Spectrum allocation optimization method for coexistence of satellite flash and Wi-Fi in Internet of Vehicles

Abstract

The invention belongs to the technical field of short-distance wireless communication of the Internet of Vehicles, and relates to a spectrum allocation optimization method for coexistence of satellite flash and Wi-Fi in the Internet of Vehicles, which comprises the following steps: setting a coexistence scene of satellite flash and Wi-Fi in the Internet of Vehicles; the satellite flash master node periodically measures and collects observable statistic information of an MAC layer so as to establish a hidden node model; a sub-channel time allocation proportion, a hidden node correction factor, a Wi-Fi influence penalty term and a satellite flash user priority are considered, and an optimization problem of maximizing the total effective throughput of the satellite flash user is established; an original non-convex DC problem is converted into a convex problem through variable replacement and variable relaxation; decomposing the joint problem into two sub-problems of sub-channel time allocation and power control by adopting a centralized alternating iterative algorithm, and solving the two sub-problems to obtain a spectrum allocation optimization scheme; according to the invention, the interference problem during coexistence of the satellite flash and the Wi-Fi in the Internet of Vehicles can be reduced, and the total effective throughput of the satellite flash user as a high priority is improved at the same time.

Description

Technical Field

[0001] This invention belongs to the field of short-range wireless communication technology for vehicle-to-everything (V2X) networks, and relates to a spectrum allocation optimization method for the coexistence of satellite flash and Wi-Fi in V2X networks. Background Technology

[0002] With the rapid development of intelligent connected vehicles, the in-vehicle communication environment is becoming increasingly complex. The emergence of applications such as intelligent cockpits, in-vehicle entertainment, and autonomous driving assistance systems has led to the parallel operation of multiple short-range wireless communication technologies within vehicles. Among them, StarFlash, as a new generation of short-range wireless communication technology, is gradually becoming a key component of in-vehicle communication systems due to its high reliability and low latency, enabling various new in-vehicle applications. Meanwhile, traditional Wi-Fi technology continues to support high-bandwidth services such as in-vehicle entertainment and mobile terminals. Both types of wireless systems operate on the unlicensed 2.4GHz and 5GHz frequency bands and are highly coupled physically, with their nodes distributed within the enclosed metal vehicle compartment, thus forming a high-density, heterogeneous short-range communication scenario.

[0003] The StarNetwork, centered on vehicle-mounted domain control, employs centralized scheduling and time synchronization mechanisms, making it particularly suitable for real-time applications such as sensor data and control commands. In contrast, the Wi-Fi network, based on Carrier Sense Multiple Access with Collision Avoidance (CSMA / CA), is better suited for bursty high-bandwidth services. Due to the enclosed space, dense node density, and significant signal obstruction within the vehicle, the signal energy detection thresholds of StarNet and Wi-Fi differ, and their detection mechanisms are incompatible, making cross-standard hidden node phenomena highly likely. When a StarNet node transmits directly according to schedule and Wi-Fi fails to detect its activity, it appears as a hidden node from Wi-Fi's perspective; conversely, it does not. When the traditional CSMA / CA mechanism struggles to effectively detect hidden nodes, frequent collisions with hidden nodes by StarNet users lead to increased ACK loss rates and signal decoding failure rates, significantly reducing the overall effective throughput of StarNet users.

[0004] In summary, for the communication environment where vehicle-to-everything (V2X) communication and Wi-Fi coexist, it is urgent to reduce the interference caused by hidden nodes, and to improve the total effective throughput of V2X users as a high priority while ensuring the basic in-vehicle service requirements of Wi-Fi. Summary of the Invention

[0005] To address the aforementioned issues, this invention proposes a spectrum allocation optimization method for the coexistence of satellite flash and Wi-Fi in vehicle-to-everything (V2X) networks, comprising the following steps:

[0006] S1. Set up a scenario where in-vehicle starlight and Wi-Fi can coexist;

[0007] S2. The master node periodically measures and collects observable statistics from the MAC layer to establish a hidden node model;

[0008] S3. Design a channel allocation scheme and establish an optimization problem to maximize the total effective throughput of star flash users;

[0009] S4. Transform the original non-convex DC problem into a convex problem by using variable substitution and variable relaxation;

[0010] S5. A centralized alternating iterative algorithm is used to decompose the joint problem into two sub-problems: sub-channel time allocation and power control, and solve them to obtain an optimized spectrum allocation scheme.

