Method and apparatus for auto encoder based precoder information compression in a wireless communication system

Abstract

The disclosure relates to a 5G or 6G communication system for supporting a higher data transmission rate. The present disclosure relates to method and system for auto encoder based precoder information compression. The method of compressing precoder information, by a distribution unit (O-DU) (102) includes selecting an Artificial Intelligence (AI) model from one or more AI models corresponding to one or more Auto-Encoder (AE) based compression techniques, based on one or more parameters related to input precoder information to be compressed. Further, applying AE based compression to compress the input precoder information using the selected AI model. Subsequently, transmitting a control plane message comprising the compressed input precoder information and configuration information related to the compressed input precoder information to a radio unit (O-RU) (104) of the O-RAN. The auto encoder based precoder information compression exploits the possible correlation within the precoder data, thereby decreasing the bandwidth requirement for transferring the precoder information over the fronthaul.

Description

METHOD AND APPARATUS FOR AUTO ENCODER BASED PRECODER INFORMATION COMPRESSION IN A WIRELESS COMMUNICATION SYSTEM

[0001] The present disclosure relates to the field of wireless communication. More particularly, the disclosure relates to method, apparatus, and system for auto encoder based precoder information compression.

[0002] 5G mobile communication technologies define broad frequency bands such that high transmission rates and new services are possible, and can be implemented not only in “Sub 6GHz” bands such as 3.5GHz, but also in “Above 6GHz” bands referred to as mmWave including 28GHz and 39GHz. In addition, it has been considered to implement 6G mobile communication technologies (referred to as Beyond 5G systems) in terahertz (THz) bands (for example, 95GHz to 3THz bands) in order to accomplish transmission rates fifty times faster than 5G mobile communication technologies and ultra-low latencies one-tenth of 5G mobile communication technologies.

[0003] At the beginning of the development of 5G mobile communication technologies, in order to support services and to satisfy performance requirements in connection with enhanced Mobile BroadBand (eMBB), Ultra Reliable Low Latency Communications (URLLC), and massive Machine-Type Communications (mMTC), there has been ongoing standardization regarding beamforming and massive MIMO for mitigating radio-wave path loss and increasing radio-wave transmission distances in mmWave, supporting numerologies (for example, operating multiple subcarrier spacings) for efficiently utilizing mmWave resources and dynamic operation of slot formats, initial access technologies for supporting multi-beam transmission and broadbands, definition and operation of BWP (BandWidth Part), new channel coding methods such as a LDPC (Low Density Parity Check) code for large amount of data transmission and a polar code for highly reliable transmission of control information, L2 pre-processing, and network slicing for providing a dedicated network specialized to a specific service.

[0004] Currently, there are ongoing discussions regarding improvement and performance enhancement of initial 5G mobile communication technologies in view of services to be supported by 5G mobile communication technologies, and there has been physical layer standardization regarding technologies such as V2X (Vehicle-to-everything) for aiding driving determination by autonomous vehicles based on information regarding positions and states of vehicles transmitted by the vehicles and for enhancing user convenience, NR-U (New Radio Unlicensed) aimed at system operations conforming to various regulation-related requirements in unlicensed bands, NR UE Power Saving, Non-Terrestrial Network (NTN) which is UE-satellite direct communication for providing coverage in an area in which communication with terrestrial networks is unavailable, and positioning.

[0005] Moreover, there has been ongoing standardization in air interface architecture / protocol regarding technologies such as Industrial Internet of Things (IIoT) for supporting new services through interworking and convergence with other industries, IAB (Integrated Access and Backhaul) for providing a node for network service area expansion by supporting a wireless backhaul link and an access link in an integrated manner, mobility enhancement including conditional handover and DAPS (Dual Active Protocol Stack) handover, and two-step random access for simplifying random access procedures (2-step RACH for NR). There also has been ongoing standardization in system architecture / service regarding a 5G baseline architecture (for example, service based architecture or service based interface) for combining Network Functions Virtualization (NFV) and Software-Defined Networking (SDN) technologies, and Mobile Edge Computing (MEC) for receiving services based on UE positions.

[0006] As 5G mobile communication systems are commercialized, connected devices that have been exponentially increasing will be connected to communication networks, and it is accordingly expected that enhanced functions and performances of 5G mobile communication systems and integrated operations of connected devices will be necessary. To this end, new research is scheduled in connection with eXtended Reality (XR) for efficiently supporting AR (Augmented Reality), VR (Virtual Reality), MR (Mixed Reality) and the like, 5G performance improvement and complexity reduction by utilizing Artificial Intelligence (AI) and Machine Learning (ML), AI service support, metaverse service support, and drone communication.

[0007] Furthermore, such development of 5G mobile communication systems will serve as a basis for developing not only new waveforms for providing coverage in terahertz bands of 6G mobile communication technologies, multi-antenna transmission technologies such as Full Dimensional MIMO (FD-MIMO), array antennas and large-scale antennas, metamaterial-based lenses and antennas for improving coverage of terahertz band signals, high-dimensional space multiplexing technology using OAM (Orbital Angular Momentum), and RIS (Reconfigurable Intelligent Surface), but also full-duplex technology for increasing frequency efficiency of 6G mobile communication technologies and improving system networks, AI-based communication technology for implementing system optimization by utilizing satellites and AI (Artificial Intelligence) from the design stage and internalizing end-to-end AI support functions, and next-generation distributed computing technology for implementing services at levels of complexity exceeding the limit of UE operation capability by utilizing ultra-high-performance communication and computing resources.

[0008] The present disclosure relates to method, apparatus, and system for auto encoder based precoder information compression.

[0009] According to an aspect of an exemplary embodiment, there is provided a communication method in a wireless communication system.

[0010] Aspects of the present disclosure provide efficient communication methods in a wireless communication system.

[0011] The accompanying drawings, which are incorporated in and constitute a part of this disclosure, illustrate exemplary embodiments and, together with the description, explain the disclosed principles. In the figures, the left-most digit(s) of a reference number identifies the figure in which the reference number first appears. The same numbers are used throughout the figures to reference like features and components. Some embodiments of system and / or methods in accordance with embodiments of the present subject matter are now described, by way of example only, and regarding the accompanying figures, in which:

[0012] FIG. 1 shows an exemplary architecture for auto encoder based precoder information compression, in accordance with some embodiments of the present disclosure;

[0013] FIG. 2A shows a detailed block diagram of a distribution unit (O-DU) of an Open Radio Access Network (O-RAN) architecture, for auto encoder based precoder information compression, in accordance with some embodiments of the present disclosure;

[0014] FIG. 2B illustrates the SE 1 frame for transmitting the precoder information to the O-RU 104, in accordance with some embodiments of the present disclosure;

[0015] FIG. 3 shows a detailed block diagram of an O-RU of an O-RAN architecture, for auto encoder based precoder information decompression, in accordance with some embodiments of the present disclosure;

[0016] FIG. 4A illustrates an exemplary control plane signaling flow between O-DU 102 and O-RU, in accordance with the embodiments of the present disclosure;

[0017] FIG. 4B shows an exemplary illustration of AE based compression of precoder information, by an O-DU of an O-RAN architecture, in accordance with the embodiments of the present disclosure;

[0018] FIG. 5A shows an exemplary flowchart illustrating a method of compressing precoder information, by an O-DU of an O-RAN architecture, in accordance with some embodiments of the present disclosure;

[0019] FIG. 5B shows an exemplary flowchart illustrating a method of decompressing precoder information, by an O-RU of an O-RAN architecture, in accordance with some embodiments of the present disclosure;

[0020] FIG. 6 illustrates a general computer system architecture, in accordance with some embodiments of the present disclosure;

[0021] FIG. 7 illustrates a block diagram of a UE according to various embodiments of the present disclosure; and

[0022] FIG. 8 illustrates a block diagram of a base station or a network entity according to various embodiments of the present disclosure.

[0023] It should be appreciated by those skilled in the art that any block diagrams herein represent conceptual views of illustrative systems embodying the principles of the present subject matter. Similarly, it will be appreciated that any flow charts, flow diagrams, state transition diagrams, pseudo code, and the like represent various processes which may be substantially represented in computer readable medium and executed by a computer or processor, whether such computer or processor is explicitly shown.

[0024] Open Radio Access Network (O-RAN) provides an open interface between the Remote Radio Unit (RRU) and the Distributed Unit (DU), resulting in a modular architecture. This architecture allows service providers to mix and match components from different vendors without being locked to one vendor for all components, thus creating an open RAN network. In the O-RAN 7.2 architecture, most Layer 1 (L1) processing occurs in the Open Distributed Unit (O-DU), thus simplifying the processing at the Open Radio Unit (O-RU). The real-time aspects of control and user plane communication with the O-RU are managed by the O-DU over the fronthaul link (FH). The O-RAN 7.2 split architecture supports two categories of O-RU: Category A, where the O-RU performs analog-to-digital conversion, time domain processing, Cyclic Prefix (CP) removal, Fast Fourier Transform (FFT), and combining / compression; and Category B, where the O-RU performs these functions along with downlink (DL) precoding. This O-RAN 7.2 split architecture requires a significant amount of data transfer from the O-RU to the O-DU, resulting in higher fronthaul overhead. Additionally, with Category B O-RU, precoder / beamforming weights are transferred from the O-DU to the O-RU.

[0025] In the O-RAN 7.2 split architecture, the precoder weights (or precoder information) are computed at the O-DU and then communicated to the O-RU. The limited processing performed at O-RU requires the O-DU to compute the precoder weights and then communicate to the O-RU. To transfer the precoder information, the O-RAN supports compression techniques, which are based on one-dimensional (1D) or two-dimensional (2D) Discrete Fourier Transform (DFT) codebooks, similar to the 3rd Generation Partnership Project New Radio (3GPP NR) Type II codebooks.

