An Uplink and Downlink Joint Dynamic Resource Allocation Method for 6G Fully Decoupled Networks

Abstract

A method for uplink and downlink joint dynamic resource allocation for a 6G fully decoupled network. 1) The uplink base station collects channel estimations of the entire network, including pilot signals sent by users and channel estimations made by users based on pilot signals of the downlink base station, and transmits the channel information of the entire network back to the edge center for dynamic resource allocation. 2) According to the channel information between users and uplink / downlink base stations and different sub-channels, assuming that the base station and users adopt the method of equal power sharing, the three-dimensional matching problem of user-base station-sub-channel is modeled as a large-scale many-to-many matching model. 3) According to the matching relationship of user-base station-sub-channel, based on the theory of successive convex approximation, for the uplink coherent transmission of multiple users and the downlink coherent transmission of multiple base stations respectively, an uplink and downlink power control algorithm is adopted. 4) Steps 2 and 3 are both calculated in parallel to obtain the uplink and downlink spectrum division ratio that maximizes the system capacity, and the dynamic resource allocation scheme is sent to the uplink users and downlink base stations for implementation.

Description

Technical Field

[0001] The present invention belongs to the field of 6G full decoupled network resource allocation, and relates to an uplink and downlink joint dynamic resource allocation method for a 6G full decoupled network. Background Art

[0002] 5G still faces three fundamental challenges: scarce spectrum resources, increasing demand for high-quality network services, and soaring network operation costs. In the 6G full decoupled radio access network (FD-RAN) titled "A Fully-Decoupled RAN Architecture for 6G Inspired by Neurotransmission" published by Professor Yu Quan et al. in the Journal of Communications and Information Networks in 2019, the base station is physically decoupled into a control base station (UBS), an uplink base station (UBS), and a downlink base station (DBS). Through the complete physical decoupling of the base station, the network will become extremely flexible, facilitating better network and resource cooperation in order to achieve better spectral efficiency and energy efficiency. At the same time, how to efficiently utilize the spectrum resources released by the new FD-RAN architecture has become a new challenge.

[0003] After searching the existing literature, it is found that few literatures propose joint dynamic scheduling of power, base stations, and sub-channels. Mamoun Guenach et al. published "Joint Power Control and Access Point Scheduling in Fronthaul-Constrained Uplink Cell-Free Massive MIMO Systems" in the IEEE Transactions on Communications, which proposed a joint power control and access point scheduling algorithm for an uplink transmission system considering fronthaul capacity constraints. However, this article does not consider the multi-sub-channel allocation problem. Therefore, for the multi-sub-channel scenario, there will also be a power allocation problem between users and base stations on multiple sub-channels. The 6G full decoupled network spectrum allocation method proposed in this paper, compared with FDD, adaptively allocates the uplink and downlink frequency bands by the edge cloud, dynamically adjusts the allocation of the uplink and downlink frequency bands in the face of the rapid change of user service requirements, and can greatly improve the flexibility of spectrum utilization.

[0004] In summary, the problems existing in the prior art are as follows: (1) The traditional TDD and FDD spectrum allocation methods cannot fully unleash the potential of spectrum resource utilization in 6G fully decoupled networks. (2) Different from the user-base station single connection mode and sub-channel allocation method based on round-robin scheduling in existing mobile communication networks, the multi-user-multi-base station-multi-sub-channel allocation in 6G fully decoupled networks is a super-large-scale non-convex integer programming problem, and a low-complexity 0-1 integer programming algorithm needs to be designed. (3) The power allocation between users and base stations in uplink and downlink transmissions is a non-convex optimization problem, with the weighted rate of users as the objective function, and a low-complexity power control algorithm needs to be redesigned. The significance of solving the above technical problems lies in: For the service requirements and scientific problems of 5G / B5G wireless networks, the proposed solution helps the country to layout 6G network technology reserves and provides theoretical guidance and reference for 6G architecture and algorithm design. Summary of the Invention

[0005] Aiming at the problems existing in the prior art, the object of the present invention is to propose a joint dynamic resource allocation method for the uplink and downlink transmissions in the context of 6G fully decoupled networks, targeting three types of resources: spectrum, base stations, and power.

