Non-ground network resource scheduling method for coexistence of multiple types of terminals

Abstract

The invention provides a non-ground network resource scheduling method for coexistence of multiple types of terminals, and belongs to the technical field of 6G Internet. According to the method, the time delay characteristic of satellite communication is considered, and a time slot scheduling method for coexistence of an FD-FDD terminal and an HD-FDD terminal is provided for a conflict scene of SSB and uplink transmission in a satellite internet system based on a 6G system; the neighbor cell beam switching rule is formulated for the FD-FDD terminal in the multi-terminal coexistence scene, so that the resource utilization rate is improved, and the processing load of the satellite-borne base station is reduced; in order to adaptively meet the maximum capacity constraint and the time delay constraint between the satellite and the ground, a target function with a penalty strategy is designed and added, and the gradient of a critical region of a constraint condition is increased, so that the scheduling complexity of a satellite-borne base station is further reduced, and the satellite-borne base station is better adapted to large-connection-oriented satellite internet application.

Description

Technical Field

[0001] The present invention relates to the field of 6G Internet technology, and in particular to a non-terrestrial network resource scheduling method for coexistence of multiple types of terminals, which can be used for uplink and downlink scheduling and resource allocation of 6G satellite Internet and space-based Internet. Background Art

[0002] In traditional Internet of Things (IoT) applications, duplex communication methods often need to consider issues such as resource constraints, spectrum allocation, and interference. As a compromise solution, half-duplex communication allows devices to perform two-way communication at different time periods, but cannot perform uplink and downlink communication simultaneously. In 3GPP R17, half-duplex frequency division duplex (HD-FDD) was introduced for RedCap. Compared with traditional full-duplex frequency division duplex (FD-FDD) communication, HD-FDD can reduce the complexity of devices to a certain extent, thereby achieving the purpose of cost control. Among them, Type A and Type B HD-FDD can reduce costs by 7% and 10% respectively. In actual terrestrial cellular networks, there are often full-duplex and half-duplex terminals coexisting. For example, smartphones can support full-duplex communication, while sensor nodes may use half-duplex communication. Regarding the conflict handling between HD-FDD RedCap terminals and FD-FDD terminals, referring to the conflict handling mechanism in the case of single carrier / single cell of NR TDD in 3GPP R15 and R16 versions, existing research mainly discusses the handling solutions for different scenarios based on the mechanism of terrestrial IoT, lacking research on conflict handling in the non-terrestrial network IoT system.

[0003] Full-duplex and half-duplex communication methods each have their own advantages and limitations, and the coexistence of duplex methods brings a series of challenges to satellite Internet systems. Therefore, effective resource allocation and management are required to adapt to satellite Internet systems. In this context, base station resource scheduling plays a key role in helping to manage the access, resource allocation, interference coordination, etc. of terminals in these two duplex methods. With the rapid growth of communication service demands, the resource allocation problem of high-throughput satellite systems has become a major research hotspot in the field of space communication technology today. Currently, the academic community has conducted more research on the optimal allocation of single-bandwidth resources. Traditional bandwidth allocation for multi-beam satellite systems adopts a fixed average allocation strategy. Among the commonly used allocation algorithms, more research has been done on convex optimization methods such as Lagrangian duality and subgradient methods. When using traditional meta-heuristic algorithms: quantum particle swarm algorithms for joint bandwidth-power optimization allocation, less research has been done on how to handle constraints. Due to the limited and expensive communication resources of satellite platforms, it is particularly important to optimize the allocation of satellite communication resources according to the communication demands and channel conditions of different beams to meet communication demands and improve resource utilization.

