OUTPUT (MIMO) DETECTION PROCESS

FIELD

The present disclosure relates to wireless communications, and in particular, to determining initial vectors usable to perform an iterative detection process in massive multiple input- multiple output (MIMO) systems.

BACKGROUND

The Third Generation Partnership Project (3 GPP) has developed and is developing standards for Fourth Generation (4G) (also referred to as Long Term Evolution (LTE)), Fifth Generation (5G) (also referred to as New Radio (NR)), and Sixth Generation (6G) wireless communication systems. Such systems provide, among other features, broadband communication between network nodes (NNs), such as base stations, and mobile wireless devices (WD) such as user equipment (UE), as well as communication between network nodes and between WDs.

One of the pillars of future wireless communication systems, e.g., cellular networks, is massive multiple-input multiple-output (MIMO) systems. Massive MIMO refers to a large antenna array at a network node (e.g., base station (BS )) that potentially provides high throughput and spectral efficiency for a large number of users. Due to the large number of users and antennas in a massive MIMO system, the complexity of the optimal detector, based on the maximum likelihood (ML) metric, is prohibitive.

Alternative linear detection techniques, such as zero-forcing (ZF) and minimum mean square error (MMSE) need a matrix inverse calculation, with complexity of O^K ³ ') for K users, and calculating a Gram matrix, with complexity of Ofj K ² ') for K users and N antennas at the network node (NN) (e.g., BS).

In some cases, a K X N massive MIMO system may be used, where a massive MIMO NN (e.g., massive MIMO BS) is equipped with N antennas serving K users (e.g., K may refer to the number of users saved/served by the network node (e.g., base station), N may refer to the number of antennas serving at the network node) . Both N and K may be large. In recent years, different iterative techniques have been proposed to avoid the full matrix inversion. The iterative detection techniques are a low complexity approximation to the MMSE detectors. In one approach, to derive an iterative detection formula, the following regularized optimization problem is solved:

1 s = arg min \\ x - Hs II2 + 777777 II s H sec ^K SNR

Here s is the detected signal, H is the channel matrix, x is the received signal at the BS, and SNR is the signal-to-noise-ratio. Different techniques such as Conjugate

Gradient (CG) or Orthogonal Coordinate Descent (OCD) have been proposed to solve the above equation. Another approach is to approximate the matrix inversion in the

MMSE technique:

An example of this approach is the Gauss-Seidel (GS) technique. Both approaches result in an iterative solution of the following form to estimate the transmitted signal: where i is the iteration number, and b and A represent a vector and matrix that depend on the specific technique used. Usually, a zero vector is chosen as the starting point of the iterative detectors.

The complexity of iterative detectors depends on the number of iterations needed to converge to an acceptable symbol error rate (SER) performance such as in MMSE. Although the choice of the initial vector may not affect the convergence point, the choice has a significant impact on the convergence rate and, in turn, on the complexity of the iterative methods. Therefore, different initializing methods have been introduced, such as using the first I terms of the Neumann series (NS), e.g.,: where I is either zero or one, and GMMSE i ^s decomposed into a diagonal matrix D and a hollow matrix E. Although both of these choices for initialization may be used to improve the convergence rate of the iterative methods in independent, identically distributed (i.i.d.) Rayleigh fading channels, they are far from optimal. Further, it is not clear which one to use in different situations. Another type of massive MIMO detector uses learning-based techniques. This type of detector unfolds an iterative detector over different layers of a deep neural network (DNN) and introduces a few trainable parameters to each layer. Some of the learning-based techniques are very complex and include matrix inversion, e.g., orthogonal approximate message passing network (OAMPNet), a complexity of (7(A ³ ) which is much higher than the complexity of the MMSE approach.

In some cases, learning-based detectors may only include matrix-vector multiplications, i.e., low complexity learning-based detectors. The mainstream learningbased detector of this type is MMNet (a deep learning MIMO detection scheme). There may be two types of MMNet approaches. The first type is for the task of signal detection in i.i.d. Rayleigh fading channels, which is called MMNet-i.i.d. and has two scalar trainable parameters per layer. The other approach is for signal detection in realistic 3GPP channels, called MMNet, which has K(2N + 1) trainable parameters per layer. Both types of MMNet may be ten-layers deep. The complexity of each layer of MMNet (both types) is of order O NK ² ).

Further, iterative techniques typically need several iterations before their SER performance converges to the MMSE. In the simple case of i.i.d. Rayleigh fading channels, where the elements of the channel matrix are distributed normally, the convergence rate depends heavily on (i = K /N, i.e., the ratio between K, the number of users, and N, the number of antennas at the NN (e.g., BS). When (i is small, it takes about two to three iterations for the iterative methods to converge to the MMSE’s SER performance. On the other hand, when (i ~ 0.25, the iterative methods need several more iterations to converge. This increase in needed iterations directly affects the complexity of these techniques until the point that their complexity reaches and even surpasses that of the MMSE.

On the other hand, although the SER performance of a low complexity learningbased detector like MMNet is close to that of the ML approach, it lacks the generalizability and robustness needed for wireless communications channels. FIG. 1 shows MMNet-iid symbol error rate (SER) performance vs. SNR for 16 x64 massive MIMO system and 64-QAM (quadrature amplitude modulation) in an i.i.d. Rayleigh fading channel. MMNet-iid (i.e., no Mismatch) is trained on the same massive MIMO system and MMNet-iid (i.e., mismatch) is trained on a 64-QAM 32x64 massive MIMO system. Put differently, FIG. 1 shows an example SER performance of MMNet-iid when there is a mismatch between trained and tested massive MIMO systems in an i.i.d. Rayleigh fading channel. In this case, the MMNet-iid is trained on a 64-QAM 32 X 64 massive MIMO system and tested on a 64-QAM 16 X 64 one. As shown, the MMNet- iid’ s SER performance degrades dramatically when there is a mismatch between trained and tested massive MIMO systems, e.g., even in this relatively simple channel.

To keep the SER performance at a steady state for all massive MIMO configurations, the learning-based detector may be trained from scratch whenever the number of users changes (which is a lengthy process). To decrease the training time, the NN (e.g., BS) may require to be equipped with large computational resources. This requirement conflicts with the demand for sustainability in cellular networks and the need to reduce the carbon footprint of the NNs (e.g., BSs).

The dominant sources of carbon footprint in cellular networks may be NNs (e.g., BSs). The efforts in 5G and 6G are to decrease the carbon footprint of the cellular network. It may not be reasonable to equip the NNs (BSs) with unnecessary computation resources to use a near-optimum detection scheme for the massive MIMO systems. This makes learning-based techniques for massive MIMO detection vulnerable when considering sustainability in cellular networks. Furthermore, there is also the issue of maintenance of the NNs (e.g., BSs). As the complexity of the NNs (e.g., BSs) increases, it becomes harder to properly maintain them.

SUMMARY

Some embodiments advantageously provide methods, systems, and apparatuses for determining initial vectors usable to perform an iterative detection process in massive multiple input multiple output (MIMO) systems.

In some embodiments, the NN (e.g., BSs) may be kept as simple and sustainable as possible by using a detector (e.g., a suboptimal detector) with acceptable performance. In some other embodiments, an efficient and adaptive detector for massive MIMO systems is described, e.g., to address the high number of iterations needed by typical iterative detectors and lack of adaptability of typical learning-based detectors.

In an embodiment, a smart initialization technique (i.e., Hint) is described. The smart initialization technique (i.e., Hint) may be implemented in network node, WD, or any other device/node. Further, the smart initialization technique may work on top of an iterative technique and/or provide an initial vector (e.g., good initial vector) to the interactive technique. The initial vector may be tailored for the current environment such as using a learning-based technique. In addition, Hint may be configured for a choice (e.g., good choice) for a regularization parameter in the detection optimization problem such as to increase the convergence rate even more (e.g., when compared to typical systems). Hint may further be configured to combine learning-based and iterative techniques such as to increase the convergence rate of the iterative techniques and keep the generalizability of the combination at an acceptable level.

In some embodiments, simulation results show that using Hint in combination with a typical iterative technique such as GS may increase its convergence rate in both i.i.d. Rayleigh fading and realistic 3GPP channels.

