Resource scheduling method and device, communication equipment and readable storage medium

Abstract

The invention discloses a resource scheduling method and device, communication equipment and a readable storage medium, and belongs to the technical field of wireless. The resource scheduling method in the embodiment of the invention comprises the following steps: acquiring a historical active state of each terminal in a plurality of terminals, wherein the plurality of terminals are associated with a plurality of access points (AP); according to the historical active state of each terminal, determining the active state of each terminal in each coherent time block of a future first time period; according to the active state of each terminal in each coherent time block, determining a transmission mode of each AP in the plurality of APs in the first time period, the transmission mode being an uplink transmission mode or a downlink transmission mode; and scheduling uplink and / or downlink transmission of the plurality of APs according to the transmission mode of each AP in the first time period. Therefore, the AP duplex mode optimization does not need to be operated once in each coherent time block, so that the AP switching time delay is reduced.

Description

Technical Field

[0001] The present application belongs to the field of wireless technology, and specifically relates to a resource scheduling method, apparatus, communication equipment, and readable storage medium. Background Art

[0002] In a network-assisted full-duplex (NAFD) system based on a cellular-free architecture, multiple access points (APs) are equipped with a large number of antennas and densely distributed throughout an area. Each AP has two service modes: uplink and downlink. In each time slot, each AP can independently select either the uplink or downlink transmission mode. To reduce system latency, AP duplex mode optimization is typically performed once within a coherent time block. This involves assuming that the channel state remains unchanged within a coherent time block and finding the optimal AP uplink and downlink transmission mode selection vector under these channel conditions. In this case, once the channel conditions change, AP duplex mode optimization must be performed again in the next coherent time block. This results in frequent AP duplex mode optimization runs and frequent switching of AP transmission modes, leading to significant switching latency. Summary of the Invention

[0003] The purpose of the embodiments of the present application is to provide a resource scheduling method, apparatus, communication device and readable storage medium to solve the problem of long AP switching delay in related technologies.

[0004] In order to solve the above technical problems, this application is implemented as follows:

[0005] In a first aspect, a resource scheduling method is provided, comprising:

[0006] Obtaining a historical activity status of each terminal among a plurality of terminals, wherein the plurality of terminals are associated with a plurality of access points AP;

[0007] Determining, according to the historical activity state of each terminal, the activity state of each terminal in each coherent time block of a future first time period;

[0008] Determining a transmission mode of each of the plurality of APs in the first time period according to an activity state of each terminal in each coherent time block; wherein the transmission mode is an uplink transmission mode or a downlink transmission mode;

[0009] Schedule uplink and / or downlink transmission of the multiple APs according to the transmission mode of each AP in the first time period.

[0010] Optionally, determining, according to the activity state of each terminal in each coherent time block, a transmission mode of each AP in the multiple APs in the first time period includes:

[0011] According to the active state of each terminal in each coherent time block, with the goal of minimizing the sum of delays of the multiple terminals in the first time period, the transmission mode of each AP in the first time period is predicted.

[0012] Optionally, the predicting, based on the activity state of each terminal in each coherent time block, the transmission mode of each AP in the first time period with the goal of minimizing the sum of delays of the multiple terminals in the first time period, includes:

[0013] Determining a delay violation probability of each active terminal in each coherent time block according to the activity status of each terminal in each coherent time block; wherein the delay violation probability of each active terminal is: a probability that the delay of each active terminal exceeds a preset threshold;

[0014] Determining a maximum delay violation probability of active terminals within each coherent time block according to the delay violation probability of each active terminal within each coherent time block;

[0015] According to the maximum delay violation probability of the active terminals in each coherent time block, with the goal of minimizing the sum of the maximum delay violation probabilities corresponding to the first time period, the transmission mode of each AP in the first time period is predicted.

[0016] Optionally, determining, according to the activity state of each terminal in each coherent time block, a transmission mode of each of the multiple APs in the first time period includes:

[0017] In a first coherent time block in the first time period, a transmission mode of each AP in the first time period is determined according to an activity state of each terminal in each coherent time block.

[0018] Optionally, before scheduling uplink and / or downlink transmissions of the multiple APs according to the transmission mode of each AP in the first time period, the method further includes:

[0019] Determining a power allocation strategy within each coherent time block according to the transmission mode of each AP within the first time period and the channel state information within each coherent time block;

[0020] The step of scheduling uplink and / or downlink transmission of the multiple APs according to the transmission mode of each AP in the first time period includes:

[0021] The uplink and / or downlink transmissions of the multiple APs are scheduled according to the transmission mode of each AP in the first time period and the power allocation strategy.

[0022] Optionally, determining the power allocation strategy in each coherent time block according to the transmission mode of each AP in the first time period and the channel state information in each coherent time block includes:

[0023] According to the transmission mode of each AP in the first time period and the channel state information in each coherent time block, with the goal of maximizing the spectrum efficiency of each coherent time block, a power allocation strategy in each coherent time block is predicted.

[0024] Optionally, determining the power allocation strategy in each coherent time block according to the transmission mode of each AP in the first time period and the channel state information in each coherent time block includes:

[0025] In each coherent time block, a power allocation strategy in each coherent time block is determined according to the transmission mode of each AP in the first time period and the channel state information in each coherent time block.

[0026] Optionally, after determining the transmission mode of each of the multiple APs in the first time period, the method further includes:

[0027] Sending instruction information to a central processor, where the instruction information is used to indicate a transmission mode of each AP in the first time period;

[0028] receiving an evaluation result of the central processor on the transmission mode of each AP in the first time period;

[0029] Determine whether to adjust the transmission mode of each AP in the first time period according to the evaluation result.