[0004] Due to the special nature of satellite communication, including signal propagation delay and channel delay, the coexistence of duplex methods becomes extremely challenging in satellite communication, which requires more complex resource allocation and time slot planning to ensure that communication between different terminals is not significantly affected by delay problems. In the 3GPP R17 research on the coexistence of different duplex methods between terrestrial RedCap IoT terminals and ordinary UEs, because the delay of terrestrial data transmission is small, the coexistence scheme formulated does not consider the impact of propagation delay, but this is obviously not applicable in the satellite Internet system. Even when satellite IoT terminals are usually relatively insensitive to delay, the space-ground delay cannot be ignored. Due to the non-negligible delay of satellite communication, the solution to avoid conflicts between HD-FDD Redcap IoT terminals and ordinary FD-FDD UE terminals in the ground network during SSB and uplink transmission: the priority SSB scheme is not applicable to satellite Internet communication systems, and this scheme will further increase the waiting delay of uplink transmission, resulting in untimely communication, affecting the user experience, and reducing the access efficiency of RedCap IoT terminals.

[0005] At the same time, the traditional fixed average allocation strategy results in equal capacity provided by each beam, which does not conform to the actual situation where the demands of each beam vary greatly, and the resource utilization is also low. Several traditional allocation algorithms cannot well balance effectiveness and fairness, and the factors considered in different multi-beam satellite resource allocation models also vary greatly. When using meta-heuristic algorithms for bandwidth optimization allocation, especially for satellite Internet systems facing wide-area IoT applications, considering the constraint of space-ground propagation delay will lead to an increase in computational complexity and it is difficult for the algorithm to converge. Summary of the Invention

[0006] In view of this, the present invention proposes a non-terrestrial network resource scheduling method for coexistence of multiple types of terminals. In the scenario of SSB and uplink transmission, the present invention designs a satellite Internet base station scheduling scheme applicable to the coexistence of FD-FDD and HD-FDD duplex mode terminals, which can overcome the problem of high complexity of resource utilization rate in the satellite Internet system by the commonly used resource allocation algorithms in the prior art.

[0007] In order to achieve the above object of the invention, the present invention provides the following technical solutions:

[0008] A non-terrestrial network resource scheduling method for coexistence of multiple types of terminals, which is applied to a satellite Internet communication system. The system includes satellites, HD-FDD terminals and FD-FDD terminals. The method includes the following steps:

[0009] Step 1: Initialize the priority of the on-board base station scheduling time slots for HD-FDD terminals;

[0010] Step 2: Before the terminals access the on-board base station, evenly distribute half of the total bandwidth resource B total and power resource P total of the satellite network to M satellite beams as initial values;

[0011] Step 3: When the HD-FDD terminals and FD-FDD terminals initially access the satellite network, the HD-FDD terminals first synchronize information according to the SSB downlink transmission, and then perform uplink data transmission, while the FD-FDD terminals perform uplink transmission and downlink reception simultaneously;

[0012] Step 4: When there are multiple terminals accessing the same on-board base station at a certain moment, for the FD-FDD terminals inside and outside the beam coverage area, screen the RSRP values of each cell and select the cell with the highest RSRP value for access;

[0013] Step 5: Dynamically adjust the total bandwidth resource B total and power resource P total of the satellite network, and calculate the optimal bandwidth B j and power resource P j allocated to each satellite beam in different time slots;

[0014] Step 6: In the connected state, when the spaceborne base station needs to perform beam measurement within a fixed cycle range to configure SSB-SMTC, transmit SSB with the highest priority. In non-measurement cases, swap the priorities of the uplink and downlink transmissions of HD-FDD terminals, and periodically increase the priority of SSB transmission by time Tc. When the priority of SSB is consistent with the uplink data transmission, perform downlink transmission of SSB in a time slot and reset the priority to the state of step S1; FD-FDD terminals still schedule the uplink transmission of the terminals according to the configured resources of the terminals.

[0015] Step 7: When an inter-satellite handover occurs to the terminal in the mobile satellite Internet of Things, repeat steps 1 - 6 to achieve the communication coexistence of HD-FDD terminals and FD-FDD terminals in the satellite Internet system.

[0016] Furthermore, the specific method of step 1 is as follows:

[0017] Considering the battery life and energy efficiency of the terminal, set the priorities from low to high as levels 1 to N according to the remaining energy characteristics of the terminal; the spaceborne base station adopts the same priority division method as the terminal; among them, the uplink transmission priority is w1, and the priority of SSB downlink transmission is w2, where w1 < w2, and other priorities are determined based on w1 and w2.