In some other embodiments, the smart initialization technique (i.e., Hint) may be a smart tool that looks for the best initial vector and regularization parameter for an iterative massive MIMO detector.

In an embodiment, the training process of the smart initialization technique (i.e., Hint) is simple, quick, and/or may be performed using one or more processors. In another embodiment, complexity may be added to the iterative detector. The added complexity to the iterative detector may be very small in comparison with the complexity of the underlying (i.e., typical) iterative detector. The training complexity, which does not impact real time complexity, may be ignored.

Using the smart initialization technique (i.e., Hint) in combination with an iterative detector increases the convergence rate of the iterative detectors above one or more thresholds.

In some embodiments, due to the different conditions of the channel, using one rigid initial vector for the iterative detectors may not an efficient choice. However, the smart initialization technique (i.e., Hint) has the power to tailor an initial vector for the underlying channel situation. Consequently, the suggested initial vector works well in all channel situations.

Some embodiments may include one or more of the following:

1- A system and method for iterative detection of multiuser massive MIMO systems where the initial estimate of the detected signal is parametrized by a small number of parameters that are obtained using learning-based techniques.

2- The iterative detector may utilize the Gauss-Seidel (GS) technique to solve the MMSE multiuser massive MIMO detection problem. The initial vector may be a linear combination of first terms of the Neumann series of a matrix inverse component of the MMSE solution. Linear combination coefficients may be obtained using learning-based techniques. 3- A diagonal loading factor of the MMSE detector may be obtained using learning-based techniques.

4- Trained parameters may be dependent on the number of active users and the number of base antennas. In this case, different values of the parameters may be obtained through training for each possible scenario and stored to be used for the corresponding scenario.

5- The parameters may be retrained if a major change in a channel is detected. Such detection may include monitoring the average means squared error (MSE) or the Error Vector Magnitude (EVM) and declaring a major change in the channel if the average MSE or EVM is larger than a preconfigured threshold.

One or more embodiments provide an increase of the convergence rate of the underlying iterative detector (thereby decreasing the overall complexity of the massive MIMO detector) and address a need for sustainability. In addition, some embodiments provide a robust solution, i.e., the initial vector and regularization parameter calculated for one channel condition may be reused when there is a change in the channel because the drop in SER performance may be negligible. Further, the training process may be short and performed using one or more processors, and retraining may be performed without the need for equipping the NNs (e.g., BS) with complex computation resources.

According to one aspect, a network node configured to communicate with a wireless device (WD) using a massive multiple input multiple output (MIMO) process and a minimum mean square error (MMSE) process is provided. The network node is configured to determine an initial estimate of a signal, the determined initial estimate being usable to perform an iterative detection process and being parameterized by a set of parameters, the set of parameters being based at least in part on a learning-based process.

According to this aspect, in some embodiments, at least one of: the performed iterative detection process utilizes a Gauss-Seidel (GS) technique; and the determined initial estimate of the signal is an initial vector, the initial vector being a linear combination of first terms of a Neumann series of a matrix inverse component of the MMSE process, the linear combination being associated with linear combination coefficients obtained using the learning-based process. In some embodiments, the network node is configured to determine a diagonal loading factor of the MMSE process using the learning-based process to at least one of determine the initial estimate and perform the interactive detection process. In some embodiments, at least one parameter of the set of parameters is based at least in part on a first quantity of active users and a second quantity of base antennas, the at least one parameter being determined using the learningbased process for each scenario of a plurality of scenarios, the at least one parameter being used to determine the initial estimate based at least in part on a corresponding scenario. In some embodiments, the network node is configured to at least one of: detect that a major change in a channel has occurred when an average MSE is greater than a preconfigured threshold; and retrain, using the learning-based process, at least one parameter of the set of parameters if the major change in the channel is detected. In some embodiments, the network node is configured to perform the iterative detection process based at least in part on the determined initial estimate of the signal.

According to another aspect, a method implemented in a network node configured to communicate with a wireless device (WD) using a massive multiple input multiple output (MIMO) process and a minimum mean square error (MMSE) process is provided. The method includes determining an initial estimate of a signal, the determined initial estimate being usable to perform an iterative detection process and being parameterized by a set of parameters, the set of parameters being based at least in part on a learning-based process.

According to this aspect, in some embodiments, at least one of: the performed iterative detection process utilizes a Gauss-Seidel (GS) technique; and the determined initial estimate of the signal is an initial vector, the initial vector being a linear combination of first terms of a Neumann series of a matrix inverse component of the MMSE process, the linear combination being associated with linear combination coefficients obtained using the learning-based process. In some embodiments, the method includes determining a diagonal loading factor of the MMSE process using the learningbased process to at least one of determine the initial estimate and perform the interactive detection process. In some embodiments, at least one parameter of the set of parameters is based at least in part on a first quantity of active users and a second quantity of base antennas, the at least one parameter being determined using the learning-based process for each scenario of a plurality of scenarios, the at least one parameter being used to determine the initial estimate based at least in part on a corresponding scenario. In some embodiments, the method includes at least one of: detecting that a major change in a channel has occurred when an average MSE is greater than a preconfigured threshold; and retraining, using the learning-based process, at least one parameter of the set of parameters if the major change in the channel is detected. In some embodiments, the method includes performing the iterative detection process based at least in part on the determined initial estimate of the signal.

According to yet another aspect, a wireless device (WD) configured to communicate with a network node using a massive multiple input multiple output (MIMO) process and a minimum mean square error (MMSE) process, the WD configured to interoperate and/or communicate with and/or signal to/from the network node in accordance with any of the methods disclosed herein is provided.

According to another aspect, method implemented in a wireless device (WD) configured to communicate with a network node using a massive multiple input multiple output (MIMO) process and a minimum mean square error (MMSE) process, the method comprising interoperating and/or communicating with and/or signaling to/from the WD in accordance with any of the methods disclosed herein is provided.

BRIEF DESCRIPTION OF THE DRAWINGS

A more complete understanding of the present embodiments, and the attendant advantages and features thereof, will be more readily understood by reference to the following detailed description when considered in conjunction with the accompanying drawings wherein:

FIG. 1 shows an example SER performance of MMNet-iid when there is a mismatch between trained and tested massive MIMO systems, i.e. the number of served users by the network node (base station), K, changed;

FIG. 2 is a schematic diagram of an example network architecture illustrating a communication system connected via an intermediate network to a host computer according to the principles in the present disclosure;

FIG. 3 is a block diagram of a host computer communicating via a network node with a wireless device over an at least partially wireless connection according to some embodiments of the present disclosure;

FIG. 4 is a flowchart illustrating example methods implemented in a communication system including a host computer, a network node and a wireless device for executing a client application at a wireless device according to some embodiments of the present disclosure;

FIG. 5 is a flowchart illustrating example methods implemented in a communication system including a host computer, a network node and a wireless device for receiving user data at a wireless device according to some embodiments of the present disclosure;

FIG. 6 is a flowchart illustrating example methods implemented in a communication system including a host computer, a network node and a wireless device for receiving user data from the wireless device at a host computer according to some embodiments of the present disclosure;

FIG. 7 is a flowchart illustrating example methods implemented in a communication system including a host computer, a network node and a wireless device for receiving user data at a host computer according to some embodiments of the present disclosure;

FIG. 8 is a flowchart of an example process in a network node according to some embodiments of the present disclosure;

FIG. 9 shows an example massive MIMO system with N antennas providing service for K users according to some embodiments of the present disclosure;

FIG. 10 shows an example process for providing an initial vector and/regularization parameter to the iterative detector according to some embodiments of the present disclosure;

FIG. 11 shows example SER performance on i.i.d. Rayleigh fading channel for different massive MIMO configurations and 64-QAM modulation based on a combination of the smart initialization technique (i.e., Hint) and GS according to some embodiments of the present disclosure;

FIG. 12 shows example SER performance in 3GPP realistic channel for different massive MIMO configurations and different QAM modulations based on a combination of the smart initialization technique (i.e., Hint) and GS according to some embodiments of the present disclosure;

FIG. 13 shows example SER performance on i.i.d. Rayleigh fading when there is a mismatch between the trained and test massive MIMO systems, i.e., the number of served users by the network node (e.g. base station), K, changed, according to some embodiments of the present disclosure;

FIG. 14 shows an example SER performance on 3GPP realistic channel when there is a mismatch between the trained and test massive MIMO systems (in terms of number of users) according to some embodiments of the present disclosure; FIG. 15 shows an example SER performance on 3 GPP realistic channel when there is a mismatch between the trained and test massive MIMO systems (in terms of channel correlation) according to some embodiments of the present disclosure;

FIG. 16 shows a flowchart of an example detection process in massive MIMO system according to some embodiments of the present disclosure; and

FIG. 17 shows a flowchart of another detection process in massive MIMO system according to some embodiments of the present disclosure.