[0030] In a second aspect, a resource scheduling device is provided, comprising:

[0031] an acquisition module, configured to acquire a historical activity status of each terminal among a plurality of terminals, wherein the plurality of terminals are associated with a plurality of APs;

[0032] A first determining module is configured to determine, based on a historical activity state of each terminal, an activity state of each terminal in each coherent time block in a first future time period;

[0033] a second determining module, configured to determine a transmission mode of each of the plurality of APs in the first time period according to an activity state of each terminal in each coherent time block; wherein the transmission mode is an uplink transmission mode or a downlink transmission mode;

[0034] A scheduling module is configured to schedule uplink and / or downlink transmission of the multiple APs according to a transmission mode of each AP in the first time period.

[0035] In a third aspect, a communication device is provided, comprising a processor, a memory, and a program or instruction stored in the memory and executable on the processor, wherein the program or instruction, when executed by the processor, implements the steps of the method described in the first aspect.

[0036] In a fourth aspect, a readable storage medium is provided, on which a program or instruction is stored. When the program or instruction is executed by a processor, the steps of the method described in the first aspect are implemented.

[0037] In a fifth aspect, a computer program product is provided, comprising computer instructions, which, when executed by a processor, implement the steps of the method described in the first aspect.

[0038] In an embodiment of the present application, by obtaining the historical activity status of each terminal in a plurality of terminals, the plurality of terminals are associated with a plurality of APs, and based on the historical activity status of each terminal, the activity status of each terminal in each coherent time block in a first time period in the future is determined, and based on the activity status of each terminal in each coherent time block, the transmission mode of each AP in a plurality of APs in the first time period is determined, and based on the transmission mode of each AP in the first time period, the uplink and / or downlink transmission of the plurality of APs is scheduled, thereby implementing an AP allocation strategy that enables the terminal to obtain relatively satisfactory service over a long time scale, without running an AP duplex mode optimization once in each coherent time block, thereby reducing the amount of calculation, AP switching delay and energy loss. BRIEF DESCRIPTION OF THE DRAWINGS

[0039] Figure 1 This is a flow chart of a resource scheduling method provided by an embodiment of the present application;

[0040] Figure 2 is a schematic diagram of a scheduling scheme in a specific embodiment of the present application;

[0041] Figure 3 is a flowchart of the resource scheduling process in a specific embodiment of the present application;

[0042] Figure 4This is a schematic diagram of the LETM network deployment structure in a specific embodiment of the present application;

[0043] Figure 5 This is a schematic diagram of the structure of a resource scheduling device provided in an embodiment of the present application;

[0044] Figure 6 It is a structural diagram of a communication device provided in an embodiment of the present application. DETAILED DESCRIPTION

[0045] The following will be combined with the drawings in the embodiments of this application to clearly and completely describe the technical solutions in the embodiments of this application. Obviously, the embodiments described are part of the embodiments of this application, not all of them. Based on the embodiments in this application, all other embodiments obtained by ordinary technicians in this field without making creative efforts are within the scope of protection of this application.

[0046] The terms "first," "second," and the like in the specification and claims of this application are used to distinguish similar objects, and are not used to describe a specific order or precedence. It should be understood that the terms used in this manner are interchangeable where appropriate, so that the embodiments of this application can be implemented in an order other than that illustrated or described herein, and that the objects distinguished by "first," "second," and the like are generally of the same type, and do not limit the number of objects; for example, the first object can be one or more. In addition, the term "and / or" in the specification and claims refers to at least one of the connected objects, and the character " / " generally indicates that the objects connected are in an "or" relationship.

[0047] The resource scheduling method, apparatus, communication device, and readable storage medium provided in the embodiments of the present application are described in detail below with reference to the accompanying drawings through specific embodiments and their application scenarios.

[0048] See Figure 1 , Figure 1 This is a flow chart of a resource scheduling method provided by an embodiment of the present application, which is applied to a communication device, such as a central processing unit (CPU) or an edge expansion node. Figure 1 As shown, the method includes the following steps:

[0049] Step 11: Obtain a historical activity status of each terminal among a plurality of terminals, where the plurality of terminals are associated with a plurality of APs;

[0050] Step 12: determining the activity state of each terminal in each coherent time block of a first time period in the future according to the historical activity state of each terminal;

[0051] Step 13: determining a transmission mode of each AP in the plurality of APs in the first time period according to an activity state of each terminal in each coherent time block;

[0052] Step 14: Schedule uplink and / or downlink transmissions of the multiple APs according to the transmission mode of each AP in the first time period.

[0053] In the embodiment of the present application, the length of the first time period can be determined based on actual needs and can include multiple coherent time blocks, for example, 10 coherent time blocks, without limitation. The first time period can be selected as a superframe, i.e., the transmission mode of the corresponding AP is determined for a superframe, and a superframe includes, for example, 10 coherent time blocks.

[0054] Optionally, when executing step 11, the activity state of each terminal in the first Z (Z≥1) coherent time blocks may be extracted from the historical data pool as the historical activity state of each terminal. The activity state may be inactive or active.

[0055] Optionally, after obtaining the historical activity status of each terminal, a long short-term memory (LSTM) network or other pre-trained models can be used to analyze the activity patterns of each terminal in each coherent time block based on the historical activity status of each terminal, and then predict the activity status of each terminal in each coherent time block in the first time period in the future.

[0056] Optionally, the transmission mode may be an uplink transmission mode (or referred to as an uplink service mode) or a downlink transmission mode (or referred to as a downlink service mode).

