A Vehicle-to-Everything (V2V2I) Routing Method Based on Adaptive Variable Beacon Power

Abstract

This invention discloses a V2V2I routing method for vehicular networks based on adaptive variable beacon power. To improve the link stability between vehicles and their RSUs, the vehicle first collects neighbor beacons to analyze the signal coverage status of the current location, adaptively adjusts the vehicle's current beacon power, and then analyzes available neighbor vehicle nodes through improved Q-learning to find the optimal neighbor node and select it as the next-hop route, thus completing the V2V2I data routing. This invention not only adaptively adjusts beacon power according to different situations, saving vehicle energy consumption, but also achieves good technical results in reducing vehicle selfishness, average vehicle RSU connection time, routing hop count, and link hold-up time when the beacon power variation range is ±30% of the standard power.

Description

Technical Field

[0001] This invention relates to the field of computer technology, and in particular to a vehicle-to-everything (V2V2I) routing method based on adaptive variable beacon power. Background Technology

[0002] Due to its advantages of proximity to users, low transmission latency, and high processing speed, Motorized Access Centers (MECs) have been extensively studied in vehicle-to-everything (V2I) communication in recent years. Vehicles communicate with Roadside Units (RSUs) to offload data to nearby MECs for processing; this is known as V2I communication. However, due to the limited signal coverage of vehicles, in scenarios where the distance between the vehicle and the RSU is significant, communication must rely on intermediate vehicle routing, i.e., V2V2I communication. In V2V2I communication, the most crucial aspect is the V2V link selection problem. Currently, domestic and international scholars have conducted extensive research on V2V routing, with the selection of the next-hop vehicle as the routing node being a primary consideration. Key factors considered include latency, hop count, bandwidth, distance, vehicle selfishness, security, energy, and mobility.

[0003] Existing technical solutions include: 1. Vehicle beacon power is fixed; 2. By default, all vehicles participate in vehicle routing without conditions; 3. The routing algorithm adopts the classic Q-learning algorithm.

[0004] However, existing technical solutions have the following shortcomings: 1. They cannot flexibly adjust beacon power according to different vehicle environments. 2. In reality, they may not meet the needs of individuals who are unwilling to participate in routing for personal reasons. 3. While classic Q-learning can improve V2V link conditions to some extent, it cannot adapt well to the rapid movement of vehicles. Summary of the Invention

[0005] To address the shortcomings of existing technologies, this invention provides a vehicle-to-everything (V2V2I) routing method based on adaptive variable beacon power, which includes the following steps:

[0006] Step 1: Collect beacon information of neighboring vehicles within one hop of signal range;

[0007] Step 2: Analyze whether there are any usable vehicles among the neighboring vehicles;

[0008] Step 3: If there is one or more available vehicles, proceed to step 5; otherwise, continue to step 4.

[0009] Step 4: Increase beacon power according to the adaptive formula and collect available vehicles; if the beacon power of a vehicle is at its maximum value and the vehicle search timeout occurs, trigger event 1 and end the unloading work of the current stage in advance.

[0010] Step 5: Analyze the states of available vehicles and the current vehicle. If the states of available vehicles and the current vehicle are not in state S4, they need to be dynamically adjusted to state S4.

[0011] Step 6: Collect beacon information of available vehicles, and comprehensively consider factors such as vehicle selfishness, vehicle mobility, and link stability to determine the next-hop route based on improved Q-learning.

[0012] Step 7: Offload data according to the selected route.

[0013] Preferably, the state of the vehicle includes:

[0014] Status S1: Vehicles not connected to the RSU can receive beacon information from vehicles V1 that are connected to the RSU, but vehicle V1 cannot receive beacons from V2;

[0015] State S2: Unconnected RSU vehicle V2 does not receive any beacons from neighboring vehicles or RSUs, while connected RSU vehicle V1 can receive beacons from vehicle V2;

[0016] Status S3: Vehicle beacon signals do not overlap with each other;

[0017] Status S4: Normal communication status.

[0018] Preferably, when the vehicle is continuously in an environment with few vehicles or in a traffic jam, power saving events 1 and 2 can be triggered, wherein,

[0019] Event 1: On roads with sparse traffic, if a vehicle is at its maximum beacon power value for an extended period and cannot find any available vehicles, Event 1 will be automatically triggered when the number of vehicles in the neighbor list is less than the threshold for an extended period. This will reduce the vehicle beacon power value to the minimum, thereby reducing unnecessary energy consumption.