DETAILED DESCRIPTION

Before describing in detail example embodiments, it is noted that the embodiments reside primarily in combinations of apparatus components and processing steps related to determining initial vectors usable to perform an iterative detection process in massive multiple input multiple output (MIMO) systems. Accordingly, components have been represented where appropriate by conventional symbols in the drawings, showing only those specific details that are pertinent to understanding the embodiments so as not to obscure the disclosure with details that will be readily apparent to those of ordinary skill in the art having the benefit of the description herein. Like numbers refer to like elements throughout the description.

As used herein, relational terms, such as “first” and “second,” “top” and “bottom,” and the like, may be used solely to distinguish one entity or element from another entity or element without necessarily requiring or implying any physical or logical relationship or order between such entities or elements. The terminology used herein is for the purpose of describing particular embodiments only and is not intended to be limiting of the concepts described herein. As used herein, the singular forms “a”, “an” and “the” are intended to include the plural forms as well, unless the context clearly indicates otherwise. It will be further understood that the terms “comprises,” “comprising,” “includes” and/or “including” when used herein, specify the presence of stated features, integers, steps, operations, elements, and/or components, but do not preclude the presence or addition of one or more other features, integers, steps, operations, elements, components, and/or groups thereof.

In embodiments described herein, the joining term, “in communication with” and the like, may be used to indicate electrical or data communication, which may be accomplished by physical contact, induction, electromagnetic radiation, radio signaling, infrared signaling or optical signaling, for example. One having ordinary skill in the art will appreciate that multiple components may interoperate, and modifications and variations are possible of achieving the electrical and data communication.

In some embodiments described herein, the term “coupled,” “connected,” and the like, may be used herein to indicate a connection, although not necessarily directly, and may include wired and/or wireless connections.

The term “network node” used herein may be any kind of network node comprised in a radio network which may further comprise any of base station (BS), radio base station, base transceiver station (BTS), base station controller (BSC), radio network controller (RNC), g Node B (gNB), evolved Node B (eNB or eNodeB), Node B, multistandard radio (MSR) radio node such as MSR BS, multi-cell/multicast coordination entity (MCE), integrated access and backhaul (IAB) node, relay node, donor node controlling relay, radio access point (AP), transmission points, transmission nodes, Remote Radio Unit (RRU) Remote Radio Head (RRH), a core network node (e.g., mobile management entity (MME), self-organizing network (SON) node, a coordinating node, positioning node, MDT node, etc.), an external node (e.g., 3 ^rd party node, a node external to the current network), nodes in distributed antenna system (DAS), a spectrum access system (SAS) node, an element management system (EMS), etc. The network node may also comprise test equipment. The term “radio node” used herein may be used to also denote a wireless device (WD) such as a wireless device (WD) or a radio network node.

In some embodiments, the non-limiting terms wireless device (WD) or a user equipment (UE) are used interchangeably. The WD herein may be any type of wireless device capable of communicating with a network node or another WD over radio signals, such as wireless device (WD). The WD may also be a radio communication device, target device, device to device (D2D) WD, machine type WD or WD capable of machine to machine communication (M2M), low-cost and/or low-complexity WD, a sensor equipped with WD, Tablet, mobile terminals, smart phone, laptop embedded equipped (LEE), laptop mounted equipment (LME), USB dongles, Customer Premises Equipment (CPE), an Internet of Things (loT) device, or a Narrowband loT (NB-IOT) device, etc.

Also, in some embodiments the generic term “radio network node” is used. It may be any kind of a radio network node which may comprise any of base station, radio base station, base transceiver station, base station controller, network controller, RNC, evolved Node B (eNB), Node B, gNB, Multi-cell/multicast Coordination Entity (MCE), IAB node, relay node, access point, radio access point, Remote Radio Unit (RRU) Remote Radio Head (RRH). Note that although terminology from one particular wireless system, such as, for example, 3GPP LTE and/or New Radio (NR), may be used in this disclosure, this should not be seen as limiting the scope of the disclosure to only the aforementioned system. Other wireless systems, including without limitation Wide Band Code Division Multiple Access (WCDMA), Worldwide Interoperability for Microwave Access (WiMax), Ultra Mobile Broadband (UMB) and Global System for Mobile Communications (GSM), may also benefit from exploiting the ideas covered within this disclosure.

Note further, that functions described herein as being performed by a wireless device or a network node may be distributed over a plurality of wireless devices and/or network nodes. In other words, it is contemplated that the functions of the network node and wireless device described herein are not limited to performance by a single physical device and, in fact, may be distributed among several physical devices.

Unless otherwise defined, all terms (including technical and scientific terms) used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this disclosure belongs. It will be further understood that terms used herein should be interpreted as having a meaning that is consistent with their meaning in the context of this specification and the relevant art and will not be interpreted in an idealized or overly formal sense unless expressly so defined herein.

Referring again to the drawing figures, in which like elements are referred to by like reference numerals, there is shown in FIG. 2 a schematic diagram of a communication system 10, according to an embodiment, such as a 3 GPP-type cellular network that may support standards such as LTE and/or NR (5G), which comprises an access network 12, such as a radio access network, and a core network 14. The access network 12 comprises a plurality of network nodes 16a, 16b, 16c (referred to collectively as network nodes 16), such as NBs, eNBs, gNBs or other types of wireless access points, each defining a corresponding coverage area 18a, 18b, 18c (referred to collectively as coverage areas 18). Each network node 16a, 16b, 16c is connectable to the core network 14 over a wired or wireless connection 20. A first wireless device (WD) 22a located in coverage area 18a is configured to wirelessly connect to, or be paged by, the corresponding network node 16a. A second WD 22b in coverage area 18b is wirelessly connectable to the corresponding network node 16b. While a plurality of WDs 22a, 22b (collectively referred to as wireless devices 22) are illustrated in this example, the disclosed embodiments are equally applicable to a situation where a sole WD is in the coverage area or where a sole WD is connecting to the corresponding network node 16. Note that although only two WDs 22 and three network nodes 16 are shown for convenience, the communication system may include many more WDs 22 and network nodes 16.

Also, it is contemplated that a WD 22 may be in simultaneous communication and/or configured to separately communicate with more than one network node 16 and more than one type of network node 16. For example, a WD 22 may have dual connectivity with a network node 16 that supports LTE and the same or a different network node 16 that supports NR. As an example, WD 22 may be in communication with an eNB for LTE/E-UTRAN and a gNB for NR/NG-RAN.

The communication system 10 may itself be connected to a host computer 24, which may be embodied in the hardware and/or software of a standalone server, a cloud- implemented server, a distributed server or as processing resources in a server farm. The host computer 24 may be under the ownership or control of a service provider or may be operated by the service provider or on behalf of the service provider. The connections 26, 28 between the communication system 10 and the host computer 24 may extend directly from the core network 14 to the host computer 24 or may extend via an optional intermediate network 30. The intermediate network 30 may be one of, or a combination of more than one of, a public, private or hosted network. The intermediate network 30, if any, may be a backbone network or the Internet. In some embodiments, the intermediate network 30 may comprise two or more sub-networks (not shown).