[0057] Optionally, after determining the transmission mode of each AP in the first time period, the transmission mode may be indicated to the corresponding AP to schedule uplink and / or downlink transmission of the AP.

[0058] Optionally, the applicable scenarios of the embodiments of the present application include but are not limited to NAFD systems based on non-cellular architecture, etc.

[0059] The resource scheduling method of an embodiment of the present application obtains the historical activity status of each terminal in a plurality of terminals, wherein the plurality of terminals are associated with a plurality of APs, determines the activity status of each terminal in each coherent time block in a first time period in the future according to the historical activity status of each terminal, determines the transmission mode of each AP in a plurality of APs in the first time period according to the activity status of each terminal in each coherent time block, and schedules the uplink and / or downlink transmission of the plurality of APs according to the transmission mode of each AP in the first time period. This can implement an AP allocation strategy that enables the terminal to obtain relatively satisfactory service over a long time scale, without running an AP duplex mode optimization once in each coherent time block, thereby reducing the amount of calculation, AP switching delay and energy loss.

[0060] Optionally, the process of determining the transmission mode of each of the multiple APs in the first time period according to the activity status of each terminal in each coherent time block may include:

[0061] Based on the activity status of each terminal within each coherent time block, the transmission mode of each AP within the first time period is predicted, for example, uplink or downlink transmission mode, with the goal of minimizing the sum of the latencies of the multiple terminals within the first time period. This allows for an optimized AP mode with reduced latency, thereby achieving an AP allocation strategy that enables terminals to obtain more satisfactory service.

[0062] It should be noted that, in order to minimize the sum of delays of multiple terminals in a first future time period, the optimization variable is the transmission mode of each AP in the first time period, so as to achieve the optimization goal through the AP mode optimization solution.

[0063] In embodiments of the present application, to facilitate characterizing the optimization objective, namely, minimizing the sum of the delays of multiple terminals within a first future time period, a delay violation probability may be introduced. This delay violation probability may be defined as the probability that the delay of an active terminal exceeds a preset threshold. This allows the transmission pattern of each AP within the first time period to be predicted using an optimization problem established based on the delay violation probability. The preset threshold may be pre-set based on actual needs and is not subject to limitation.

[0064] Optionally, the process of predicting the transmission mode of each AP in the first time period based on the activity state of each terminal in each coherent time block with the goal of minimizing the sum of delays of the multiple terminals in the first time period may include:

[0065] Determine the delay violation probability of each active terminal in each coherent time block based on the activity status of each terminal in each coherent time block, that is, determine the probability that the delay of each active terminal in each coherent time block exceeds a preset threshold;

[0066] Determining a maximum delay violation probability of active terminals in each coherent time block based on the delay violation probability of each active terminal in each coherent time block;

[0067] According to the maximum delay violation probability of active terminals in each coherent time block, with the goal of minimizing the sum of the maximum delay violation probabilities corresponding to the first time period, the transmission mode of each AP in the first time period is predicted; wherein the optimization variable is the transmission mode of each AP in the first time period.

[0068] In this way, by means of the optimization problem established based on the delay violation probability, the transmission mode of each AP in each coherent time block in the first time period in the future can be easily predicted.

[0069] Optionally, the determining, according to the activity state of each terminal in each coherent time block, the transmission mode of each AP in the multiple APs in the first time period may include:

[0070] In the first coherent time block of the first time period, the transmission mode of each AP in the first time period is determined based on the activity status of each terminal in each coherent time block. In other words, in the first coherent time block of the future first time period, the transmission mode of each AP in each coherent time block of the first time period is determined. Thus, the transmission mode of each AP in subsequent coherent time blocks can be determined in the first coherent time block, eliminating the need to perform AP duplex mode optimization in subsequent coherent time blocks. This reduces computational complexity and reduces AP handover latency and energy loss.

[0071] Considering that power allocation is another effective resource scheduling solution for improving system performance, before scheduling uplink and / or downlink transmissions of the multiple APs according to the transmission mode of each AP in the first time period, the resource scheduling method in this embodiment may further include:

[0072] A power allocation strategy within each coherent time block is determined based on the transmission mode of each AP within the first time period and channel state information within each coherent time block, wherein the channel state information can be obtained through channel estimation in each coherent time block.

[0073] The above-mentioned scheduling of uplink and / or downlink transmissions of the multiple APs according to the transmission mode of each AP in the first time period includes: scheduling the uplink and / or downlink transmissions of the multiple APs according to the transmission mode of each AP in the first time period and the power allocation strategy.

[0074] Therefore, resource allocation scheduling can be effectively implemented by combining the transmission mode of each AP in the first time period and the power allocation strategy in each coherent time block.

[0075] Optionally, determining the power allocation strategy in each coherent time block according to the transmission mode of each AP in the first time period and the channel state information in each coherent time block may include:

[0076] Based on the transmission mode of each AP during the first time period and the channel state information within each coherent time block, a power allocation strategy within each coherent time block is predicted with the goal of maximizing the spectral efficiency and sum of each coherent time block. The optimization variable is the power allocation strategy within each coherent time block. In this way, a power allocation strategy that minimizes the spectral efficiency and sum of each coherent time block can be obtained, thereby effectively implementing resource allocation and scheduling.

[0077] In addition, in addition to the above-mentioned goal of maximizing the sum of the spectrum efficiency of each coherent time block, other optimization problems may also be adopted, such as maximizing the sum of the transmission efficiency of each coherent time block, which is not limited.