[0026] However, in the traditionally used compression approach sharing precoder information between the O-DU and O-RU requires a significantly high payload size to achieve the required accuracy at the O-RU, leading to large fronthaul bandwidth usage. Opting for lower payload sizes results in poor error performance, causing communication failures for users in the downlink. Further, the traditionally used compression approach quantizes the real and imaginary part of the precoder coefficients individually, in conjunction with transformation to the DFT domain, increasing the fronthaul bandwidth usage.

[0027] Additionally, with the increasing number of antennas in beyond 5G and 6G systems, the overhead on the fronthaul link for sharing precoder information is expected to grow significantly. The currently deployed fiber links may fail to support the required data rates, necessitating the creation of new fiber infrastructure, which will be costly.

[0028] Therefore, there exists a need to decrease the bandwidth requirement for transferring the precoder information over the fronthaul, in the O-RAN.

[0029] The information disclosed in this background of the disclosure section is only for the enhancement of understanding of the general background of the invention and should not be taken as an acknowledgment or any form of suggestion that this information forms the prior art already known to a person skilled in the art.

[0030] In the present document, the word “exemplary” is used herein to mean “serving as an example, instance, or illustration.” Any embodiment or implementation of the present subject matter described herein as “exemplary” is not necessarily to be construed as preferred or advantageous over other embodiments.

[0031] While the disclosure is susceptible to various modifications and alternative forms, specific embodiment thereof has been shown by way of example in the drawings and will be described in detail below. It should be understood, however, that it is not intended to limit the disclosure to the specific forms disclosed, but on the contrary, the disclosure is to cover all modifications, equivalents, and alternatives falling within the scope of the disclosure.

[0032] The terms “comprises”, “comprising”, “includes”, or any other variations thereof, are intended to cover a non-exclusive inclusion, such that a setup, device, or method that comprises a list of components or steps does not include only those components or steps but may include other components or steps not expressly listed or inherent to such setup or device or method. In other words, one or more elements in a system or apparatus proceeded by “comprises... a” does not, without more constraints, preclude the existence of other elements or additional elements in the system or method.

[0033] The term “precoding” is used herein to refer to a signal processing technique that calculates and applies weights to the transmitted data streams. Precoding helps focus the beams of signals toward particular users, maximizing signal strength and reducing interference.

[0034] The term “Open RAN Radio Unit (O-RU),” is used herein to refer to an Open RAN Radio Unit or O-RAN RU, that resides closest to the user, typically mounted on cell towers or base stations. The O-RU is configured to perform the critical task of converting radio signals into digital data and vice versa. The RU also performs radio functions including a portion of physical layer (“PHY”) functions according to an LLS lower-layer split option. The RU performs conversions between radio frequency (“RF”) signals and baseband signals.

[0035] The term “Open RAN Distribution Unit (O-DU),” also referred to as Baseband Unit (BBU) is used herein to refer to the component of the 5G network architecture that connects to multiple radio units (RUs) in the southbound interface. The O-DU is responsible for managing and controlling the RUs, which are typically located at the cell sites. The O-DU performs baseband processing. In an open radio access network (“O-RAN”), the BBU and RU can be referred to as O-DU and O-RU, respectively. The DU is responsible for real-time layer 1 (L1, physical layer) and lower layer 2 (L2) which contains the data link layer and scheduling function. The O-DU is communicatively connected to the O-RU through a Fronthaul (FH) interface.

[0036] The term “channel information” is used herein to mean, for example, information about channel properties carried by the channel values. The term channel value / data can also be used herein to refer to, for example, one or a set of complex values representing the amplitude and phase of the channel coefficients in frequency domain. The channel values are related to the frequency response of the wireless channel.

[0037] The term “Channel State Information” (CSI) is used herein to refer to, detailed information about the current state of a communication channel. CSI includes complex values representing the amplitude and phase of the channel coefficients in the frequency domain.

[0038] The term “beam” is used herein to refer to a directional beam formed by multiplying a signal with different weights, in frequency-domain, at multiple antennas such that the energy of the wanted signal is concentrated to a certain direction and / or the energy of the interference signal is nulled at a certain direction. The term “beamforming” is used herein to refer to a technique which multiplies a signal with different weights (in the frequency domain) at multiple antennas, which enables the signal energy to be sent in space with a desired beam pattern by forming a directional beam concentrating on certain direction or forming nulling in certain direction, or a combination of both.

[0039] The term “Beamforming Weight” (“BFW”) is used herein to refer to a set of one or more complex weights, each set is multiplied with a signal of one user layer at a subcarrier or a group of subcarriers. The weighted signals of different user layers towards the same antenna or transmit beam are combined linearly.

[0040] As discussed in the background section, there is a need to decrease the bandwidth requirement for transferring the precoder information over the fronthaul to the O-RU. The existing predefined compression schemes in the O-RAN specification rely on individually quantizing the real and imaginary parts of the precoder coefficients. Hence, the existing techniques do not fully exploit the possible correlation present within the precoder data, increasing the bandwidth requirement for transferring the precoder information over the fronthaul. Hence, the present disclosure discloses Auto-Encoder (AE) based compression and decompression approach for transferring the precoder information from O-DU to O-RU.

[0041] Further, the present disclosure provides an apparatus (i.e., O-DU of an O-RAN architecture) for compressing the precoder information. Furthermore, the present disclosure provides an apparatus (i.e., O-RU of an O-RAN architecture) for decompressing precoder information. The present disclosure fully exploits the possible correlation within the precoder data, thereby decreasing the bandwidth requirement for transferring the precoder information over the fronthaul.

[0042] The proposed AI-based AEs are well known for their ability to compress data by exploiting the Frequency Domain (FD) correlation. The auto encoder based compression approach achieves higher precoder reconstruction accuracy at the O-RU compared to the existing compression techniques, thereby significantly reducing the fronthaul overhead. The present disclosure not only mitigates the fronthaul overhead but also results in better precoder reconstruction accuracy at the O-RU, leading to improved downlink performance. Additionally, the present disclosure proposes using necessary configuration parameters at the O-DU to communicate the specifics of the compression configuration to the O-RU. Moreover, the proposed control signaling and compression method is useful for Time Division Duplex (TDD) systems for transferring precoder information estimated using Sounding Reference Signal (SRS) receptions.

[0043] In the following detailed description of the embodiments of the disclosure, reference is made to the accompanying drawings that form a part hereof, and in which are shown by way of illustration specific embodiments in which the disclosure may be practiced. These embodiments are described in sufficient detail to enable those skilled in the art to practice the disclosure, and it is to be understood that other embodiments may be utilized and that changes may be made without departing from the scope of the present disclosure. The following description is, therefore, not to be taken in a limiting sense.

[0044] OVERVIEW

[0045] Generally, in traditional 1G / 2G systems, signals received from antennas are processed in a Base Station (BS) unit located close to a tower. The signals received from the antennas are transmitted to the BS unit via Radio Frequency (RF) cables where the entire processing (analog and digital signal processing) happens in the BS unit. However, the RF cables between Rx antennas and the BS unit are lossy connected. To minimize the losses due to RF cables, during 3G deployments, some techniques include a distributed architecture for the BS. Herein, the RF processing is performed in a Remote Radio Head (RRH), located close to the antennas while baseband processing is performed in a Base Band Unit (BBU). Then, the signals received are transmitted from RRH to BBU unit via optical cables using CPRI / OBSAI / ORI protocol.

[0046] Conventionally, the distributed architecture for the BS is configured using Centralized Radio Access Network (CRAN), Virtualized Radio Access Network (VRAN), Open Radio Access Network (ORAN).

[0047] In CRAN along with advancement of 5th Generation (5G), the distributed architecture evolved with large scale centralized base station deployments. Further, many Remote Radio Units (RRUs) connected to a centralized Baseband unit (BBU) pool and RRUs are connected to the BBU-pool via the front-haul network.

[0048] The VRAN decouples a software from hardware by virtualizing network functions. The C-RAN uses proprietary hardware while the VRAN uses network functions on the server platform. A Hardware (HW) / a Software (SW) decoupling offers flexibility and scalability and reduces hardware costs.

[0049] The VRAN is a closed network, as RRU and distributed unit (DU), must be bought from a same vendor. However, the ORAN provides Open interface between RRU and DU resulting in a modular architecture. The customer can mix and match components from different vendors without being locked to one vendor for all these components, thus resulting in an open RAN network.

[0050] The below Table A gives comparison between the traditional RAN / CRAN, the VRAN, and the ORAN.

[0051]

[0052] A functional split determines amount of functions performed locally at the antenna site (O-RU), and the amount of functions centralized at a high processing powered data centre (O-DU). 3rd Generation Partnership Project (3GPP) has proposed eight functional split options including several sub-options.

[0053] The functional split are two types, such as lower functional splits and higher functional splits.

[0054] In the lower functional splits, simpler functionalities at the O-RU making it simple for implementation. Most processing is done at O-DU, harnessing the benefits of centralized processing. Hence, high front-haul overhead occurs.

[0055] In the higher functional splits, most processing is done at the O-RU, making the O-RU computation intensive. The lower front-haul bandwidth is required. Hence, benefits of centralized processing such as interference management is lost.

[0056] To keep the O-RU implementation and maintenance simple and benefit from centralized processing, split options 8-6 are explored the most in the literature. Most L1 processing in 7.2 architecture happens in the O-DU, simplifying the processing at O-RU. The real time aspects of control and user plane communication with the O-RU are controlled by the O-DU over fronthaul link (FH).

[0057] In 7.2 split architecture, two categories of O-RU supported:

[0058] - Category A: O-RU performs analog-to-digital conversion, time domain processing, CP removal, FFT, combing / compression.