The communication system of FIG. 2 as a whole enables connectivity between one of the connected WDs 22a, 22b and the host computer 24. The connectivity may be described as an over-the-top (OTT) connection. The host computer 24 and the connected WDs 22a, 22b are configured to communicate data and/or signaling via the OTT connection, using the access network 12, the core network 14, any intermediate network 30 and possible further infrastructure (not shown) as intermediaries. The OTT connection may be transparent in the sense that at least some of the participating communication devices through which the OTT connection passes are unaware of routing of uplink and downlink communications. For example, a network node 16 may not or need not be informed about the past routing of an incoming downlink communication with data originating from a host computer 24 to be forwarded (e.g., handed over) to a connected WD 22a. Similarly, the network node 16 need not be aware of the future routing of an outgoing uplink communication originating from the WD 22a towards the host computer 24. A network node 16 is configured to include a NN management unit 32 which is configured to perform any step and/or task and/or process and/or method and/or feature described in the present disclosure, e.g., determine an initial estimate of a signal, the determined initial estimate being usable to perform an iterative detection process and parameterized by a set of parameters, the set of parameters being based on a learningbased process. A wireless device 22 is configured to include a WD management unit 34 which is configured to perform any step and/or task and/or process and/or method and/or feature described in the present disclosure, e.g., WD functions associated with the network node 16.

Example implementations, in accordance with an embodiment, of the WD 22, network node 16 and host computer 24 discussed in the preceding paragraphs will now be described with reference to FIG. 2. In a communication system 10, a host computer 24 comprises hardware (HW) 38 including a communication interface 40 configured to set up and maintain a wired or wireless connection with an interface of a different communication device of the communication system 10. The host computer 24 further comprises processing circuitry 42, which may have storage and/or processing capabilities. The processing circuitry 42 may include a processor 44 and memory 46. In particular, in addition to or instead of a processor, such as a central processing unit, and memory, the processing circuitry 42 may comprise integrated circuitry for processing and/or control, e.g., one or more processors and/or processor cores and/or FPGAs (Field Programmable Gate Array) and/or ASICs (Application Specific Integrated Circuitry) adapted to execute instructions. The processor 44 may be configured to access (e.g., write to and/or read from) memory 46, which may comprise any kind of volatile and/or nonvolatile memory, e.g., cache and/or buffer memory and/or RAM (Random Access Memory) and/or ROM (Read-Only Memory) and/or optical memory and/or EPROM (Erasable Programmable Read-Only Memory).

Processing circuitry 42 may be configured to control any of the methods and/or processes described herein and/or to cause such methods, and/or processes to be performed, e.g., by host computer 24. Processor 44 corresponds to one or more processors 44 for performing host computer 24 functions described herein. The host computer 24 includes memory 46 that is configured to store data, programmatic software code and/or other information described herein. In some embodiments, the software 48 and/or the host application 50 may include instructions that, when executed by the processor 44 and/or processing circuitry 42, causes the processor 44 and/or processing circuitry 42 to perform the processes described herein with respect to host computer 24. The instructions may be software associated with the host computer 24.

The software 48 may be executable by the processing circuitry 42. The software 48 includes a host application 50. The host application 50 may be operable to provide a service to a remote user, such as a WD 22 connecting via an OTT connection 52 terminating at the WD 22 and the host computer 24. In providing the service to the remote user, the host application 50 may provide user data which is transmitted using the OTT connection 52. The “user data” may be data and information described herein as implementing the described functionality. In one embodiment, the host computer 24 may be configured for providing control and functionality to a service provider and may be operated by the service provider or on behalf of the service provider. The processing circuitry 42 of the host computer 24 may enable the host computer 24 to observe, monitor, control, transmit to and/or receive from the network node 16 and or the wireless device 22. The processing circuitry 42 of the host computer 24 may include a host management unit 54 configured to perform any step and/or task and/or process and/or method and/or feature described in the present disclosure, e.g., enable the service provider to observe/monitor/ control/transmit to/receive from the network node 16 and or the wireless device 22.

The communication system 10 further includes a network node 16 provided in a communication system 10 and including hardware 58 enabling it to communicate with the host computer 24 and with the WD 22. The hardware 58 may include a communication interface 60 for setting up and maintaining a wired or wireless connection with an interface of a different communication device of the communication system 10, as well as a radio interface 62 for setting up and maintaining at least a wireless connection 64 with a WD 22 located in a coverage area 18 served by the network node 16. The radio interface 62 may be formed as or may include, for example, one or more RF transmitters, one or more RF receivers, and/or one or more RF transceivers. Radio interface 62 may also include one or more antennas 76 (e.g., MIMO antennas, MIMO system such as a massive MIMO). The communication interface 60 may be configured to facilitate a connection 66 to the host computer 24. The connection 66 may be direct or it may pass through a core network 14 of the communication system 10 and/or through one or more intermediate networks 30 outside the communication system 10.

In the embodiment shown, the hardware 58 of the network node 16 further includes processing circuitry 68. The processing circuitry 68 may include a processor 70 and a memory 72. In particular, in addition to or instead of a processor, such as a central processing unit, and memory, the processing circuitry 68 may comprise integrated circuitry for processing and/or control, e.g., one or more processors and/or processor cores and/or FPGAs (Field Programmable Gate Array) and/or ASICs (Application Specific Integrated Circuitry) adapted to execute instructions. The processor 70 may be configured to access (e.g., write to and/or read from) the memory 72, which may comprise any kind of volatile and/or nonvolatile memory, e.g., cache and/or buffer memory and/or RAM (Random Access Memory) and/or ROM (Read-Only Memory) and/or optical memory and/or EPROM (Erasable Programmable Read-Only Memory).

Thus, the network node 16 further has software 74 stored internally in, for example, memory 72, or stored in external memory (e.g., database, storage array, network storage device, etc.) accessible by the network node 16 via an external connection. The software 74 may be executable by the processing circuitry 68. The processing circuitry 68 may be configured to control any of the methods and/or processes described herein and/or to cause such methods, and/or processes to be performed, e.g., by network node 16. Processor 70 corresponds to one or more processors 70 for performing network node 16 functions described herein. The memory 72 is configured to store data, programmatic software code and/or other information described herein. In some embodiments, the software 74 may include instructions that, when executed by the processor 70 and/or processing circuitry 68, causes the processor 70 and/or processing circuitry 68 to perform the processes described herein with respect to network node 16. For example, processing circuitry 68 of the network node 16 may include NN management unit 32 which is configured to perform any step and/or task and/or process and/or method and/or feature described in the present disclosure, e.g., determine an initial estimate of a signal, the determined initial estimate being usable to perform an iterative detection process and parameterized by a set of parameters, the set of parameters being based on a learningbased process.

The communication system 10 further includes the WD 22 already referred to. The WD 22 may have hardware 80 that may include a radio interface 82 configured to set up and maintain a wireless connection 64 with a network node 16 serving a coverage area 18 in which the WD 22 is currently located. The radio interface 82 may be formed as or may include, for example, one or more RF transmitters, one or more RF receivers, and/or one or more RF transceivers. The radio interface may also be configured as radio interface 62 and/or comprise any other components of radio interface 62. The hardware 80 of the WD 22 further includes processing circuitry 84. The processing circuitry 84 may include a processor 86 and memory 88. In particular, in addition to or instead of a processor, such as a central processing unit, and memory, the processing circuitry 84 may comprise integrated circuitry for processing and/or control, e.g., one or more processors and/or processor cores and/or FPGAs (Field Programmable Gate Array) and/or ASICs (Application Specific Integrated Circuitry) adapted to execute instructions. The processor 86 may be configured to access (e.g., write to and/or read from) memory 88, which may comprise any kind of volatile and/or nonvolatile memory, e.g., cache and/or buffer memory and/or RAM (Random Access Memory) and/or ROM (Read-Only Memory) and/or optical memory and/or EPROM (Erasable Programmable Read-Only Memory).

Thus, the WD 22 may further comprise software 90, which is stored in, for example, memory 88 at the WD 22, or stored in external memory (e.g., database, storage array, network storage device, etc.) accessible by the WD 22. The software 90 may be executable by the processing circuitry 84. The software 90 may include a client application 92. The client application 92 may be operable to provide a service to a human or non-human user via the WD 22, with the support of the host computer 24. In the host computer 24, an executing host application 50 may communicate with the executing client application 92 via the OTT connection 52 terminating at the WD 22 and the host computer 24. In providing the service to the user, the client application 92 may receive request data from the host application 50 and provide user data in response to the request data. The OTT connection 52 may transfer both the request data and the user data. The client application 92 may interact with the user to generate the user data that it provides.

