CN113316159B

CN113316159B - 5G network system based on heterogeneous physical layer

Info

Publication number: CN113316159B
Application number: CN202110587585.1A
Authority: CN
Inventors: 温向明; 朱子珅; 章晨宇; 王鲁晗; 郑伟; 蒋秋萍
Original assignee: Beijing University of Posts and Telecommunications
Current assignee: Beijing University of Posts and Telecommunications
Priority date: 2021-05-27
Filing date: 2021-05-27
Publication date: 2022-09-02
Anticipated expiration: 2041-05-27
Also published as: CN113316159A

Abstract

The present disclosure provides a 5G network system based on heterogeneous physical layer, including: the system comprises a heterogeneous physical layer, a media access control layer, a radio link control layer, a packet data convergence protocol layer and a radio resource control layer, wherein the heterogeneous physical layer comprises: the system comprises a functional sublayer, an adaptation sublayer and a heterogeneous sublayer, wherein the functional sublayer comprises at least one functional module, and the functional module is used for generating a corresponding processing task; the adaptation sublayer is used for performing task allocation and interface adaptation on the heterogeneous sublayer based on the processing task; the heterogeneous sublayer comprises at least one computing module, and the computing module is used for performing corresponding computing processing based on the tasks allocated by the adaptation sublayer. According to the method and the device, the physical layer throughput and the computing rate of the 5G network can be improved, and meanwhile, the data processing advantages of other layers in the 5G network are reserved, so that the performance of a 5G network system is improved.

Description

5G network system based on heterogeneous physical layer

Technical Field

The disclosure relates to the field of communication technologies, and in particular, to a 5G network system.

Background

The physical layer (PHY) in the 5G network architecture is the lowest layer of the wireless access system, and provides services to the upper layer by using the transmission channel as an interface, so the computational performance of the physical layer will have an important influence on the performance of the 5G network. With the increasing traffic, the increasing throughput demands and the increasing speed demands put a tremendous strain on the 5G network, especially the physical layer. This may result in existing physical layer performance that cannot meet such high throughput and computation requirements. In the prior art, some algorithms with high throughput in a wireless network physical layer are optimized, and a hardware coding circuit is designed to improve the coding rate of the physical layer so as to improve the processing speed of the physical layer, but the designed hardware circuit is not universal enough and cannot be flexibly configured according to specific scenes; the physical layer mechanism and the physical layer hardware are used for accelerating the data packet to improve the processing speed of the physical layer, but the physical layer acceleration is applied to the 5G wireless communication network, and the physical layer acceleration does not focus on the signal processing and the signal generation of the physical layer, so that the performance of an air interface cannot be effectively improved.

Disclosure of Invention

In view of the above, the present disclosure is directed to a 5G network system based on heterogeneous physical layers.

In view of the above, according to a first aspect of the present disclosure, there is provided a heterogeneous physical layer based 5G network system, including: the system comprises a heterogeneous physical layer, a media access control layer, a radio link control layer, a packet data convergence protocol layer and a radio resource control layer, wherein the heterogeneous physical layer comprises: a functional sublayer, an adaptation sublayer, and a heterogeneous sublayer, wherein,

the functional sub-layer comprises at least one functional module, and the functional module is used for generating a corresponding processing task;

the adaptation sublayer is used for performing task allocation and interface adaptation on the heterogeneous sublayer based on the processing task;

the heterogeneous sublayer comprises at least one computing module, and the computing module is used for performing corresponding computing processing on the basis of the tasks distributed by the adaptation sublayer.

Optionally, the at least one computing module comprises a CPU and at least one of a GPU, an FPGA, or a DSP.

Optionally, the heterogeneous sublayer comprises a first computation module and a second computation module, wherein,

the first computing module generates data to be processed based on the tasks distributed by the adaptation sublayer and sends the data to be processed to the second computing module;

the second computing module performs corresponding processing on the data to be processed to obtain a processing result, and sends the processing result to the first computing module;

and the first calculation module transmits the processing result to the media access control layer.