[0011] Preferably, step S2 specifically includes:

[0012] S21. The master node of the starburst pattern performs coordination and control functions, and each slave node compiles its own statistics within an observation window. , , The system periodically reports to the master node on the control channel. After summarizing the data, the master node calculates the probability of hidden nodes in the system according to the formula, generates a correction factor, and updates the spectrum allocation and power control strategy accordingly. This mechanism maintains the architectural advantages of centralized scheduling of star-flash systems and achieves adaptive optimization for Wi-Fi coexistence interference.

[0013] S22. The probability of hidden nodes can be derived from the observable statistics of the MAC layer. Let's define nodes within an observation window. The probability of a hidden node is:

[0014]

[0015] in For StarShine master node and StarShine slave node (i.e., Star Flash users) The probability of successfully receiving an ACK frame. For the probability of node transmission attempts, To exclude the correction factor for the proportion of its own launch time, i.e., idle time, The observed channel idle probability. This indicates that the base station is in one transmission cycle The number of frames successfully sent within the time limit. This indicates the number of times the base station detected the channel as busy. The first term represents the total number of detections, i.e., the total number of energy detection windows; the second term represents the success rate that the link should have under the theoretical condition of no hidden nodes. The larger the value, the better for the user equipment. More likely to be affected by nearby undetectable devices;

[0016] S23. Assume there is a certain tolerance threshold for hidden nodes. ,Will As a hidden node correction factor, it is used to adjust the equivalent throughput of unlicensed frequency bands, characterizing the confidence of the node under interference perception, thereby reducing unfair interference caused by hidden nodes. When the threshold is exceeded, the link is temporarily silenced in the unlicensed frequency band to avoid disrupting Wi-Fi. The correction factor can be expressed as:

[0017]

[0018] Preferably, step S3 specifically includes:

[0019] S31. Star Flash User In unlicensed frequency band sub-channels Signal-to-interference-to-noise ratio :

[0020]

[0021] in, , For Star Flash users In sub-channel Transmit power and channel gain; For Wi-Fi users For Starflash user equipment Interference; The noise power spectral density; Sub-channel Bandwidth; Because the internal links of the StarSpark system are scheduled by the master node and strictly enforce time orthogonality, there are no same-layer interference items within a single StarSpark network;

[0022] S32. The StarScan master node periodically performs energy detection to assess the idle ratio of unlicensed frequency band channels, and determines the optimal time allocation based on the signal quality of each link and the probability of hidden nodes, thereby realizing the allocation of unlicensed frequency band channel resources for StarScan users. In sub-channel throughput for:

[0023]

[0024] in, For Star Flash users In sub-channel The time allocation is centrally distributed by the master node based on real-time channel sensing information and interference status.

[0025] S33. Considering the weight correction of hidden nodes, the effective throughput is:

[0026]

[0027] in, Correction factor for hidden nodes;

[0028] S34. Therefore, Star Flash users The total throughput achieved on the unlicensed frequency band is:

[0029]

[0030] S35. Introduction Considering actual Xingshan users The impact of surrounding hidden nodes on system resource allocation can not only avoid interference between satellite flash and Wi-Fi APs, but also fully utilize unlicensed frequency bands, further optimize spectrum allocation between frequency bands, and improve the throughput of satellite flash users. The optimization problem is to maximize the total effective throughput of satellite flash users within the frequency band by allocating sub-channel time and controlling power in the unlicensed frequency band while ensuring service quality, and to reflect the in-vehicle service requirements of satellite flash priority and Wi-Fi backup, by introducing weights. Construct a weighted optimal solution:

[0031]

[0032] in, , , For Star Flash users Effective throughput, in sub-channels The objective function considers the signal-to-noise ratio and the time allocated to unlicensed frequency bands. The first part of the objective function is the primary optimization objective: the weighted total throughput of the Wi-Fi user equipment; the second part is the Wi-Fi impact penalty term, representing the impact of Wi-Fi on the channel. The energy impact of the transmission on Wi-Fi devices is calculated by significantly penalizing the target value of the starburst when the interference received by Wi-Fi exceeds a threshold, by averaging the square of the excess. , For Star Flash users In sub-channel Effective throughput and minimum throughput that meets QoS requirements; For Star Flash users In sub-channel The percentage of time allocated above; This is the penalty coefficient; Channel gain for Wi-Fi users on unlicensed sub-channels; It is a StarFlash user in an unlicensed subchannel The maximum percentage of time that can be used; , For Star Flash users Transmit power and maximum transmit power on unlicensed channels; For StarShine User Equipment In sub-channel On the transmit power; For Star Flash users To Wi-Fi users Channel gain between; For Wi-Fi users For Star Flash users The maximum tolerable interference threshold; constraint C1 ensures fair coexistence of satellite flash and Wi-Fi users in unlicensed frequency bands; constraint C2 ensures that at most one satellite flash user transmits on the channel at any given time, and that the transmission time meets the boundary conditions; constraint C3 ensures that the transmit power does not exceed the master node's allowable range; constraint C4 ensures the minimum reliable transmission rate of the link; weighting coefficients. The priority of StarFlash users is reflected and not treated as an optimization variable. The weight of each user can be preset to a reasonable range according to the business priority, with a larger weight for key control links, a medium weight for ordinary communication links, and a lower weight for low-priority services.

[0033] The beneficial effects of this invention are as follows: This invention comprehensively considers the potential risk of hidden nodes in heterogeneous systems of StarFlash and Wi-Fi, designs a hidden node estimation mechanism, and introduces the probability of hidden nodes into the throughput model as a correction factor. The effective throughput of StarFlash users is used as the optimization objective, and priority weights are configured to achieve differentiated service and fair allocation. Simultaneously, a penalty term based on time-averaged interference and protective constraints with slack variables are introduced into the objective, ensuring an effective trade-off between improving the total effective throughput of StarFlash users and suppressing Wi-Fi interference in the subsequent solution of the two sub-problems. Attached Figure Description

[0034] Figure 1 This is a model diagram of a heterogeneous network communication system in a connected vehicle where satellite flash and Wi-Fi coexist in an embodiment of the present invention;

[0035] Figure 2 This is a flowchart illustrating the steps of an embodiment of the present invention; Detailed Implementation

[0036] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0037] In the embodiments provided by this invention, a heterogeneous network communication model for in-vehicle networking where satellite flash and Wi-Fi coexist is constructed, such as... Figure 1 As shown, in the 5GHz unlicensed frequency band, the scenario includes satellite flash users, Wi-Fi users, and multiple Wi-Fi access points (APs). Wi-Fi users access the channel using the IEEE 802.11 Distributed CSMA / CA protocol, while satellite flash users use a time-synchronization (TDMA) mechanism based on centralized scheduling and control for channel-aware access. In the system model, it is assumed that satellite flash users are high-priority services that continuously compete for the channel, and the usage time of Wi-Fi users in the unlicensed frequency band needs to be limited to ensure the priority of satellite flash users.

[0038] like Figure 2 As shown, this invention provides an embodiment of a spectrum allocation optimization method for the coexistence of satellite flash and Wi-Fi in vehicle-to-everything (V2X) networks, specifically including the following steps:

[0039] S1. Set up a scenario where in-vehicle starlight and Wi-Fi can coexist;

[0040] S2. The master node periodically measures and collects information to establish a hidden node model;

[0041] S3. Design a channel allocation scheme and establish an optimization problem to maximize the total effective throughput of star-flash users;

[0042] S4. Transform the original non-convex DC problem into a convex problem by using variable substitution and variable relaxation;

[0043] S5. A centralized alternating iterative algorithm is used to decompose the joint problem into two sub-problems: sub-channel time allocation and power control, and solve them to obtain an optimized spectrum allocation scheme.