[0059] - Category B: O-RU performs analog-to-digital conversion, time domain processing, CP removal, FFT, combing / compression along with DL precoding.

[0060] However, this requires huge amount of data transfer from O-RU to O-DU, resulting in higher front-haul overhead. Moreover, with category B O-RU, precoder / beamforming weights are also transferred from O-DU to O-RU.

[0061] In ORAN 7.2, precoder is computed at the DU and then communicated to the RU. For transferring the precoding data, O-RAN based on 7.2 supports the following algorithms (along with some classic compression techniques):

[0062] - Beamspace compression type I (BCT-I)

[0063] - Beamspace compression type II (BCT-II).

[0064] The BCT algorithms are based on 1D or 2D DFT codebook, similar to the 3GPP NR Type II codebooks. However, DFT algorithms do not fully exploit the correlation data during compression as each of the real and imaginary coefficients of the precoder are compressed separately.

[0065] Existing O-RAN precoder compression algorithms are based on quantizing the real and imaginary parts of the precoder coefficients separately. Hence, they do not exploit the correlation within the data to the full extent. This leads to larger overheads in precoding transfer in the current O-RAN implementations, hence presents an opportunity to optimize the approach.

[0066] For signalling between the O-DU and O-RU. AS per ORAN specifications, the data flows to exchange information between O-DU and O-RU and are categorized as follows:

[0067] - User Plane

[0068] - Control Plane

[0069] - Synchronization Plane

[0070] In the User Plane (U Plane), the main purpose of these messages is to carry the IQ data, (both DL and UL) including user data, PRACH and control channels between O-DU and O-RU.

[0071] In the Control Plane (C Plane), the main purpose of these messages is to transmit data-associated control information required for processing of user data (e.g., scheduling and beamforming commands) if such information is not provided via M-Plane. Section Types (STs) are the main C-Plane messages that are sent from O-DU to O-RU in order to control the operations of O-RU. Section Extensions (SEs) provide an extensibility for section parameters without the need to continually redefine the section header or create new STs to accommodate future fronthaul specification needs.

[0072] In the Synchronization Plane (S Plane), the main purpose is to provide time and frequency synchronization between the O-DU and O-RU.

[0073] The present disclosure provides a method and a system for Auto-Encoder (AE) Precoder compression. Conventionally, Artificial intelligence (AI) based Auto-encoders are well known for its abilities to compress the data. Also, in 3GPP Rel-18 / 19, AEs are being studied for its application to compress the CSI data during feedback from UE to the BS.

[0074] In the present disclosure, a new AE based precoder compression method is proposed to be introduced in to O-RAN for transfer of precoding information from DU to RU. Particularly, the method allows full exploitation of the correlation within the precoding data to reduce the precoding transfer overhead requirements.

[0075] The method provides an approach where the O-DU compresses the precoding information of a slot using an AI based encoder and quantizes it. Further, the latent space quantized data is received by the O-RU, wherein a corresponding AI based decoder is selected to reconstruct the precoding information from the latent space.

[0076] The proposed method applies the AE-encoder compression at the O-DU to compress the precoding information to latent space. Once the data is received at the O-RU, using an AE-decoder it reconstructs precoders.

[0077] The present disclosure provides another alternative method for signalling the Auto-Encoder Precoder Compression. The existing O-RAN fields bfwCompHdr (bfwCompMeth & bfwIqWidth) and bfwCompParam are used to signal the usage and configuration information for AE based precoder compression.

[0078] In the alternative solution, the bfwCompMeth field may be used for AE based precoder compression.

[0079] At the O-DU, following fields are used to indicate the custom AE compression configuration:

[0080] - Using the bfwCompParam field, communicate the AE config info as selected by the O-DU for compression

[0081] - Using the contents of the bfwIqWidth field, communicate the payload size based on the pre-defined LUT shared with O-RU via M-Plane.

[0082] At the O-RU, the bfwIqWidth field may be used to deduce the size info of the compressed precoder, and bfwCompParam to identify the decoder AI model to be used for decompressing the data. In this alternative method, the AE encoder and decoder model pairs configurations are defined as per O-DU and O-RU vendor requirements, respectively.

[0083] In the alternative method, the bfwCompHdr and the bfwCompParam may be used to signal the selection and configuration information for AE based precoder compression between O-DU and O-RU.

[0084] The present disclosure provides another alternative method for Auto-Encoder Precoder Compression using Standardization in the O-RAN. In this alternative method, the same fields, using the bfwCompHdr and the bfwCompParam, the proposal could be standardized. In this method, the set of AE encoder-decoder models are predefined as per the standard, at least to the extent of defining a universal model identify, per encoder-decoder pair.

[0085] At the O-DU, bfwCompParam is used to indicate the AE model identify, thus the AE compression configuration.

[0086] - Using the field, communicate the AE model identify and configuration to the O-RU.

[0087] Using the bfwCompParam field, the O-RU may deduce the model for identifying the AE encoder at the O-DU, and respectively selects the pre-processing and AE decoder models for data reconstruction.

[0088] The present disclosure performs signalling for ORAN supported compression methods in C plane. When ORAN 7.2x split is implemented using category B O-RU, the precoding for the DL signal takes place in O-RU. In order to support this precoding feature at O-RU, the relevant precoders / beamforming weights are sent from O-DU to O-RU over the fronthaul link using C-Plane messages. Typically, ST1 is used for scheduling and beamforming information along with SE 1 or SE 11 which are used for transfer of these precoder / beamforming weights.

[0089] In a parameter called bfwCompHdr that specifies the bit width for the beamforming weights(I / Q) and the compression method used at the O-DU. The parameter bfwCompMeth may be a 4 bit field and only has certain defined compression methods in the ORAN standard while others are reserved. Thus, one of the combinations from reserved can be used to inform O-RU about the proposed AE based compression method in a proprietary implementation.

[0090] So, as an example, when the O-DU uses AE based compression method, it sets the bfwCompMeth to say ‘0110’ and bfwIqWidth specifies the size of the payload according to pre-defined LUT.

[0091] The parameter bfwCompParam may be used to indicate the AI model Identify in order for O-RU to accordingly invoke the AE decoder models for data reconstruction. When a proprietary O-RU receives this C-Plane message, it can invoke the AE based decompression method to reconstruct the beamforming weights. The proposed signaling and compression method of the present disclosure may also be introduced into the standards if agreed upon by the O-RAN Alliance.

[0092] The present disclosure teaches about detectability. The present disclosure provides a method for the O-RAN systems as the interfaces between O-DU and O-RU nodes are open.

[0093] Therefore, any entity that has the capability to read the open interface can detect the C-Plane message carrying the precoder information. Particularly, when it parses the parameters bfwCompHdr and bfwCompParam from the C-Plane message, it will be evident that the specifications for the compression technique are different than the existing definitions in the ORAN specifications.

[0094] FIG. 1 shows an exemplary architecture of auto-encoder based precoder information compression, in accordance with some embodiments of the present disclosure.

[0095] The exemplary architecture 100 comprises a Distribution Unit (O-DU) 102 of an Open Radio Access Network (O-RAN) architecture and a Radio Unit (O-RU) 104 of the O-RAN architecture communicatively connected to the O-DU 102. In an embodiment, the O-DU 102 may be a server optimized to run real-time RAN functions located below split 2 and to connect with the O-RU 104 through a Fronthaul (FH) interface based on O-RAN split 7-2x. In an embodiment, the O-RU may function as a physical interface between User Equipments (UEs) and the O-DU 102. The primary functionality of the O-RU 104 is to handle the transmission and reception of radio signals, converting digital data into analog signals for wireless transmission and vice versa. Additionally, the O-RU 104 may also manage radio resources, including power control, interference management, handover procedures, and scheduling algorithms, with the aid of the O-DU. In an embodiment, the O-RU 104 may be used to convert radio signals sent to and from an antenna into a digital baseband signal, and O-RU 104 may be connected to the DU over the O-RAN split 7-2x fronthaul interface.

[0096] In an embodiment, the O-DU 102 may be configured to compute precoder information or precoder matrix (the terms ‘precoding information’, ‘precoder information’, and ‘precoder matrix’ are interchangeably used throughout the description). Instead of directly transmitting the full precoder matrix, which can be multidimensional and resource-intensive, the O-DU 102 may compress the precoder information. In an embodiment, for compressing the precoder information the O-DU 102 selects an Artificial Intelligence (AI) model from one or more AI models corresponding to one or more Auto-Encoder (AE) based compression techniques, based on one or more parameters related to input precoder information to be compressed.

[0097] In an embodiment, the AI models corresponding to one or more AE based compression techniques, maybe a trained neural network. In some embodiments, the AI models may be self-supervised machine learning models that consist of an encoder-decoder pair, typically trained together. The encoder compresses input data into a lower-dimensional latent space, and the decoder reconstructs the original data from this compressed representation. In an embodiment, the encoder may be associated with the O-DU 102 and the decoder may be associated with the O-RU 104. The training of the AI models may involve minimizing the difference between the input and the reconstructed output using a loss function, such as Mean Squared Error (MSE), and adjusting the AE model’s weights through backpropagation to improve reconstruction accuracy. In an embodiment, the one or more AI models may include, without limitation, a Bi-LSTM based auto-encoder.

[0098] In some embodiments, the O-DU 102 may quantize the compressed input precoder information using a quantizer. The quantizer may further reduce the data size by discretizing the compressed input precoder information. In an embodiment, the quantizer may be configured to convert the compressed input precoder information to a discrete set of values that represent the quantized version of the compressed input precoder information to be transmitted through a communication channel. For example, a quantizer may convert a real number (e.g., an integer) representation of compressed input precoder information to a binary bit stream. This binary bit stream may be then transmitted through a physical uplink or downlink channel. Similarly, the O-RU 104 may use a dequantizer to convert the bit stream to reconstruct the compressed input precoder information. In some embodiments, a quantizer or dequantizer may be considered part of a Machine Learning (ML) model. For example, a generation model may include an encoder and a corresponding quantizer, and / or a reconstruction model may include a corresponding dequantizer.