Optionally, the first computing module is a CPU, and the second computing module is a GPU, where the first computing module stores the data to be processed in a memory, and sends the data to be processed in the memory to the second computing module;

and the second calculation module stores the data to be processed in a video memory and correspondingly processes the data to be processed in the video memory to obtain the processing result.

Optionally, the at least one functional module comprises: at least one of a mapping module, an interlayer mapping module, a modulation module, a channel estimation module, a cyclic redundancy check module, a precoding module, an LDPC coding/decoding module, a hybrid automatic repeat request module, a rate matching module, or a data segmentation module.

Optionally, the computing modules communicate with each other through a PCIe bus.

Optionally, the heterogeneous physical layer communicates with the medium access control layer through an nFAPI interface.

Optionally, the adapting the sub-layer for interface adaptation includes: for implementing interface adaptation between the functional modules or between the computing modules.

Optionally, the adaptation sublayer is further configured to implement data format adaptation between the functional modules.

Optionally, the system further includes an antenna unit configured to communicate with a user terminal, and the heterogeneous physical layer communicates with the antenna unit through an O-RAN interface.

As can be seen from the foregoing, the 5G network system based on the heterogeneous physical layer provided by the present disclosure accelerates the processing and transmission of data through the high-performance computing processing capability of the heterogeneous sublayers in the heterogeneous physical layer, can improve the physical layer throughput and the computing rate of the 5G network, and simultaneously retains the data processing advantages of other layers in the 5G network, thereby improving the performance of the 5G network system.

Drawings

In order to more clearly illustrate the technical solutions in the present disclosure or related technologies, the drawings needed to be used in the description of the embodiments or related technologies are briefly introduced below, and it is obvious that the drawings in the following description are only embodiments of the present disclosure, and for those skilled in the art, other drawings can be obtained according to these drawings without creative efforts.

Fig. 1 is a schematic block diagram of a heterogeneous physical layer based 5G network system according to an embodiment of the present disclosure. (ii) a

Fig. 2 is a schematic block diagram of a heterogeneous physical layer in accordance with an embodiment of the present disclosure;

FIG. 3 is a schematic diagram of a heterogeneous physical layer according to an embodiment of the present disclosure;

fig. 4 is a diagram illustrating decoding speed comparison of heterogeneous physical layers with physical layers in a conventional manner according to the disclosed embodiments.

Detailed Description

For the purpose of promoting a better understanding of the objects, aspects and advantages of the present disclosure, reference is made to the following detailed description taken in conjunction with the accompanying drawings.

It is to be noted that technical terms or scientific terms used in the embodiments of the present disclosure should have a general meaning as understood by those having ordinary skill in the art to which the present disclosure belongs, unless otherwise defined. The use of "first," "second," and similar terms in the embodiments of the disclosure is not intended to indicate any order, quantity, or importance, but rather to distinguish one element from another. The word "comprising" or "comprises", and the like, means that the element or item listed before the word covers the element or item listed after the word and its equivalents, but does not exclude other elements or items. The terms "connected" and the like are not limited to physical or mechanical connections, but may include electrical connections, and when the absolute position of the object to be described changes, the relative position may also change accordingly.

The 5G network is a long-term network constructed by the evolution and development of a complementary new technology of the existing wireless technology, provides limit experience to meet different requirements of users by integrating various wireless access technologies, and is a general term of solutions after various novel wireless access technologies and the existing wireless access technologies are integrated. The 5G network provides services with multiple scenes such as ultrahigh access rate, zero-delay use experience, billions of equipment connection capacity, ultrahigh flow density, ultrahigh connection number density, ultrahigh mobility and the like for users. This determines that 5G networks are facing increasing throughput and computational challenges. The radio interface of the 5G network includes a physical layer (L1), a data link layer (L2), and an application layer (L3), wherein the physical layer is the lowest layer of the radio access system, which interfaces with transport channels, and has a main task of providing data transport services for higher layers. However, the increasing throughput and computation challenges necessarily place a great deal of stress on the physical layer, which may result in the inability of existing physical layer performance to meet such large throughput and computation requirements. In the prior art, although the data processing speed of the physical layer is improved by optimizing an algorithm or a hardware circuit in the physical layer, the performance of an air interface cannot be effectively improved, and flexible configuration cannot be performed according to a specific scene.