[0078] Optionally, a power allocation strategy can be determined for each coherent time block. The above-mentioned determining the power allocation strategy within each coherent time block based on the transmission mode of each AP in the first time period and the channel state information in each coherent time block may include: determining the power allocation strategy within each coherent time block based on the transmission mode of each AP in the first time period and the channel state information in each coherent time block.

[0079] Considering that in current NAFD systems based on a non-cellular architecture, all APs are connected to a single CPU, and all AP duplex mode optimization is performed on this CPU, which places a huge amount of computation on the CPU and reduces the scalability of the system. In the embodiments of the present application, a multi-level collaborative non-cellular architecture is proposed. By clustering APs and terminals, distributed management is performed by edge expansion nodes (SENs), and the CPU performs the top-level collaborative management of the SENs.

[0080] Optionally, after determining the transmission mode of each AP in the first time period, the resource scheduling method in this embodiment may further include:

[0081] Sending instruction information to a central processing unit (CPU), where the instruction information is used to indicate a transmission mode of each AP in the first time period;

[0082] Receiving an evaluation result of the central processor on the transmission mode of each AP during the first time period; for example, whether the AP transmission mode determined by each SEN is appropriate may be evaluated by comprehensively considering the AP transmission modes determined by multiple SENs;

[0083] According to the evaluation result, it is determined whether to adjust the transmission mode of each AP in the first time period. If it is determined to be adjusted, the above operation is performed again to determine the transmission mode of each AP in the future first time period.

[0084] The present application is described below with reference to specific embodiments.

[0085] In the specific embodiments of the present application, a long-term resource scheduling method with high reliability and low latency in a flexible duplex scenario is proposed, which completes AP duplex mode selection and power allocation from a dual time scale; and considering the scalability of the actual deployment of the system, a multi-level / hierarchical collaborative cellular-free architecture is adopted, and distributed management is performed by clustering APs and terminals, thereby reducing the computational workload of each computing node and reducing processing latency; at the same time, this technology can still be used in NAFD systems based on the original cellular-free architecture, that is, corresponding to the special case where the number of clusters is 1.

[0086] In the specific embodiment of the present application, a new long time scale is added, which corresponds to the first time period mentioned above, or called a super frame. Assuming that T = 10 coherent time blocks are a super frame, the AP mode allocation scheme / strategy that can achieve better performance in the next super frame is optimized at a time on the long time scale, and the optimal power allocation scheme at this time is selected through channel estimation based on the selected AP duplex mode on the short time scale, i.e., each coherent time block. The specific schematic diagram is shown as follows: Figure 2 As shown in the lower middle part, Figure 2 The upper middle part shows the current AP duplex mode optimization that is run once in each coherent time block.

[0087] Optionally, the algorithms involved in the specific embodiments of this application specifically include the following three algorithms:

[0088] (1) Long-term active user prediction algorithm: Using the LSTM network, based on the historical activity status of each terminal, the activity pattern of each terminal in each coherent time block is analyzed, and the activity level / activity status of each terminal in each coherent time block of the next super frame is predicted in the coherent time block at the beginning of each super frame.

[0089] (2) Long-term AP mode optimization algorithm: At the beginning of each superframe, based on the terminal activity status predicted by the long-term active user prediction algorithm, the long-term AP mode is selected with the goal of minimizing the maximum delay violation probability of each coherent time block. The obtained long-term AP mode allocation scheme can provide low-latency services to each active terminal at that moment in the next superframe.

[0090] (3) Short-term power optimization algorithm: Based on the long-term AP mode allocation scheme selected at the start coherent time block of the superframe and the channel information obtained in the channel estimation stage of each coherent time block, downlink power allocation optimization is performed with the goal of maximizing the spectral efficiency of this coherent time block.

[0091] Optionally, for the multi-level collaborative non-cellular architecture in this solution, the CPU performs the top-level collaborative management, while S edge extension nodes SEN are responsible for the lower-level collaborative management. The range of SEN is randomly distributed with M half-duplex APs, K ul potential uplink UEs and K dl potential downlink UEs, and K ul +K dl = K. It is assumed here that the positions of all APs and UEs remain unchanged, and the channel state information and the activity state of each terminal in each coherent time block remain unchanged. However, the small-scale fading of the channel and the activity state of the terminal in different coherent time blocks may vary.

[0092] Each AP is equipped with N antennas and connects to the nearest SEN via a backhaul link. The specific transmission mode selected is determined by the SEN to which it is connected. APs connected to the same SEN automatically form a regional cluster, and each potential terminal is associated with the cluster closest to the cluster center. Therefore, the APs and UEs in the system form S non-overlapping regional clusters. Each SEN is responsible for the management of a cluster, while the CPU manages the SENs of multiple clusters.

[0093] Optionally, a block fading model is used for channel modeling. The length of each coherent time block is τ. The channel matrix between the kth UE and the mth AP in the tth coherent time block can be expressed as: Among them, β m,k represents the large-scale fading coefficient, h m,k (t)~CN(0,I N ) represents small-scale fast fading, and the large-scale fading coefficient changes slowly and is assumed to be known by the CPU. The active state of each UE in the t-th coherent time block can be obtained by δ(t) = [δ1(t),…,δ k (t),…,δ K (t)] represents, δ k(t) = 0 means that the kth UE is not active in the tth coherent time block, δ k (t) = 1 means that the kth UE is active in the tth coherent time block. These active UEs include UEs with uplink requirements and UEs with downlink requirements. In the tth coherent time block, the long-term optimal AP mode allocation scheme obtained by the long-term AP mode selection algorithm is: AP switches to uplink mode, The AP switches to downlink mode, and Uplink APs are combined into Uplink UEs are provided with services. Downlink APs are combined into downlink UEs jointly transmit signals.