[0018] Furthermore, the specific method of step 5 is as follows:

[0019] Step 51: Construct the relationship between the delay constraint and the beam capacity constraint of the satellite Internet of Things:

[0020] ,

[0021] where is the delay constraint coefficient;

[0022] Step 52: Increase the gradient of the constraint condition critical area by adding a penalty term to obtain the calculation method of the fitness value with the penalty strategy for each beam j:

[0023] ,

[0024] where and are penalty coefficients;

[0025] Step 53: Use the quantum particle swarm method to solve the system capacity optimization problem, and the specific method is as follows:

[0026] Step 531: Set the particle swarm size N, the maximum number of iterations N imax , and the particle dimension D = 2M;

[0027] Generate the initial particle population , where particle i is represented as:

[0028] ,

[0029] Among them, the first M dimensions represent the particle positions of each beam allocation bandwidth, represented by the vector , and the last M dimensions represent the particle positions of each beam allocation power, represented by the vector ;

[0030] The initial position of each particle is affected by the minimum and maximum particle movement ranges and satisfies , where r is a random number and U(0,1) is a uniform distribution on the interval (0,1);

[0031] Step 532: Given the total bandwidth resource B total and power resource P total of the satellite network, perform a conversion of the resource allocation matrix for the particle population to obtain the power allocation vector and bandwidth allocation vector of each beam of particle i, where j is the beam number of particle i, j = 1, 2,..., M;

[0032] Define the resource allocation vector of particle i as , and calculate the system capacity vector of each particle;

[0033] Step 533: According to the resource allocation vector , calculate the fitness value of each particle as:

[0034] ,

[0035] where the calculation method of usc ij is as shown in Step 52 and is represented as:

[0036] ,

[0037] Step 534: After the t-th iteration, update the individual optimal position x isd (t), the global optimal position x gd (t) and the local attractor position x id (t) to realize the evolutionary iteration of the particle population;

[0038] Step 535: If the position change of all particles is less than the convergence threshold ε, it means that the particle swarm has converged, and the global optimal position of the particles is output; otherwise, repeat steps S532 to S534 until the maximum number of iterations is reached, and then output the global optimal position of the particles;

[0039] Step 54: Calculate the optimal power allocation vector of the satellite network beam in different time slots according to the global optimal position of the particles and the optimal bandwidth allocation vector .

[0040] The beneficial effects of the present invention are as follows:

[0041] 1. Solve the coexistence problem of two types of duplex terminals in the conflict scenario between SSB and uplink transmission in the non-terrestrial network. Consider the influence of satellite-ground propagation delay on time slot allocation, plan the time slots of the two transmission methods, and avoid conflicts by specifically setting the priority order.

[0042] 2. Solve the problem of excessive processing load of the spaceborne base station for multiple terminals, and formulate beam switching rules to reserve more space for resource allocation of the spaceborne base station.

[0043] 3. Solve the problem of obtaining invalid solutions when allocating bandwidth and power resources of the system using the metaheuristic algorithm under the conditions of satellite-ground delay and capacity constraints, reduce the base station scheduling complexity, improve the time-frequency resource utilization rate by better managing resources and time slots, and thus provide better connections for wide-area IoT terminals. Description of the Drawings

[0044] Figure 1 is a flowchart of the method of the present invention;

[0045] Figure 2 is a schematic diagram of the application scenario of the method of the present invention;

[0046] Figure 3 is a schematic diagram of accessing adjacent area beams in the scenario of multiple terminals of the method of the present invention; Detailed Embodiments

[0047] The present invention will be further described in detail below with reference to the drawings and specific embodiments.