[0020] Event 2: When vehicles are in congested areas, a large number of vehicle broadcasts will cause wireless channel congestion; when a vehicle detects that the number of neighbors in its list exceeds the threshold for an extended period of time, Event 2 will be automatically triggered, which means reducing the beacon power while ensuring normal connectivity.

[0021] The present invention has the following beneficial effects:

[0022] Existing V2V routing algorithms have provided solutions for dynamic topology changes caused by vehicle movement, but they do not comprehensively consider vehicle beacon coverage status, variable beacon power control, and vehicle selfishness. To address this, this invention comprehensively considers these influencing factors, analyzes different beacon signal coverage states in detail, and implements adaptive beacon power control by combining on-demand control and trigger events. Finally, an improved Q-learning method selects the optimal next hop from the neighbor list, completing V2V2I data offloading. Attached Figure Description

[0023] Figure 1 This is a V2V2I system framework diagram of the present invention;

[0024] Figure 2 This is a schematic diagram of the beacon broadcasting and data transmission timeline of the present invention;

[0025] Figure 3 This is a schematic diagram illustrating the classification of vehicle signal coverage status in this invention;

[0026] Figure 4 This is a flowchart of the overall beacon power adaptive control of the present invention;

[0027] Figure 5 The figure shows the simulation results of the method of the present invention on the HighD dataset. Detailed Implementation

[0028] To make the objectives, technical solutions, and advantages of this invention clearer, the invention will be further described in detail below with reference to specific embodiments and the accompanying drawings. It should be understood that these descriptions are merely exemplary and not intended to limit the scope of the invention. Furthermore, descriptions of well-known structures and techniques are omitted in the following description to avoid unnecessarily obscuring the concept of the invention.

[0029] In traditional vehicle-to-vehicle-to-roadside unit (V2V2I) routing algorithms, the beacon power of vehicles is fixed, making dynamic adjustment impossible by incorporating event-driven and on-demand operation methods. This results in poor vehicle dynamic adaptability and significant energy waste. To improve the flexibility of beacon power adjustment in dynamic environments, and considering practical factors such as signal coverage status, selfishness, and vehicle motion characteristics, an adaptive variable beacon power-based V2V2I routing algorithm, ABPR, is proposed. Specifically, to improve the link stability between vehicles and RSUs, the vehicle first collects neighbor beacons to analyze the current signal coverage status and adaptively adjusts its beacon power. Then, through improved Q-learning, it analyzes available neighbor vehicle nodes, finds the optimal neighbor node, and selects it as the next-hop route to complete the V2V2I data routing.

[0030] Figure 1 This is a diagram of the V2V2I system framework. The V2V2I system framework mainly consists of two parts: V2V and V2I. V2V is responsible for data transmission between vehicles, while V2I is responsible for data transmission between vehicles and RSUs.

[0031] The present invention provides a vehicle-to-everything (V2V2I) routing method based on adaptive variable beacon power, comprising the following steps:

[0032] Step 1: Collect beacon information of neighboring vehicles within one hop of signal range;

[0033] Step 2: Analyze whether there are any usable vehicles among the neighboring vehicles;

[0034] Step 3: If there is one or more available vehicles, proceed to step 5; otherwise, continue to step 4.

[0035] Step 4: Increase beacon power according to the adaptive formula and collect available vehicles; if the beacon power of a vehicle is at its maximum value and the vehicle search timeout occurs, trigger event 1 and end the unloading work of the current stage in advance.

[0036] Step 5: Analyze the states of available vehicles and the current vehicle. If the states of available vehicles and the current vehicle are not in state S4, they need to be dynamically adjusted to state S4.

[0037] Step 6: Collect beacon information of available vehicles, and comprehensively consider factors such as vehicle selfishness, vehicle mobility, and link stability to determine the next-hop route based on improved Q-learning.

[0038] Step 7: Offload data according to the selected route.

[0039] The vehicle's broadcast beacon parameters include not only the vehicle's current motion characteristics, but also its connectivity status, RSU connectivity status, neighbor vehicle list, and selfishness parameters. The vehicle's broadcast beacon parameters are shown in Table 1.

[0040] Table 1 Vehicle Broadcast Beacon Parameters

[0041]

[0042]

[0043] In addition to exchanging information with each other via broadcast beacons, vehicles also need to exchange beacon information with the RSU. The broadcast beacon parameters of the RSU are shown in Table 2.