[0006] The present invention is implemented as follows. The joint dynamic resource allocation method for the uplink and downlink in the 6G fully decoupled network includes the following steps:

[0007] Step 1: The uplink base station collects the channel estimates of the entire network, including the pilot signals sent by users and the channel estimates made by users based on the pilot signals of the downlink base station, and forwards the channel information of the entire network to the edge center for dynamic resource allocation.

[0008] Step 2: According to the channel information between users and uplink / downlink base stations and different sub-channels, assuming that the base stations and users adopt the method of equal power sharing, the three-dimensional matching problem of multi-user-multi-base station-multi-sub-channel is modeled as a large-scale many-to-many matching model, and a many-to-many matching algorithm based on maximizing the system weighted rate is designed.

[0009] Step 3: According to the matching relationship of user-base station-sub-channel, considering the coherent transmission of users and base stations on the same sub-channel, based on the theory of successive convex approximation, uplink and downlink power control algorithms are designed respectively for the uplink coherent transmission of multiple users and the downlink coherent transmission of multiple base stations.

[0010] Step 4: Through parallel computing, the uplink and downlink spectrum division ratios that maximize the system capacity are obtained, and the dynamic resource allocation scheme is sent to the uplink users and downlink base stations for implementation.

[0011] Furthermore, we dynamically divide the total sub-channels into uplink sub-channels and downlink sub-channels to provide services for uplink and downlink transmissions respectively.

[0012] Furthermore, for the uplink transmission of the 6G fully decoupled network, the uplink base station selected by each user first performs minimum mean square error (MMSE) combining on the information received by all its antennas, combines the multi-antenna information into one piece of information and uploads it to the edge cloud. Then, the edge cloud performs a second MMSE combining on the multiple pieces of information sent by multiple cooperative uplink base stations it has collected. This method is called the two-level information combining method.

[0013] Furthermore, for the downlink transmission of the 6G fully decoupled network, multiple base stations use the beamforming direction of maximum ratio transmission (MRT) and use power control to avoid interference to form efficient beamforming.

[0014] The described spectrum dynamic allocation method is different from both the existing frequency division duplexing (FDD) and time division duplexing (TDD) spectrum allocation methods. Compared with FDD, the edge cloud controls the adaptive allocation of uplink and downlink frequency bands. Facing the rapid changes in user service requirements, it dynamically adjusts the allocation of uplink and downlink frequency bands, which can more effectively improve the flexibility of spectrum utilization. Compared with TDD, it does not require a guard time slot for the conversion of user uplink and downlink transmissions, which can save spectrum usage time.

[0015] In the 6G fully decoupled network, the messages sent by uplink users can be jointly received by multiple uplink base stations and uploaded to the edge cloud. Downlink users can be cooperatively transmitted by multiple downlink base stations, and both uplink users and downlink base stations can use multiple sub-channels to transmit data: through the mapping relationship between base stations and sub-channels, a three-dimensional matching of multi-user - multi-base station - multi-sub-channel is modeled into a many-to-many matching of users to base station - sub-channel pairs. In the matching stage, based on the principles of equal power sharing and maximizing the total weighted rate of users, matching connections are made.

[0016] Furthermore, during the matching process, in each round of matching, for each user, the addition strategy and exchange strategy are tried. Based on the user preference list, the most suitable user is assigned to each base station - sub-channel pair for connection, and only the strategy that can improve the system capacity will be accepted.

[0017] Furthermore, in the uplink transmission matching, an uplink user may be connected by multiple uplink base station - uplink sub-channel pairs, and in the downlink transmission matching, a downlink base station - downlink sub-channel pair may also serve multiple users simultaneously.

