A handover decision and power allocation coordination method for a mobile phone direct connection low earth orbit satellite network

Abstract

This invention provides a collaborative method for handover decision-making and power allocation in mobile phone direct connection to low-Earth orbit satellite networks, belonging to the field of mobile phone direct connection to low-Earth orbit satellite communication. This invention decouples the original handover problem into two sub-problems: the handover decision of the source satellite and the downlink power allocation of the target satellite. Specifically, the source satellite uses a genetic algorithm to make decisions on user handover timing and target, thereby improving overall handover performance; after the user connects to the target satellite, the latter uses a continuous convex approximation algorithm to jointly optimize the downlink power of both new and existing users, thereby maximizing the system's downlink throughput. This method can effectively improve the handover success rate, reduce handover latency, achieve system load balancing, and significantly improve the downlink throughput of the target satellite after user handover, realizing overall optimization of the handover process in mobile phone direct connection to low-Earth orbit satellite networks.

Description

Technical Field

[0001] This invention relates to the field of low-Earth orbit satellite communication, specifically a collaborative method for handover decision-making and power allocation for mobile phones directly connecting to low-Earth orbit satellite networks. Background Technology

[0002] With the continuous development of 6G mobile communication technology, non-terrestrial networks have become a core research direction in the field of communications. Among them, low-Earth orbit (LEO) satellites, with their low orbital altitude, have short propagation delays and low path losses, making them a key infrastructure for achieving seamless global communication coverage. In particular, direct connection between mobile phones and LEO satellite networks breaks through the dependence of traditional satellite communication on dedicated terminals. Ordinary smartphones can directly establish connections with satellite networks, significantly lowering the barrier to entry for users and enabling satellite communication technology to truly enter the mass market.

[0003] Compared to traditional low-Earth orbit (LEO) satellite communication networks, direct mobile phone connections to LEO satellites operate at even lower altitudes, exhibiting more pronounced high dynamic characteristics. This results in shorter visibility times for users, leading to more frequent handovers. Furthermore, the high-speed movement of LEO satellites causes significant Doppler shift, noticeably degrading the quality of the satellite-to-ground communication link. In scenarios involving direct mobile phone connections to LEO satellites, the antenna gain of ground users is far lower than that of dedicated satellite communication terminals, resulting in weaker received downlink signals. This further strains the link budget and reduces the stability of the satellite-to-ground link. Therefore, the rapidly changing and poor-quality satellite-to-ground link in direct mobile phone connections to LEO satellite networks necessitates optimization of the user handover process to ensure a continuous and stable connection between users and the satellite.

[0004] Current research on low-Earth orbit (LEO) satellite handover mainly focuses on two dimensions: handover decision-making and resource scheduling. Regarding handover decision-making, existing methods typically implement hard handover based solely on a single indicator such as received signal strength. This is a typical threshold-triggered algorithm, simple to implement but easily affected by link fluctuations. Furthermore, it generally fails to consider multi-dimensional factors such as satellite load status, remaining visibility time, and link stability, making it difficult to achieve handover selection oriented towards overall system performance. In terms of resource scheduling, while existing schemes can perform beam management or power allocation, most are based on static assumptions and do not fully consider the dynamics brought about by new user access after handover. In particular, they lack specific designs for how to efficiently redistribute downlink power after handover to balance the service quality for both new and old users with the system's downlink throughput.

[0005] Therefore, it is essential to design handover decisions and power allocation in a coordinated manner: during the handover phase, multi-dimensional system states should be considered to optimize target satellite selection; after user access, the target satellite should redistribute downlink power, taking into account both the service quality for new and old users and the system's downlink throughput. Executing these two phases step-by-step and coordinating the optimization of the user handover process can improve overall handover performance. Summary of the Invention

[0006] To address the challenges of high complexity in handover management schemes, significant overall optimization difficulties, and the inability to accommodate dynamic updates for both new and existing users in the context of mobile phone direct connection to low-Earth orbit satellites, this invention proposes a collaborative method for handover decision-making and power allocation in mobile phone direct connection to low-Earth orbit satellite networks. This method decomposes the handover problem into two sub-problems: source satellite handover decision-making and target satellite downlink power allocation.

[0007] A collaborative method for handover decision-making and power allocation for mobile phones directly connecting to low-Earth orbit satellite networks includes the following steps:

[0008] Step 1: Construct a communication system that includes direct connections between users and mobile phones to low-Earth orbit satellites, initialize system parameters, and construct the comprehensive utility function for each user-satellite pair.