[0044] Preferably, in step S1, initializing the scenario of coexistence of vehicle-to-everything (V2X) in-vehicle Wi-Fi and satellite flashing specifically includes:

[0045] Define the StarShine user set as Wi-Fi user set is In the 5GHz unlicensed band, stroboscopic users and Wi-Fi users share the channel. To facilitate multi-user concurrent modeling, it is divided into M unlicensed sub-channels with uniform bandwidth, and the set of sub-channels is defined as follows: Each bandwidth is To avoid strong interference between different star-flash links, it is assumed that the unlicensed frequency band... The sub-channels are orthogonal to each other, and different star links will share different unlicensed sub-channels.

[0046] Preferably, step S2 specifically includes:

[0047] The probability of a hidden node occurring can be determined by the node transmission attempt probability. ACK success rate and channel idle probability It is deduced that the three required statistical quantities can be obtained from the link layer or physical layer of their respective protocols. For Wi-Fi nodes, these three quantities are derived from the transmission attempt count, ACK feedback statistics, and CCA energy detection results, respectively. For star-flash nodes, an equivalent hidden node model can be designed. It can be determined by the scheduling table. Statistics can be obtained from feedback from the link. This is obtained from the energy detection signal of the master node. All of the above statistics can be calculated within a fixed time window and fed back to the central controller, thereby realizing real-time estimation of the probability of hidden nodes and optimization of spectrum sharing.

[0048] The master node of the starburst pattern undertakes coordination and control functions. Since the formation of hidden nodes is closely related to the overall spatial topology and power differences of the system, a single terminal cannot independently perceive the global hidden state. Therefore, this study adopts a centralized aggregation and computation mechanism: each slave node statistically analyzes its own data within an observation window. , , The system periodically reports these data to the master node via the control channel. The master node then aggregates the data, calculates the probability of hidden nodes in the system using a formula, generates a correction factor, and updates the spectrum allocation and power control strategies accordingly. This mechanism maintains the architectural advantages of centralized star-flash scheduling while also achieving adaptive optimization to Wi-Fi coexistence interference.

[0049] Therefore, the probability of hidden nodes can be derived from the observable statistics of the MAC layer. Let's define nodes within an observation window. The probability of a hidden node is:

[0050]

[0051] in For StarShine master node and StarShine slave node (i.e., Star Flash users) The probability of successfully receiving an ACK frame. For the probability of node transmission attempts, To exclude the correction factor for the proportion of its own launch time, i.e., idle time, The observed channel idle probability. This indicates that the base station is in one transmission cycle The number of frames successfully sent within the time limit. This indicates the number of times the base station detected the channel as busy. The first term represents the total number of detections, i.e., the total number of energy detection windows. The second term represents the success rate that the link should have under the theoretical condition of no hidden nodes. The larger the value, the better for the user equipment. It is more likely to be affected by nearby undetectable devices.

[0052] Assuming there is a certain tolerance threshold for hidden nodes. ,Will As a hidden node correction factor, it is used to adjust the equivalent throughput of unlicensed frequency bands, characterizing the confidence of the node under interference perception, thereby reducing unfair interference caused by hidden nodes. When the threshold is exceeded, the link is temporarily silenced in the unlicensed frequency band to avoid disrupting Wi-Fi. The correction factor can be expressed as:

[0053]

[0054] Preferably, step S3 specifically includes:

[0055] Star Flash users In unlicensed frequency band sub-channels Signal-to-interference-to-noise ratio :

[0056]

[0057] in, , For Star Flash users In sub-channel Transmit power and channel gain; For Wi-Fi users For Starflash user equipment Interference can usually be estimated by the measured mean channel energy, or set to a constant value; For noise power spectral density, Sub-channel Bandwidth. Because the internal links of the StarSpark system are scheduled by the master node and strictly adhere to time orthogonality, there are no intra-layer interference items within a single StarSpark network.