[0099] In an embodiment, upon quantization, the O-DU 102 may transmit a control plane message comprising the compressed input precoder information and configuration information related to the compressed input precoder information to the O-RU 104 of the O-RAN, over the O-RAN FH.

[0100] In an embodiment, the O-RU 104 may receive the control plane message from the O-DU 102. In an embodiment, the control plane message comprises compressed input precoder information and configuration information related to the compressed input precoder information. In some embodiments, the O-RU 104 may preprocess the received control plane message before providing the control plane message to the AE decoder, associated with the O-RU 104. In an embodiment, the preprocessing may include operations like, without limitation, reshaping the input features for the AE decoder to understand In an embodiment, the O-RU 104 may further determine the AI model from one or more AI models corresponding to one or more AE based decompression techniques, based on the configuration information, for decompressing the received compressed input precoder information. Subsequently, the O-RU 104 may apply AE based decompression to decompress the compressed input precoder information using the determined AI model. In an embodiment, the O-RU 104 may calculate beamforming matrices from the decompressed precoder information. Subsequently, the O-RU 104 may apply the beamforming matrices and perform digital beamforming or analog beamforming.

[0101] FIG. 2A shows a detailed block diagram of a distribution unit (O-DU) of an Open Radio Access Network (O-RAN) architecture, for auto encoder based precoder information compression, in accordance with some embodiments of the present disclosure.

[0102] In an embodiment, the O-DU 102 for auto encoder based precoder information compression, may include a processor 202 of the O-DU 102, a memory 204 of the O-DU 102 and an I / O interface 206 of the O-DU 102. In an embodiment, the memory 204 of the O-DU 102 may be communicatively coupled to the processor 202 of the O-DU 102. The processor 202 of the O-DU 102 may be configured to perform one or more functions of the O-DU 102, using the data 208 of the O-DU 102 and the one or more modules 218 of the O-DU 102. In an embodiment, the memory 204 may store the data 208 of the O-DU 102.

[0103] In some embodiments, the data 208 stored in the memory 204 of the O-DU 102 may include, without limitation, precoder information 210 of the O-DU 102, AE data 212 of the O-DU 102, control plane data 214 of the O-DU 102 and other data 216 of the O-DU 102. In some implementations, the data 208 of the O-DU 102 may be stored within the memory 204 in the form of various data structures. Additionally, the data 208 may be organized using data models, such as relational or hierarchical data models. The other data 216 may include various temporary data and files generated by the one or more modules 218 of the O-DU 102.

[0104] In an embodiment, precoder information 210 of the O-DU 102 may include a precoder matrix (e.g., a precoder). In an embodiment, the precoder matrix may include I (In-phase) and Q (Quadrature) components, which are essential for representing the complex signals used in digital communication. In an embodiment, I and Q components may be derived by the O-DU 102 from the Channel State Information (CSI). As an example, in a 2x2 Multiple-Input Multiple-Output (MIMO) system, the precoder matrix may be:

[0105]

[0106] Here, each element (e.g., P11) may be a complex number representing the I and Q components.

[0107] In an embodiment, the AE data 212 of the O-DU 102 may include one or more parameters related to input precoder information to be compressed. In an embodiment, the one or more parameters may include, without limiting to, a target Squared Generalized Cosine Similarity (SGCS) value corresponding to the AI model. The SCGS value is a metric used to evaluate the accuracy of data reconstruction by the AI models, particularly in the context of precoder information. SGCS measures the similarity between the original precoder information compressed by the AI model and reconstructed precoder information after performing the decompression by the corresponding AI model. The similarity between the original precoder information and reconstructed precoder information may be measured by comparing their respective eigenvectors. Mathematically, SCGS value is defined as the squared cosine of the angle between the original eigenvector and the reconstructed eigenvector, normalized by their magnitudes. The formula for SGCS is given by:

[0108]

[0109] where P is the compressed precoder matrix or original eigenvector, P' is the reconstructed precoder matrix or reconstructed eigenvector, and PHdenotes the Hermitian transpose of P. In some embodiments, to evaluate the accuracy of data reconstruction by the AI models, any other metric as well can be used.

[0110] In an embodiment, the AE data 212 of the O-DU 102, may also include data related to the AI models corresponding to one or more AE based compression techniques. In an embodiment, the one or more AI models may include, without limitation, a Bidirectional Long Short-Term Memory (Bi-LSTM) based auto-encoder, Polar-DenseNet based auto-encoder, Convolutional Neural Network (CNN) based auto-encoder, and the like. In some embodiments, the one or more AI models may include any trained neural network configured to function as an auto-encoder. Although the AE-based compression technique described in this disclosure is explained using a Bi-LSTM-based autoencoder, it should not be construed as a limitation. In an embodiment, the Bi-LSTM based auto-encoder processes sequential data in both forward and backward directions, capturing dependencies from past and future data points, making it ideal for compressing time-series data. The Bi-LSTM model may compress the input precoder information into a latent representation, which is then further compressed and later reconstructed with minimal information loss.

[0111] In an embodiment, the AE data 212 of the O-DU 102, as shown in Table 1, may include the identity of the one or more AI models, the payload size achieved by the encoder, and the corresponding SCGS value of each AI model. As an example, the AI model identity “0000” may correspond to a first AI model of Bi-LSTM based auto-encoder, the payload size achieved by the Bi-LSTM based auto-encoder is 55 bits, and the corresponding SCGS value is 0.8065. Similarly, the AI model identity ‘0001’ may indicate a second AI model of Bi-LSTM based auto-encoder, with a payload size of 84 bits and an SCGS value of 0.8448. The AI model identity ‘0010’ may represent another AI model of Bi-LSTM auto-encoder with a payload size of 100 bits and an SCGS value of 0.8538.

[0112]

[0113] In an embodiment, the control plane data 214 of the O-DU 102, may include the compressed input precoder information and configuration information related to the compressed input precoder information. In an embodiment, the configuration information may include, but not limited to, size of the compressed input precoder information, indication of type of compression technique used for compressing the input precoder information, and identifier of the selected AI model when the type of compression technique is the AE based compression technique.

[0114] In an embodiment, the control plane message may be included in Section Type 11 (ST1), which may be used for transmitting the precoder information to the O-RU 104. The ST1 may be used for scheduling the beamforming information along with Section Extension 1 (SE 1) or Section Extension 11 (SE 11), which are used for the transfer of the precoder / beamforming weights.

[0115] Fig. 2B illustrates the SE 1 frame for transmitting the precoder information to the O-RU 104, in accordance with the embodiments of the present disclosure.

[0116] As shown in Fig. 2B, the SE1 frame may include the field “bfwCompHdr, this field holds the size of the compressed input precoder information, which is indicated by the “bfwIqWidth” field. The “bfwCompHdr” specifies the bit width for the beamforming weights(I / Q), according to a pre-defined Look Up Table (LUT) and the compression method used at the O-DU (as shown below in Table 2).

[0117]

[0118] In an embodiment, the type of compression technique (method) used for compressing the precoder information may be indicated in the “bfwCompMeth” field (as shown below in Table 3). As an example, a specific reserved value (e.g., '0110') may be used to signify the AE-based compression technique.

[0119]

[0120] The SE 1 frame may also include the “bfwCompParam”, this field holds the AI model identifier, which allows the O-RU 104 to choose the appropriate AE-based decoder model for data reconstruction.

[0121] In an embodiment, the SE 1 frame may also include “bfwI” and “bfwQ” fields which comprise the compressed I and Q components of the beamforming weights. The compressed precoder information, or payload, essentially consists of the beamforming weights themselves, with their real and imaginary components represented in binary form. These components of the beamforming weights are preceded by configuration parameters within the SE 1 message. In some embodiments, the User plane (U-plane) message may comprise the DL data, which is compressed I and Q components of the user data.

[0122] In an embodiment, the data 208 may be processed by the one or more modules 218 of the O-DU 102. In some implementations, the one or more modules 218 may be communicatively coupled to the processor 202 for performing one or more functions of the O-DU 102. In an implementation, the one or more modules 218 of the O-DU 102 may include, without limiting to, compression module 220, training module 222 of the O-DU 102, transceiver module 224 of the O-DU 102, and miscellaneous modules 226 of the O-DU 102.

[0123] As used herein, the term module may refer to an Application Specific Integrated Circuit (ASIC), an electronic circuit, a hardware processor (shared, dedicated, or group) and memory that execute one or more software or firmware programs, a combinational logic circuit, and / or other suitable components that provide the described functionality. In an implementation, each of the one or more modules 218 of the O-DU 102 may be configured as stand-alone hardware computing unit. In an embodiment, the miscellaneous modules 226 of the O-DU 102 may be used to perform various miscellaneous functionalities of the O-DU 102. It will be appreciated that such one or more modules 218 may be represented as a single module or a combination of different modules.

[0124] In an embodiment, the compression module 220 of the O-DU 102 may be configured for compressing the input precoder information. In an embodiment, the compression module 220 may select an AI model from one or more AI models corresponding to one or more AE based compression techniques, based on the one or more parameters related to the input precoder information to be compressed.