Based on the above consideration, the embodiment of the present disclosure provides a 5G network system based on a heterogeneous physical layer. Referring to fig. 1, fig. 1 illustrates a schematic block diagram of a heterogeneous physical layer based 5G network system according to an embodiment of the present disclosure. As shown in fig. 1, the 5G network system based on heterogeneous physical layers includes: a heterogeneous physical layer 110, a medium access control layer 120, a radio link control layer 130, a packet data convergence protocol layer 140, and a radio resource control layer 150. Wherein the medium access control layer 120, the radio link control layer 130, the packet data convergence protocol layer 140, and the radio resource control layer 150 may operate in a CPU.

In some embodiments, referring to fig. 1, the heterogeneous physical layer based 5G network system may further include: an antenna unit 160 for communicating with the user terminal UE. Further, in some embodiments, the heterogeneous physical layer communicates with the antenna units over an O-RAN interface.

In some embodiments, the heterogeneous physical layer communicates with the medium access control layer over an nFAPI interface. The nFAPI interface may be a logical interface that represents a data path between the heterogeneous physical layer and the media access control layer, among other things.

Specifically, the nFAPI interface may be composed of two entities, and when the heterogeneous physical layer data in the GPU is processed, the data is passed through the entity in the GPU, then transmitted to the entity in the CPU, and then handed to the media access control layer to implement the data path.

Specifically, referring to fig. 1, a user terminal UE transmits user data to an antenna unit (RU)160, the antenna unit (RU)160 processes the user data into first data, and forwards the first data to a Heterogeneous Physical Layer (H-PHY) 110 through an Open-Radio Access Network (Open-Radio Access Network) interface, the Heterogeneous Physical Layer 110 processes the first data and then sequentially transmits the first data to an upper Layer, second Data output by the Radio Resource Control layer 150 is transmitted back to the Core Network through a Medium Access Control (MAC) layer 120 in a Data Link layer (i.e., L2 layer), a Radio Link Control (RLC) layer 130, a Packet Data Convergence Protocol (PDCP) layer 140, and a Radio Resource Control (RRC) layer 150 in an application layer (i.e., L3). Similarly, the Core Network may transmit the third data back to the radio resource control layer 150, the radio resource control layer 150 sequentially processes the third data through the packet data convergence protocol layer 140, the radio link control layer 130, the medium access control layer 120, and the heterogeneous physical layer 110 outputs fourth data and forwards the fourth data to the antenna unit 160 through the O-RAN interface. The antenna unit 160 processes the fourth data and transmits the processed fourth data to the user equipment UE.

Therefore, according to the 5G network system based on the heterogeneous physical layer, the heterogeneous accelerated heterogeneous physical layer is introduced to accelerate the data processing speed of the existing physical layer, and the physical layer throughput and the calculation rate of the 5G network can be improved. Meanwhile, the protocol sublayer in the L2 layer is prevented from being changed, and the advantages of the CPU on the L2 layer and the flow processing are reserved.

Referring to fig. 2, fig. 2 shows a schematic block diagram of a heterogeneous physical layer according to an embodiment of the present disclosure. As shown in fig. 2, the heterogeneous physical layer 110 includes: a functional sublayer 111, an adaptation sublayer 112, and a heterogeneous sublayer 113, wherein,

the functional sub-layer 111 includes at least one functional module, and the functional module is configured to generate a corresponding processing task;

the adaptation sublayer 112 is configured to perform task allocation and interface adaptation on the heterogeneous sublayer 113 based on the processing task;

the heterogeneous sublayer 113 includes at least one computing module, and the computing module is configured to perform corresponding computing processing based on the task allocated by the adaptation sublayer 112.