[0058] The master node periodically performs energy detection to assess the idle ratio of unlicensed frequency band channels and determines the optimal time duty cycle configuration by combining the signal quality of each link with the probability of hidden nodes, thereby achieving dynamic channel resource allocation in the unlicensed frequency band. When a user's probability of being a hidden node increases, their transmission efficiency per unit time decreases, and the master node will correspondingly reduce their time allocation to avoid resource waste; conversely, links with fewer hidden nodes and higher reliability will receive a larger time allocation.

[0059] Star Flash users In sub-channel throughput for:

[0060]

[0061] in, For Star Flash users In sub-channel The allocated time percentage is centrally distributed by the master node based on real-time channel awareness information and interference status. It's important to note that this time percentage only describes the time-domain resource allocation within the StarSpark system on that sub-channel. Wi-Fi is an independent distributed system and does not follow StarSpark's scheduling commands. Its nodes randomly access channels in unlicensed frequency bands according to their own CCA / LBT mechanisms. Therefore, when hidden nodes exist or StarSpark fails to detect Wi-Fi, Wi-Fi may independently transmit at the same time and frequency band. Thus, when calculating the throughput of StarSpark users, cross-system interference from Wi-Fi to StarSpark must still be considered.

[0062] The correction factor reflects the effective transmission reliability of the link in a specific frequency band, enabling the time allocation strategy to adaptively adjust according to channel interference conditions, achieving efficient utilization of spectrum resources and fairness in coexistence. Considering the hidden node weight correction, the effective throughput is:

[0063]

[0064] in, This is a correction factor for hidden nodes.

[0065] Therefore, Star Flash users The total throughput achieved on the unlicensed frequency band is:

[0066]

[0067] Introduction Considering actual Xingshan users The impact of surrounding hidden nodes on system resource allocation can not only avoid interference between satellite flash and Wi-Fi APs, but also fully utilize unlicensed frequency bands, further optimizing spectrum allocation between frequency bands and improving satellite flash user throughput. The optimization problem is to maximize the throughput of all satellite flash users within the frequency band while ensuring service quality, through unlicensed frequency band sub-channel time allocation and power control, reflecting the in-vehicle service requirements of satellite flash priority and Wi-Fi backup, and introducing weights. Construct a weighted optimal solution.

[0068]

[0069] in, , , For Star Flash users Effective throughput, in sub-channels The objective function considers the signal-to-noise ratio and the time allocated to unlicensed frequency bands. The first part of the objective function is the primary optimization objective: the weighted total throughput of the Wi-Fi user equipment; the second part is the Wi-Fi impact penalty term, representing the impact of Wi-Fi on the channel. The energy impact of the transmission on Wi-Fi devices is calculated by significantly penalizing the target value of the starburst when the interference received by Wi-Fi exceeds a threshold, by averaging the square of the excess. , For Star Flash users In sub-channel Effective throughput and minimum throughput that meets QoS requirements; For Star Flash users In sub-channel The percentage of time allocated above; This is the penalty coefficient; Channel gain for Wi-Fi users on unlicensed sub-channels; It is a StarFlash user in an unlicensed subchannel The maximum percentage of time that can be used; , For Star Flash users Transmit power and maximum transmit power on unlicensed channels; For StarShine User Equipment In sub-channel On the transmit power; For Star Flash users To Wi-Fi users Channel gain between; For Wi-Fi users For Star Flash users The maximum tolerable interference threshold; constraint (1) ensures fair coexistence of Starflash and Wi-Fi users in unlicensed frequency bands; constraint (2) ensures that at most one Starflash user transmits on the channel at the same time, and the transmission time meets the boundary conditions; constraint (3) ensures that the transmit power does not exceed the allowable range of the master node; constraint (4) ensures the minimum reliable transmission rate of the link; weighting coefficient The priority of StarFlash users is reflected and not treated as an optimization variable. The weight of each user can be preset to a reasonable range according to the business priority, with a larger weight for key control links, a medium weight for ordinary communication links, and a lower weight for low-priority services.