[0125] In an embodiment, the one or more AI models may include, without limitation, the Bi-LSTM based auto-encoder, the Polar-DenseNet based auto-encoder, and the like. In an embodiment, the selected AI model may identify the level of correlation between the elements of the input precoder information to be compressed and apply AE based compression to compress the input precoder information. For example, in some embodiments (e.g., with any of the frameworks disclosed herein), the input precoder information (KxNs) to be compressed having I and Q components (or coefficients) may be input to a pre-processor that may apply a transform (e.g., Discrete Fourier Transform / Inverse DFT (DFT / IDFT), Discrete Cosine Transform (DCT) / Inverse DCT (IDCT), and / or the like) to the input. The transformed signal may then be input to the selected AI model for compression. The selected AI model may use the correlation between different frequency bands, or Sub-Bands (SBs), in the frequency domain of the I and Q coefficients. For Single-User (SU) scheduling, where all the SBs are scheduled to the same user, there may exist a high correlation among SBs in the frequency domain, achieving a higher compression ratio compared to Multiple-User (MU) scheduling. Conversely, for MU scheduling, since the SBs may be scheduled to different users in the same slot, the correlation in the frequency domain decreases resulting in a lower compression ratio compared to SU scheduling.

[0126] In an embodiment, the training module 222 of the O-DU 102 may be configured to train the encoder-decoder pair of the auto-encoder model. In an embodiment, the encoder-decoder pair of the one or more AI models may be trained jointly.

[0127] In an embodiment, the encoder and the decoder may be trained with a training data set that includes precoding information that matches the corresponding channel quality information. In some embodiments, the training of the encoder-decoder pair may be performed simultaneously, where both the encoder and the decoder are trained together on the same training data set. In an embodiment, the training data set may be precoder information. In an embodiment, the training module 222 of the O-DU 102 may use the training data set for training the encoder placed at the O-DU 102. The training module 222 of the O-DU 102 may train the encoder to compress the precoder information into a binarized data stream in the AE latent space, using an AI model. The encoder placed at the O-DU 102, may compress the precoder information to generate a binarized data stream in the AE latent space. This compressed data stream is then sent to the O-RU 104 for training the O-RU 104 for decompressing the compressed the precoder information.

[0128] In some other embodiments, the training of the encoder-decoder pair may be performed alternately, with one model being frozen while the other is trained or using different but related training data sets for each model. In an embodiment, the training module 222 of the O-DU 102 may use a performance metric to evaluate the accuracy of the encoder design, configuration, and / or training of the encoder. In an embodiment, the performance metric to evaluate the accuracy of the encoder design may be SGCS value corresponding to the AI model. The SCGS value is a metric used to evaluate the accuracy of data reconstruction by the AI models.

[0129] In an embodiment, the transceiver module 224 of the O-DU 102 may be configured for transmitting the control plane (C-Plane) message comprising the compressed input precoder information and configuration information related to the compressed input precoder information to a O-RU 104 of the O-RAN. In an embodiment, the compressed input precoder information is transmitted over the fronthaul link using C-Plane messages. In an embodiment, the control plane message may be included in the SE 1, which may be used for transmitting the precoder information to the O-RU 104.

[0130] FIG. 3 shows a detailed block diagram of a radio unit (O-RU) of an O-RAN architecture, for auto encoder based precoder information decompression, in accordance with some embodiments of the present disclosure.

[0131] In an embodiment, the O-RU 104 for auto encoder based precoder information decompression, may include a processor 302 of the O-RU 104, a memory 304 of the O-RU 104 and an I / O interface 306 of the O-RU 104. In an embodiment, the memory 304 of the O-RU 104 may be communicatively coupled to the processor 302 of the O-RU 104. The processor 302 of the O-RU 104 may be configured to perform one or more functions of the O-RU 104, using the data 308 of the O-RU 104 and the one or more modules 316 of the O-RU 104. In an embodiment, the memory 304 may store the data 308 of the O-RU 104.

[0132] In some embodiments, the data 308 stored in the memory 304 of the O-RU 104 may include, without limitation, control plane data 310 of the O-RU 104, AE data 312 of the O-RU 104, and other data 314 of the O-RU 104. In some implementations, the data 308 of the O-RU 104 may be stored within the memory 304 of the O-RU 104, in the form of various data structures. Additionally, the data 308 may be organized using data models, such as relational or hierarchical data models. The other data 314 may include various temporary data and files generated by the one or more modules 316 the O-RU 104.

[0133] In an embodiment, control plane data 310 of the O-RU 104 may include the control plane message received from the O-DU 102 of the O-RAN architecture. In an embodiment, the control plane message may include, but not limited to, compressed input precoder information and configuration information related to the compressed input precoder information. In an embodiment, the configuration information comprises size of the compressed input precoder information. In some embodiments the configuration information may also comprises at least one of, indication of type of compression technique used for compressing the input precoder information, and identifier of the selected AI model when the type of compression technique is the AE based compression technique. In an embodiment, the O-RU 104 may extract the control plane message included in the SE 1, which may be received at the O-RU 104 from the O-DU 102. In an embodiment, fields included in the control plane message may allow the O-RU 104 to choose the appropriate AE-based decoder model for data reconstruction.

[0134] In an embodiment, the AE data 312 of the O-RU 104, may include data related to the AI models corresponding to one or more AE based compression techniques. In an embodiment, the one or more AI models may include, without limitation, a Bidirectional Long Short-Term Memory (Bi-LSTM) based auto-encoder, Polar-DenseNet based auto-encoder, and the like. In an embodiment, the AE data 312 of the O-RU 104, may be used for decompressing the received compressed input precoder information. In an embodiment, AE data 312 of the O-RU 104, may include, the identity of the one or more AI models. As an example the AI model identity “0000” may correspond to a first AI model of the Bi-LSTM based auto-encoder and the AI model identity ‘0001’ may correspond to another AI model of the Bi-LSTM based auto-encoder. In an embodiment, the O-RU 104 may determine the AI model from one or more AI models corresponding to one or more AE based on the identity of the one or more AI models received in the C-Plane message.

[0135] In an embodiment, the data 308 may be processed by the one or more modules 316 the O-RU 104. In some implementations, the one or more modules 316 the O-RU 104 may be communicatively coupled to the processor 302 for performing one or more functions of the O-RU 104. In an implementation, the one or more modules 316 of the O-RU 104 may include, without limiting to, transceiver module 318 of the O-RU 104, decompression module 320, training module 322 of the O-RU 104, and miscellaneous modules 324 of the O-RU 104.

[0136] As used herein, the term module may refer to an Application Specific Integrated Circuit (ASIC), an electronic circuit, a hardware processor (shared, dedicated, or group) and memory that execute one or more software or firmware programs, a combinational logic circuit, and / or other suitable components that provide the described functionality. In an implementation, each of the one or more modules 316 the O-RU 104 may be configured as stand-alone hardware computing unit. In an embodiment, the miscellaneous modules 324 of the O-RU 104 may be used to perform various miscellaneous functionalities of the of the O-RU 104. It will be appreciated that such one or more modules 316 of the O-RU 104 may be represented as a single module or a combination of different modules.

[0137] In an embodiment, the transceiver module 318 of the O-RU 104 may be configured for receiving the control plane message from the O-DU 102 of the O-RAN architecture. In an embodiment, the control plane message may include, but not limited to, compressed input precoder information and configuration information related to the compressed input precoder information. In an embodiment, the configuration information may include, but not limited to, size of the compressed input precoder information, indication of type of compression technique used for compressing the input precoder information, and identifier of the selected AI model when the type of compression technique is the AE based compression technique. In some embodiment, the transceiver module 318 of the O-RU 104 may receive the SE 1 frame which comprises the control plane message, from the O-DU 102.

[0138] In an embodiment, the decompression module 320 of the O-RU 104 may be configured to determine the AI model from one or more AI models corresponding to one or more AE based decompression techniques, based on the configuration information, for decompressing the received compressed input precoder information. In an embodiment, the configuration information comprises size of the compressed input precoder information. In some embodiments, without limiting to, the configuration information may also comprise indication of type of compression technique used for compressing the input precoder information, and identifier of the selected AI model when the type of compression technique is the AE based compression technique.

[0139] For example, the AI model identity may be given as “0000,” which indicates that the AI model is a Bi-LSTM-based autoencoder, that results in a payload size of 55 bits and with SCGS value of 0.80. Therefore, when the AI model used for compression of the input precoder information is a Bi-LSTM-based autoencoder, the O-RU 104 may determine that the AI model to be used for the decompression technique is also a Bi-LSTM-based autoencoder, based on the configuration information. In an embodiment, the decompression module 320 of the O-RU 104 may apply the AE based decompression to decompress the compressed input precoder information using the determined AI model. As an example, the Bi-LSTM-based auto encoder may decompress the received compressed precoder information in the latent representation. The decompression process involves reconstructing the input precoder information from the latent representation with minimal information loss, leveraging the model’s ability to capture dependencies from both past and future data points.

[0140] In an embodiment, the training module 322 of the O-RU 104 may be configured to train the encoder-decoder pair of the auto-encoder model. In an embodiment, the encoder-decoder pair of the one or more AI models may be trained jointly.

[0141] In an embodiment, the training module 322 of the O-RU 104 may train the encoder and the decoder with a training data set that includes precoding information that matches the corresponding channel quality information. In some embodiments, the training of the encoder-decoder pair may be performed simultaneously, where both the encoder and the decoder are trained together on the same training data set. In an embodiment, the training data set may be precoder information. For example, the training module 322 of the O-RU 104 may use the training data set for training the decoder placed at the O-RU 104. The training module 322 of the O-RU 104 may train the decoder to decompress binarized data stream in the AE latent space, using an AI model. The decoder placed at the O-RU 104, performs the reconstruction of the precoder information from the compressed data.

[0142] In some other embodiments, the training of the encoder-decoder pair may be performed alternately, with one model being frozen while the other is trained, or using different but related training data sets for each model. In an embodiment, the training module 322 of the O-RU 104 may use a performance metric to evaluate the accuracy of the decoder design, configuration, and / or training of the decoder.