[0094] like Figure 3 As shown in Figure 2, the specific resource scheduling process includes:

[0095] Step 1: Adopt the long-term active user prediction algorithm. That is, after entering a super frame, extract the activity status of each user (i.e., terminal UE) in the previous L super frames (i.e., LT coherent time blocks) from the historical data pool, and use the LSTM network to predict the activity status of each UE in the next super frame (i.e., T coherent time blocks).

[0096] According to the actual scene detection, in the long term, the UE’s active state in each time slot is δ k (t) follows a Bernoulli distribution and has a certain distribution pattern. Therefore, an LSTM network is used in the long-term active user prediction algorithm to predict the activity status of each UE in the next super frame based on the activity status of each UE in the previous L super frames.

[0097] At the same time, considering that centralized prediction of the future active status of all UEs requires a relatively complex LSTM network structure, and the computational complexity of the prediction algorithm will increase linearly as the number of potential UEs in the system increases, the advantages of the collaborative service cell-free architecture are fully utilized to deploy the LSTM prediction network on each SEN, and the prediction network on each SEN only predicts the future active status of UEs in the cluster it manages. The LSTM network deployment structure deployed in each SEN can be as follows: Figure 4 As shown, it includes input layer, hidden layer and input layer, and the corresponding input historical data is as follows Figure 4 As shown:

[0098] Step 2: Based on the predicted UE activity in each coherent time block, calculate the delay violation probability of each active UE in each future coherent time block.

[0099] When MRT precoding is used for downlink transmission, in the tth coherent time block, if the downlink UE1 is active, the signal-to-noise and interference ratio (SINR) of the UE can be expressed as:

[0100]

[0101]

[0102]

[0103]

[0104]

[0105] in, It indicates the interference caused by the active downlink UE at this time, which is predicted by the same pilot as the downlink UE1. It indicates the interference caused by the downlink UE active at this time using a different pilot signal from the downlink UE1. represents the interference caused by channel estimation error, Indicates the predicted cross-link interference caused by the signals sent by the active uplink UE at this time; X d is a diagonal matrix with M rows and M columns used to represent AP duplex mode selection, where M represents the number of APs. If the mth diagonal element is 1, it means that the mth AP is working in downlink mode, and if the mth diagonal element is 0, it means that the mth AP is working in uplink mode. N represents an N-row, N-column unit matrix, where N is the number of antennas on the AP. The multiplication sign inside the circle represents the Kronecker product; |·| represents the absolute value. represents the estimated channel vector between the lth downlink UE and all APs; represents the channel estimation error vector between the lth downlink UE and all APs; ρ dl,k is the downlink transmission power of the kth downlink UE, ρ ul,i is the uplink transmit power of the i-th uplink UE; g t,j,i represents the interference channel vector between the i-th uplink UE and the j-th downlink UE; is the variance of the downlink channel noise; when using maximum ratio transmission, the downlink precoding vector of the lth downlink UE is Defined as: in is the estimated channel vector of the actual channel of the lth downlink UE obtained in the channel estimation phase; |||| represents the 2-norm of the vector; the superscript H represents the conjugate transpose; represents the downlink precoding vector of the l′th downlink UE; represents the downlink precoding vector of the k-th downlink UE; represents the downlink UE index set of the t-th coherent time block; represents the uplink UE index set of the tth coherent time block; P tl Indicates the UE index set that uses the same pilot as the lth downlink UE.

[0106] Therefore, according to the SINR of the downlink UE1, the spectrum efficiency corresponding to the downlink UE1 can be expressed as:

[0107]

[0108] Where τ represents the number of symbols per coherent time block, τ C Indicates the number of pilot symbols.

[0109] Consider setting a service delay limit T D For UEk, if the time it takes to receive the required service data exceeds T D , then it is a "delay violation" event and does not meet the transmission requirements of URLLC. Therefore, the delay violation probability of UEk in the tth coherent time block is defined as:

[0110]

[0111] in, The number of bits required to meet the Quality of Service (QoS) of UEk is W, and W is the transmission bandwidth.

[0112] When using the MRC receive vector for uplink transmission, in the tth coherent time block, if the uplink UEu is active, the SINR of the UE can be expressed as:

[0113]

[0114] in, Indicates the interference caused by the active uplink UE at this time, which is predicted by the same pilot as the uplink UEu. Indicates the interference caused by the active uplink UE at this time using a different pilot signal from the uplink UEu. represents the interference caused by channel estimation error, Indicates the predicted cross-link interference caused by the signals sent by the active uplink UEs at this time; represents the maximum ratio combined reception vector for the u-th uplink UE; the meanings of other symbols are similar to those of the symbols in the above-mentioned downlink UE SINR, and are not repeated here.

[0115] Similarly, based on the SINR of uplink UEu, the spectrum efficiency corresponding to uplink UEu can be expressed as:

[0116]

[0117] The delay violation probability corresponding to the uplink UEu can be expressed as:

[0118]

[0119] Step 3: Use the long-term AP mode optimization algorithm to obtain the AP mode allocation plan. The specific process includes:

[0120] Step 31: Extract the maximum delay violation probability of the active UE predicted in each of the calculated future T coherent time blocks, and establish an optimization problem with the goal of minimizing the maximum delay violation probability in the future T coherent time blocks.

[0121] Taking T=10 as an example, based on the predicted UE activity status, each SEN calculates the delay violation probability of the active UE in each coherent time block of the next superframe, and obtains:

[0122]

[0123] in, and They respectively represent the number of active uplink UEs and downlink UEs in the t-th coherent time block predicted based on the predicted UE activity status.