The functional sublayer 111 in the heterogeneous physical layer 110 allocates the data processing task to the heterogeneous sublayer 113 having the computing module through the adaptation sublayer 112 to perform computation, and accelerates processing and transmission of data by using the high-performance computing processing capability of the heterogeneous sublayer 113, so that the physical layer throughput and the computing rate of the 5G network can be improved, and meanwhile, the data processing advantages of other layers in the 5G network are retained, thereby improving the performance of the 5G network system. The method is suitable for being widely applied to 5G networks.

Optionally, referring to fig. 2, at least one functional module in the functional sub-layer 111 may include: at least one of a Mapping (MAP) module, an inter-Layer mapping (Layer-MAP) module, a Modulation (MOD) module, a channel estimation (Chan-est) module, a Cyclic Redundancy Check (CRC) module, a precoding (Pre-code) module, an LDPC coding/decoding module, a Hybrid Automatic Repeat Request (HARQ) module, a Rate matching (Rate-Mat) module, or a data segmentation (Seg) module.

The functional modules are independent of each other and are regarded as different tasks to be processed in the heterogeneous physical layer.

In some embodiments, the Modulation (MOD) module may comprise a QAM modulation module.

Optionally, the adaptation sublayer 112 may also be used to implement data format adaptation between the functional modules.

The adaptation sublayer 112, as an intermediate layer, needs to support the functional sublayer and the heterogeneous sublayer, and its functions include a task allocation mechanism for adaptively allocating the heterogeneous sublayer according to the characteristics of different functional modules, a task management mechanism for recording and managing mappings between different functional modules and different computing modules in the heterogeneous sublayer to coordinate serial processing flows between the computing modules, interface adaptations between different functional modules and different computing modules, and data format adaptations between the functional modules.

In some embodiments, the adapting the sub-layer performs task allocation on the heterogeneous sub-layer based on the processing task, which may include: and allocating the tasks of the functional sub-layer to the corresponding computing modules in sequence based on the task priority.

Further, in some embodiments, the adapting sublayer performs task allocation on the heterogeneous sublayer based on the processing task, and may further include:

and when the total computing power of the tasks distributed by the corresponding computing modules exceeds the preset computing power threshold value of the corresponding computing modules, distributing the tasks to other computing modules.

Specifically, assume that the functional sub-layer in the heterogeneous physical layer has N processing tasks, each of which is denoted as j _i ，j _i E.j, (i ═ 1, 2.. N). Wherein J represents a task set, i.e.

In the representation of each task, q _i Indicating the kind of computing modules in the heterogeneous sub-layer to which the task is adapted,

and

respectively representing the maximum and minimum computing power required for the task, f _i For indicating the priority of the task.

The heterogeneous sublayer is provided with M computing modules, which are marked as p _j ，p _j E.p, (i ═ 1, 2.... M). Where P represents a collection of computing modules,

wherein q is _j A flag indicating the number of the calculation modules is indicated,

is the total computing power of the computing module. Then it is determined that,

expressed as the total power of the heterogeneous sublayers. Using c in the adaptation sublayer _i Representing the computational power ultimately assigned by task i. Then, the adaptation sublayer may pair f _i The tasks are sequenced to obtain a priority order, and the tasks with high priority are distributed to the q corresponding to the tasks on the premise of meeting the preset requirement _i In (1). Wherein the preset requirements may include all tasks having

(

i

1, 2.... N), and for all orientations the same p _j All have

If a certain p _j If the preset requirement is not met due to excessive tasks being distributed, the q of the task with low priority is determined _i Randomly changing to another different Q ∈ Q, and so on, until Q is not needed _i And (6) changing. Optimizing according to the preset requirements

The computational power of the optimal task assignment and the matched equipment can be obtained.