[0143] Fig. 4A illustrates an exemplary control plane signaling flow between O-DU 102 and O-RU, in accordance with the embodiments of the present disclosure.

[0144] In an embodiment, at step (1) the O-DU 102 may transmit the Control Plane (C-Plane) message (ST1 + SE1 / 11) comprising the compressed input precoder information and configuration information related to the compressed input precoder information to a O-RU 104 of the O-RAN. In an embodiment, the configuration information comprises size of the compressed input precoder information. In some embodiments, the configuration information may also comprises at least one of, the indication of type of compression technique used for compressing the input precoder information, and the identifier of the selected AI model when the type of compression technique is the AE based compression technique. In some embodiments, the control plane message may also include information indicating the starting and ending subcarriers for which the beamforming weights are being transmitted.

[0145] In an embodiment, the control plane message may include the field “bfwCompHdr, this field holds the size of the compressed input precoder information, The “bfwCompHdr” specifies the bit width for the beamforming weights(I / Q) and the compression method used at the O-DU 102. The “bfwCompHdr” also specifies the particular AI model used for beamforming weight compression. In an embodiment, the C-Plane message also includes a “bfwCompParam” field. This parameter contains information related to the compression method used for the beamforming weights.

[0146] At step (2) the O-RU 104 may receive the C-Plane message. The O-RU 104 may read the C-Plane message and extract the relevant information, including the AI model identifier and the bfwCompParam. Further, based on the AI model identifier, the O-RU 104 may reconstruct the original beamforming weights from the compressed representation. In an embodiment, the reconstructed beamforming weights may be then applied to the downlink data to form the desired beam. Finally, the O-RU transmits the beamformed data over the air interface.

[0147] Fig. 4B shows an exemplary illustration of AE based compression of precoder information, by an O-DU of an O-RAN architecture, in accordance with the embodiments of the present disclosure.

[0148] In an embodiment, the O-DU 102, at step 402, may select an AI model from one or more AI models corresponding to one or more AE based compression techniques, based on one or more parameters related to input precoder information to be compressed. The input precoder information may be a precoder matrix P (KxNs matrix). The KxNs matrix indicates that there are NSnumber of sub bands in the precoder matrix P. Further, the O-DU 102 may apply AE based compression to compress the input precoder information using the selected AI model. At step 402, the selected AI model may process the input precoder matrix, learning the underlying structure and sub-band frequency correlation of the input precoder matrix. Subsequently, at step 406, the selected AI model may compress the input precoder matrix, resulting in final payload bits = B. The AE based compression techniques enables the encoder to represent the original input precoder information in a compressed form, significantly reducing the amount of precoder information that needs to be transmitted between the O-DU 102 and O-RU 104. This compression not only reduces overhead and improves spectral efficiency but also contributes to lower latency by minimizing the time required for precoder information transmission.

[0149] FIG. 5A shows a flowchart illustrating a method of compressing precoder information, by a distribution unit (O-DU) of an Open Radio Access Network (O-RAN) architecture, in accordance with some embodiments of the present disclosure.

[0150] As illustrated in FIG. 5A, the method 500a may include one or more blocks illustrating a method of compressing precoder information, by an O-DU of an O-RAN architecture. The method 500a may be described in the general context of computer executable instructions. Generally, computer executable instructions can include routines, programs, objects, components, data structures, procedures, modules, and functions, which perform specific functions or implement specific abstract data types.

[0151] The order in which the method 500a is described is not intended to be construed as a limitation, and any number of the described method blocks can be combined in any order to implement the method. Additionally, individual blocks may be deleted from the methods without departing from the scope of the subject matter described herein. Furthermore, the method can be implemented in any suitable hardware, software, firmware, or combination thereof.

[0152] At block 502, the method 500a includes selecting, by a processor 202 of the O-DU 102 an AI model from one or more AI models corresponding to one or more AE based compression techniques, based on one or more parameters related to input precoder information to be compressed. In an embodiment, the one or more parameters comprise a target Squared Generalized Cosine Similarity (SGCS) value corresponding to the AI model. In an embodiment, the one or more AI models are AE based models comprising an encoder-decoder pair that are trained jointly.

[0153] At block 504, the method 500a includes applying, by the processor 202 of the O-DU 102, AE based compression to compress the input precoder information using the selected AI model.

[0154] At block 506, the method 500a includes transmitting, by the processor 202 of the O-DU 102, a control plane message comprising the compressed input precoder information and configuration information related to the compressed input precoder information to a radio unit (O-RU) of the O-RAN. In an embodiment, the configuration information comprises size of the compressed input precoder information, indication of type of compression technique used for compressing the input precoder information, and identifier of the selected AI model when the type of compression technique is the AE based compression technique.

[0155] Fig. 5B shows an exemplary flowchart illustrating a method of decompressing precoder information, by an O-RU of an O-RAN architecture, in accordance with some embodiments of the present disclosure.

[0156] As illustrated in FIG. 5B, the method 500b may include one or more blocks illustrating a method of compressing precoder information, by an O-DU 102 of an O-RAN architecture. The method 500b may be described in the general context of computer executable instructions. Generally, computer executable instructions can include routines, programs, objects, components, data structures, procedures, modules, and functions, which perform specific functions or implement specific abstract data types.

[0157] The order in which the method 500b is described is not intended to be construed as a limitation, and any number of the described method blocks can be combined in any order to implement the method. Additionally, individual blocks may be deleted from the methods without departing from the scope of the subject matter described herein. Furthermore, the method can be implemented in any suitable hardware, software, firmware, or combination thereof.

[0158] At block 508, the method 500b includes receiving, by a processor 302 of the O-RU 104, a control plane message from the O-DU 102 of the O-RAN architecture. In an embodiment, the control plane message comprises compressed input precoder information and configuration information related to the compressed input precoder information. In an embodiment, the configuration information comprises size of the compressed input precoder information, indication of type of compression technique used for compressing the input precoder information, and identifier of the selected AI model when the type of compression technique is the AE based compression technique.

[0159] At block 510, the method 500b includes determining, by the processor 302 of the O-RU 104 an AI model from one or more AI models corresponding to one or more AE based decompression techniques, based on the configuration information, for decompressing the received compressed input precoder information. In an embodiment, each of the one or more AI models are AE based models comprising an encoder-decoder pair that are trained jointly.

[0160] At block 512, the method 500b includes applying, by the processor 302 of the O-RU 104, AE based decompression to decompress the compressed input precoder information using the determined AI model.

[0161] Computer System

[0162] FIG. 6 illustrates a block diagram of an exemplary computer system 600 for implementing embodiments consistent with the present disclosure. In an embodiment, the computer system 600 may be the O-DU 102 (as shown in FIG. 6). In an embodiment, when the computer system 600 is the O-DU 102, the computer system 600 may be connected to the O-RU 104, via a communication network 609. In an alternate embodiment, the computer system 600 may be the O-RU 104 (not shown in FIG. 6). In an embodiment, when the computer system 600 is the O-RU 104, the computer system 600 may be communicatively connected to a UE1to UEN(collectively referred to as one or more UEs). The computer system 600 may include a central processing unit (“CPU” or “processor” or “memory controller”) 602.

[0163] The processor 602 may comprise at least one data processor for executing program components for executing user- or system-generated host application execution processes. A user may include an administrator, a network manager, an application developer, a programmer, an organization, or any system / sub-system being operated parallelly to the computer system 600. The processor 602 may include specialized processing units such as integrated system (bus) controllers, memory controllers / memory management control units, floating point units, graphics processing units, digital signal processing units, etc.

[0164] The processor 602 may be disposed in communication with one or more Input / Output (I / O) devices (611 and 612) via I / O interface 601. The I / O interface 601 may employ communication protocols / methods such as, without limitation, audio, analog, digital, stereo, IEEE®-139 6, serial bus, Universal Serial Bus (USB), infrared, PS / 2, BNC, coaxial, component, composite, Digital Visual Interface (DVI), high-definition multimedia interface (HDMI), Radio Frequency (RF) antennas, S-Video, Video Graphics Array (VGA), IEEE® 802.n / b / g / n / x, Bluetooth, cellular (e.g., Code-Division Multiple Access (CDMA), High-Speed Packet Access (HSPA+), Global System For Mobile Communications (GSM), Long-Term Evolution (LTE) or the like), etc. Using the I / O interface 601, the computer system 600 may communicate with one or more I / O devices 611 and 612.

[0165] In some embodiments, the processor 602 may be disposed in communication with a network 609 via a network interface 603. The network interface 603 may communicate with the network 609. The network interface 603 may employ connection protocols including, without limitation, direct connect, Ethernet (e.g., twisted pair 10 / 100 / 1000 Base T), Transmission Control Protocol / Internet Protocol (TCP / IP), token ring, IEEE® 802.11a / b / g / n / x, etc.

[0166] In an implementation, the preferred network 609 may be implemented as one of the several types of networks, such as intranet or Local Area Network (LAN) and such within the organization. The preferred network 609 may either be a dedicated network or a shared network, which represents an association of several types of networks that use a variety of protocols, for example, Hypertext Transfer Protocol (HTTP), Transmission Control Protocol / Internet Protocol (TCP / IP), Wireless Application Protocol (WAP) etc., to communicate with each other. Further, the network 609 may include a variety of network devices, including routers, bridges, servers, computing devices, storage devices, etc. Using the network interface 603 and the network 609, the computer system 600 may communicate with the O-RU 104, when the computer system 600 is the O-DU 102. In another embodiment, using the network interface 603 and the network 609, the computer system 600 may communicate with the one or more UEs, when the computer system 600 is the O-RU 104.