[0044] Table 2 RSU Broadcast Beacon Parameters

[0045]

[0046] Specifically, the definition of variable beacon power is explained below. Compared to traditional fixed beacon power, variable beacon power offers greater flexibility in vehicle energy consumption adjustment and search range. When a vehicle is in a suburban area with low vehicle density, its beacon power can be increased to search for more available vehicles. When a vehicle is on a congested road with a large number of available vehicles, its beacon power can be decreased to reduce channel congestion and save energy. Both increases and decreases in vehicle beacon power are based on a base beacon power, and the formula for the range of beacon power variation is defined as follows:

[0047] P adap =P base *Δad

[0048] Δad∈{N,1,2-N},(1>N>0) (1)

[0049] Where P adap P represents the range of varying power. base The value represents the base beacon power, and Δad represents the increment factor. Δad can only be one of three numbers: a maximum value of 2-N, a minimum value of N, or 1. The value of N needs to be determined based on the hardware conditions of the actual communication. When the increment is 1, it means maintaining the standard beacon power.

[0050] Since vehicles need to maintain a stable connection link during data routing, and frequent changes in beacon power can lead to link instability, this paper uses maximum, minimum and standard beacon power values ​​to perform step-by-step changes. This satisfies the vehicle's beacon power value change requirements while reducing the impact of link stability degradation caused by frequent beacon power changes.

[0051] Specifically, regarding vehicle signal coverage under variable beacon power: For vehicles to communicate with each other, their signal coverage must be within the coverage area of ​​the other vehicle. For example... Figure 3 As shown in (a), vehicle V2, which is not connected to the RSU, can receive beacon information from vehicle V1. However, the beacon power of V2 is low, and the signal range cannot cover V1, so normal wireless communication cannot be achieved.

[0052] Figure 3 This is a schematic diagram illustrating the classification of vehicle signal coverage states in this invention. Although it is generally understood that vehicle V2 needs to increase its beacon power to ensure signal coverage of vehicle V1 and achieve normal data offloading, in actual vehicle operation, both vehicle V1 and V2 are independent vehicles. There is no higher-level system to inform them of their current state; the vehicles can only infer their state from the received beacon information. The main vehicle state classifications are as follows:

[0053] (1) State S1: Vehicles not connected to the RSU can receive beacon information from vehicles V1 that are connected to the RSU, but vehicles V1 cannot receive beacons from V2.

[0054] After receiving a beacon message from a neighboring vehicle, the vehicle will... neber It records all information about neighboring vehicles. Figure 3 In state S1 shown in (a), for vehicle V2, V2 can receive the beacon from V1, which contains the neighbor list parameter L of V1. neber Therefore, vehicle V2 extracts the L value from vehicle V1. neber Analysis revealed that vehicle V2 was not in vehicle V1's neighbor list. Therefore, vehicle V2 could determine that it was currently in state S1 with vehicle V1 and needed to actively increase its beacon power. Once vehicle V2's beacon power was increased enough to cover vehicle V1, and V2 received vehicle V1's beacon again, vehicle V2 would be added to its neighbor list, ultimately achieving normal communication.

[0055] For vehicle V1, it only receives beacon information from the RSU, and the neighbor vehicle parameter L in the RSU beacon parameters... car It includes V2 vehicles, enabling normal communication without the need to adjust beacon power.

[0056] (2) State S2: Unconnected RSU vehicle V2 does not receive any beacons from neighboring vehicles and RSUs, while connected RSU vehicle V1 can receive beacons from vehicle V2.

[0057] like Figure 3 As shown in (b), for vehicle V1, although vehicle V1 is able to communicate normally with the RSU, it receives a beacon from vehicle V2 indicating that V2 is not connected to the RSU, and the neighbor list L... neber There is no V1 vehicle in the system. Therefore, the V1 vehicle needs to actively expand the beacon power coverage until the signal range can cover the V2 vehicle and complete normal communication.

[0058] (3) Status S3: Vehicle beacon signals do not overlap with each other.

[0059] like Figure 3 As shown in (c), when neither vehicle can cover the other, the beacon will not change actively because vehicle V1 has already established a connection with the RSU and there are no unconnected vehicles in the neighbor list. Vehicle V2, without an RSU connection and with no available vehicles in its neighbor list, will actively increase its beacon power to search for more vehicles. Once V2's beacon power increases, if the signal can cover vehicle V1, the state will transition to state S2 to resolve the issue.

[0060] If a vehicle's V2 beacon power is at its maximum for a period of time and no neighboring vehicles are detected, it will enter a power-saving state with the minimum beacon power until the next control cycle restarts detection.

[0061] (4) Status S4: Normal communication status.

[0062] For the first three signal coverage states, the final result must be achieved. Figure 3 Only the state of S4 shown in (d) can complete normal V2V2I communication.