[0124] The predicted maximum delay violation probability of active UEs in each coherence time block It can be expressed as:

[0125]

[0126] Then, an optimization problem is established with the goal of minimizing the maximum delay violation probability in the next T coherent time blocks, and the optimization variable is the AP mode selection matrix x u and x d :

[0127]

[0128] in, and Respectively represent the AP maximum downlink power consumption constraint and the UE maximum uplink power consumption constraint, the binary matrix variable x u and x d Represent the uplink and downlink state vectors of the AP respectively. and They represent the minimum communication rate required by the lth downlink UE and the minimum communication rate required by the uth uplink UE respectively.

[0129] Step 32: Leveraging the advantages of the hierarchical cooperative cell-free architecture, the MADDPG algorithm is used to solve the above optimization problem. The input is the long-term optimal AP mode allocation plan for the next superframe, that is, the transmission mode of each AP in the next superframe. The specific solution process is divided into two parts: network training and network call:

[0130] (1) Network training treats each SEN as an agent. Each agent learns the optimal action based on feedback from the environment, and finds the optimal duplex mode for the APs in the cluster managed by the SEN. At the same time, the learning feedback values of each agent in the MADDPG algorithm are interconnected, so the optimal duplex mode for the APs in the cluster output by the SEN is combined to approximate the global optimal solution.

[0131] (2) After the SEN network parameter training is completed, the current system state can be input into the SEN and the network can be directly called to automatically output the AP uplink and downlink mode allocation vector that can achieve the optimal balance between the utility function and the system spectrum efficiency under the current system state. and

[0132] It should be noted that in this solution, historical data is used for network training, and the trained network is directly called after the network training is completed. In this way, each SEN can select the network based on the training action and quickly find the optimal duplex mode allocation scheme for the APs in the cluster.

[0133] Step 4: Perform channel estimation for each coherent time block, obtain channel state information and the actual UE activity state, and store the actual UE activity state for this coherent time block into the historical data pool.

[0134] A short-term power optimization algorithm is used to optimize downlink power allocation for each coherent time block based on the estimated channel state information and a given optimal AP mode allocation scheme to maximize the user sum rate (i.e., spectral efficiency and rate). The short-term optimal power allocation scheme for this coherent time block is output, that is, the power allocation strategy within this coherent time block. Subsequently, uplink and downlink data transmission are carried out using this long-term optimal AP mode allocation scheme and short-term optimal power allocation scheme until the end of the corresponding super frame.

[0135] In this step, a short-term optimization problem is established with spectrum efficiency and as the optimization objectives. The optimization variable is the power allocated by the AP to each active downlink UE:

[0136]

[0137] Alternatively, a genetic algorithm can be used to solve this optimization problem, for example: encoding each possible combination of downlink power allocation into a binary row vector as a coded chromosome individual, and randomly generating N pop The initial population of individuals (i.e., downlink power combination) is generated, and the fitness value of each individual is calculated according to the optimization target of P2; individuals with high fitness are selected, hybridized, and mutated under a certain probability to produce a new offspring population; after generating N maxgen After the population is inherited, the algorithm terminates and the individual with the highest fitness corresponds to the optimal solution, i.e., the optimal short-term power allocation scheme for this coherent time block.

[0138] In summary, the specific embodiments of the present application can achieve the following beneficial effects:

[0139] 1) This application proposes a dual-time-scale resource scheduling scheme for long-term resource allocation scheduling in a NAFD system based on a non-cellular architecture. This scheme combines a long-term, large-time-scale AP mode optimization algorithm with the optimization goal of minimizing the maximum delay violation probability and a short-time-scale power allocation algorithm with the optimization goal of maximizing the sum rate. This approach can reduce the complexity of the system resource scheduling algorithm while meeting the high-reliability and low-latency service requirements of the terminal.

[0140] 2) This application proposes a prediction-then-optimization approach, namely a "long-term active user prediction algorithm-long-term AP mode optimization algorithm," to find an AP allocation strategy that can ensure that UEs receive relatively satisfactory service over a long period of time, thus avoiding the problem of frequent AP mode changes in short-term AP mode optimization solutions.

[0141] 3) This application proposes a resource allocation scheme for a NAFD system based on a hierarchical collaborative cellular-free architecture. Different from the existing resource allocation scheme of a NAFD system based on the original cellular-free architecture, this resource allocation scheme fully considers the scalability of the system in actual deployment, introduces SEN as an edge computing unit, and distributes the LSTM network in the long-term active user prediction algorithm and the intelligent agent deep learning network in the long-term AP mode optimization algorithm in each SEN to perform partition management of APs and UEs in a cluster, sharing the huge computational workload of resource allocation of all APs and UEs in the CPU management system in the existing resource allocation scheme, thereby improving computing efficiency.

[0142] It should be noted that the resource scheduling method provided in the embodiments of the present application can be executed by a resource scheduling device, or a control module in the resource scheduling device for executing the resource scheduling method. In the embodiments of the present application, the resource scheduling device provided in the embodiments of the present application is described by taking the resource scheduling device executing the resource scheduling method as an example.