[0018] Furthermore, for the uplink and downlink power control problems after the matching is completed, since the uplink total rate and downlink total rate in the objective function are non-convex, the continuous convex approximation theory is used to approximate the non-convex part as convex, and then a convex optimization algorithm is used to solve it to obtain the final power allocation scheme.

[0019] When performing sub-channel allocation, considering the quality of service guarantee for users and the capacity constraint of the base station fronthaul, it is required to allocate at least one base station and sub-channel to each user for connection service, and there is a limit on the number of user services for each base station.

[0020] After determining the matching relationship of multi-user-multi-base station-multi-sub-channel, in order to further avoid interference and improve system capacity, it is necessary to reallocate and control the power of users and base stations, and perform Taylor expansion on the non-convex part of the objective function, thus proposing an uplink and downlink power control algorithm based on successive convex approximation.

[0021] Compared with the prior art, the beneficial effects of the present invention are as follows: First, the uplink and downlink joint dynamic resource allocation method for the 6G fully decoupled network can effectively improve the network-wide system capacity and resource utilization rate; Second, considering the specific modes of uplink transmission and downlink transmission in the 6G fully decoupled network, we re-modeled the information transmission and processing methods and obtained the rate modeling; Third, in the face of the matching problem of multi-user-multi-base station-multi-sub-channel, we proposed a low-complexity matching game algorithm based on the many-to-many matching model, which can quickly obtain a stable solution; Finally, in the face of the power allocation problem of users and base stations on multiple sub-channels, we proposed a power control algorithm based on successive convex approximation, which can further improve the system capacity.

[0022] Based on the successive convex approximation theory, the present invention designs uplink and downlink power control algorithms respectively for the uplink cooperative transmission of multiple users and the downlink coherent transmission of multiple base stations, based on the matching relationship of user-base station-sub-channel, to further improve the system capacity. Compared with the sub-channel allocation based on round-robin scheduling and the single-user-single-base station connection method of equal power sharing in the traditional network, flexible multi-base station cooperation and efficient dynamic resource allocation algorithms can effectively improve the network capacity and improve the quality of service of mobile user communication. Brief Description of the Drawings

[0023] Figure 1 It is a scenario diagram of uplink and downlink joint dynamic resource allocation for the 6G fully decoupled network faced by the embodiments of the present invention.

[0024] Figure 2 It is a block diagram of the implementation of the multi-user-multi-base station-multi-sub-channel matching algorithm based on many-to-many matching proposed by the embodiments of the present invention.

[0025] Figure 3 It is a block diagram of the implementation of the power control algorithm based on successive convex approximation proposed by the embodiments of the present invention. Detailed Embodiment

[0026] To make the objectives, technical solutions, and advantages of the present invention more clearly understood, the following provides a detailed description of the embodiments of the present invention in conjunction with the accompanying drawings: These embodiments are implemented on the premise of the technical solutions of the present invention, and detailed implementation manners and specific operation processes are given. It should be understood that the specific examples described herein are only used to explain the present invention, but the protection scope of the present invention is not limited to the following embodiments.

[0027] The 6G fully decoupled network architecture considered in this embodiment is shown as Figure 1 consisting of D downlink base stations (DBSs), U uplink base stations (UBSs), H downlink users (DUEs) at random positions, and M uplink users (UUEs) at random positions. Assume that each UBS has L U receiving antennas, each DBS has L D transmitting antennas, and each user has one receiving antenna and one transmitting antenna. We represent the sets of DBSs and UBSs as and respectively, and the sets of DUEs and UUEs are represented as and respectively. Assume that there are a total of N subchannels (SCs) for uplink and downlink allocation, and the bandwidth of each subchannel is B 0 . We allocate C uplink subchannels to the uplink users and S downlink subchannels to the downlink base stations, which means C + S = N. Correspondingly, we represent the uplink and downlink subchannel sets as and