[0167] In some embodiments, the processor 602 may be disposed in communication with a memory 605 (e.g., RAM 613, ROM 616, etc. as shown in FIG. 6) via a storage interface 604. The storage interface 604 may connect to memory 605 including, without limitation, memory drives, removable disc drives, etc., employing connection protocols such as Serial Advanced Technology Attachment (SATA), Integrated Drive Electronics (IDE), IEEE-139 6, Universal Serial Bus (USB), fiber channel, Small Computer Systems Interface (SCSI), etc. The memory drives may further include a drum, magnetic disc drive, magneto-optical drive, optical drive, Redundant Array of Independent Discs (RAID), solid-state memory devices, solid-state drives, etc.

[0168] The memory 605 may store a collection of program or database components, including, without limitation, user / application interface 606, an operating system 607, a web browser 608, and the like. In some embodiments, computer system 600 may store user / application data, such as the data, variables, records, etc. as described in this invention. Such databases may be implemented as fault-tolerant, relational, scalable, secure databases such as Oracle® or Sybase®.

[0169] The operating system 607 may facilitate resource management and operation of the computer system 600. Examples of operating systems include, without limitation, APPLE® MACINTOSH® OS X®, UNIX®, UNIX-like system distributions (E.G., BERKELEY SOFTWARE DISTRIBUTION® (BSD), FREEBSD®, NETBSD®, OPENBSD, etc.), LINUX® DISTRIBUTIONS (E.G., RED HAT®, UBUNTU®, KUBUNTU®, etc.), IBM® OS / 2®, MICROSOFT® WINDOWS® (XP®, VISTA® / 7 / 8, 10 etc.), APPLE® IOS®, GOOGLE TM ANDROID TM, BLACKBERRY® OS, or the like.

[0170] The user interface 606 may facilitate display, execution, interaction, manipulation, or operation of program components through textual or graphical facilities. For example, the user interface 606 may provide computer interaction interface elements on a display system operatively connected to the computer system 600, such as cursors, icons, check boxes, menus, scrollers, windows, widgets, and the like. Further, Graphical User Interfaces (GUIs) may be employed, including, without limitation, APPLE® MACINTOSH® operating systems’ Aqua®, IBM® OS / 2®, MICROSOFT® WINDOWS® (e.g., Aero, Metro, etc.), web interface libraries (e.g., ActiveX®, JAVA®, JAVASCRIPT®, AJAX, HTML, ADOBE® FLASH®, etc.), or the like.

[0171] The web browser 608 may be a hypertext viewing application. Secure web browsing may be provided using Secure Hypertext Transport Protocol (HTTPS), Secure Sockets Layer (SSL), Transport Layer Security (TLS), and the like. The web browsers 608 may utilize facilities such as AJAX, DHTML, ADOBE® FLASH®, JAVASCRIPT®, JAVA®, Application Programming Interfaces (APIs), and the like. Further, the computer system 600 may implement a mail server stored program component. The mail server may utilize facilities such as ASP, ACTIVEX®, ANSI® C++ / C#, MICROSOFT®, .NET, CGI SCRIPTS, JAVA®, JAVASCRIPT®, PERL®, PHP, PYTHON®, WEBOBJECTS®, etc. The mail server may utilize communication protocols such as Internet Message Access Protocol (IMAP), Messaging Application Programming Interface (MAPI), MICROSOFT® exchange, Post Office Protocol (POP), Simple Mail Transfer Protocol (SMTP), or the like. In some embodiments, the computer system 600 may implement a mail client stored program component. The mail client may be a mail viewing application, such as APPLE® MAIL, MICROSOFT® ENTOURAGE®, MICROSOFT® OUTLOOK®, MOZILLA® THUNDERBIRD®, and the like.

[0172] FIG. 7 illustrates a block diagram of a UE according to various embodiments of the present disclosure. Furthermore, the UE of FIG. 7 corresponds to the UE referred by the present disclosure.

[0173] As shown in FIG. 7, the UE according to an embodiment may include a transceiver 710, a memory 720, and a processor 730. The transceiver 710, the memory 720, and the processor 730 of the UE may operate according to a communication method of the UE described above. However, the components of the UE are not limited thereto. For example, the UE may include more or fewer components than those described above. In addition, the processor 730, the transceiver 710, and the memory 720 may be implemented as a single chip. Also, the processor 730 may include at least one processor.

[0174] The transceiver 710 collectively refers to a UE receiver and a UE transmitter, and may transmit / receive a signal to / from a base station or a network entity. The signal transmitted or received to or from the base station or a network entity may include control information and data. The transceiver 710 may include a RF transmitter for up-converting and amplifying a frequency of a transmitted signal, and a RF receiver for amplifying low-noise and down-converting a frequency of a received signal. However, this is only an example of the transceiver 710 and components of the transceiver 710 are not limited to the RF transmitter and the RF receiver.

[0175] Also, the transceiver 710 may receive and output, to the processor 730, a signal through a wireless channel, and transmit a signal output from the processor 730 through the wireless channel.

[0176] The memory 720 may store a program and data required for operations of the UE. Also, the memory 720 may store control information or data included in a signal obtained by the UE. The memory 720 may be a storage medium, such as read-only memory (ROM), random access memory (RAM), a hard disk, a CD-ROM, and a DVD, or a combination of storage media.

[0177] The processor 730 may control a series of processes such that the UE operates as described above. For example, the transceiver 710 may receive a data signal including a control signal transmitted by the base station or the network entity, and the processor 730 may determine a result of receiving the control signal and the data signal transmitted by the base station or the network entity.

[0178] FIG. 8 illustrates a block diagram of a base station or a network entity according to various embodiments of the present disclosure. Furthermore, the base station or the network entity of FIG. 8 corresponds to the gNB or network entity referred by the present disclosure.

[0179] As shown in FIG. 8, the base station(or the network entity receiver) according to an embodiment may include a transceiver 810, a memory 820, and a processor 830. The transceiver 810, the memory 820, and the processor 830 of the base station(or the network entity receiver) may operate according to a communication method of the base station(or the network entity receiver) described above. However, the components of the base station(or the network entity receiver) are not limited thereto. For example, the base station may include more or fewer components than those described above. In addition, the processor 830, the transceiver 810, and the memory 820 may be implemented as a single chip. Also, the processor 830 may include at least one processor.

[0180] The transceiver 810 collectively refers to the base station(or the network entity receiver) and a base station(or the network entity) transmitter, and may transmit / receive a signal to / from a terminal or a network entity or a base station. The signal transmitted or received to or from the terminal or a network entity or the base station may include control information and data. The transceiver 810 may include a RF transmitter for up-converting and amplifying a frequency of a transmitted signal, and a RF receiver for amplifying low-noise and down-converting a frequency of a received signal. However, this is only an example of the transceiver 810 and components of the transceiver 810 are not limited to the RF transmitter and the RF receiver.

[0181] Also, the transceiver 810 may receive and output, to the processor 830, a signal through a wireless channel, and transmit a signal output from the processor 830 through the wireless channel.

[0182] The memory 820 may store a program and data required for operations of the base station(or the network entity receiver). Also, the memory 820 may store control information or data included in a signal obtained by the base station. The memory 820 may be a storage medium, such as read-only memory (ROM), random access memory (RAM), a hard disk, a CD-ROM, and a DVD, or a combination of storage media.

[0183] The processor 830 may control a series of processes such that the base station(or the network entity receiver) operates as described above. For example, the transceiver 810 may receive a data signal including a control signal transmitted by the terminal or the network entity or the base station, and the processor 830 may determine a result of receiving the control signal and the data signal transmitted by the terminal or the network entity or the base station.

[0184] Furthermore, one or more computer-readable storage media may be utilized in implementing embodiments consistent with the present invention. A computer-readable storage medium refers to any type of physical memory on which information or data readable by a processor may be stored. Thus, a computer-readable storage medium may store instructions for execution by one or more processors, including instructions for causing the processor(s) to perform steps or stages consistent with the embodiments described herein. The term “computer-readable medium” should be understood to include tangible items and exclude carrier waves and transient signals, i.e., non-transitory. Examples include Random Access Memory (RAM), Read-Only Memory (ROM), volatile memory, nonvolatile memory, hard drives, Compact Disc (CD) ROMs, Digital Video Disc (DVDs), flash drives, disks, and any other known physical storage media.

[0185] Disclosed herein is a method of compressing precoder information, by a distribution unit (O-DU) of an Open Radio Access Network (O-RAN) architecture. The method includes selecting an Artificial Intelligence (AI) model from one or more AI models corresponding to one or more Auto-Encoder (AE) based compression techniques, based on one or more parameters related to input precoder information to be compressed. Further, the method includes applying AE based compression to compress the input precoder information using the selected AI model. Subsequently, the method includes transmitting a control plane message comprising the compressed input precoder information and configuration information related to the compressed input precoder information to a radio unit (O-RU) of the O-RAN.

[0186] Further, the present disclosure discloses a method of decompressing precoder information, by a radio unit (O-RU) of an Open Radio Access Network (O-RAN) architecture. The method includes receiving a control plane message from a distribution unit (O-DU) of the O-RAN architecture. The control plane message comprises compressed input precoder information and configuration information related to the compressed input precoder information. Further, the method includes determining an Artificial Intelligence (AI) model from one or more AI models corresponding to one or more Auto-Encoder (AE) based decompression techniques, based on the configuration information, for decompressing the received compressed input precoder information. Subsequently, the method includes applying AE based decompression to decompress the compressed input precoder information using the determined AI model.

[0187] Further, the present disclosure discloses a distribution unit (O-DU) of an Open Radio Access Network (O-RAN) architecture for compressing precoder information. The O-DU comprises a processor and a memory. The memory is communicatively coupled to the processor and stores instructions, which on execution, cause the processor to select an Artificial Intelligence (AI) model from one or more AI models corresponding to one or more Auto-Encoder (AE) based compression techniques, based on one or more parameters related to input precoder information to be compressed. The processor further applies AE based compression to compress the input precoder information using the selected AI model. Subsequently, the processor transmits a control plane message comprising the compressed input precoder information and configuration information related to the compressed input precoder information to a radio unit (O-RU) of the O-RAN.

