A 6G-oriented intelligent, efficient, and fully decoupled network architecture

Abstract

A 6G-oriented intelligent, simple, efficient, and fully decoupled network architecture is based on base stations, users, and edge clouds, wherein the data base stations further include downlink base stations and uplink base stations; the edge cloud is connected to each base station with an optical fiber; the transmission delay of the control channel is lower than that of the data channel; the data channel is distributed and collaboratively sent, and transmission, processing, and storage are received in an integrated manner; the edge cloud is the entity for centralized processing of data in the entire collaborative area; the fully decoupled network is a network architecture in which multiple base stations collaborate in a local area to provide wireless access services to users; the control plane and data plane of the fully decoupled network are physically separated; the uplink and downlink of the data plane are also physically separated; and service transmission on the data plane is achieved through multi-point collaboration; the present invention can fully decouple the network, effectively promote the collaboration of multi-dimensional resources, improve the spectrum efficiency of the entire network, provide users with personalized communication services, and reduce networking costs.

Description

Technical Field

[0001] The present invention belongs to the field of 6G wireless mobile communications, and specifically relates to a 6G-oriented intelligent, simple, efficient and fully decoupled network architecture. Background Art

[0002] Future wireless mobile communication networks will feature integrated sky-ground heterogeneous networking, access to trillions of terminals, and collection and distribution of massive sensor data. The network structure will be more complex, the service types will be more diverse, and the application scenarios will be more varied.

[0003] Mobile communications are constantly evolving, but each generation of network technology requires the allocation of new spectrum resources. As spectrum allocations become increasingly narrow, high-quality spectrum resources for wireless communications are practically exhausted. Furthermore, the fragmented spectrum usage and rigid duplex technology of current mobile communication networks significantly constrain the development of broadband mobile communications. As network deployments move toward higher frequency bands, high-frequency carriers further limit coverage areas, leading to ultra-high-density network deployments. The increased cost per site associated with these new technologies has dramatically increased wireless network deployment costs. Furthermore, the excessive power load per base station significantly increases operators' operations and maintenance costs. It is estimated that to achieve the same coverage goals, the base station deployment density of 5G networks will be approximately three to four times that of 4G, and the overall network energy consumption will increase by almost an order of magnitude. Traditional networks that couple control and service, and uplink and downlink, enable rapid and economical deployment when network density and scale are low. However, with the depletion of low-frequency resources, the explosive growth in the number of terminals, and the dramatic increase in network energy consumption, existing network architectures will face significant challenges in terms of resource efficiency, scalability, and cost-effectiveness.

[0004] To address these challenges, researchers have proposed various technologies, such as distributed antenna systems (DAS), cloud radio access networks (C-RAN), and cell-free networks. In DAS, high-power single-antenna base stations are replaced by a group of low-power antenna elements distributed over the same area, providing better coverage while reducing transmit power and improving reliability. R. Heath et al. describe the application of DAS for downlink transmission in cellular networks in the paper "A current perspective on distributed antenna systems for the downlink of cellular systems." In C-RAN, a series of remote radio heads (RRHs) perform radio functions such as up / down conversion, analog-to-digital (A / D), and digital-to-analog (D / A) conversion. Furthermore, the baseband signals from the RRHs are processed by a centralized baseband unit (BBU) pool located in the edge cloud. A. Checko et al. detail the basic features of the C-RAN architecture and related research progress in the paper "Cloud RAN for mobile networks—a technology overview." In a cell-free network architecture, a group of distributed access points (APs) collaborate to use the same frequency-time resources to simultaneously serve all active users within the network coverage area. HQNgo, et al. verified in the paper "Cell-free massive MIMO versus small cells" that the cell-free architecture brings a huge improvement in frequency efficiency in wireless communications.

[0005] However, in the technologies described above, DAS still uses a cellular network architecture, simply distributing existing cellular antennas across different locations. C-RAN primarily utilizes a unified BBU pool, and the cell-free architecture still couples uplink and downlink, as well as control and services. This coupling of control and services, as well as uplink and downlink, significantly limits network resource flexibility and hinders widespread collaboration between base stations.