[0070] Preferably, step S4 specifically includes:

[0071] Obviously, due to In product form, For non-convex structures, the penalty term is related to... If the function is convex, then OP1 is a non-convex difference problem. A centralized alternating solution scheme is adopted, where the master node periodically collects channel and interference information and then jointly optimizes the two sub-problems of sub-channel time allocation and power control. To solve OP1, a new variable is defined. To avoid bilinearity. For each Introducing slack variables Make the penalty term equal to Then OP1 is equivalent to OP2:

[0072]

[0073] Then, SCA is used to linearize the penalty term, transforming each subproblem into a convex optimization problem.

[0074] Preferably, step S5 specifically includes:

[0075] S51. Initialization and calculate Set tolerance With maximum number of iterations ,cycle ;

[0076] S52. Assume linearization penalty weights ;

[0077] S53. At the current iteration point Using first-order Taylor to define convex functions The linear lower bound is replaced by constant term The optimization is negligible, so removing the penalty term from the objective function makes it approximately a linear function. Thus, the objective function is bounded to be a convex function, and maximizing the concave function is equivalent to minimizing its negative value, which is convex and solvable.

[0078] S54. For the sub-channel time allocation subproblem, fix the current iteration point. In solving about The convex problem is obtained ;

[0079] S55. For the power control subproblem, fixed Solve for the optimal sub-solution ;

[0080] S56. Update linearization points Return to linearization and solution in S52 until the subproblem converges;

[0081] S57. If the target value ,or Then stop; the former indicates convergence to a stationary point, and the latter indicates termination due to the iteration limit, requiring further adjustment of parameters such as the maximum number of iterations; otherwise... Return to S52 and continue the iterative solution.

[0082] Preferably, the above solution to the two subproblems yields the optimal solution. and Value, i.e., spectrum allocation scheme.

[0083] Although embodiments of the invention have been shown and described, it will be understood by those skilled in the art that various changes, modifications, substitutions and alterations can be made to these embodiments without departing from the principles and spirit of the invention, the scope of which is defined by the appended claims and their equivalents.

Claims

1. A method for spectrum allocation optimization for coexistence of starlink and Wi-Fi in Internet of Vehicles, characterized in that, The method comprises the following steps: S1. setting a V2X in-vehicle star flash and Wi-Fi coexistence scene; S2. periodically measuring and collecting MAC observable statistical quantity information by a master node, thereby establishing a hidden node model; S3. designing a channel allocation scheme, and establishing an optimization problem of maximizing the total effective throughput of star flash users; S4. converting the original non-convex DC optimization problem into a convex problem by using variable substitution and variable relaxation; S5. adopting a centralized alternating iteration algorithm to decompose the joint problem into two subproblems of time allocation and power control for solving, and obtaining a spectrum allocation optimization scheme. 2.The method of claim 1, wherein, The setting of the in-vehicle starlink and Wi-Fi coexistence scene includes: S starlink users, W Wi-Fi users, and a plurality of Wi-Fi wireless access points (APs). The starlink user set is defined as , the Wi-Fi user set is defined as , the starlink users and the Wi-Fi users share a channel in the 5GHz unlicensed frequency band, in order to facilitate multi-user concurrent modeling, the channel is divided into M uniform bandwidth unlicensed sub-channels, and the sub-channel set is defined as , each bandwidth is , in order to avoid strong interference between different starlink links, it is assumed that the M sub-channels of the unlicensed frequency band are mutually orthogonal, and different starlink links share different unlicensed sub-channels. 3.The method of claim 1, wherein, The step S2 specifically comprises: S21. The star flash master node assumes the coordination control function, and each slave node counts its own confirmation information (ACK, Acknowledgement) success rate in an observation window , channel idle probability , node transmission attempt probability , and periodically reports to the master node on the control channel. The master node calculates the system hidden node probability according to the formula after summarizing, generates a correction factor, and updates the spectrum allocation and power control strategy accordingly. This mechanism not only maintains the architectural advantages of star flash centralized scheduling, but also realizes adaptive optimization of Wi-Fi coexistence interference. S22. The probability of hidden nodes can be derived from the observable statistics of the MAC layer. Let's define nodes within an observation window. The probability of a hidden node is: ,in For StarShine master node and StarShine slave node (i.e., Star Flash users) The probability of successfully receiving an ACK frame. For the probability of node transmission attempts, To exclude the correction factor for the proportion of its own launch time, i.e., idle time, The observed channel idle probability. This indicates that the base station is in one transmission cycle The number of frames successfully sent within the time limit. This indicates the number of times the base station detected the channel as busy. The first term represents the total number of detections, i.e., the total number of energy detection windows. The second term represents the success rate that the link should have under the theoretical condition of no hidden nodes. The larger the value, the better for the user equipment. More likely to be affected by nearby undetectable devices; S23. Assume there is a certain tolerance threshold for hidden nodes. ,Will As a hidden node correction factor, it is used to adjust the equivalent throughput of unlicensed frequency bands, characterizing the confidence of the node under interference perception, thereby reducing unfair interference caused by hidden nodes. When the threshold is exceeded, the link is temporarily silenced in the unlicensed frequency band to avoid disrupting Wi-Fi. The correction factor can be expressed as: .