[0009] The communication system includes A mobile phone directly connects to low-orbit satellites and For each ground user, the total system bandwidth is [number] bandwidth. All satellites use the same frequency for multiplexing. The core network or source satellites obtain the current time slot. The set of candidate visible satellites for all users, for each user-satellite pair Calculate the five-dimensional sub-utility, including: the payload of each satellite. Downlink signal strength Satellite elevation angle Remaining visible time Interstellar jumps Normalize each sub-utility to construct a comprehensive utility function;

[0010] The comprehensive utility function is in the form of a weighted summation, and a load penalty mechanism is introduced to suppress overload, specifically expressed as follows:

[0011]

[0012] Jointly Adjusted Time-Varying Weight Vector and load penalty factor This enables the algorithm to adaptively balance multiple optimization objectives under different operating scenarios, thereby meeting diverse switching needs.

[0013] Step 2: Based on the comprehensive utility function of each user-satellite pair, the source satellite uses a genetic algorithm to generate a handover decision scheme for its users;

[0014] The switching decision problem can be represented as an average utility maximization model:

[0015]

[0016] Among them, binary decision variables Indicates user and satellite In the time slot The following connections;

[0017] The above model is subject to four types of constraints:

[0018] Constraint (1): Any user in each time slot It can only connect to one satellite, that is ;

[0019] Constraint (2): Satellite No connection may be made exceeding its carrying capacity. users, i.e. ;

[0020] Constraint (3): A connection is only allowed when the satellite is visible to the user, i.e. ,in As a visual indicator, it is defined as follows:

[0021]

[0022] Constraint (4): For 0-1 integer variables, if A value of 1 indicates that the user is connected to this satellite; otherwise, it indicates that the user is not connected.

[0023] The average utility maximization model described above is solved using a genetic algorithm to obtain the switching allocation decision:

[0024] (1) Randomly generate an initial population, with each individual being a feasible user satellite mapping;

[0025] (2) Calculate individual fitness (total utility minus capacity overload penalty);

[0026] (3) Perform selection, crossover, and mutation operations, and repair infeasible solutions;

[0027] (4) Iterate until convergence and output the optimal switching scheme. ;

[0028] Step 3: After the user completes the handover and connects to the target satellite, the target satellite reallocates downlink power to the old and new users.

[0029] First, the power allocation problem between the target satellite and its users is modeled as follows:

[0030]

[0031] In the above formula, Indicates noise spectral density; Indicates user With satellite Link gain between; Indicates to users Users with the same frequency Compared to Interference gain; This represents the downlink power variable to be optimized. Indicates to users Users with the same frequency The allocated power is subject to the following constraint: the total transmission power of the target satellite is less than its maximum transmission power. ; target satellite to users Distributed power Non-negative.

[0032] For each target satellite Based on its new user set, a continuous convex approximation algorithm is used to solve the power allocation problem:

[0033] (1) Initialize the power allocation vector ;

[0034] (2) Perform a first-order Taylor expansion on the non-convex objective function at the current point to construct a convex approximation subproblem;

[0035] (3) Solve the subproblem using a convex optimization tool (such as CVX) and update the power vector;

[0036] (4) Repeat the iteration until the power change is less than the preset threshold, and output the optimal power allocation strategy. ;

[0037] Step 4: By coordinating the above-mentioned handover decision and power allocation strategies, the handover efficiency between users and mobile phones directly connected to low-Earth orbit satellites and the downlink throughput of the system are jointly improved.

[0038] The beneficial effects of the technical solution of this invention are as follows:

[0039] This method effectively decomposes the joint optimization problem of handover decision and power allocation in mobile phone direct connection to low-Earth orbit satellite networks into two sub-problems that can be solved efficiently, significantly reducing the algorithm complexity. Compared with deep learning methods, this scheme does not require a large amount of training and has a fast convergence speed. Compared with traditional threshold strategies, this scheme significantly improves the handover success rate, system downlink throughput and load balancing through multi-dimensional utility fusion and dynamic power redistribution, making it more suitable for actual deployment scenarios of mobile phone direct connection to low-Earth orbit satellite networks. Attached Figure Description

[0040] Figure 1 This is a schematic diagram of a mobile phone directly connecting to a low-orbit satellite network switching scenario considered in this embodiment of the invention;

[0041] Figure 2 This is a schematic diagram of the process for switching mobile phones to low-Earth orbit satellite networks based on genetic algorithms and continuous convex approximation in an embodiment of the present invention. Detailed Implementation

[0042] To facilitate the understanding of the technical solution of this patent by those skilled in the art, and to make the technical purpose, technical solution and beneficial effects of this patent clearer, and to fully support the scope of protection of the claims, the technical solution of this patent will be further and more detailed below in the form of specific cases.