The processing circuitry 84 may be configured to control any of the methods and/or processes described herein and/or to cause such methods, and/or processes to be performed, e.g., by WD 22. The processor 86 corresponds to one or more processors 86 for performing WD 22 functions described herein. The WD 22 includes memory 88 that is configured to store data, programmatic software code and/or other information described herein. In some embodiments, the software 90 and/or the client application 92 may include instructions that, when executed by the processor 86 and/or processing circuitry 84, causes the processor 86 and/or processing circuitry 84 to perform the processes described herein with respect to WD 22. For example, the processing circuitry 84 of the wireless device 22 may include a WD management unit 34 which is configured to perform any step and/or task and/or process and/or method and/or feature described in the present disclosure, e.g., WD functions associated with the network node 16.

In some embodiments, the inner workings of the network node 16, WD 22, and host computer 24 may be as shown in FIG. 3 and independently, the surrounding network topology may be that of FIG. 2.

In FIG. 3, the OTT connection 52 has been drawn abstractly to illustrate the communication between the host computer 24 and the wireless device 22 via the network node 16, without explicit reference to any intermediary devices and the precise routing of messages via these devices. Network infrastructure may determine the routing, which it may be configured to hide from the WD 22 or from the service provider operating the host computer 24, or both. While the OTT connection 52 is active, the network infrastructure may further take decisions by which it dynamically changes the routing (e.g., on the basis of load balancing consideration or reconfiguration of the network).

The wireless connection 64 between the WD 22 and the network node 16 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 the WD 22 using the OTT connection 52, in which the wireless connection 64 may form the last segment. More precisely, the teachings of some of these embodiments may improve the data rate, latency, and/or power consumption and thereby provide benefits such as reduced user waiting time, relaxed restriction on file size, better responsiveness, extended battery lifetime, etc.

In some embodiments, a measurement procedure may be provided for the purpose of monitoring data rate, latency and other factors on which the one or more embodiments improve. There may further be an optional network functionality for reconfiguring the OTT connection 52 between the host computer 24 and WD 22, in response to variations in the measurement results. The measurement procedure and/or the network functionality for reconfiguring the OTT connection 52 may be implemented in the software 48 of the host computer 24 or in the software 90 of the WD 22, or both. In embodiments, sensors (not shown) may be deployed in or in association with communication devices through which the OTT connection 52 passes; the sensors may participate in the measurement procedure by supplying values of the monitored quantities exemplified above or supplying values of other physical quantities from which software 48, 90 may compute or estimate the monitored quantities. The reconfiguring of the OTT connection 52 may include message format, retransmission settings, preferred routing etc.; the reconfiguring need not affect the network node 16, and it may be unknown or imperceptible to the network node 16. Some such procedures and functionalities may be known and practiced in the art. In certain embodiments, measurements may involve proprietary WD signaling facilitating the host computer’s 24 measurements of throughput, propagation times, latency and the like. In some embodiments, the measurements may be implemented in that the software 48, 90 causes messages to be transmitted, in particular empty or ‘dummy’ messages, using the OTT connection 52 while it monitors propagation times, errors, etc.

Thus, in some embodiments, the host computer 24 includes processing circuitry 42 configured to provide user data and a communication interface 40 that is configured to forward the user data to a cellular network for transmission to the WD 22. In some embodiments, the cellular network also includes the network node 16 with a radio interface 62. In some embodiments, the network node 16 is configured to, and/or the network node’s 16 processing circuitry 68 is configured to perform the functions and/or methods described herein for preparing/initiating/maintaining/supporting/ending a transmission to the WD 22, and/or preparing/terminating/maintaining/supporting/ending in receipt of a transmission from the WD 22.

In some embodiments, the host computer 24 includes processing circuitry 42 and a communication interface 40 that is configured to a communication interface 40 configured to receive user data originating from a transmission from a WD 22 to a network node 16. In some embodiments, the WD 22 is configured to, and/or comprises a radio interface 82 and/or processing circuitry 84 configured to perform the functions and/or methods described herein for preparing/initiating/maintaining/supporting/ending a transmission to the network node 16, and/or preparing/terminating/maintaining/supporting/ending in receipt of a transmission from the network node 16.

Although FIGS. 2 and 3 show various “units” such as NN management unit 32, and WD management unit 34 as being within a respective processor, it is contemplated that these units may be implemented such that a portion of the unit is stored in a corresponding memory within the processing circuitry. In other words, the units may be implemented in hardware or in a combination of hardware and software within the processing circuitry.

FIG. 4 is a flowchart illustrating an example method implemented in a communication system, such as, for example, the communication system of FIGS. 2 and 3, in accordance with one embodiment. The communication system may include a host computer 24, a network node 16 and a WD 22, which may be those described with reference to FIG. 3. In a first step of the method, the host computer 24 provides user data (Block S100). In an optional substep of the first step, the host computer 24 provides the user data by executing a host application, such as, for example, the host application 50 (Block S102). In a second step, the host computer 24 initiates a transmission carrying the user data to the WD 22 (Block S104). In an optional third step, the network node 16 transmits to the WD 22 the user data which was carried in the transmission that the host computer 24 initiated, in accordance with the teachings of the embodiments described throughout this disclosure (Block S106). In an optional fourth step, the WD 22 executes a client application, such as, for example, the client application 92, associated with the host application 50 executed by the host computer 24 (Block S108).

FIG. 5 is a flowchart illustrating an example method implemented in a communication system, such as, for example, the communication system of FIG. 2, in accordance with one embodiment. The communication system may include a host computer 24, a network node 16 and a WD 22, which may be those described with reference to FIGS. 2 and 3. In a first step of the method, the host computer 24 provides user data (Block S 110). In an optional substep (not shown) the host computer 24 provides the user data by executing a host application, such as, for example, the host application 50. In a second step, the host computer 24 initiates a transmission carrying the user data to the WD 22 (Block SI 12). The transmission may pass via the network node 16, in accordance with the teachings of the embodiments described throughout this disclosure. In an optional third step, the WD 22 receives the user data carried in the transmission (Block S 114).

FIG. 6 is a flowchart illustrating an example method implemented in a communication system, such as, for example, the communication system of FIG. 2, in accordance with one embodiment. The communication system may include a host computer 24, a network node 16 and a WD 22, which may be those described with reference to FIGS. 2 and 3. In an optional first step of the method, the WD 22 receives input data provided by the host computer 24 (Block S 116). In an optional substep of the first step, the WD 22 executes the client application 92, which provides the user data in reaction to the received input data provided by the host computer 24 (Block S 118). Additionally or alternatively, in an optional second step, the WD 22 provides user data (Block S120). In an optional substep of the second step, the WD provides the user data by executing a client application, such as, for example, client application 92 (Block S122). In providing the user data, the executed client application 92 may further consider user input received from the user. Regardless of the specific manner in which the user data was provided, the WD 22 may initiate, in an optional third substep, transmission of the user data to the host computer 24 (Block S124). In a fourth step of the method, the host computer 24 receives the user data transmitted from the WD 22, in accordance with the teachings of the embodiments described throughout this disclosure (Block S126).

FIG. 7 is a flowchart illustrating an example method implemented in a communication system, such as, for example, the communication system of FIG. 2, in accordance with one embodiment. The communication system may include a host computer 24, a network node 16 and a WD 22, which may be those described with reference to FIGS. 2 and 3. In an optional first step of the method, in accordance with the teachings of the embodiments described throughout this disclosure, the network node 16 receives user data from the WD 22 (Block S128). In an optional second step, the network node 16 initiates transmission of the received user data to the host computer 24 (Block S130). In a third step, the host computer 24 receives the user data carried in the transmission initiated by the network node 16 (Block S132).