4. The spectrum allocation optimization method for coexistence of starburst and Wi-Fi in the Internet of Vehicles according to claim 1, characterized in that, The step S3 specifically comprises: S31. Star Flash User In unlicensed frequency band sub-channels Signal-to-interference-to-noise ratio ,in, , For Star Flash users In sub-channel Transmit power and channel gain; For Wi-Fi users For Starflash user equipment Interference can usually be estimated from the measured mean channel energy. For noise power spectral density, Sub-channel Bandwidth: Because the internal links of the StarSpark system are scheduled by the master node and strictly enforce time orthogonality, there are no interference items at the same layer within a single StarSpark network. S32. periodically performing energy detection by the star flash master node to evaluate the idle proportion of the unlicensed frequency band channel, and combining the signal quality and hidden node probability of each link to determine the optimal time proportion configuration, thereby realizing dynamic channel resource allocation of the unlicensed frequency band; when the hidden node probability of a user increases, the transmission efficiency per unit time of the user decreases, and the master node correspondingly reduces the time proportion of the user to avoid resource waste; on the contrary, the link with less hidden nodes and higher reliability will obtain more time allocation proportion; S33. StarFlash users The throughput on the subchannel is: wherein, is the time ratio allocated to StarFlash users on the subchannel , which is centrally allocated by the master node according to real-time channel sensing information and interference status; it should be noted that this time ratio only describes the time domain resource allocation of StarFlash system on this subchannel, Wi-Fi is an independent distributed system and does not follow the scheduling command of StarFlash, and its nodes will randomly access the channel according to the Listen-Before-Talk (LBT) mechanism in the unlicensed frequency band, so when there are hidden nodes or StarFlash does not detect Wi-Fi, Wi-Fi may independently transmit in the same time and same frequency band, therefore, the cross-system interference of Wi-Fi on StarFlash still needs to be considered when calculating the throughput of StarFlash users.​ S34. The correction factor reflects the effective transmission reliability of the link on a specific frequency band, so that the time allocation strategy can be adaptively adjusted with the channel interference state, realizing efficient utilization of spectrum resources and coexistence fairness; considering the hidden node weight correction, the effective throughput is: wherein, is the hidden node correction factor; S35. Therefore StarFlash users The total effective throughput achieved on the unlicensed band is: ; S36. Introducing Considering the actual star flash user The influence of the surrounding hidden nodes on the allocation of system resources not only avoids the interference between star flash and Wi-Fi AP, but also fully utilizes the unlicensed frequency band, further optimizes the spectrum allocation between frequency bands, improves the star flash user throughput, and optimizes the problem. Under the condition of ensuring service quality, the total effective throughput of star flash users in the frequency band is maximized through time allocation and power allocation of unlicensed frequency band subchannels, and the star flash priority and Wi-Fi bottom line of in-vehicle business demand are reflected. Weight is introduced , and the weighted optimal is constructed: wherein, , , are the total effective throughput of the Starlink users, the signal-to-noise ratio on the subchannel and the time ratio allocated to the Starlink users in the unlicensed frequency band, respectively; the first part of the objective function is the main optimization target: the weighted total throughput of the Starlink user equipment; the second part is the Wi-Fi impact penalty term, which represents the