[0028] For the uplink transmission of the 6G fully decoupled network, we define a three-dimensional 0-1 variable to represent whether UUE i selects subchannel j and is served by UBS k. If so, takes 1, otherwise takes 0. Correspondingly, the uplink power control variable is defined as P ij , representing the power allocation of user i on subchannel j. First, for the first-level combining at each uplink base station, for UUE i using subchannel j in UBS k, the local signal received by the base station is where is the channel vector between UUE i and UBS k at subchannel j, is the set of UUEs that select subchannel j, is the noise vector, According to the minimum mean square error (MMSE) criterion, the signal combining vector of the base station is where represents the number of subchannels selected by UUE i'. Therefore, the signal aggregated by user i at subchannel j is:

[0029]

[0030] where is the base station cooperation set of user i on subchannel j, are the first-level combining MMSE coefficients of these uplink base stations. Define the effective channel between UUE i and UBS k under subchannel j as The interference channel from user i' to UBS k is Define and According to the MMSE criterion, we can obtain the optimal combining coefficient at the edge cloud as where represents the interference signal. Thus, we obtain the uplink signal-to-interference-plus-noise ratio of user i at subchannel j as

[0031] For the downlink transmission of the 6G fully decoupled network, we define a three-dimensional 0-1 variable to indicate whether DUE i selects subchannel j and is served by DBS k. If so, takes 1, otherwise takes 0. Correspondingly, we define to represent the power allocated by DBS k to DUE i at subchannel j. For the downlink transmission, define the signal received by a certain DUE as where is the channel vector of the effective signal, is the desired signal vector, is the desired signal from DBS i, and its power is is the interference channel vector, is the interference signal vector, is the interference signal from DBS j, is the Gaussian white noise. Considering that each DBS has L U transmit antennas and each DUE has one receive antenna, we denote as the channel vector between DUE i and DBS k in subchannel j, and the corresponding transmission precoding is Similarly, define Therefore, the signal received by DUE i on subchannel j is

[0032] According to the maximum ratio transmission (MRT) technology, the precoding coefficient of the signal x of the channel vector g is We have Therefore, the signal-to-interference-plus-noise ratio of DUE i at subchannel j is where

[0033] For the dynamic resource allocation in the 6G fully decoupled network, define the variables indicating whether UUE i is served by UBS k. If so take 1, otherwise take 0. Similarly indicating whether DUE i is served by DBS k. If so take 1, otherwise take 0. Define and as the maximum transmit power of UUE and DBS. Therefore, the optimal resource allocation problem (ORSP) for the uplink and downlink transmissions in the 6G fully decoupled network can be modeled as:

[0034]

[0035] s.t. C + S = N, (19a)

[0036]

[0037] Among them, constraints (19c) and (19h) ensure that at least one subchannel is allocated to the user, constraints (19d) and (19i) are the fronthaul capacity limitations for each UBS (DBS), and constraints (19f) and (19k) are the maximum transmit power constraints of UUE and DBS respectively.

[0038] First, we assume that the users and base stations adopt the power equalization method and first solve the matching problem of multi-user - multi-base station - multi-subchannel. First, form base station - subchannel pairs by combining base stations and subchannels, and convert the three-dimensional matching problem of multi-user - multi-base station - multi-subchannel into a many-to-many matching problem between users and base station - subchannel pairs, and solve it using the matching model. The specific flowchart is shown in Figure 2 . The matching is divided into two parts: In the initialization part, based on the channel quality, the users establish a preference list of base station - subchannel pairs, and match the most favorite base station - subchannel pair for each user under the condition of meeting the connection constraints and fronthaul capacity; during the matching process, in each round of matching, for each user, try the addition strategy and the exchange strategy, and based on the user preference list, allocate the most suitable user to each base station - subchannel pair for connection, and only the strategy that can improve the system capacity will be accepted. This process will continue to iterate until the matching relationship no longer changes or reaches the maximum number of iterations, and then the algorithm terminates.

