wherein, The central processing system (504) includes: Network interface (1208); and The processing circuitry (1204) associated with the network interface (1208) is configured to cause the central processing system (504) to: For each of two or more access points (APs) (502-1, ..., 502-M) in the distributed cellless MIMO network, receive (904) long-term channel state information (CSI) for at least one UE (506) from the AP (502-m); Based on the long-term CSI received from the two or more APs (502-1, ..., 502-M) for the at least one UE (506), calculate (908) the precoding vector w for the at least one UE (506) on all APs in the two or more APs (502-1, ..., 502-M); and Transmit (910) the precoding vector w to the two or more APs (502-1, ..., 502-M). When calculating the precoding vector w, the processing circuit (1204) is configured to cause the central processing system (504) to further perform the following steps: (a) In iteration 0, the precoding vector w is initialized to provide the precoding vector w(0) for iteration 0; (b) Calculate the local average signal-to-interference-plus-noise ratio (SINR) of the at least one UE (506); (c) Identify the weakest UE from the at least one UE (506) based on the calculated local average SINR; (d) In iteration n+1, based on the long-term CSI obtained from the at least one UE (506) for the weakest UE, the precoding vector w is updated to provide a precoding vector w(n+1) for iteration n+1. (e) Normalize the precoded vector w(n+1) used for iteration n+1; and (f) Repeat steps (b) to (e) until the stopping criterion is met, so that the normalized precoded vector for the last iteration is provided as the precoded vector w.

13. A method for broadcasting or multicasting data to a user equipment (UE) at an access point (AP) in a distributed, cell-free, massive MIMO network, the network comprising two or more APs, the method comprising: Obtain (904) Long-term Channel State Information (CSI) for at least one UE (506); Transmit (904) the long-term CSI for the at least one UE (506) to the central processing system (504); A precoding vector w is obtained from the central processing system (504), the precoding vector w being used for the at least one UE (506) on all APs of two or more APs (502-1, ..., 502-M); Based on the precoding vector w, the data to be broadcast or multicast to the at least one UE (506) is precoded (912); as well as Broadcast or multicast the precoded (912) data to the at least one UE (506). The central processing system calculates the precoding vector w by performing the following steps: (a) In iteration 0, the precoding vector w is initialized to provide the precoding vector w(0) for iteration 0; (b) Calculate the local average signal-to-interference-plus-noise ratio (SINR) of the at least one UE (506); (c) Identify the weakest UE from the at least one UE (506) based on the calculated local average SINR; (d) In iteration n+1, based on the long-term CSI obtained from the at least one UE (506) for the weakest UE, the precoding vector w is updated to provide a precoding vector w(n+1) for iteration n+1. (e) Normalize the precoded vector w(n+1) used for iteration n+1; and (f) Repeat steps (b) to (e) until the stopping criterion is met, so that the normalized precoded vector for the last iteration is provided as the precoded vector w.

14. The method according to claim 13, wherein, The at least one UE (506) is two or more UEs (506-1, ..., 506-K).

15. The method according to any one of claims 13 to 14, wherein, The long-term CSI includes estimated path loss.

16. The method according to any one of claims 13 to 14, wherein, The long-term CSI has a coherence interval spanning more than 1,000 symbols, and: Obtaining (904) the long-term CSI for the at least one UE (506) includes: estimating the long-term CSI for the at least one UE (506) over a fixed time interval based on the transmission of the at least one UE (506) to a dedicated pilot.

17. An access point (AP) (502-m) for a distributed, cell-free, massive MIMO network (500), used to broadcast or multicast data to a user equipment (UE) (506), wherein, The distributed cellless massive MIMO network (500) includes two or more APs (502-1 to 502-M), said AP (502-m) being adapted to perform the method according to any one of claims 13 to 16.

18. An access point (AP) (502-m) for a distributed, cell-free, massive MIMO network (500), used to broadcast or multicast data to a user equipment (UE) (506), wherein, The distributed, cell-free, massive MIMO network (500) comprises two or more access points (502-1 to 502-M), wherein the access points (502-m) include: Network interface (1008); At least one transmitter (1012); At least one receiver (1014); and A processing circuit (1004) associated with the network interface (1008), the at least one transmitter (1012), and the at least one receiver (1014), wherein the processing circuit (1004) is configured such that the AP (502-m): Obtain (904) Long-term Channel State Information (CSI) for at least one UE (506); Transmit (904) the long-term CSI for the at least one UE (506) to the central processing system (504); A precoding vector w is obtained from the central processing system (504), the precoding vector w being used for the at least one UE (506) on all APs of two or more APs (502-1, ..., 502-M); Based on the precoding vector w, the data to be broadcast or multicast to the at least one UE (506) is precoded (912); and Broadcast or multicast precoded (912) data to the at least one UE (506), wherein the central processing system calculates the precoded vector w by performing the following steps: (a) In iteration 0, the precoding vector w is initialized to provide the precoding vector w(0) for iteration 0; (b) Calculate the local average signal-to-interference-plus-noise ratio (SINR) of the at least one UE (506); (c) Identify the weakest UE from the at least one UE (506) based on the calculated local average SINR; (d) In iteration n+1, based on the long-term CSI obtained from the at least one UE (506) for the weakest UE, the precoding vector w is updated to provide a precoding vector w(n+1) for iteration n+1. (e) Normalize the precoded vector w(n+1) used for iteration n+1; and (f) Repeat steps (b) to (e) until the stopping criterion is met, so that the normalized precoded vector for the last iteration is provided as the precoded vector w.

Images