[0006] The research significance of this invention lies in: proposing a 6G-oriented intelligent, simple, efficient and fully decoupled network architecture, which realizes flexible resource collaboration through network architecture decoupling, and achieves efficient utilization of network resources, personalized user services and low-cost networking. Summary of the Invention

[0007] The purpose of the present invention is to propose an access network architecture for wireless mobile communications, which fully decouples the network architecture, protocols and resources to achieve efficient utilization of network resources, personalized user services and low-cost networking.

[0008] The present invention is implemented as follows: a 6G-oriented intelligent, simple, efficient, and fully decoupled network architecture includes base stations, users, and edge clouds, wherein the data base stations further include downlink base stations and uplink base stations; the edge cloud is connected to each base station with an optical fiber; the transmission delay of the control channel is much lower than the transmission delay of the data channel; the data channel is distributed and collaboratively sent, and the transmission, processing, and storage are integrated and received; the edge cloud is the entity that centrally processes data in the entire collaborative area; the fully decoupled network is a network architecture in which multiple base stations collaborate in a local area to provide wireless access services to users; the control plane and data plane of the fully decoupled network are physically separated; the uplink and downlink of the data plane are also physically separated; and the service transmission of the data plane is achieved through multi-point collaboration;

[0009] The physical separation of the data plane and the control plane means that the functions of the data plane and the control plane on the wireless side are carried by the data base station and the control base station respectively;

[0010] The control plane latency is much smaller than the data plane latency. The physical separation of the data and control planes means that the data plane functions and the control plane functions on the wireless side are carried by the data base station and the control base station respectively.

[0011] The control base station operates in a low frequency band and has a coverage area much larger than that of the data base station, providing control plane support for service assurance for all users in the entire collaborative area.

[0012] The data base station is divided into an uplink base station and a downlink base station; the uplink base station constitutes an uplink network, carrying the user's uplink business; the downlink base station constitutes a downlink network, carrying the downlink business; the connection between the control base station and the user is bidirectional, implemented in a frequency division duplex manner; the connection between the data base station and the user is unidirectional, and the unidirectional business is implemented through a simplex channel between the user and the base station;

[0013] The multi-point collaboration on the data plane connects the edge cloud and the base station through a high-speed, low-latency link, and the base stations are quasi-synchronized. The user's uplink service is realized through distributed reception of multiple uplink base stations in the uplink network, and the user's downlink service is realized through collaborative transmission by multiple downlink base stations in the downlink network.

[0014] The multi-point collaboration on the data plane connects the edge cloud and the base station through a high-speed, low-latency link, and the base stations are quasi-synchronized. The user's uplink service is realized through distributed reception of multiple uplink base stations in the uplink network, and the user's downlink service is realized through collaborative transmission by multiple downlink base stations in the downlink network.

[0015] The above-mentioned intelligent, simple, efficient and fully decoupled network architecture for 6G, wherein: the service transmission includes the following steps:

[0016] Step 1, state perception stage: The state perception is mainly performed in the non-data transmission stage, which is characterized by: the controller at the edge cloud periodically sends instructions to the data base station and the control base station to detect the status of the data channel; the user will transmit the status information to the control base station through the control channel, and the control base station and the data base station will transmit their collected status to the edge cloud through optical fiber; the edge cloud will obtain the status information of the entire network within the control area.

[0017] Step 2, demand perception stage: The demand perception controls the base station to maintain a quasi-real-time connection with the user; when the user has a data transmission demand, the demand information will be transmitted to the control base station through the control channel, and the control base station will transmit the demand information to the edge cloud controller through a fast wired link; through the demand reporting of all users in the control domain, the edge cloud will obtain the demand information of users in the control domain.

[0018] Step 3, decision-making stage; the control function entity on the edge cloud side will allocate resources in the resource pool to users based on the overall network status and user needs in the domain through the resource allocation algorithm in the algorithm library, and generate corresponding resource allocation decisions; the resource allocation algorithm here includes a series of methods to achieve resource allocation, including traditional proportional fairness, optimization, and machine learning.

[0019] Step 4, decision distribution phase: The edge cloud distributes resource allocation decision information to the control base station and data base station through high-speed wired links. The control base station sends user-related decision information to the user through wireless channels.