FIG. 8 is a flowchart of an example process in a network node 16. One or more blocks described herein may be performed by one or more elements of network node 16 such as by one or more of processing circuitry 68 (including the NN management unit 32), processor 70, radio interface 62 (including antennas 76) and/or communication interface 60. Network node 16 such as via processing circuitry 68 and/or processor 70 and/or radio interface 62 and/or communication interface 60 is configured to determine (Block S134) an initial estimate of a signal, where the determined initial estimate is usable to perform an iterative detection process and is parameterized by a set of parameters, the set of parameters being based on a learning-based process.

In some embodiments, at least one of the performed iterative detection process utilizes a Gauss-Seidel (GS) technique, and the determined initial estimate of the signal is an initial vector. The initial vector is a linear combination of first terms of a Neumann series of a matrix inverse component of the MMSE process. The linear combination is associated with linear combination coefficients obtained using a learning-based process.

In some embodiments, the method further includes determining a diagonal loading factor of the MMSE process using the learning-based process to at least one of determine the initial estimate and perform the interactive detection process.

In some embodiments, at least one parameter of the set of parameters is based on a first quantity of active users and a second quantity of base antennas. The at least one parameter is determined using the learning-based process for each scenario of a plurality of scenarios. The at least one parameter is used to determine the initial estimate based on a corresponding scenario.

In some embodiments, at least one of the network node and the processing circuitry is configured to at least one of detect that a major change in a channel has occurred when an average MSE is greater than a preconfigured threshold; and retrain, using the learning-based process, at least one parameter of the set of parameters if the major change in the channel is detected.

In some embodiments, the method further includes performing the iterative detection process based on the determined initial estimate of the signal.

Having described the general process flow of arrangements of the disclosure and having provided examples of hardware and software arrangements for implementing the processes and functions of the disclosure, the sections below provide details and examples of arrangements for determining initial vectors usable to perform an iterative detection process in massive multiple input multiple output (MIMO) systems. In some embodiments, Hint refers to one or more steps performed by any one of the NN 16 (and/or any of its components), WD 22 (and/or any of its components), and/or any other component of communication system 10.

FIG. 9 shows an example massive MIMO system with N antennas providing service for K users according to some embodiments of the present disclosure. Network node 16 may include N antennas 76 and be configured to communicate with one or more WDs 22a, 22b, 22c.

FIG. 10 shows an example process for providing an initial vector and/regularization parameter to the iterative detector to form G _MMSr . More specifically, NN management unit 32 may be configured to perform one or more steps associated with the smart initialization technique (i.e., Hint).

In multiuser massive MIMO systems, K users are served by N antennas at the NN 16 such as a base station (FIG. 9). A user may refer to a WD 22 and/or K users may refer to K WDs 22. N and K may be larger (e.g., substantially larger) than in traditional networks. In the uplink, each user transmits a symbol from a constellation set (1 . To simplify the discussion, all users are assumed to use the same constellation set, which is not expected to materially impact results. After transmission through a wireless channel H and corrupted by noise n, signals from all WDs 22 (e.g., users) may simultaneously reach the NN 16 (e.g., BS), forming the received vector x = [x ₁₍ x ₂ , ■■■ > x _w ] ^T . The received signal at the NN 16 (e.g., BS) may be described as x = Hs + n, where x 6 C ^N is the received signal vector at NN 16 (e.g., BS) and s = [s ₁₍ s ₂ , ... , s ] ^T is the vector of transmitted symbols by each user. Here, H and n denote the channel matrix and noise vector, respectively. The noise is white and Gaussian with variance i.e., n ~

Here, Hint is introduced, where Hint may be a technique inspired by the impact of the initial vector on the convergence rate of the iterative techniques. Hint is a simple and efficient learning-based approach that is trained in a few iterations and requires low computing power. Hint may use a learning-based approach such as to accelerate the convergence rate of the underlying iterative technique by finding the best initial solution and regularization parameter for the underlying massive MIMO system.

Using the first two terms of the Neumann series, ¹ E) ⁱ D~ ¹ )H ^H x as the initial vector for the GS technique is efficient in massive MIMO systems when -> 0. The initialization technique is used as the building block for Hint. To find an initial solution (e.g., a good initial solution), two scalar trainable parameters are introduced, i.e., a, and 8, to the two first terms of the NS and they are fed as the initial vector to an iterative detector:

Simultaneously, using Hint, a regularization parameter (e.g., a best regularization parameter) is determined by introducing a trainable parameter, y, to the GMMSE calculation formula, i.e.:

In other words, in one or more embodiments, Hint has only three trainable parameters, (a, 8, y), and one layer. In the final step of Hint, the generated initial vector is clipped to be in the range of the tightest convex polytope that contains/includes the constellation set. In some embodiments, this process may not add significantly to the complexity of the technique and/or only includes a comparison step. The output of Hint, i.e., the initial vector and regularization parameter, may be fed to a low complexity linear iterative technique to detect the transmitted signal in a massive MIMO system (e.g., in FIG. 10). To train Hint, Hint may be combined with an iterative detector. Then, the user data vector, s, of length K and noise vector, n, of length N, may be sampled from the constellation set and C/V(0, o _n ), respectively. The received vector x may be calculated, and x, H, and s may be fed to Hint-iterative detector combination to start the training process. In some embodiments, using one or more steps of an optimizer, Hint may only need ten epochs of training with a batch size of 500. For each training batch, the SNR may be selected randomly and uniformly from the desired operating interval of the SNR. In some embodiments, the training process may be repeated for three times the integer length of the desired SNR interval in dB scale, i.e.: number of iterations = 3| I-SNR l _m lli..A„ — SNR _m 11U; _n UJ I

Furthermore, the mean squared error loss function may be used, i.e.: L =11 s - s \\ ² where s is the true vector of transmitted signal and s is the vector of the detected signal at the output of the iterative detector.

Hint complexity: Putting the training complexity aside, Hint may share the calculation of the Gram matrix, GMMSE? ^anc l matched filter, H ^H X, with the underlying iterative method. Consequently, the added complexity to the underlying iterative detector is equal to K ² + 14/f. To show the effectiveness of Hint, Hint may be used in combination with the GS technique, i.e., an initial vector, and the regularization parameter for the GS technique using Hint may be produced. The GS techniques may be described using decomposed into a diagonal matrix, D, and a lower triangular matrix, L, i.e., GMMSE ⁼ D + L + L ^H .

After each iteration of the GS, the output vector may be clipped to be in the tightest convex polytope that includes the constellation set, as in the final step of the Hint. Further, the SER performance of the Hint-GS combination in i.i.d. Rayleigh fading and realistic 3GPP channels may be evaluated. More specifically, a block-fading Monte Carlo simulation is performed, where a block of length 1000 is considered in each simulation, and the results are averaged over experiments on 1000 channel realizations. Note that one or more Hint-GS combinations may be used.

The added computational complexity of using Hint: In this paragraph, the computational complexity (CC) of using Hint in combination with GS is analyzed. Table 1 shows the computation complexity of the Hint-GS combination (without considering the training period of the Hint), GS with three initializing techniques, and MMSE. As is shown, the complexity of GS as an iterative detector depends on I, the number of needed iterations for acceptable SER performance. By using a good initializing vector, the total complexity of the iterative detectors may be decreased. As shown in the following sections, using Hint as the initial vector dramatically reduces the needed number of iterations. Table 1: Computational Complexity (CC) analysis between different techniques. In this table, I is the number of iterations.

FIG. 11 shows example SER performance on an i.i.d. Rayleigh fading channel for different massive MIMO configurations and 64-QAM modulation based on a combination of the smart initialization technique (i.e., Hint) and GS according to some embodiments of the present disclosure.

SER performance on LED. Rayleigh fading MIMO Channel: In this section, the SER performance of the Hint-GS combination in i.i.d. Rayleigh fading channels is compared with MMSE, and the GS initialized by three different vectors, the zero vector, the first term of the NS, and the first two terms of the NS. As illustrated FIG. 11, in various configurations of the massive MIMO systems, using Hint on top of the GS improves the GS’s SER performance. The Hint-GS combination may achieve the MMSE’s SER performance with just one and two iterations in cases of 4 X 64 and

8 X 64 massive MIMO systems, respectively, while other initializations of the GS techniques need more iterations. In one case (e.g., an extreme case) of = 0.25, the Hint- GS combination converges to MMSE performance in just three iterations. As shown in FIG. 11, using other initialization techniques may require more iterations to reach the SER performance of the MMSE approach.