[0143] See Figure 5, Figure 5 This is a schematic diagram of the structure of a resource scheduling device provided by an embodiment of the present application, which is applied to a communication device, such as a central processing unit (CPU) or an edge expansion node. Figure 5 As shown, the resource scheduling device 50 includes:

[0144] An acquisition module 51 is configured to acquire a historical activity status of each terminal among a plurality of terminals, wherein the plurality of terminals are associated with a plurality of APs;

[0145] A first determining module 52 is configured to determine, based on the historical activity status of each terminal, the activity status of each terminal in each coherent time block of a first time period in the future;

[0146] a second determining module 53, configured to determine a transmission mode of each of the plurality of APs in the first time period according to an activity state of each terminal in each coherent time block; wherein the transmission mode is an uplink transmission mode or a downlink transmission mode;

[0147] The scheduling module 54 is configured to schedule uplink and / or downlink transmissions of the multiple APs according to the transmission mode of each AP in the first time period.

[0148] Optionally, the second determining module 53 is specifically configured to:

[0149] According to the active state of each terminal in each coherent time block, with the goal of minimizing the sum of delays of the multiple terminals in the first time period, the transmission mode of each AP in the first time period is predicted.

[0150] Optionally, the second determining module 53 is specifically configured to:

[0151] Determining a delay violation probability of each active terminal in each coherent time block according to the activity status of each terminal in each coherent time block; wherein the delay violation probability of each active terminal is: a probability that the delay of each active terminal exceeds a preset threshold;

[0152] Determining a maximum delay violation probability of active terminals within each coherent time block according to the delay violation probability of each active terminal within each coherent time block;

[0153] According to the maximum delay violation probability of the active terminals in each coherent time block, with the goal of minimizing the sum of the maximum delay violation probabilities corresponding to the first time period, the transmission mode of each AP in the first time period is predicted.

[0154] Optionally, the second determining module 53 is specifically configured to:

[0155] In a first coherent time block in the first time period, a transmission mode of each AP in the first time period is determined according to an activity state of each terminal in each coherent time block.

[0156] Optionally, the resource scheduling device 50 further includes:

[0157] a third determining module, configured to determine a power allocation strategy within each coherent time block according to the transmission mode of each AP within the first time period and the channel state information within each coherent time block;

[0158] The scheduling module 54 is further configured to schedule uplink and / or downlink transmissions of the multiple APs according to the transmission mode of each AP in the first time period and the power allocation strategy.

[0159] Optionally, the third determining module is specifically configured to:

[0160] According to the transmission mode of each AP in the first time period and the channel state information in each coherent time block, with the goal of maximizing the spectrum efficiency of each coherent time block, a power allocation strategy in each coherent time block is predicted.

[0161] Optionally, the third determining module is specifically configured to:

[0162] In each coherent time block, a power allocation strategy in each coherent time block is determined according to the transmission mode of each AP in the first time period and the channel state information in each coherent time block.

[0163] Optionally, the resource scheduling device 50 further includes:

[0164] a sending module, configured to, after determining the transmission mode of each AP in the first time period, send indication information to a central processor, wherein the indication information is used to indicate the transmission mode of each AP in the first time period;

[0165] a receiving module, configured to receive an evaluation result of the central processor on the transmission mode of each AP in the first time period;

[0166] A fourth determining module is configured to determine, based on the evaluation result, whether to adjust the transmission mode of each AP within the first time period.

[0167] The resource scheduling device 50 of the embodiment of the present application can realize the above Figure 1The various processes of the method embodiment shown can achieve the same technical effect, and to avoid repetition, they will not be described again here.

[0168] The present application also provides a computer program product including computer instructions, which can achieve the above-mentioned Figure 1 The various processes of the method embodiment shown can achieve the same technical effect, and to avoid repetition, they will not be described here.

[0169] Optional, such as Figure 6 As shown, an embodiment of the present application also provides a communication device 60, including a processor 61, a memory 62, and a program or instruction stored in the memory 62 and executable on the processor 61. When the program or instruction is executed by the processor 61, each process of the above-mentioned resource scheduling method embodiment is implemented, and the same technical effect can be achieved. To avoid repetition, it will not be repeated here.

[0170] An embodiment of the present application also provides a readable storage medium on which a program or instruction is stored. When the program or instruction is executed by a processor, the various processes of the above-mentioned resource scheduling method embodiment can be implemented and the same technical effect can be achieved. To avoid repetition, it will not be repeated here.

[0171] Computer-readable media includes both permanent and non-permanent, removable and non-removable media, and can be implemented using any method or technology for information storage. Information can be computer-readable instructions, data structures, program modules, or other data. Examples of computer storage media include, but are not limited to, phase-change RAM (PRAM), static random access memory (SRAM), dynamic random access memory (DRAM), other types of random access memory (RAM), read-only memory (ROM), electrically erasable programmable read-only memory (EEPROM), flash memory or other memory technologies, compact disc read-only memory (CD-ROM), digital versatile disc (DVD) or other optical storage, magnetic cassettes, magnetic tape, magnetic disk storage or other magnetic storage devices, or any other non-transmission media that can be used to store information that can be accessed by a computing device. As defined herein, computer-readable media does not include transitory computer-readable media, such as modulated data signals and carrier waves.

[0172] It should be noted that, in this document, the terms "comprises," "includes," or any other variations thereof are intended to encompass non-exclusive inclusion, such that a process, method, article, or apparatus comprising a series of elements includes not only those elements but also other elements not explicitly listed, or elements inherent to such process, method, article, or apparatus. In the absence of further limitations, an element defined by the phrase "comprising a ..." does not exclude the presence of other identical elements in the process, method, article, or apparatus comprising the element.

[0173] The serial numbers of the above embodiments of the present application are for description only and do not represent the advantages or disadvantages of the embodiments.