[0020] Step 5, decision execution phase: The user and base station interpret and execute the decision information, the control base station and the user exchange control plane information in real time, and the data base station and the user transmit uplink and downlink service information through a simplex data channel.

[0021] As described above, in a 6G-oriented intelligent, simple, efficient, and fully decoupled network architecture, in the uplink service transmission, the user interprets the received decision information and transmits data with multiple uplink base stations through the allocated resources and corresponding resource usage methods; multiple base stations on the uplink network side receive data based on the resource allocation decision and transmit the received information to the central data processing unit on the data plane of the edge cloud side; the central data processing unit merges the data of multiple base stations to realize uplink service transmission.

[0022] As described above, a 6G-oriented intelligent, simple, efficient, and fully decoupled network architecture, the downlink service transmission is characterized by: the central data processing unit inside the edge cloud distributes the data to be transmitted to multiple allocated downlink base stations based on decision information; multiple downlink base stations send information to users in a collaborative manner on the allocated wireless resources based on the decision information; users merge the information from multiple base stations based on the decision information to realize downlink service transmission.

[0023] To sum up, the advantages and positive effects of the present invention are as follows: compared with the existing technology, the present invention will fully decouple the network architecture, protocols, and resources, effectively promote the collaboration of multi-dimensional resources, thereby improving the spectrum efficiency of the entire network, and realizing efficient utilization of network resources, personalized user services, and low-cost networking. BRIEF DESCRIPTION OF THE DRAWINGS

[0024] Figure 1 It is a logical block diagram of downlink information transmission for a 6G-oriented intelligent, simple, efficient, and fully decoupled network architecture provided by an embodiment of the present invention.

[0025] Figure 2 This is a business information transmission flow chart of a 6G-oriented intelligent, simple, efficient, and fully decoupled network architecture provided by an embodiment of the present invention.

[0026] Figure 3 This is a schematic diagram of downlink multi-point collaborative transmission of a 6G-oriented intelligent, simple, efficient, and fully decoupled network architecture provided by an embodiment of the present invention.

[0027] Figure 4 This is a schematic diagram of the 6G fully decoupled network architecture of the present invention. DETAILED DESCRIPTION

[0028] To further clarify the objectives, technical solutions, and advantages of the present invention, the present invention is further described in detail below with reference to the following embodiments. It should be understood that the specific examples described herein are intended only to illustrate the present invention and are not intended to limit the present invention. The present invention proposes a method for calculating uplink transmit power based on proportional planning, using channel statistical information uploaded by base stations. This method can efficiently calculate user transmit power and improve the frequency and energy efficiency of wireless mobile communication networks.

[0029] The application principle of the present invention is described in detail below with reference to the accompanying drawings.

[0030] First, the logical framework of the 6G-oriented intelligent, efficient, and fully decoupled network architecture is as follows: Figure 1As shown, the control plane and data plane of the fully decoupled network are physically separated; the uplink and downlink of the data plane are also physically separated; the physical separation of the data plane and the control plane means that the functions of the data plane and the control plane on the wireless side are carried by the data base station and the control base station respectively.

[0031] The control channel is duplex, with the data base stations divided into separate uplink and downlink base stations. The data channel is simplex, with the uplink base stations forming the uplink network, carrying uplink services for users, and the downlink base stations forming the downlink network, carrying downlink services. The control plane latency is significantly lower than the data plane latency; data plane service transmission is achieved through multi-point collaboration. The connection between the control base station and the user is bidirectional, implemented via frequency division duplexing; the connection between the data base station and the user is unidirectional, with unidirectional services carried over the simplex channel between the user and the base station.

[0032] The multi-point collaboration on the data plane connects the edge cloud and base stations via high-speed, low-latency links. Base stations are quasi-synchronized, achieved through the high-precision Beidou / GPS system. User uplink services are distributed across multiple uplink base stations in the uplink network, while user downlink services are collaboratively transmitted across multiple downlink base stations in the downlink network. Collaborative transmission includes single-point and multi-point transmission. Single-point transmission refers to data being sent to users via a single downlink base station, including but not limited to collaborative link adaptation and dynamic selection of transmitting nodes. Multi-point transmission refers to information being sent to users via multiple base stations, including but not limited to cross-layer and cross-base station collaboration at the physical layer, such as beamforming, power allocation, and spectrum resource allocation. The control base station operates in a low-frequency band and has a coverage area significantly larger than that of the data base station. It provides control plane support for service assurance for all users within the collaborative area. The connection between the control base station and users is bidirectional, implemented through frequency division duplexing. The connection between the data base station and users is unidirectional, with unidirectional services carried out via simplex channels between users and base stations.