[0188] Furthermore, the present disclosure discloses a radio unit (O-RU) of an Open Radio Access Network (O-RAN) architecture for decompressing precoder information. The O-RU comprises a processor and a memory. The memory is communicatively coupled to the controller and stores instructions, which on execution, causes the processor to receive a control plane message from a distribution unit (O-DU) of the O-RAN architecture. The control plane message comprises compressed input precoder information and configuration information related to the compressed input precoder information. The processor further determines an Artificial Intelligence (AI) model from one or more AI models corresponding to one or more Auto-Encoder (AE) based decompression techniques, based on the configuration information, for decompressing the received compressed input precoder information. Subsequently, the processor applies AE based decompression to decompress the compressed input precoder information using the determined AI model.

[0189] Furthermore, the present disclosure discloses a system comprising a distribution unit (O-DU) of an Open Radio Access Network (O-RAN) architecture configured to perform the method of compressing precoder information, and a radio unit (O-RU) of the O-RAN architecture communicatively connected to the O-DU, configured to perform the method of decompressing precoder information.

[0190] The above embodiments are illustrative only and are not intended to be in any way limiting. In addition to the illustrative aspects, embodiments, and features described above, further aspects, embodiments, and features will become apparent by reference to the drawings and the following detailed description.

[0191] Advantages of the embodiments of the present disclosure are illustrated herein

[0192] The present disclosure provides a method, apparatus and system for auto encoder based precoder information compression.

[0193] The present disclosure discloses Auto-Encoder (AE) based compression and decompression approach for transferring the precoder information from O-DU to O-RU. The present disclosure fully exploits the possible correlation within the precoder data, thereby decreasing the bandwidth requirement for transferring the precoder information over the fronthaul. The proposed AI-based AEs compress data by exploiting the Frequency Domain (FD) correlation. The auto encoder based compression approach achieves higher precoder reconstruction accuracy at the O-RU compared to the existing compression techniques, thereby significantly reducing the fronthaul overhead. The present disclosure not only mitigates the fronthaul overhead but also results in better precoder reconstruction accuracy at the O-RU, leading to improved downlink performance.

[0194] The proposed auto encoder based compression approach reduces overhead, improves spectral efficiency and contributes to lower latency by minimizing the time required for precoder information transmission.

[0195] Additionally, the present disclosure proposes using necessary configuration parameters at the O-DU to communicate the specifics of the compression configuration to the O-RU. Moreover, the proposed control signaling and compression method is useful for Time Division Duplex (TDD) systems for transferring precoder information estimated using Sounding Reference Signal (SRS) receptions.

[0196] The autoencoder-based compression approach of the present disclosure provides flexibility in terms of scaling the number of antennas in beyond 5G and 6G systems without having to replace the existing fiber link infrastructure between O-RU and O-DU. The increase in the number of antennas allows a higher number of users and data rates to be supported with the existing infrastructure. The present disclosure also improves the reliability of communication between users and the O-RAN base station by achieving higher accuracy at lower payload sizes required for sharing the precoder information from O-DU to O-RU.

[0197] In light of the technical advancements provided by the disclosed method, the claimed steps, as discussed above, are not routine, conventional, or well-known aspects in the art, as the claimed steps provide the aforesaid solutions to the technical problems existing in the conventional technologies. Further, the claimed steps clearly bring an improvement in the functioning of the system itself, as the claimed steps provide a technical solution to a technical problem.

[0198] The terms "an embodiment", "embodiment", "embodiments", "the embodiment", "the embodiments", "one or more embodiments", "some embodiments", and "one embodiment" mean "one or more (but not all) embodiments of the invention(s)" unless expressly specified otherwise.

[0199] The terms "including", "comprising", “having” and variations thereof mean "including but not limited to", unless expressly specified otherwise.

[0200] The enumerated listing of items does not imply that any or all the items are mutually exclusive, unless expressly specified otherwise. The terms "a", "an" and "the" mean "one or more", unless expressly specified otherwise.

[0201] A description of an embodiment with several components in communication with each other does not imply that all such components are required. On the contrary, a variety of optional components are described to illustrate the wide variety of possible embodiments of the invention.

[0202] When a single device or article is described herein, it will be clear that more than one device / article (whether they cooperate) may be used in place of a single device / article. Similarly, where more than one device / article is described herein (whether they cooperate), it will be clear that a single device / article may be used in place of the more than one device / article or a different number of devices / articles may be used instead of the shown number of devices or programs. The functionality and / or features of a device may be alternatively embodied by one or more other devices which are not explicitly described as having such functionality / features. Thus, other embodiments of invention need not include the device itself.

[0203] Finally, the language used in the specification has been principally selected for readability and instructional purposes, and it may not have been selected to delineate or circumscribe the inventive subject matter. It is therefore intended that the scope of the invention be limited not by this detailed description, but rather by any claims that issue on an application based here on. Accordingly, the embodiments of the present invention are intended to be illustrative, but not limiting, of the scope of the invention, which is set forth in the following claims.

[0204] While various aspects and embodiments have been disclosed herein, other aspects and embodiments will be apparent to those skilled in the art. The various aspects and embodiments disclosed herein are for purposes of illustration and are not intended to be limiting, with the true scope and spirit being indicated by the following claims.

[0205] Referral Numerals:

[0206]

[0207]

Claims

1.A method of compressing precoder information, performed by a distribution unit (O-DU) of an Open Radio Access Network (O-RAN) architecture in a wireless communication system, the method comprising:selecting an Artificial Intelligence (AI) model from one or more AI models corresponding to one or more Auto-Encoder (AE) based compression techniques, based on one or more parameters related to input precoder information to be compressed;applying AE based compression to compress the input precoder information using the selected AI model; andtransmitting a control plane message comprising the compressed input precoder information and configuration information related to the compressed input precoder information to a radio unit (O-RU) of the O-RAN.2.The method of claim 1, wherein the configuration information comprises size of the compressed input precoder information, indication of type of compression technique used for compressing the input precoder information, and identifier of the selected AI model when the type of compression technique is the AE based compression technique.3.The method of claim 1, wherein the one or more parameters comprise a target Squared Generalized Cosine Similarity (SGCS) value corresponding to the AI model.4.The method of claim 3, wherein SCGS value is a metric used to evaluate the accuracy of data reconstruction by the AI model.5.The method of claim 1, wherein each of the one or more AI models are AE based models comprising an encoder-decoder pair that are trained jointly.6.A method of decompressing precoder information, performed by a radio unit (O-RU) of an Open Radio Access Network (O-RAN) architecture in a wireless communication system, the method comprising:receiving a control plane message from a distribution unit (O-DU) of the O-RAN architecture, wherein the control plane message comprises compressed input precoder information and configuration information related to the compressed input precoder information;determining an Artificial Intelligence (AI) model from one or more AI models corresponding to one or more Auto-Encoder (AE) based decompression techniques, based on the configuration information, for decompressing the received compressed input precoder information; andapplying AE based decompression to decompress the compressed input precoder information using the determined AI model.7.The method of claim 6, wherein the configuration information comprises size of the compressed input precoder information, indication of type of compression technique used for compressing the input precoder information, and identifier of the selected AI model when the type of compression technique is the AE based compression technique.8.The method of claim 6, wherein each of the one or more AI models are AE based models comprising an encoder-decoder pair that are trained jointly.9.A distribution unit (O-DU) of an Open Radio Access Network (O-RAN) architecture for compressing precoder information in a wireless communication system, the O-DU comprising:a processor; anda memory communicatively coupled to the processor, wherein the memory stores the processor-executable instructions, which, on execution, causes the processor to:select an Artificial Intelligence (AI) model from one or more AI models corresponding to one or more Auto-Encoder (AE) based compression techniques, based on one or more parameters related to input precoder information to be compressed;apply AE based compression to compress the input precoder information using the selected AI model; andtransmit a control plane message comprising the compressed input precoder information and configuration information related to the compressed input precoder information to a radio unit (O-RU) of the O-RAN.10.The O-DU of claim 9, wherein the configuration information comprises size of the compressed input precoder information, indication of type of compression technique used for compressing the input precoder information, and identifier of the selected AI model when the type of compression technique is the AE based compression technique.11.The O-DU of claim 9, wherein the one or more parameters comprise a target Squared Generalized Cosine Similarity (SGCS) value corresponding to the AI model.12.The O-DU of claim 9, wherein each of the one or more AI models are AE based models comprising an encoder-decoder pair that are trained jointly.13.A radio unit (O-RU) of an Open Radio Access Network (O-RAN) architecture for decompressing precoder information in a wireless communication system, the O-RU comprising:a processor; anda memory communicatively coupled to the processor, wherein the memory stores the processor-executable instructions, which, on execution, causes the processor to:receive a control plane message from a distribution unit (O-DU) of the O-RAN architecture, wherein the control plane message comprises compressed input precoder information and configuration information related to the compressed input precoder information;determine an Artificial Intelligence (AI) model from one or more AI models corresponding to one or more Auto-Encoder (AE) based decompression techniques, based on the configuration information, for decompressing the received compressed input precoder information; andapply AE based decompression to decompress the compressed input precoder information using the determined AI model.14.The O-RU of claim 13, wherein the configuration information comprises size of the compressed input precoder information, indication of type of compression technique used for compressing the input precoder information, and identifier of the selected AI model when the type of compression technique is the AE based compression technique.15.The O-RU of claim 13, wherein each of the one or more AI models are AE based models comprising an encoder-decoder pair that are trained jointly.

Images