[0048] A non-terrestrial network resource scheduling method for coexistence of multiple types of terminals, which is applied to a satellite Internet communication system. The system includes satellites, HD-FDD terminals and FD-FDD terminals, and includes the following steps:

[0049] Step 1: Establish a satellite base station scheduling time slot priority system for HD-FDD terminals. Considering the battery life and energy efficiency of the terminals, according to the remaining energy characteristics of the terminals, the priorities are set from low to high as levels 1 to N; the satellite base station adopts the same priority division method as the terminals; among them, the uplink transmission priority is w1, and the priority of SSB downlink transmission is w2, where w1 < w2, and other priorities are determined based on w1 and w2;

[0050] Step 2: Before the terminal accesses the satellite base station, regardless of the actual requirements of the terminal, evenly allocate half of the total bandwidth resource B total and power resource P total of the satellite network to M satellite beams as the initial value, that is, each beam is allocated bandwidth resource and power resource;

[0051] Step 3: When the HD-FDD terminal and the FD-FDD terminal initially access the satellite network, the HD-FDD terminal first synchronizes information according to the SSB downlink transmission, and then performs uplink data transmission, while the FD-FDD terminal simultaneously performs uplink transmission and downlink reception;

[0052] Step 4: When there are multiple terminals accessing the same satellite base station at a certain moment, for the FD-FDD terminals inside and outside the beam coverage area, screen the RSRP values of each cell and select the cell with the highest RSRP value for access;

[0053] Step 5: Once the terminal accesses, establish a low-complexity scheduling scheme for the satellite base station resources of the two types of duplex terminals. Through this scheduling scheme, dynamically adjust the total bandwidth resource B total and power resource P total of the satellite network, and finally calculate the optimal bandwidth B j and power resource P j allocated to each satellite beam in different time slots;

[0054] Step 6: In the connected state, when the satellite base station needs to perform beam measurement within a fixed cycle range to configure SSB-SMTC, transmit the SSB with the highest priority. In the non-measurement case, swap the uplink and downlink transmission priorities of the HD-FDD terminal, that is, w1 > w2, give priority to the uplink transmission of the terminal, and periodically increase the priority of SSB transmission according to time Tc. After (w1 - w2)∙Tc time, the priority of SSB is the same as that of uplink data transmission. At this time, perform the downlink transmission of SSB in the time slot and reset the priority to the state of step S1; the FD-FDD terminal still schedules the uplink transmission of the terminal according to the configured resources of the terminal;

[0055] Step 7: When the terminal undergoes inter-satellite handover in the mobile satellite Internet of Things, repeat Steps 1 - 6 to achieve the communication coexistence of HD-FDD terminals and FD-FDD terminals in the satellite Internet of Things system.

[0056] Among them, the specific method of Step 5 is as follows:

[0057] Step 51: Establish the relationship between the delay constraint and the beam capacity constraint of the satellite Internet of Things: , Among them, is the delay constraint coefficient; Step 52: By adding a penalty term to increase the gradient of the constraint condition critical region, obtain the fitness value calculation method for each beam j with a penalty strategy: , Among them, and are the penalty coefficients; Step 53: Use the quantum particle swarm optimization method to solve the system capacity optimization problem, and the specific method is as follows: Step 531: Set the particle swarm size N, the maximum number of iterations N imax , and the particle dimension D = 2M; Generate the initial particle population , among which, particle i is represented as: , Among them, the first M dimensions represent the particle positions of the bandwidth allocated to each beam, represented by the vector , and the last M dimensions represent the particle positions of the power allocated to each beam, represented by the vector ; The initial position of each particle is affected by the minimum and maximum particle movement ranges , satisfying , r is a random number, and U(0,1) is a uniform distribution on the interval (0,1); Step 532: Given the total bandwidth resource B total and power resource P total of the satellite network, perform the conversion of the resource allocation matrix for the particle population to obtain the power allocation vector and bandwidth allocation vector of each beam of particle i, where j is the beam number of particle i, j = 1, 2,..., M; Define the resource allocation vector of particle i as , and calculate the system capacity vector of each particle; Step 533: According to the resource allocation vector , calculate the fitness value of each particle as follows: , wherein, usc ij is calculated as shown in Step 52, expressed as: , Step 534: After the t-th iteration, update the individual optimal position x isd (t), the global optimal position x gd (t) and the local attractor position x id (t) to realize the evolutionary iteration of the particle swarm; Step 535: If the position change amount of all particles is less than the convergence threshold ε, it means that the particle swarm has converged, and output the global optimal position of the particles; otherwise, repeat Steps S532 to S534 until the maximum number of iterations is reached, and output the global optimal position of the particles; Step 54: According to the global optimal position of the particles, calculate the optimal power allocation vector and the optimal bandwidth allocation vector of the satellite network beam in different time slots.