[0039] For the power allocation problem in the uplink transmission of the 6G fully decoupled network, we adopt the successive convex approximation technique. For the uplink transmission, the problem is:

[0040]

[0041] where Denote the uplink weighted total rate, Define the function and its first-order Taylor expansion expression is where e i is a vector with the i-th element being 1 and others being 0, and the (k, l)-th element of Θ is If k≠i and l=j, Otherwise A = 0. Therefore, we can obtain If then the equation holds. The corresponding algorithm execution process is as Figure 3 shown.

[0042] For the power allocation problem in the downlink transmission of the 6G fully decoupled network, we adopt the successive convex approximation technique. For the downlink transmission, the problem is:

[0043]

[0044] where denotes the downlink signal-to-interference-plus-noise ratio. First, we define the function and the function of DUE i at sub-channel j We have:

[0045]

[0046] where represents the power allocation of all D downlink base stations, represents the power allocation matrix of downlink base station D. Then the function G i,j at the first-order Taylor expansion is:

[0047]

[0048] We can obtain and when the equation holds. The corresponding algorithm execution process is as Figure 3 shown.

[0049] The above are only the preferred embodiments of the present invention and are not intended to limit the present invention. Any modifications, equivalent replacements, and improvements made within the spirit and principles of the present invention shall be included within the protection scope of the present invention.

Claims

1. A method for uplink and downlink joint dynamic resource allocation for a 6G fully decoupled network, characterized in that: Adopt an uplink and downlink joint dynamic resource allocation method based on multi-point cooperative transmission, and the steps are as follows: Step 1: The uplink base station collects the channel estimates of the whole network, including the pilot signals sent by users and the channel estimates made by users based on the pilot signals of the downlink base station, and transmits the channel information of the whole network back to the edge center for dynamic resource allocation; Step 2: According to the channel information between users and uplink and downlink base stations and different sub-channels, assuming that the base station and users adopt the method of equal power sharing, model the three-dimensional matching problem of user-base station-sub-channel as a large-scale many-to-many matching model, and design a many-to-many matching based on the maximization of the system weighted rate, that is, a multi-user-multi-base station-multi-sub-channel matching algorithm based on the many-to-many matching model; Step 3: According to the matching relationship of user-base station-sub-channel, considering the coherent transmission of users and base stations on the same sub-channel, based on the theory of successive convex approximation, adopt uplink and downlink power control algorithms for the uplink coherent transmission of multiple users and the downlink coherent transmission of multiple base stations respectively; Step 4: Both Steps 2 and 3 are calculated in parallel to obtain the uplink and downlink spectrum division ratio that maximizes the system capacity, and send the dynamic resource allocation scheme to the uplink users and downlink base stations for implementation; After determining the matching relationship of multi-user-multi-base station-multi-sub-channel, reallocate and control the power of users and base stations, and perform Taylor expansion on the non-convex part of the objective function, thus proposing an uplink and downlink power control algorithm based on successive convex approximation; Dynamic resource allocation in a 6G fully decoupled network, define variables Indicates whether UUE i is served by UBS k, if so Take 1, otherwise take 0; Similarly Indicates whether DUE i is served by DBS k, if so Take 1, otherwise take 0; Define and As the maximum transmit power of UUE and DBS, therefore, the optimal resource allocation problem (ORSP) for uplink and downlink transmission in a 6G fully decoupled network is modeled as: s.t. C+S=N, (19a) Among them, constraints (19c) and (19h) ensure that at least one sub-channel is allocated to the user, constraints (19d) and (19i) are the fronthaul capacity limits for each UBS (DBS), and constraints (19f) and (19k) are the maximum transmit power constraints of UUE and DBS respectively; First, assume that users and base stations adopt the method of equal power sharing, and first solve the matching problem of multi-user-multi-base station-multi-sub-channel: First, form base station-sub-channel pairs by combining base stations and sub-channels, and convert the three-dimensional matching problem of multi-user-multi-base station-multi-sub-channel into a many-to-many matching problem of users and base station-sub-channel pairs, and use the matching model to solve it; The matching is divided into two parts: In the initialization part, based on the channel quality, users establish a preference list of base station-sub-channel pairs, and match the most favorite base station-sub-channel pair for each user under the condition of meeting the connection constraint and fronthaul capacity; In the matching process, in each round of matching, for each user, try the addition strategy and the exchange strategy, and based on the user preference list, allocate the most suitable user to each base station-sub-channel pair for connection, and the strategy that can improve the system capacity will be accepted. This process will continue to iterate until the matching relationship no longer changes or reaches the maximum number of iterations, and then the algorithm terminates; For the power allocation problem of the uplink transmission of the 6G fully decoupled network, adopt the successive convex approximation technology; for the uplink transmission, the problem is: Among them represents the uplink weighted total rate Define the function and its first-order Taylor expansion expression is where e i is a vector with the i-th element being 1 and others being 0, and the (k, l) element of Θ is if k≠i, l = j otherwise A = 0; thus, we get If then the equation holds For the power allocation problem of the downlink transmission of the 6G fully decoupled network, adopt the successive convex approximation technology; for the downlink transmission, the problem is: Among them represents the downlink signal-to-interference-plus-noise ratio; First, define the function and the function of DUE i at subchannel j There is Among them represents the power allocation of all D downlink base stations represents the power allocation matrix of downlink base station D; then the function G i,j at The first-order Taylor expansion is as follows: obtain and when the equation holds.