Further, each functional module of the functional sub-layer in the heterogeneous physical layer substantially exhibits a serial processing relationship, and therefore the adaptation sub-layer also needs to record and manage mapping between different functional modules and computing modules of the heterogeneous sub-layer to coordinate efficient operation of the serial processing flow. For each task, the management module of the adaptation sublayer may record information of the task using the task ID, the task timing, and the task window W. The task ID can be used for uniquely identifying the current task for the management module to identify and operate. Each task is assigned a task timing for identifying the timing relationship of the current task with other tasks. When the adjacent time sequence tasks are executed serially, the computing modules of the heterogeneous sublayers where the tasks are located can directly carry out efficient interaction to avoid the time delay of data transmission. The task timing is dynamically adjusted in the management module of the adaptation sublayer to accommodate changes in computational requirements. The window W is used for recording the unresponsive time of the task, and when the window is overtime, the management module can sleep the task to release resources, so that the high-efficiency utilization of the resources is achieved.

Furthermore, there may be a problem of non-uniform interfaces between different functional modules or between computing modules of heterogeneous sublayers, so that the adaptation sublayer is required to perform uniform adaptation on differentiated module interfaces, thereby implementing an interface adaptation function of the adaptation sublayer.

Furthermore, the computation modules of different heterogeneous sublayers need to adopt different data processing formats to achieve the best performance, for example, a DSP is more suitable for performing fixed point number computation of Q1.15 class; NVIDIA GPU requires the use of the floating point number format of HALF _ 2. Therefore, the adaptation sublayer needs to adapt the data format between different computing modules.

Therefore, the adaptation sublayer allocates the functional sublayer, and the heterogeneous physical layers are all on the computation modules such as the GPU and the DSP in the heterogeneous sublayer, so that the adaptation sublayer allocates resources to the computation modules. The distribution program prestored in the CPU by the adaptation sublayer can carry out dynamic distribution according to the condition of the heterogeneous sublayer and the functional sublayer needing to be operated at the moment, and distribute the corresponding functional module to the corresponding hardware; when a functional module is established, a mark is established for the module, and the mark prompts an interval of the size of the resource needing to be allocated by an allocation program and an optimally matched platform. The method can also be used for carrying out interface adaptation and data structure adaptation based on two programs and used for converting data interfaces, for example, single-precision floating point numbers are needed for processing complex floating point numbers in a CPU, half2 floating point numbers are needed in a GPU, and therefore conversion adaptation of data formats is needed. The hardware result transmission between the computing modules is to transmit data through PCIe interfaces between different hardware.

Optionally, at least one computing module in the heterogeneous sublayer 113 includes a CPU and at least one of a GPU, an FPGA, or a DSP.

The CPU, the GPU, the FPGA or the DSP respectively and specifically have respective computing processing advantages, so that when the task of the functional module is completed, the adaptation sublayer can adapt the proper computing module according to the characteristics of the task of the functional module to be completed, and the acceleration of the data processing speed of the physical layer is realized by fully utilizing the computing processing advantages of the computing module.

In some embodiments, the computing modules communicate therebetween over a PCIe bus. The computing modules can communicate with each other through an actual connection channel, namely a physical interface, namely a PCIe interface.

and the second calculation module carries out corresponding processing on the data to be processed to obtain a processing result.

Further, in some embodiments, the first computing module is a CPU, and the second computing module is a GPU, where the first computing module stores the data to be processed in a memory, and sends the data to be processed in the memory to the second computing module;

In some embodiments, the second calculation module further receives the data to be processed based on a tensor container and performs format definition on the data to be processed.