[0058] The following is a more specific example:

[0059] A satellite Internet resource scheduling method for coexistence of multiple types of terminals, as Figure 1 shown, includes the following steps.

[0060] S1: In the coexistence scenario of several HD-FDD RedCap terminals such as smart glasses and FD-FDD terminals shown in Figure 2 , first establish the scheduling time slot priority of the spaceborne base station for HD-FDD RedCap terminals. According to the energy consumption characteristics of the terminal devices, the priority is divided into 10 categories, and the priority is set from low to high as 1 to 10 levels. Among them, the initial priority of the uplink transmission is 2, and the priority of the SSB downlink transmission is 9. Obviously, the energy resources of the spaceborne base station are greater than those of the ground terminals in the initial state, so the downlink transmission priority is higher;

[0061] S2: When the terminal initially accesses the satellite network, the HD-FDD RedCap terminal first performs SSB downlink transmission synchronization information with a higher priority, and then the terminal performs uplink data transmission, while the FD-FDD terminal simultaneously performs uplink transmission and downlink reception;

[0062] S3: In the same beam projected by the satellite to the ground or the air, any terminal UE in a certain user group is set as the first terminal UE. The first terminal UE determines its own location information through its own GNSS positioning ability. The location information includes, but is not limited to, longitude, latitude, and altitude. The first terminal UE obtains data such as the satellite ephemeris at the latest moment by receiving the pilot signal transmitted by the satellite, and then determines the satellite location information and the beam center location information at the current moment. The location information includes longitude, latitude, and altitude; the pilot signal includes satellite ephemeris information, satellite beam sub-satellite point location information, etc. The purpose of the ground terminal receiving the pilot signal is to confirm the beam center location information of the satellite and the first ground terminal UE.

[0063] When multiple terminals access the same spaceborne base station simultaneously, the difference between the own location information of the first terminal UE and the beam center location information exceeds the FD-FDD terminal with a beam coverage radius switches to an adjacent beam by selecting the cell corresponding to the highest RSRP value, as Figure 3 shown. Specifically, the ground or air terminal refers to a processing device used to communicate with the satellite within the satellite coverage beam range, and the first terminal UE refers to any one of several ground or air terminals;

[0064] S4: In the connected state, when the spaceborne base station needs to configure SSB-SMTC for beam measurement within a fixed cycle range, it will transmit SSB with the highest priority. In non-measurement cases, the uplink and downlink transmission priorities of the HD-FDD RedCap terminal are swapped, and the uplink transmission of the terminal is prioritized, that is, the uplink transmission priority is 9, and the SSB downlink transmission priority is 2. And in non-measurement cases, the HD-FDD RedCap terminal will increase the SSB transmission priority periodically every 3 minutes. After a cycle time of 21 minutes, when the priority of SSB is consistent with the uplink data transmission, the SSB is downlink transmitted in this time slice, and the priority is reset to the initial state, while the FD-FDD terminal still schedules the uplink transmission of the terminal according to the configured resources of the terminal;

[0065] S51: Establish the relationship between the delay constraint and the beam capacity constraint of the satellite Internet of Things;

[0066] Considering that the resource allocation of the satellite Internet system is affected by the propagation delay, set the total number of beams M to 5, and the capacity requirements of beam j are [100, 120, 140, 160, 180] Mbps, the total bandwidth resource is 600 MHz, the total power resource is 300 W, C jIndicates the actual amount of satellite system capacity allocation. The packet error rate of the communication link of the beam is 0.01, and the total number of terminals N within the coverage range of each beam j is 4 for all, the beam radius R is 2 km, and r n is the distance from terminal n to the center of the beam, and the delay constraint coefficient is calculated as follows:

[0067] ,

[0068] The calculation formula of the objective function USC of this optimization problem can be expressed as follows:

[0069] ,

[0070] S52: Add a penalty part to the unmet system capacity (USC) to obtain the fitness value calculation method with penalty strategy for each beam j:

[0071] ,

[0072] In this example, the quantum particle swarm algorithm with penalty strategy processes the delay constraint and proceeds as follows:

[0073] Step 531: Set the particle swarm size N, the maximum number of iterations N imax , and the particle dimension D = 2M;

[0074] Generate the initial population of particles , where particle i is represented as:

[0075] ,

[0076] Among them, the first M dimensions represent the particle positions of the bandwidths allocated to each beam, represented by the vector The last M dimensions represent the particle positions of the powers allocated to each beam, represented by the vector ;

[0077] The initial position of each particle is affected by the minimum and maximum particle movement ranges and satisfies , where U(0,1) is a uniform distribution on the interval (0,1).

[0078] Specifically, set the algorithm parameters, including the particle swarm size N to 50, the maximum number of iterations N imax to 400, and the particle dimension D to 10, and generate the initial population of particles.

[0079] Particle i can be represented as: , the first 5 dimensions represent the particle positions of each beam allocation bandwidth, and the last 5 dimensions represent the particle positions of each beam allocation power. The initial particle population can be expressed as , where the initial position of each particle is affected by the particle motion range . It will satisfy .

[0080] The complexity of this step is constant time because the amount of calculation involved does not increase with the increase of the problem scale.

[0081] S532: Perform the conversion of the resource allocation matrix for the particle population; the power allocation vector of particle i for each beam j can be calculated as:

[0082] ,

[0083] where the maximum value p max of the power component that a single beam can use is set to 90W, and the minimum value p min is set to 30W. is the minimum power vector of each beam;

[0084] is the scaling ratio of the power vector. This scaling factor will limit the beam power allocation to meet the maximum power constraint condition, expressed as:

[0085] ,

[0086] where represents the temporary power allocated to beam j after the transformation of particle i;

[0087] The bandwidth allocation vector is calculated as:

[0088] ,

[0089] where the maximum value b max of the bandwidth component that a single beam can use is set to 140MHz, and the minimum value b minx is set to 100MHz. is the minimum bandwidth vector of each beam;

[0090] is the scaling ratio of the bandwidth vector. This scaling factor will limit the beam bandwidth allocation to meet the maximum bandwidth constraint condition, expressed as:

[0091] ,

[0092] where Denote the temporary bandwidth allocated to beam j after the transformation of particle i;

[0093] Define the resource allocation vector of particle i as , and calculate the system capacity vector of each particle ; The calculation formula for the actual amount of system capacity allocation of particle i to beam j can be expressed as:

[0094] ,

[0095] where, set is the channel attenuation coefficient of 10 -3 , and the white noise power spectral density N0 of the downlink channel is 10 -15 W / Hz.

[0096] This step involves converting the particle population into a resource allocation matrix, and the complexity is proportional to the number of particles N and the resource dimension D. Therefore, the complexity is O(N × D).

[0097] S533: Calculate the fitness value. According to the resource allocation vector , calculate the fitness value of each particle as:

[0098] ,

[0099] where, usc ij is calculated as shown in S52 and is expressed as:

[0100] ,

[0101] The complexity of this calculation is proportional to the number of particles N and the resource dimension D. Therefore, the complexity is O(N×D).

[0102] S534: Particle population evolution. After the t-th iteration, update the individual optimal position x isd (t), the global optimal position x gd (t) and the local attractor position x id (t) to achieve the evolutionary iteration of the particle population;

[0103] The goal is to minimize the fitness value of the particle. After obtaining the fitness value usc i of the particle, determine whether to update the individual optimal position x isd (t), that is: if the fitness value of the current particle is lower than its historical optimal fitness, then update the individual optimal position x isd (t) to the position X i (t) of the current particle; At the same time, traverse the individual optimal positions x isd (t) of all particles and find the particle with the lowest fitness value to update the global optimal position xgd (t).