[0033] Further, refer to Figure 1 and Figure 2 Taking user A watching an online video and user B downloading a file as an example, the downlink transmission process includes the following steps:

[0034] S101: Perceive network situation.

[0035] S102: Sense the transmission needs of users A, B, and all other users.

[0036] S103: The edge cloud generates a network decision.

[0037] S104: Network decision sent to data base station and all users

[0038] S105: User A, user B, and the data base station execute network decision.

[0039] Furthermore, step S101 is the state perception stage: the state perception is mainly performed in the non-data transmission stage, and is characterized in that: the controller at the edge cloud periodically sends control signaling to the downlink base station and the control base station, and the control base station further sends the control signaling to the user through the control channel; the user performs periodic or event-triggered detection on the state of the data channel according to the control signaling; next, the user will transmit the channel state information to the control base station through the control channel; the control base station will transmit the collected wide-area coverage channel state information to the edge cloud through optical fiber; the edge cloud will obtain the state information of the entire network within the control area.

[0040] Furthermore, step S102 is the demand perception stage: the demand perception controls the base station to maintain a quasi-real-time connection with the user; when user A has an online video transmission demand and user B has a file download demand, the demand information will be transmitted to the control base station through the network control channel; the control base station will collect the demand information of the entire network, and according to the latency level and service level of all users, transmit the online video demand of user A, the file download demand of user B, and the demand information of all other users in real time to the edge cloud controller through a fast wired link; through the demand reporting of all users in the control domain, the edge cloud will obtain the demand information of users in the control domain.

[0041] Furthermore, step S103 is the decision-making stage; the control function entity on the edge cloud side will allocate resources in the resource pool to users based on the overall network status and user needs in the domain, targeting the different business needs of user A and user B, that is, user A requires real-time medium-bandwidth QoS protection, while user B requires large-bandwidth non-real-time QoS protection, and generate corresponding resource allocation decisions through the resource allocation algorithm in the algorithm library. In this embodiment, we use the currently popular deep reinforcement learning method:

[0042] Q(S t ,A t ;θ)←Q(S t ,A t ;θ)+α[R t+1 +γmax a′ Q(S t+1 ,a′;θ)-Q(S t ,A t ;θ)]

[0043] Among them, A t represents the resource allocation decision of the network at time t, S t Indicates the network status at that moment, R t+1 Indicates decision At The return value when Q(S t ,A t ) represents the cumulative reward of the network at this moment, 0<γ<1 represents the attenuation coefficient of the reward value Q, α represents the step size of the Q value update, and θ represents the network parameters. Through the continuous learning of the network θ, continuous decision-making of the optimal Q value is achieved.

[0044] The resource allocation algorithms here include but are not limited to proportional fairness, average, maximum throughput optimization, maximum minimum throughput optimization, deep learning, and enhanced learning resource allocation algorithms. Figure 3 As shown in Figure 1, the implementation of these algorithms is cross-layer and cross-base station execution, mainly scheduling multiple base stations at the same time, allocating resources such as bandwidth, power, and time to complete user control signaling and arranging base station execution to achieve efficient and reliable information delivery.

[0045] Furthermore, step S104 is the decision distribution stage: the edge cloud sends the decision information of resource allocation to the control base station and the downlink base station respectively through a high-speed optical fiber link; the control base station sends the decision information related to user A and user B to the user through different control channel resources through the wireless channel, that is, it schedules multiple base stations at the same time, allocates bandwidth, power, time and other resources, and the base station executes specific control signaling; the wireless channel here is an independent spectrum resource different from the data channel, which is used for wireless signaling interaction between users and control base stations in a fully decoupled network. It is not specifically stated that the communication mode of the base station in the present invention can adopt the existing 4G and 5G communication methods;