[0063] Figure 2 This is a schematic diagram of the beacon broadcasting and data transmission timeline of the present invention. During vehicle travel on the road, the acquisition of beacon information and the routing nodes for data transmission have certain time constraints, and the broadcast times of different vehicles also differ. Therefore, it is necessary to acquire and utilize beacon information promptly. The processing of beacon information and data transmission are performed in two independent parts within the vehicle's control unit.

[0064] In the independent task 1 of beacon broadcasting, the vehicle's own information is broadcast, beacon reception within one hop signal is completed, and the neighbor vehicle information table is updated within each broadcast period T.

[0065] In the independent task 2 for data transmission, the vehicle will perform adaptive beacon power control and route selection based on the information collected from the beacon broadcast in task 1, and finally complete the data transmission according to the selected next route until the next transmission cycle begins.

[0066] Specifically, selfishness based on energy consumption is defined as follows: selfish nodes in a network tend to pursue individual interests and are unwilling to cooperate or share resources, thus negatively impacting network performance and effectiveness. Based on this definition, selfish nodes in a network can be categorized into static and dynamic selfish nodes.

[16] That is, when the vehicle has sufficient remaining resources, it exhibits general cooperation, but when the remaining resources are insufficient, it switches to selfishness. As new energy vehicles gradually become the future trend, the remaining battery power of a vehicle directly determines its driving range. When the vehicle's battery power is too low, other power-consuming functions of the vehicle can be disabled to increase the driving range. This paper uses the remaining battery power of the vehicle as the basis for judging the selfishness of the vehicle, as shown in formula (2).

[0067]

[0068] in, This represents the remaining energy of the vehicle at time i. This represents an energy threshold. When the vehicle's remaining energy falls below this threshold, the vehicle's selfish behavior is triggered, maintaining the beacon power at a minimum. The selfish vehicle no longer actively participates in routing, retaining only its own offloading function, until the vehicle's remaining battery power exceeds the threshold. It will then participate in vehicle data routing again.

[0069] On the other hand, vehicle battery power decreases over time due to driving and data transmission. Many factors influence battery power, such as frequent acceleration and deceleration, excessive application usage, and prolonged high-power signal transmission. Some parameters also vary depending on user habits. In vehicle-to-everything (V2X) communication is the most energy-intensive operation. This paper primarily considers the impact of beacon transmission power on energy consumption. In reality, the relationship between beacon transmission power and energy consumption is more complex, involving a range of other factors such as communication system efficiency, signal propagation environment, and hardware performance. Therefore, to simplify the model, this paper categorizes all other similar influencing factors under basic energy consumption. The simplified formula is defined as follows:

[0070]

[0071] in, P is the total energy consumed in the i-th cycle; i It represents the signal transmission power during the i-th cycle; It represents the basic energy consumption required for the vehicle to operate during the i-th cycle.

[0072] Figure 4 This is a flowchart illustrating the overall adaptive beacon power control of the present invention. For vehicles not connected to an RSU, the beacon power needs to be adjusted as needed to control the signal coverage in three different unconnected states (S1, S2, and S3), with the final state being S4, where signals overlap. For vehicles already connected to an RSU, in situations such as traffic congestion or sparse traffic in suburban areas, the beacon power needs to be controlled individually based on different events.

[0073] Beacon power control on demand: For vehicles not connected to an RSU and vehicles already connected to an RSU, beacon power will be increased according to different situations. The power increase value is defined in formula (1), and the beacon power will be continuously increased until the maximum value is reached. For vehicles not connected to an RSU, since there are no available vehicles or RSUs within the current signal range, the beacon power needs to be actively increased to search for available vehicles, which solves the situation of vehicles not connected to an RSU in states S1 and S3. For vehicles already connected to an RSU, if a neighboring vehicle is detected not connected to an RSU, the beacon power needs to be passively increased, which solves the situation of vehicles already connected to an RSU in state S2.

[0074] Event-triggered beacon power control: In order to further reduce the energy consumption of beacon power, when the vehicle is in an environment with few vehicles or in a traffic jam, power saving events 1 and 2 can be triggered.

[0075] Event 1: On sparsely populated roads, a vehicle, having been at maximum beacon power for an extended period and unable to find any available vehicles, [the situation arises when the vehicle detects its neighbor list L]. neber The quantity is less than the threshold for a long time In this state, event 1 will be automatically triggered, reducing the vehicle beacon to the minimum power value and reducing unnecessary energy consumption.

[0076] Event 2: When a vehicle is in a congested area, there may be many available vehicles within the same signal range. Simultaneously, numerous vehicle broadcasts can also cause wireless channel congestion. Therefore, when a vehicle detects the neighbor list L... neber The number is greater than the threshold for a long time In this state, event 2 will be automatically triggered, which means reducing the beacon power while ensuring a normal connection.