[0104] After the t-th iteration, affected by the individual optimal position and the global optimal position, calculate the local attraction factor x of particle i id (t) as: , where . Let C(t) represent the average optimal position coordinate of the particles, that is ; and the potential well length L id is regulated by the distance of the particle from the average optimal position, thus restricting the search range of the particle, expressed as: . Among them, is the contraction-expansion coefficient. When takes a smaller value, the convergence accuracy of the algorithm can be improved, and at the same time, it has better local refinement ability; when α takes a larger value, it can get rid of the particle being trapped in the local optimum and has better space expansion ability. In order to meet the changing search requirements, an adaptive contraction-expansion coefficient is constructed as:

[0105] ,

[0106] where represents the i-th iteration, and are constants.

[0107] After the t-th iteration, the update of each particle position is represented by the following formula:

[0108] ,

[0109] where the random number .

[0110] This step involves updating the individual optimal position, the global optimal position and the local attractor position of the particle to achieve evolutionary iteration. The complexity of this operation is proportional to the number of iterations N imax and the number of particles N, so the complexity is O(N × N imax ).

[0111] S535: Judge the termination condition. If the position change amount of all particles is less than the convergence threshold 1x10 -6 , then output the optimal particle position vector x gd ; if the termination condition is not satisfied, then repeat steps S532 to S534 until the maximum number of iterations is satisfied.

[0112] The complexity of the operation of judging the termination condition in this step is constant time.

[0113] S54: Calculate the optimization result. After obtaining the optimal particle position, use the matrix transformation formula in S532 to calculate the optimal bandwidth allocation vector of the satellite network beam and the optimal power allocation vector , and calculate the capacity allocation result of each beam based on this:

[0114] .

[0115] In this step, calculate the corresponding bandwidth allocation vector according to the optimal particle position vector, and calculate the capacity allocation result based on this. The complexity of this operation is proportional to the resource dimension D, so the complexity is O(D).

[0116] The complexity of the entire algorithm mainly depends on steps S532, S533, and S534, and its complexity is O(Nimax×N×D).

[0117] Table 1 conducts a comparative analysis of the complexity of the quantum particle swarm algorithm of the present invention and traditional meta-heuristic algorithms. Among them, the traditional meta-heuristic algorithms select genetic algorithms and multi-objective optimization algorithms (such as NSGA-II). It can be seen that the quantum particle swarm algorithm of the present invention has a lower complexity. In Table 1, L is the number of decision variables.

[0118]

[0119] The present invention considers the time-delay characteristics of satellite communication and proposes a time-slot scheduling method for the coexistence scenario of FD-FDD terminals and HD-FDD terminals in the conflict scenario between SSB and uplink transmission in a satellite Internet system based on the 6G system. By formulating different priority rules for SSB and uplink transmission in the initial access and connection states of the two types of terminals, the possibility of conflict during coexistence is avoided.

[0120] The present invention formulates a neighboring cell beam switching rule for FD-FDD terminals in the multi-terminal coexistence scenario, improving resource utilization and reducing the processing load of the on-board base station.

[0121] The present invention designs a joint bandwidth and power optimization algorithm for the on-board base station based on quantum particle swarm for the entire satellite Internet system, which is applicable to 6G wide-area Internet of Things technology. In order to adaptively meet the maximum capacity constraint and the time-delay constraint between the satellite and the ground, a penalty strategy is designed to be added to the objective function and the gradient of the constraint condition critical area is increased, further reducing the scheduling complexity of the on-board base station, so as to better adapt to the satellite Internet application facing large connections.

[0122] In summary, to overcome the problem of coexistence between HD-FDD terminals and FD-FDD terminals in the satellite communication system, the present invention proposes a low-complexity time slot scheduling and resource allocation method for 6G satellite Internet that supports coexistence of multiple types of terminals based on the conflict scenario between SSB and uplink transmission, which can better adapt to and support the coexistence of terminals with different duplex modes, so as to effectively expand the application scenarios of satellite Internet for 6G applications.