[0046] Furthermore, step S105 is the decision execution stage: the downlink base station receives the decision information from the edge cloud, and then schedules its own resources to send information to user A and user B through the wireless side; after user A and user B receive the decision information from the control base station, they receive the information sent by the downlink base station on the corresponding wireless resources. Here, due to the different QoS requirements of user A and user B, the network will schedule time-continuous medium-bandwidth resources for user A for data transmission; for user B, the network will call on large-bandwidth resources to transmit information to the user, but these large-bandwidth resources can be continuous (low network load), discontinuous (medium network load), or even temporarily interrupted (high network load). When receiving downlink information, the user will interact with the base station for real-time control plane information; when sending information, the downlink base station focuses on transmitting downlink data information, and is decoupled from the functions of the control base station and the uplink base station.

[0047] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions and improvements made within the spirit and principles of the present invention should be included in the scope of protection of the present invention.

Claims

1. A 6G-oriented intelligent, simple, efficient, and fully decoupled network architecture, characterized by: It includes base stations, users, and edge clouds. Data base stations include downlink base stations and uplink base stations. The edge cloud and base stations are connected by optical fiber. The transmission delay of the control channel is lower than that of the data channel. The data channel is distributed and collaboratively sent, with integrated transmission, processing, and storage. The edge cloud is the entity that centrally processes data within the entire collaborative area. The fully decoupled network is a network architecture in which multiple base stations collaborate within a local area to provide wireless access services to users. The control plane and data plane of the fully decoupled network are physically separated, as are the uplink and downlink of the data plane. Service transmission on the data plane is achieved through multi-point collaboration. The physical separation of the data plane and the control plane means that the functions of the data plane and the control plane on the wireless side are carried by the data base station and the control base station respectively; The control plane latency of a fully decoupled network is much smaller than the data plane latency. The physical separation of the data and control planes means that on the wireless side, the data plane functions and the control plane functions are carried by the data base station and the control base station respectively. The control base station operates in a low frequency band and has a coverage area much larger than that of the data base station, providing control plane support for service assurance for all users in the entire collaborative area. The data base station is divided into an uplink base station and a downlink base station; the uplink base station constitutes an uplink network and carries the user's uplink business; Downlink base stations form the downlink network and carry downlink services. The connection between the control base station and the user is bidirectional and is implemented through frequency division duplexing. The connection between the data base station and the user is unidirectional, and the unidirectional service is implemented through the simplex channel between the user and the base station. In the multi-point collaboration on the data plane, the edge cloud and the base station are connected by a high-speed, low-latency link; the base stations are quasi-synchronized, which is achieved through the high-precision Beidou / GPS system; the user's uplink business is realized through distributed reception by multiple uplink base stations in the uplink network, and the user's downlink business is realized through collaborative transmission by multiple downlink base stations in the downlink network.

2. According to the 6G-oriented intelligent, simple, efficient and fully decoupled network architecture of claim 1, the collaborative transmission is characterized by: Collaborative transmission includes single-point transmission and multi-point transmission; single-point transmission means that the data to be sent is sent to the user through a single downlink base station, including collaborative link adaptation and dynamic selection of transmitting nodes; multi-point transmission means that information is sent to the user through multiple base stations, including physical layer beamforming, power allocation, spectrum resource allocation cross-layer and cross-base station collaboration.

3. According to the 6G-oriented intelligent, simple, efficient and fully decoupled network architecture of claim 1, the service transmission is characterized by: The steps include: Step 1, state perception phase: This phase is mainly performed during the non-data transmission phase. The controller at the edge cloud periodically sends instructions to the data base station and the control base station to detect the state of the data channel. The user transmits the state information to the control base station via the control channel. The control base station and the data base station transmit the collected state information to the edge cloud via optical fiber. The edge cloud will obtain the status information of the entire network within the control area; Step 2, demand perception phase: This phase involves controlling the base station and maintaining a near-real-time connection with the user. When a user has a data transmission demand, the control base station transmits the demand information to the control base station via a control channel. The control base station then transmits the demand information to the edge cloud controller via a fast wired link. Through demand reporting by all users in the control domain, the edge cloud obtains demand information from users in the control domain. Step 3: Decision generation. The edge cloud control entity allocates resources from the resource pool to users based on the overall network status and user needs within the domain, using the resource allocation algorithms in the algorithm library. It also generates corresponding resource allocation decisions. These resource allocation algorithms include traditional proportional fairness, optimization, and machine learning approaches. Step 4, decision distribution phase: The edge cloud distributes resource allocation decision information to the control base station and data base station via high-speed wired links. The control base station sends user-related decision information to the user via wireless channels. Step 5, decision execution phase: The user and base station interpret and execute the decision information, the control base station and the user exchange control plane information in real time, and the data base station and the user transmit uplink and downlink service information through a simplex data channel.