SER Performance on 3GPP MIMO Channel: To investigate the Hint-GS’s SER performance in the 3GPP 3D MIMO channel model, QuadRiGa (A 3-D multi-cell channel model with time evolution for enabling virtual field trials) may be used to generate realistic channel samples. The 3GPP channel parameters are set to be the same as urban macro (Uma), e.g.: antenna and user heights are equal to 25 m and 1.5 m, respectively. The WDs 22 (e.g., users) may be distributed randomly in front of the antenna with a minimum and maximum distance of 35 m and 500 m from the antenna, respectively. The channel bandwidth may be set to be 100 MHz, and the number of sub-carriers may be 1024. In one example, Hint was tested on one sub-carrier. Other sub-carriers may be needed to train the Hint.

FIG. 12 shows example SER performance in 3GPP realistic channel for different massive MIMO configurations and different QAM modulations based on a combination of the smart initialization technique (i.e., Hint) and GS. As depicted in FIG. 12, in the case of 4 X 64 and 8 X 64 massive MIMO systems, using Hint on top of GS accelerates the convergence rate. Further, the GS may show the same SER performance as the MMSE after 9 and 14 iterations, respectively. Without using Hint, at least four more iterations are needed to achieve acceptable SER performance for the GS and next best initialization technique. In the case of a 4 X 64 massive MIMO system, putting the Hint-GS combination aside, the best SER performance of GS happens when it is initialized by the first term of the NS, whereas in the cases of 8 X 64 and 16 X 64 massive MIMO systems, the best SER performance after Hint happens when using the zero vector as the initialization. This observation shows that using a fixed initial vector for all channel situations may not be adequate for certain cases, and a robust initial vector may be needed to adapt itself to different situations. Furthermore, in FIG. 12, case of = 0.25, the channel is highly correlated. Hint finds the optimum balance between the two terms of the NS. This balance results in the Hint-GS combination having a slightly worse SER performance in lower SNRs. But the balance equips it with a powerful tool, e.g., the initial guess, to achieve superior SER performance in more practical SNRs and to perform more closely to the MMSE approach.

Robustness on Massive MIMO Configurations: Another aspect of present disclosure is the robustness of Hint’s output. Depicted in FIGS. 13 and 14 is the comparison between the SER performance of Hint-GS combination trained and tested on the same massive MIMO systems (No Mismatch) and trained and tested on different massive MIMO systems (Mismatch). More specifically, FIG. 13 shows example SER performance on i.i.d. Rayleigh fading when there is a mismatch between the trained and tested massive MIMO systems based on a combination of the smart initialization technique (i.e., Hint) and GS. The mismatched Hint-GS combination is trained on a 64- QAM 32 x64 massive MIMO system. FIG. 14 shows an example SER performance on 3 GPP realistic channel when there is a mismatch between the trained and tested massive MIMO systems (in terms of number of users) based on a combination of the smart initialization technique (i.e., Hint) and GS. The mismatched Hint-GS combination is trained on a 64-QAM 16 x64 massive MIMO system.

For the case of i.i.d. Rayleigh fading channels, i.e., FIG. 13, Hint is trained on a 32 X 64 massive MIMO system and tested on a 16 X 64 massive MIMO system (both 64-QAM). For the case of the 3GPP realistic channel, i.e., FIG. 14, the Hint-GS combination is trained for a 16 X 64 massive MIMO system and tested on a 8 X 64 massive MIMO system (both 64-QAM). The mismatch and the no-mismatch Hint-GS combination’s SER performance are very close to each other in both cases of the i.i.d. Gaussian and 3GPP realistic channels. Further, because Hint's training process is short and may be done using one or more processors, it is possible to continue using the Hint-GS combination until it is trained for the new massive MIMO configuration. To test the robustness of the Hint-GS combination on different channel realizations, i.e., different channel correlations, the Hint-GS combination may be trained on a 3GPP channel realization on a 8 X 64 massive MIMO system, 64-QAM, with randomly placed users. The SER performance of the trained Hint-GS combination on newly realized 3GPP channels with random user placement may be tested.

FIG. 15 shows an example SER performance on a 3GPP realistic channel when there is a mismatch between the trained and tested massive MIMO systems (in terms of channel correlation) based on a combination of the smart initialization technique (i.e., Hint) and GS, in some embodiments. The mismatched Hint-GS combination is trained on a different channel realization with 64-QAM 8 x64 massive MIMO system. Put differently, depicted in FIG. 15, labeled with no-mismatch is the Hint-GS SER performance trained on each new channel realization, and marked with mismatch is the Hint-GS trained for one channel realization and tested on all channel realizations. As shown, the performance gap between match and no-mismatch Hint-GS performance is negligible. It shows that Hint-GS may be trained on a randomly chosen channel realization with a fixed number of users and use it for channel detection for other channel realizations.

FIG. 16 shows a flowchart of an example detection process in a massive MIMO system (e.g., in NN 16), and FIG. 17 shows a flowchart of another detection process in the massive MIMO system. In FIG. 16, at step S200, NN 16 may be configured to determine whether major changes in the channel realization have occurred. At step S202, NN 16 may be configured to load Hint parameters from memory 72, and at step S204, detect a signal using the iterative detector.

In FIG. 17, at step S300, NN 16 may be configured to determine whether major changes in the channel realization have occurred. At step S302, NN 16 may be configured to load Hint parameters from memory 72, and at step S304, detect a signal using the iterative detector. At step S306, a maximum iteration may be determined by NN 16. At step S305, NN 16 may be configured to determine whether i is less than the maximum iteration. If not, at step S310, Hint parameters are stored in memory 72. Otherwise, at step S312, s and n are generated, and x is calculated. At step S314, x, H, and s are fed to Hint, and another iteration is started.

In other words, in FIGS. 16 and 17, the simplified and detailed flowchart for the task of data detection in massive MIMO systems using Hint on top of an iterative detector is depicted. When NN 16 (e.g., BS) receives a signal vector, the signal detection task begins. If it is the first received vector on the new channel condition, the training process of Hint begins. Upon completion of Hint training, the new trained parameters may be stored in memory (e.g., memory 72). Simultaneously, the Hint parameters are loaded and fed to the iterative detector to detect the received signal. If Hint is already trained for the underlying channel, the Hint’s training process would be skipped. Due to the robustness of Hint (for different channel realizations and number of users), it is enough to repeat the training process every once in a while, when there are major changes in the channel, and not for every channel realization.

To detect major changes in the channel, a drift detection module (i.e., comprised in and/or performed by NN 16 or any other component of communication system 10) may be added, where the drift detection module may declare a major change in the channel if the monitored average mean square error (MSE) is larger than a preconfigured threshold. This may be achieved by continuously calculating and monitoring the average MSE using one of the following example approaches: Calculating MSE, II s — s II2, where s is the true vector of transmitted signal obtained by the same random sampling procedure used for training for a desired SNR interval, and s is the vector of the detected signal at the output of the iterative detector. MSE may be averaged over the desired SNR interval.

Calculating MSE, || s — s II2, where s is the true vector of transmitted signal obtained from real decoded user data that pass CRC check, and s is the vector of the detected signal at the output of the iterative detector.

In both approaches, MSE may be averaged over time using any time-averaging technique. The MSE may be calculated periodically based on a preconfigured periodicity

Due to the robustness of the Hint algorithm, a simpler method for detecting major changes in the channel is monitoring the number of active users in the channel. Whenever there is a change in the number of active users, a major change in the channel may be declared.

To train Hint, it may be combined with an iterative detector. Then the user data vector, s, of length K and noise vector, n, of length N, is sampled from the constellation set and CW(0, cr _n ), respectively. Then, the received vector x is calculated. The x, H, and s are fed to Hint-iterative detector combination to start the training process. This process is repeated for three times the integer length of the desired SNR interval in dB scale. Then, the new trained parameters are stored in memory for future use.