Claims

1. A non-terrestrial network resource scheduling method for coexistence of multiple types of terminals, applied to a satellite Internet communication system, the system including satellites, HD-FDD terminals and FD-FDD terminals, characterized in that: The following steps are involved: Step 1: Initialize the satellite base station scheduling time slot priority of the HD-FDD terminal; Step 2: Before the terminal accesses the satellite base station, evenly distribute the total bandwidth resources of the satellite network B total With power resource P total Half of is given to M satellite beams as the initial value; Step 3: When the HD-FDD terminal and the FD-FDD terminal initially access the satellite network, the HD-FDD terminal first synchronizes information according to the SSB downlink transmission, and then transmits uplink data, while the FD-FDD terminal simultaneously transmits uplink and receives downlink; Step 4: When multiple terminals access the same satellite base station at the same time, the FD-FDD terminal in the outer circle of the beam coverage area screens the RSRP value of each cell and selects the cell with the highest RSRP value for access; Step 5: The total bandwidth resource B of the satellite network total and power resource P total Dynamically adjust and calculate the optimal bandwidth B allocated to each satellite beam in different time slots j and power resource P j ; Step 6: In the connected state, when the satellite base station needs to perform beam measurement within a fixed period range to configure SSB-SMTC, the SSB is transmitted with the highest priority. In the non-measurement state, the priorities of the uplink and downlink transmissions of the HD-FDD terminal are swapped, and the priority of the SSB transmission is periodically increased according to the time Tc. When the priority of the SSB is consistent with the uplink data transmission, the downlink transmission of the SSB is performed in the time slice, and the priority is reset to the state of step S1; the FD-FDD terminal still schedules the uplink transmission of the terminal according to the configured resources of the terminal; Step 7: When the terminal undergoes inter-satellite switching in the mobile satellite Internet of Things, repeat steps 1 to 6 to achieve communication coexistence between the HD-FDD terminal and the FD-FDD terminal in the satellite Internet system.

2. The method for scheduling non-terrestrial network resources with coexistence of multiple types of terminals according to claim 1, characterized in that: The specific method of step 1 is: Taking into account the battery life and energy efficiency of the terminal, the priority is set from 1 to N levels from low to high according to the remaining energy characteristics of the terminal; the satellite base station adopts the same priority division method as the terminal; among them, the uplink transmission priority is w1, the SSB downlink transmission priority is w2, w1<w2, and other priorities are determined based on w1 and w2.

3. The method for scheduling non-terrestrial network resources with coexistence of multiple types of terminals according to claim 1, characterized in that: The specific method of step 5 is: Step 51: Construct the relationship between the delay constraint and beam capacity constraint of the satellite Internet of Things: ， in, is the delay constraint coefficient; Step 52: By adding a penalty term to increase the gradient of the critical region of the constraint condition, the fitness value calculation method of each beam j with penalty strategy is obtained: ， in, and is the penalty coefficient; Step 53: Use the quantum particle swarm method to solve the system capacity optimization problem. The specific method is as follows: Step 531: Set the particle swarm size N and the maximum number of iterations N imax , particle dimension D = 2M; Generate initial particle population , where particle i is represented by: ， Among them, the first M dimension The particle positions of each beam allocation bandwidth are represented by vectors Indicates that the post-M dimension The particle positions that represent the power distribution of each beam are represented by vectors express; The initial position of each particle is constrained by the minimum and maximum particle motion ranges. The impact of , r is a random number, U(0,1) is a uniform distribution on the interval (0,1); Step 532: Given the total bandwidth resource B of the satellite network total and power resource P total In the case of , the resource allocation matrix of the particle population is transformed to obtain the power allocation vector of each beam of particle i and bandwidth allocation vector , where j is the beam number of particle i, j=1,2,...,M; Define the resource allocation vector of particle i as , calculate the capacity vector of each particle system ; Step 533: allocate resources according to the vector , calculate the fitness value of each particle as: ， Among them, usc ij The calculation method of is shown in step 52, which is expressed as: ， Step 534: After the tth iteration, update the individual optimal position x of the particle isd (t), the global optimal position x gd (t) and the local attractor position x id (t), realize the evolutionary iteration of particle population; Step 535: If the position change of all particles is less than the convergence threshold ε, it means that the particle swarm has converged, and the global optimal position of the particle is output; otherwise, repeat steps S532 to S534 until the maximum number of iterations is reached, and the global optimal position of the particle is output; Step 54: Calculate the optimal power allocation vector of the satellite network beam at different time slots based on the global optimal position of the particle and the optimal bandwidth allocation vector .

Images