2. A method for uplink-downlink joint dynamic resource allocation for a 6G fully decoupled network as claimed in claim 1, characterized in that: For the uplink transmission of a 6G fully decoupled network, the uplink base station selected by each user first performs minimum mean square error (MMSE) combining on the information received by all its antennas, combines the multi-antenna information into one piece of information and uploads it to the edge cloud, and then the edge cloud performs a second MMSE combining on the multiple pieces of information sent by multiple cooperative uplink base stations it has collected.

3. A method for uplink-downlink joint dynamic resource allocation for a 6G fully decoupled network as claimed in claim 1, characterized in that: In a 6G fully decoupled network, the messages sent by uplink users can be jointly received by multiple uplink base stations and uploaded to the edge cloud, downlink users can be cooperatively transmitted by multiple downlink base stations, and both uplink users and downlink base stations can use multiple sub-channels to transmit data. Through the mapping relationship between the base station and the sub-channel, a many-to-many matching of multiple users-multiple base stations-multiple sub-channels is modeled into a many-to-many matching of users to base station-sub-channel pairs. In the matching stage, based on the principles of power equalization and maximizing the total user weighted rate, matching connections are made; During the matching process, in each round of matching, for each user, an adding strategy and a swapping strategy are attempted. Based on the user preference list, the most suitable user is assigned to each base station-sub-channel pair for connection, and only the strategy that can improve the system capacity will be accepted; In the uplink transmission matching, an uplink user will be connected by multiple uplink base station-uplink sub-channel pairs, and in the downlink transmission matching, a downlink base station-downlink sub-channel pair serves multiple users simultaneously; For the uplink and downlink power control problems after the matching is completed, since the uplink total rate and the downlink total rate in the objective function are non-convex, the continuous convex approximation theory is used to approximate the non-convex part as convex, and thus a convex optimization algorithm is used to solve it to obtain the final power allocation scheme.

4. A method for uplink-downlink joint dynamic resource allocation for a 6G fully decoupled network as claimed in claim 1, characterized in that: When performing sub-channel allocation, considering the quality of service guarantee of users and the base station fronthaul capacity constraint, it is required to allocate at least one base station and sub-channel for each user for connection service, and there is a limit on the number of users served by each base station.

Images