Specifically, referring to fig. 3, fig. 3 illustrates a schematic diagram of a heterogeneous physical layer according to an embodiment of the present disclosure. As shown in fig. 3, the heterogeneous sublayer includes a first computation module CPU and a second computation module GPU, and the GPUs have strong parallel processing capability, so that tasks capable of reducing computation time through parallel processing can be allocated to the GPUs, such as codec tasks and modulation tasks. Taking LDPC encoding/decoding task as an example, the heterogeneous physical layer receives Data from the antenna unit as Input Data, and performs LDPC encoding/decoding on the Data, the LDPC encoding/decoding module generates the LDPC encoding/decoding task, the adaptation sublayer performs task allocation on the CPU and the GPU according to the LDPC encoding/decoding task, and the CPU stores the Input Data to be encoded/decoded and the encoding/decoding control pattern in the memory according to the task allocated by the adaptation sublayer. Because the CPU and the GPU in the heterogeneous sublayer communicate through the PCIe bus, the CPU transmits Input Data to be encoded/decoded and an encoding/decoding control pattern in the memory to the GPU through the PCIe bus, and the GPU may convert the received Data into Data in a GPU format based on a cudammcpy Function, generate tensor description and a workspace (workspace) for storing the converted Data in the GPU format, and deliver the Data in the GPU format to a high-speed LDPC encoding/decoding module of a Kernel Function in the GPU, such as a cuda encoding Function (cudaencodedunc)/cuda decoding Function (cudadecodedunc), for parallel processing to obtain corresponding encoded/decoded result Data. At this time, the coded/decoded data is in a GPU format, and the GPU may also convert the coded/decoded result data into result data in a CPU format through a cudaMemcpy function. Then, the CPU transmits the result data to the mac layer 120, the rlc layer 130, the pdcp layer 140, and the rrc layer 150 of the upper layer to perform corresponding data processing.

The data to be coded/decoded is sent to the GPU from the CPU through the PCIe channel at one time, so that the transmission time overhead of the data is reduced. The GPU may also receive and format the data to be encoded/decoded using tensor containers to define the format and actual meaning of the data stream. The tensor operation is suitable for the matrix operation, and the tensor container may be an implementation method of the tensor operation, such as a C + + class. Therefore, due to the high-capacity video memory of the GPU, the construction of the high-capacity workspace shortens the repeated data reading and writing time. Referring to fig. 4, fig. 4 illustrates a diagram comparing decoding speeds of heterogeneous physical layers and physical layers in a conventional manner according to the disclosed embodiments. As shown in fig. 4, the average time consumed for decoding 80 code blocks in the GPU is 27us in the actual test process, while it takes about 620us to decode 8 code blocks in the CPU. Comparing the performance of the LDPC decoding module in the GPU is far greater than that of the CPU of x86 architecture.

Therefore, compared with the conventional method in which the physical layer is calculated only by using the CPU, the heterogeneous physical layer according to the embodiment of the present disclosure may improve the matrix operation efficiency required for LDPC encoding/decoding and the floating point number calculation capability required for decoding using a tensor (tensor) calculation manner in the GPU and an acceleration method of parallel stream processing.

In summary, the 5G network system based on the heterogeneous physical layer provided by the present disclosure accelerates the processing and transmission of data through the high-performance computation processing capability of the heterogeneous sublayer in the heterogeneous physical layer, can improve the physical layer throughput and the computation rate of the 5G network, and simultaneously retains the data processing advantages of other layers in the 5G network, thereby improving the performance of the 5G network system.

Those of ordinary skill in the art will understand that: the discussion of any embodiment above is meant to be exemplary only, and is not intended to intimate that the scope of the disclosure, including the claims, is limited to these examples; within the idea of the present disclosure, also technical features in the above embodiments or in different embodiments may be combined, steps may be implemented in any order, and there are many other variations of the different aspects of the embodiments of the present disclosure as described above, which are not provided in detail for the sake of brevity.