energy impact of the transmission of the Starlink user equipment on the Wi-Fi equipment on the channel , and when the interference received by the Wi-Fi exceeds the threshold, the target value of the Starlink is greatly punished in the square of the exceeding part; , are the effective throughput of the Starlink users on the subchannel and the minimum throughput that meets the quality of service (QoS) requirement, respectively; is the time ratio allocated to the Starlink users on the subchannel ; is the penalty coefficient; is the channel gain of the Wi-Fi user on the unlicensed subchannel; is the maximum time ratio that the Starlink user can use on the unlicensed subchannel ; , are the transmission power and the maximum transmission power of the Starlink users on the unlicensed channel, respectively; is the transmission power of the Starlink user equipment on the subchannel ; is the channel gain between the Starlink users and the Wi-Fi users ; is the maximum interference threshold that the Wi-Fi users can tolerate to the Starlink users ; constraint C1 guarantees the fair coexistence of the Starlink and Wi-Fi users in the unlicensed frequency band transmission; constraint C2 ensures that at most one Starlink user transmits on the channel at the same time, and the transmission time meets the boundary condition; constraint C3 ensures that the transmission power does not exceed the allowed range of the master node; constraint C4 ensures the minimum reliable transmission rate of the link; the weight coefficient reflects the priority of the Starlink users and is not processed as an optimization variable, and the weights of each user can be preset to a reasonable range according to the business priority, wherein the key control link takes a larger weight, the ordinary communication link takes a medium weight, and the low-priority business takes a lower weight. 5.The method of claim 1, wherein, The step S4 specifically comprises: S41. Since is non-convex, the penalty term is non-convex, the penalty term is convex, OP1 is a non-convex difference-of-convex problem; a centralized alternating solution scheme is adopted, in which the master node periodically collects the channel and interference information and jointly optimizes the two sub-problems of time allocation and power allocation; to solve OP1, a new variable is defined to avoid the bilinear term; for each a relaxation variable is introduced to make the penalty term equal to , OP1 is equivalent to OP2: Then, the SCA (Successive Convex Approximation) is used for linearization of the penalty term, and each subproblem is converted into a convex optimization problem. 6.The method of claim 1, wherein, The step S5 specifically comprises: S51. initialize and compute , set tolerance and maximum number of iterations , loop ; S52. Assume linearized penalty weight ; S53. At the current iteration point , the linear lower bound of the convex function is replaced by the first order Taylor expansion , the constant term is neglected in optimization, thus removed from the objective, the penalty term is approximated as a linear function , thus the objective function is upper bounded by a convex function, maximizing a concave function is equivalent to minimizing its negative, which is convex solvable; S54. For the subchannel time allocation subproblem, fix the current iteration point In solving the convex subproblem with respect to one obtains ; S55. For the power allocation sub-problem, fix , solve for the optimal sub-solution ; S56. Update linearization point Return to S52 for linearization and solution until subproblem converges. S57. If the target value or stop, the former indicates that it has converged to a stationary point, and the latter indicates that it is terminated due to the upper limit of iteration, and further adjustment of parameters such as the maximum number of iterations is required, otherwise return to S52 to continue the alternating iteration solution; S58. The above solving of the two sub-problems obtains the optimal and value, i.e. the spectrum allocation scheme.

Images