[0043] To address the bandwidth allocation and user handover issues in direct mobile phone connections to low-Earth orbit (LEO) satellite networks, existing methods either employ simple threshold switching, ignoring multi-dimensional system states, or utilize deep reinforcement learning, which suffers from high training overhead and poor convergence. Therefore, this invention proposes a collaborative method for handover decision-making and power allocation in direct mobile phone connections to LEO satellite networks: the original problem is decomposed into two sub-problems: source satellite handover decision-making and target satellite downlink power allocation. First, considering the handover decision-making problem under a fixed system state, this problem can be modeled as a capacity-constrained generalized allocation model. For a given system state, a genetic algorithm is used to solve for the optimal handover scheme. After user access, a continuous convex approximation algorithm is used to solve the non-convex power allocation problem. Through this two-stage collaboration, both handover efficiency and system downlink throughput are jointly improved.

[0044] This invention relates to direct mobile phone connection to low-Earth orbit satellite networks, such as... Figure 1 As shown, in this network, multiple users directly access the low-Earth orbit (LEO) satellite network via ordinary smartphones, achieving communication connections in areas without terrestrial base station coverage. Due to the high-speed movement of LEO satellites, users need to frequently switch serving satellites to maintain connection continuity. After each switch, the target satellite needs to reallocate downlink power to its updated assigned users to cope with dynamic load changes and co-channel interference. The considered scenarios include... A mobile phone directly connects to low-orbit satellites and For each ground user, the total system bandwidth is [number] bandwidth. Each satellite uses full frequency reuse. Each user can only connect to one visible mobile phone direct connection low-Earth orbit satellite in any time slot, and the satellite has a maximum capacity limit for connected users.

[0045] The present invention proposes a collaborative method for handover decision-making and power allocation for direct mobile phone connection to low-Earth orbit satellite networks, the process of which is as follows: Figure 2 As shown, it includes the following steps:

[0046] Step 1: Construct a communication system that includes direct connections between users and mobile phones to low-Earth orbit satellites, initialize system parameters, and construct the comprehensive utility function for each user-satellite pair.

[0047] (1) Mobile phones can directly connect to low-Earth orbit satellite networks. The index is Low-orbit satellites, and The index is Ground users; This represents the total simulation duration; It can be decomposed into equally spaced time slots, each with a duration of [duration missing]. Its index is In each time slot ,user With satellite The association state is determined by binary variables. Indicates: If A value of 1 indicates that the user is connected to this satellite; otherwise, it indicates that the user is not connected.

[0048] (2) For each user-satellite pair Construct sub-utility functions and merge them into a comprehensive utility function.

[0049] Switching decisions must consider multiple factors, including link status, service continuity, load distribution, and service priorities. Therefore, this solution selects five key indicators: current satellite load... Downlink signal strength Satellite elevation angle Interstellar jumps and remaining viewing time After normalization, sub-utility terms are constructed for each indicator. The final comprehensive utility function is a weighted summation form, and a load penalty mechanism is introduced to suppress overload, specifically expressed as follows:

[0050]

[0051] Among them, the weight vector satisfy and Conditional constraints. Represents the load penalty factor. Jointly adjusted time-varying weight vector. and load penalty factor This allows the algorithm to adaptively balance multiple optimization objectives under different operating scenarios, thereby meeting diverse switching needs.

[0052] Step 2: Based on the comprehensive utility function of each user-satellite pair, the source satellite uses a genetic algorithm to generate a handover decision scheme for its users;

[0053] For the specific architecture of direct mobile phone connection to low-Earth orbit satellites, the fundamental purpose of handover decision optimization is to simultaneously improve received signal strength, reduce handover latency and frequency, and achieve on-board load balancing while ensuring basic user services. This problem can be formalized as a long-run average utility maximization model:

[0054]

[0055] Among them, binary decision variables Indicates user and satellite In the time slot The following connections; This represents the overall utility function for switching decisions.

[0056] The above optimization is subject to four types of constraints:

[0057] Constraint (1): Any user in each time slot It can only connect to one satellite, that is ;

[0058] Constraint (2): Satellite No connection may be made exceeding its carrying capacity. users, i.e. ;

[0059] Constraint (3): A connection is only allowed when the satellite is visible to the user, i.e. ,in As a visual indicator, it is defined as follows:

[0060]

[0061] Constraint (4): It is a variable of integers from 0 to 1.