In addition, well-known power/ground connections to Integrated Circuit (IC) chips and other components may or may not be shown in the provided figures for simplicity of illustration and discussion, and so as not to obscure the embodiments of the disclosure. Furthermore, devices may be shown in block diagram form in order to avoid obscuring embodiments of the present disclosure, and this also takes into account the fact that specifics with respect to implementation of such block diagram devices are highly dependent upon the platform within which the embodiments of the present disclosure are to be implemented (i.e., specifics should be well within purview of one skilled in the art). Where specific details (e.g., circuits) are set forth in order to describe example embodiments of the disclosure, it should be apparent to one skilled in the art that the embodiments of the disclosure can be practiced without, or with variation of, these specific details. Accordingly, the description is to be regarded as illustrative instead of restrictive.

While the present disclosure has been described in conjunction with specific embodiments thereof, many alternatives, modifications, and variations of these embodiments will be apparent to those of ordinary skill in the art in light of the foregoing description. For example, other memory architectures (e.g., dynamic ram (dram)) may use the discussed embodiments.

The disclosed embodiments are intended to embrace all such alternatives, modifications and variances which fall within the broad scope of the appended claims. Therefore, any omissions, modifications, equivalents, improvements, and the like that may be made within the spirit and principles of the embodiments of the disclosure are intended to be included within the scope of the disclosure.

Claims

1. A heterogeneous physical layer based 5G network system, comprising: the system comprises a heterogeneous physical layer, a media access control layer, a radio link control layer, a packet data convergence protocol layer and a radio resource control layer, wherein the heterogeneous physical layer comprises: a functional sublayer, an adaptation sublayer, and a heterogeneous sublayer, wherein,

the heterogeneous sublayer comprises at least one computing module, and the computing module is used for performing corresponding computing processing on the basis of the tasks distributed by the adaptation sublayer;

for each task, the adaptation sublayer records the information of the task by using a task ID, a task time sequence and a task window; the task ID is used for uniquely identifying the current task, the task time sequence is used for identifying the time sequence relation between the current task and other tasks, and when the adjacent time sequence tasks are executed in series, the computing modules of the heterogeneous sub-layers where the tasks are located directly interact; the task window is used for recording the unresponsive time of the current task, and the adaptation sublayer sleeps the current task to release resources when the window is overtime;

the adaptation sub-layer sequentially distributes the tasks of the functional sub-layer to the corresponding computing modules based on the task priority; and when the total computing power of the tasks distributed by the corresponding computing modules exceeds the preset computing power threshold value of the corresponding computing modules, distributing the tasks to other computing modules.

2. The system of claim 1, wherein the at least one computing module comprises a CPU and at least one of a GPU, FPGA, or DSP.

3. The system of claim 1, wherein the heterogeneous sublayer comprises a first computing module and a second computing module, wherein,

the second computing module correspondingly processes the data to be processed to obtain a processing result and sends the processing result to the first computing module;

4. The system of claim 3, wherein the first computing module is a CPU and the second computing module is a GPU, wherein the first computing module stores the data to be processed in a memory and sends the data to be processed in the memory to the second computing module;

5. The system of claim 1, wherein the at least one functional module comprises: at least one of a mapping module, an interlayer mapping module, a modulation module, a channel estimation module, a cyclic redundancy check module, a precoding module, an LDPC coding/decoding module, a hybrid automatic repeat request module, a rate matching module, or a data segmentation module.

6. The system of claim 1, wherein the computing modules communicate therebetween over a PCIe bus.

7. The system of claim 1, wherein the heterogeneous physical layer and the media access control layer communicate over an nAPI interface.

8. The system of claim 1, wherein the adaptation sublayer to interface adaptation comprises: for implementing interface adaptation between the functional modules or between the computing modules.

9. The system of claim 1, wherein the adaptation sublayer is further to enable data format adaptation between the computing modules.

10. The system of claim 1, wherein the system further comprises an antenna unit for communicating with a user terminal, the heterogeneous physical layer communicating with the antenna unit over an O-RAN interface.