Some embodiments may include one or more of the following:

Embodiment Al. A network node configured to communicate with a wireless device (WD) using a massive multiple input multiple output (MIMO) process and a minimum mean square error (MMSE) process, the network node configured to, and/or comprising a radio interface and/or comprising processing circuitry configured to: determine an initial estimate of a signal, the determined initial estimate being usable to perform an iterative detection process and parameterized by a set of parameters, the set of parameters being based on a learning -based process.

Embodiment A2. The network node of Embodiment Al, wherein at least one of: the performed iterative detection process utilizes a Gauss-Seidel (GS) technique; and the determined initial estimate of the signal is an initial vector, the initial vector being a linear combination of first terms of a Neumann series of a matrix inverse component of the MMSE process, the linear combination being associated with linear combination coefficients obtained using the learning-based process.

Embodiment A3. The network node of any one of Embodiments Al and A2, wherein at least one of the network node and the processing circuitry is configured to: determine a diagonal loading factor of the MMSE process using the learning-based process to at least one of determine the initial estimate and perform the interactive detection process.

Embodiment A4. The network node of any one of Embodiments Al -A3, wherein at least one parameter of the set of parameters is based on a first quantity of active users and a second quantity of base antennas, the at least one parameter being determined using the learning-based process for each scenario of a plurality of scenarios, the at least one parameter being used to determine the initial estimate based on a corresponding scenario.

Embodiment A5. The network node of any one of Embodiments A1-A4, wherein at least one of the network node and the processing circuitry is configured to at least one of: detect that a major change in a channel has occurred when an average MSE is greater than a preconfigured threshold; and retrain, using the learning-based process, at least one parameter of the set of parameters if the major change in the channel is detected.

Embodiment A6. The network node of any one of Embodiments A1-A5, wherein at least one of the network node and the processing circuitry is configured to: perform the iterative detection process based on the determined initial estimate of the signal.

Embodiment Bl. A method implemented in a network node configured to communicate with a wireless device (WD) using a massive multiple input multiple output (MIMO) process and a minimum mean square error (MMSE) process, the method comprising: determining an initial estimate of a signal, the determined initial estimate being usable to perform an iterative detection process and parameterized by a set of parameters, the set of parameters being based on a learning-based process.

Embodiment B2. The method of Embodiment B l, wherein at least one of: the performed iterative detection process utilizes a Gauss-Seidel (GS) technique; and the determined initial estimate of the signal is an initial vector, the initial vector being a linear combination of first terms of a Neumann series of a matrix inverse component of the MMSE process, the linear combination being associated with linear combination coefficients obtained using the learning-based process.

Embodiment B3. The method of any one of Embodiments B 1 and B2, wherein the method further includes: determining a diagonal loading factor of the MMSE process using the learningbased process to at least one of determine the initial estimate and perform the interactive detection process.

Embodiment B4. The method of any one of Embodiments B 1-B3, wherein at least one parameter of the set of parameters is based on a first quantity of active users and a second quantity of base antennas, the at least one parameter being determined using the learning-based process for each scenario of a plurality of scenarios, the at least one parameter being used to determine the initial estimate based on a corresponding scenario.

Embodiment B5. The method of any one of Embodiments B 1-B4, wherein the method further includes at least one of: detecting that a major change in a channel has occurred when an average MSE is greater than a preconfigured threshold; and retraining, using the learning-based process, at least one parameter of the set of parameters if the major change in the channel is detected.

Embodiment B6. The method of any one of Embodiments B 1-B5, wherein the method further includes: performing the iterative detection process based on the determined initial estimate of the signal.

Embodiment Cl. A wireless device (WD) configured to communicate with a network node using a massive multiple input multiple output (MIMO) process and a minimum mean square error (MMSE) process, the WD configured to, and/or comprising a radio interface and/or processing circuitry configured to interoperate and/or communicate with and/or signal to/from the network node in accordance with any one of the methods of Embodiments B 1-B 12.

Embodiment DI. A method implemented in a wireless device (WD) configured to communicate with a network node using a massive multiple input multiple output (MIMO) process and a minimum mean square error (MMSE) process, the method comprising interoperating and/or communicating with and/or signaling to/from the WD in accordance with any one of the methods of Embodiments B 1-B 12.

As will be appreciated by one of skill in the art, the concepts described herein may be embodied as a method, data processing system, computer program product and/or computer storage media storing an executable computer program. Accordingly, the concepts described herein may take the form of an entirely hardware embodiment, an entirely software embodiment or an embodiment combining software and hardware aspects all generally referred to herein as a “circuit” or “module.” Any process, step, action and/or functionality described herein may be performed by, and/or associated to, a corresponding module, which may be implemented in software and/or firmware and/or hardware. Furthermore, the disclosure may take the form of a computer program product on a tangible computer usable storage medium having computer program code embodied in the medium that may be executed by a computer. Any suitable tangible computer readable medium may be utilized including hard disks, CD-ROMs, electronic storage devices, optical storage devices, or magnetic storage devices.

Some embodiments are described herein with reference to flowchart illustrations and/or block diagrams of methods, systems and computer program products. It will be understood that each block of the flowchart illustrations and/or block diagrams, and combinations of blocks in the flowchart illustrations and/or block diagrams, may be implemented by computer program instructions. These computer program instructions may be provided to a processor of a general purpose computer (to thereby create a special purpose computer), special purpose computer, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, create means for implementing the functions/acts specified in the flowchart and/or block diagram block or blocks.

These computer program instructions may also be stored in a computer readable memory or storage medium that may direct a computer or other programmable data processing apparatus to function in a particular manner, such that the instructions stored in the computer readable memory produce an article of manufacture including instruction means which implement the function/act specified in the flowchart and/or block diagram block or blocks.

The computer program instructions may also be loaded onto a computer or other programmable data processing apparatus to cause a series of operational steps to be performed on the computer or other programmable apparatus to produce a computer implemented process such that the instructions which execute on the computer or other programmable apparatus provide steps for implementing the functions/acts specified in the flowchart and/or block diagram block or blocks.

It is to be understood that the functions/acts noted in the blocks may occur out of the order noted in the operational illustrations. For example, two blocks shown in succession may in fact be executed substantially concurrently or the blocks may sometimes be executed in the reverse order, depending upon the functionality/acts involved. Although some of the diagrams include arrows on communication paths to show a primary direction of communication, it is to be understood that communication may occur in the opposite direction to the depicted arrows.

Computer program code for carrying out operations of the concepts described herein may be written in an object oriented programming language such as Python, Java® or C++. However, the computer program code for carrying out operations of the disclosure may also be written in conventional procedural programming languages, such as the "C" programming language. The program code may execute entirely on the user's computer, partly on the user's computer, as a stand-alone software package, partly on the user's computer and partly on a remote computer or entirely on the remote computer. In the latter scenario, the remote computer may be connected to the user's computer through a local area network (LAN) or a wide area network (WAN), or the connection may be made to an external computer (for example, through the Internet using an Internet Service Provider).

Many different embodiments have been disclosed herein, in connection with the above description and the drawings. It will be understood that it would be unduly repetitious and obfuscating to literally describe and illustrate every combination and subcombination of these embodiments. Accordingly, all embodiments may be combined in any way and/or combination, and the present specification, including the drawings, shall be construed to constitute a complete written description of all combinations and subcombinations of the embodiments described herein, and of the manner and process of making and using them, and shall support claims to any such combination or subcombination.

Abbreviations that may be used in the preceding description include:

BS Base station

CC Computational Complexity CG Conjugate gradient

DNN Deep Neural network

GS Gauss-Seidel

LED Independent, Identically Distributed

MIMO Multiple Input Multiple Output

ML Maximum Likelihood

MMSE Minimum Mean Squared Error

MSE Mean Squared Error

NS Neumann Series

OAMPNET Orthogonal Approximate Message Passing Network

OCD Optimized Coordinate Descent

SER Symbol Error Rate

SNR Signal to Noise Ratio

ZF Zero-Forcing

It will be appreciated by persons skilled in the art that the embodiments described herein are not limited to what has been particularly shown and described herein above. In addition, unless mention was made above to the contrary, it should be noted that all of the accompanying drawings are not to scale. A variety of modifications and variations are possible in light of the above teachings without departing from the scope of the following claims.