[0062] The framework is essentially a stochastic optimization problem spanning multiple time slots. Its core objective is to effectively offset the uncertainties caused by dynamic factors such as channel fluctuations and satellite orbital motion during user handover, while maximizing the long-term expected utility of the system.

[0063] For the handover decision-making problem mentioned above, the source satellite uses a genetic algorithm to execute the handover decision for the user. The specific implementation process is as follows:

[0064] (a) Input parameter initialization: Input candidate satellite set Set of users to be switched Maximum number of iterations Population size Crossover probability Probability of mutation Comprehensive utility function weight vector and penalty factor In addition, it is necessary to obtain real-time system status information for each user, including satellite load. Downlink signal strength Angle of elevation Interstellar jumps and remaining visible time .

[0065] (b) Randomly generated Each of the following individuals represents a complete user switching plan; these individuals are grouped into an initial population. .

[0066] (c) For each individual in the population Calculate its fitness value , .in For users The overall utility value after switching to the target satellite This is a penalty imposed for violating capacity constraints.

[0067] (d) Retain the most fit individuals in the current population. Individuals are directly added to the next generation to prevent the optimal solution from being lost.

[0068] (e) After retaining elite individuals, new individuals are generated through genetic operations to supplement the population size: first, a selection strategy is used to select parent individuals from the current population; then, crossover and mutation operations are performed with preset crossover and mutation probabilities to generate offspring individuals.

[0069] (f) Compare the individual with the highest fitness in the current generation with the historical best solution. If the former is better, then update the global best solution. and its fitness .

[0070] (g) If the convergence condition is met (e.g., the optimal fitness does not improve significantly for several consecutive generations, or the maximum number of iterations is reached). If the condition is met, the algorithm terminates; otherwise, return to step (c) to continue iterating.

[0071] (h) When the algorithm ends, output the optimal switching scheme. and its corresponding optimal fitness This scheme is the user handover strategy that the source satellite should implement.

[0072] Step 3: After the user completes the handover and connects to the target satellite, the target satellite reallocates downlink power to the old and new users.

[0073] The core of modeling the downlink power allocation problem for target satellites lies in power constraints and dynamic scenario adaptation. On one hand, the total transmit power of the target satellites cannot exceed the maximum transmit power constraint. On the other hand, to ensure the model's practicality and scenario adaptability, it is necessary to effectively address issues such as changes in user clusters and disruptive topology mutations.

[0074] The core task of the above sub-problems is to optimize the target satellite's transmission power to its user set under limited target satellite transmission power. Power allocation vector The goal is to maximize the total downlink throughput of the system while ensuring user service needs are met. This sub-problem is modeled as follows:

[0075]

[0076] In the above formula, Indicates the total system bandwidth; Indicates noise spectral density; Indicates user With satellite Link gain between; Indicates to users Users with the same frequency Compared to Interference gain; This represents the downlink power variable to be optimized. Indicates to users Users with the same frequency The allocated power amount. Meanwhile, this subproblem is subject to the following constraints: the total launch power of the target satellite is less than... ; target satellite to users Distributed power Non-negative.

[0077] The target satellite performs power optimization on this subproblem based on a continuous convex approximation algorithm. The specific implementation process is as follows:

[0078] (a) Input parameter initialization: Input the current set of service users Maximum total satellite launch power Maximum number of iterations and convergence accuracy Initial power vector It can be set to equal power distribution.

[0079] (b) Set the iteration count variable Then, it enters the main loop until the convergence condition is met or the maximum number of iterations is reached.

[0080] (c) At the current iteration point At this point, for the original non-convex objective function Perform a first-order Taylor expansion to construct its concave lower bound. Then, use this linearized function as the objective function to construct a convex optimization subproblem.

[0081] (d) Solve the above convex problem using standard convex optimization tools to obtain a new power vector. .

[0082] (e) Calculate the Euclidean distance between the old and new power vectors and perform a convergence test. If the following conditions are met... The algorithm is then considered to have converged, the iteration is terminated, and the optimal power allocation is output. Otherwise, proceed to the next step.

[0083] (f) Update the iteration variable, As the starting point for the next iteration, return to step (c) and continue iterating.

[0084] (g) If the number of iterations reaches If the solution still hasn't converged, stop iterating and output the current optimal solution. As the final result.

[0085] (h) When the algorithm ends, output the optimal downlink power allocation vector. This scheme is used to guide target satellites in configuring downlink power for each user.