[0174] Through the description of the above implementation methods, those skilled in the art can clearly understand that the above-mentioned embodiment methods can be implemented by means of software plus the necessary general hardware platform, and of course can also be implemented by hardware, but in many cases the former is a better implementation method. Based on this understanding, the technical solution of the present application, or the part that contributes to the prior art, can be embodied in the form of a software product, which is stored in a storage medium (such as ROM / RAM, magnetic disk, optical disk), and includes a number of instructions for enabling a service classification device (which can be a mobile phone, computer, server, air conditioner, or network device, etc.) to execute the methods described in each embodiment of the present application.

[0175] The above is only a preferred embodiment of the present application. It should be pointed out that for ordinary technicians in this technical field, several improvements and modifications can be made without departing from the principles of the present application. These improvements and modifications should also be regarded as the scope of protection of the present application.

Claims

1. A resource scheduling method, characterized in that: include: Obtaining a historical activity status of each terminal among a plurality of terminals, wherein the plurality of terminals are associated with a plurality of access points AP; Determining, according to the historical activity state of each terminal, the activity state of each terminal in each coherent time block of a future first time period; Determining a transmission mode of each of the plurality of APs in the first time period according to an activity state of each terminal in each coherent time block; wherein the transmission mode is an uplink transmission mode or a downlink transmission mode; Schedule uplink and / or downlink transmission of the multiple APs according to the transmission mode of each AP in the first time period.

2. The method according to claim 1, characterized in that The determining, according to the activity state of each terminal in each coherent time block, a transmission mode of each AP in the plurality of APs in the first time period includes: According to the active state of each terminal in each coherent time block, with the goal of minimizing the sum of delays of the multiple terminals in the first time period, the transmission mode of each AP in the first time period is predicted.

3. The method according to claim 2, characterized in that The predicting, based on the activity state of each terminal in each coherent time block, the transmission mode of each AP in the first time period with the goal of minimizing the sum of delays of the multiple terminals in the first time period, includes: Determining a delay violation probability of each active terminal in each coherent time block according to the activity status of each terminal in each coherent time block; wherein the delay violation probability of each active terminal is: a probability that the delay of each active terminal exceeds a preset threshold; Determining a maximum delay violation probability of active terminals within each coherent time block according to the delay violation probability of each active terminal within each coherent time block; According to the maximum delay violation probability of the active terminals in each coherent time block, with the goal of minimizing the sum of the maximum delay violation probabilities corresponding to the first time period, the transmission mode of each AP in the first time period is predicted.

4. The method according to any one of claims 1 to 3, characterized in that The determining, according to the activity state of each terminal in each coherent time block, a transmission mode of each AP in the plurality of APs in the first time period includes: In a first coherent time block in the first time period, a transmission mode of each AP in the first time period is determined according to an activity state of each terminal in each coherent time block.

5. The method according to claim 1, wherein Before scheduling uplink and / or downlink transmissions of the multiple APs according to the transmission mode of each AP in the first time period, the method further includes: Determining a power allocation strategy within each coherent time block according to the transmission mode of each AP within the first time period and the channel state information within each coherent time block; The step of scheduling uplink and / or downlink transmission of the multiple APs according to the transmission mode of each AP in the first time period includes: The uplink and / or downlink transmissions of the multiple APs are scheduled according to the transmission mode of each AP in the first time period and the power allocation strategy.

6. The method according to claim 5, characterized in that The determining, according to the transmission mode of each AP in the first time period and the channel state information in each coherent time block, a power allocation strategy in each coherent time block includes: According to the transmission mode of each AP in the first time period and the channel state information in each coherent time block, with the goal of maximizing the spectrum efficiency of each coherent time block, a power allocation strategy in each coherent time block is predicted.

7. The method according to claim 5 or 6, characterized in that The determining, according to the transmission mode of each AP in the first time period and the channel state information in each coherent time block, a power allocation strategy in each coherent time block includes: In each coherent time block, a power allocation strategy in each coherent time block is determined according to the transmission mode of each AP in the first time period and the channel state information in each coherent time block.

8. The method according to claim 1, characterized in that After determining the transmission mode of each of the multiple APs in the first time period, the method further includes: Sending instruction information to a central processor, where the instruction information is used to indicate a transmission mode of each AP in the first time period; receiving an evaluation result of the central processor on the transmission mode of each AP in the first time period; Determine whether to adjust the transmission mode of each AP in the first time period according to the evaluation result.

9. A resource scheduling device, characterized in that: include: an acquisition module, configured to acquire a historical activity status of each terminal among a plurality of terminals, wherein the plurality of terminals are associated with a plurality of APs; A first determining module is configured to determine, based on a historical activity state of each terminal, an activity state of each terminal in each coherent time block in a first future time period; a second determining module, configured to determine a transmission mode of each of the plurality of APs in the first time period according to an activity state of each terminal in each coherent time block; wherein the transmission mode is an uplink transmission mode or a downlink transmission mode; A scheduling module is configured to schedule uplink and / or downlink transmission of the multiple APs according to a transmission mode of each AP in the first time period.

10. A communication device, characterized in that: The method comprises a processor, a memory, and a program or instruction stored in the memory and executable on the processor, wherein the program or instruction, when executed by the processor, implements the steps of the resource scheduling method according to any one of claims 1 to 8.

11. A readable storage medium, characterized in that: The readable storage medium stores a program or instruction, and when the program or instruction is executed by a processor, the steps of the resource scheduling method according to any one of claims 1 to 8 are implemented.

12. A computer program product, characterized in that The method comprises computer instructions, which, when executed by a processor, implement the steps of the resource scheduling method according to any one of claims 1 to 8.

Images