4. The 6G-oriented intelligent, simplified, efficient, and fully decoupled network architecture according to claim 1 is characterized by: It includes uplink business transmission. The user interprets the received decision information and transmits data with multiple uplink base stations through the allocated resources and corresponding resource usage methods. Multiple base stations on the uplink network side receive data based on the resource allocation decision and transmit the received information to the central data processing unit on the data plane of the edge cloud side. The central data processing unit merges the data of multiple base stations to realize uplink business transmission.

5. The 6G-oriented intelligent, simplified, efficient, and fully decoupled network architecture according to claim 3, including downlink service transmission, is characterized by: The central data processing unit within the edge cloud distributes the data to be transmitted to multiple allocated downlink base stations based on the decision information; the multiple downlink base stations send information to users in a collaborative manner on the allocated wireless resources based on the decision information; The user combines information from multiple base stations based on the decision information to achieve downlink service transmission.

6. The 6G-oriented intelligent, simple, efficient, and fully decoupled network architecture according to claim 3 is characterized by: The state perception phase is performed during the non-data transmission phase: the controller at the edge cloud periodically sends control signaling to the downlink base station and the control base station, and the control base station further sends the control signaling to the user via the control channel; the user performs periodic or event-triggered detection of the data channel status based on the control signaling; then, the user transmits the channel status information to the control base station via the control channel; the control base station transmits the collected wide-area coverage channel status information to the edge cloud via optical fiber; The edge cloud will obtain the status information of the entire network within the control area; During the demand perception phase, the control base station maintains a near-real-time connection with users. When user A needs to transmit online videos or user B needs to download files, they will transmit their demand information to the control base station via the network's control channel. The control base station will then collect demand information from the entire network and, based on the latency and service levels of all users, transmit user A's online video needs, user B's file download needs, and all other user demand information to the edge cloud controller via a fast wired link in real time. Through demand reporting by all users in the control domain, the edge cloud will obtain demand information from users in the control domain. The control function entity on the edge cloud side will allocate resources from the resource pool to users based on the overall network status and user needs within the domain, targeting the different business requirements of users A and B. For example, user A requires real-time medium-bandwidth QoS protection, while user B requires high-bandwidth non-real-time QoS protection. This algorithm will generate corresponding resource allocation decisions, including the use of deep reinforcement learning methods. Resource allocation decisions simultaneously schedule multiple base stations, allocating bandwidth, power, and time resources to complete user control signaling and arranging base station execution to achieve efficient and reliable information delivery; During the decision-making and distribution phase, the edge cloud sends resource allocation decision information to the control base station and downlink base station via high-speed optical fiber links. The control base station then sends the decision information related to user A and user B to the users via different control channel resources via wireless channels. This means that multiple base stations are simultaneously scheduled to allocate bandwidth, power, and time resources, and the base stations execute specific control signaling. The wireless channel here is an independent spectrum resource distinct from the data channel, and is used for wireless signaling interaction between users and the control base station in a fully decoupled network. Decision execution phase: The downlink base station receives the decision information from the edge cloud, and then schedules its own resources to send information to users A and B through the wireless side. After users A and B receive the decision information from the control base station, they receive the information sent by the downlink base station on the corresponding wireless resources. Due to the different QoS requirements of users A and B, the network will schedule time-continuous medium-bandwidth resources for user A for data transmission. For user B, the network will call on large-bandwidth resources to transmit information to the user. Large-bandwidth resources can be continuous or discontinuous, or even temporarily interrupted. When receiving downlink information, the user will interact with the base station for real-time control plane information. When sending information, the downlink base station focuses on transmitting downlink data information, and is decoupled from the functions of the control base station and the uplink base station.

Images