[0086] Step 4: By coordinating the above-mentioned handover decision and power allocation strategies, the handover efficiency between users and mobile phones directly connected to low-Earth orbit satellites and the downlink throughput of the system are jointly improved.

Claims

1. A collaborative method for handover decision-making and power allocation for mobile phones directly connecting to low-Earth orbit satellite networks, characterized in that, Includes the following steps: Step 1: Construct a communication system that includes direct connections between users and mobile phones to low-Earth orbit satellites, initialize system parameters, and construct the comprehensive utility function for each user-satellite pair; The communication system includes A mobile phone directly connects to low-orbit satellites and For each ground user, the total system bandwidth is [number] bandwidth. All satellites use the same frequency for multiplexing; the source satellite obtains the current time slot. The set of candidate visible satellites for all users, for each user-satellite pair Calculate the five-dimensional sub-utilities, normalize each sub-utility, and construct the comprehensive utility function; Step 2: Based on the comprehensive utility function of each user-satellite pair, the source satellite uses a genetic algorithm to generate a handover decision scheme for its users; The switching decision problem can be represented as an average utility maximization model: Among them, binary decision variables Indicates user and satellite In the time slot The following connections; The overall utility function representing the switching decision; Step 3: After the user completes the handover and connects to the target satellite, the target satellite reallocates downlink power to the old and new users. The power allocation problem between the target satellite and its users is modeled as follows: In the above formula, Indicates noise spectral density; Indicates user With satellite Link gain between; Indicates to users Users with the same frequency Compared to Interference gain; This represents the downlink power variable to be optimized. Indicates to users Users with the same frequency The amount of power allocated; This is the satellite's maximum transmission power; Step 4: By coordinating the above-mentioned handover decision and power allocation strategies, the handover efficiency between users and mobile phones directly connected to low-Earth orbit satellites and the downlink throughput of the system are jointly improved.

2. The collaborative method for handover decision-making and power allocation for direct mobile phone connection to low-Earth orbit satellite networks according to claim 1, characterized in that, The five-dimensional sub-utility includes: the payload status of each satellite. Downlink signal strength Satellite elevation angle Remaining visible time Interstellar jumps .

3. The collaborative method for handover decision-making and power allocation for direct mobile phone connection to low-Earth orbit satellite networks according to claim 2, characterized in that, The comprehensive utility function is in the form of a weighted summation, and a load penalty mechanism is introduced to suppress overload, specifically expressed as follows: Among them, by jointly adjusting the time-varying weight vector and load penalty factor This enables the algorithm to adaptively balance various optimization objectives under different operating scenarios.

4. The collaborative method for handover decision-making and power allocation for mobile phone direct connection to low-Earth orbit satellite networks according to claim 1, characterized in that, The four constraints in the average utility maximization model are as follows: Constraint (1): Any user in each time slot It can only connect to one satellite, that is ; Constraint (2): Satellite No connection may be made exceeding its carrying capacity. users, i.e. ; Constraint (3): A connection is only allowed when the satellite is visible to the user, i.e. ,in As a visual indicator, it is defined as follows: Constraint (4): For 0-1 integer variables, if A value of 1 indicates that the user is connected to this satellite; otherwise, it indicates that the user is not connected.

5. A collaborative method for handover decision-making and power allocation for mobile phone direct connection to low-Earth orbit satellite networks according to claim 1 or 4, characterized in that, The average utility maximization model is solved using a genetic algorithm to obtain the switching allocation decision: (1) Randomly generate an initial population, with each individual being a feasible user satellite mapping; (2) Calculate individual fitness; (3) Perform selection, crossover, and mutation operations, and repair infeasible solutions; (4) Iterate until convergence and output the optimal switching scheme.

6. The collaborative method for handover decision-making and power allocation for direct mobile phone connection to low-Earth orbit satellite networks according to claim 1, characterized in that, The constraint condition for the power allocation problem is as follows: the total transmit power of the target satellite is less than its maximum transmit power. ; target satellite to users Distributed power Non-negative.

7. The collaborative method for handover decision-making and power allocation for mobile phone direct connection to low-Earth orbit satellite networks according to claim 1, characterized in that, For each target satellite, based on its set of new and old users, a continuous convex approximation algorithm is used to solve for power allocation: (1) Initialize the power allocation vector; (2) Perform a first-order Taylor expansion on the non-convex objective function at the current point to construct a convex approximation subproblem; (3) Solve the subproblem using convex optimization tools and update the power vector; (4) Repeat the iteration until the power change is less than the preset threshold, and output the optimal power allocation strategy.

Images