[0077] The improved Q-learning routing algorithm is explained in detail below. Standard Q-learning, as a model-free method in reinforcement learning, involves an agent continuously exploring the unknown environment to maximize cumulative reward, searching for a control strategy through repeated experiments. The agent's learning process can be represented by a triple {S, A, R}, where S represents the state space, containing all possible states the agent can transition to; A represents the action space, containing the set of all possible actions the agent can perform, which are used to transition between states; and R represents the immediate reward, the positive or negative reward the agent receives for transitioning from the current state to the next state after performing an action.

[0078] Q-learning environment definition: In the Q-learning-based V2V2I routing algorithm, the environment is the entire VANET scenario, where the agent is each unloaded data packet in the vehicle, each vehicle is a routing node, and the behavior of a vehicle forwarding data packets represents an action. When a data packet arrives at a connected vehicle, it will receive a corresponding reward value.

[0079] Q-Learning reward definition: The reward value obtained by a vehicle for performing a forwarding action is formally represented as shown in equation (4).

[0080]

[0081] Among them, R i (m,k) represents the current vehicle V i Select vehicle V m The reward value obtained by reaching the k-th RSU as the next-hop routing node. If vehicle V mIf you are already connected to the nth RSU, you receive a reward of 1; otherwise, you receive 0.

[0082] Q-table definition: In Q-learning, the learning process is maintained by a Q-table with rows in the state space and columns in the action space. In this paper, for routing with multiple candidate nodes, the action space is represented by candidate RSUs, and the state space is all neighboring vehicle nodes within the one-hop signal range. The state update formula is shown in equation (5).

[0083]

[0084] Among them, Q i (m,k) represents vehicle V i Select vehicle V m The Q value corresponds to reaching the k-th RSU; α represents the learning rate, α∈[0,1], when α is 0 it means no new information is learned; β represents the discount rate, β∈[0,1], the discount rate determines whether the agent focuses more on past experience or immediate benefits; R i (m,k) represents the instant reward; Indicates vehicle V m The maximum Q value is reached when the k-th RSU is found in the neighbor list. Indicates vehicle V m The neighbor list. Learning rate based on link evaluation function: In traditional Q-learning, the learning rate α is a fixed value. However, in the complex and ever-changing VANET scenario, a fixed learning rate leads to poor algorithm adaptability. Therefore, in order to further improve the adaptability of the learning rate to different scenarios, a dynamic link evaluation function is adopted as the learning rate. The dynamic link evaluation function between vehicle i and vehicle m is defined as follows:

[0085]

[0086] in, It is a weighting factor. It is the influencing factor of vehicle motion characteristics between vehicle i and vehicle m. The more similar the motion characteristics between vehicles, the greater the possibility of maintaining the link for a long time. It is the duration that vehicle m exists in the neighbor list of vehicle i. The longer the duration, the stronger the reliability. This is a factor influencing the remaining battery power of vehicle m. The more fully charged the battery, the weaker the vehicle's selfishness, and the greater the likelihood of continued route forwarding in the future. If the vehicle's remaining battery power is about to fall below the selfishness threshold, the learning rate can be reduced in advance to avoid selecting this vehicle. and The definitions are shown in equations (7) to (9):

[0087]

[0088]

[0089] in, This represents the absolute value of the speed difference between vehicle i and vehicle m. It is the sum of the absolute values ​​of the speed differences of all vehicles in the neighbor vehicle list of vehicle i; Represents the list of neighbors of vehicle i; This indicates the remaining battery power of vehicle m. This represents the sum of the remaining battery power of all vehicles in the neighbor vehicle list of vehicle i.

[0090] Discount rate based on routing node length: In traditional Q-learning, the discount rate β is also a fixed value. According to the characteristics of V2V2I data routing, all vehicle unloading data needs to be routed to the RSU. The more routing nodes there are for vehicle unloading data, the longer the time consumption and the more unstable it becomes. Therefore, this paper sets the discount rate as a dynamic parameter of the routing node length. The discount factor is defined as shown in Equation (10).

[0091]

[0092] in, This indicates the total number of routing nodes that vehicle i will reach after selecting vehicle m as its next-hop routing node. It is the sum of all path nodes reaching the RSU from the list of neighboring vehicles of vehicle i.

[0093] Based on the improved Q-learning algorithm: In summary, the improved Q-learning algorithm is defined as shown in equation (11).

[0094]

[0095] The improved Q-learning takes into account the vehicle's state under different conditions, making it better able to adapt to the fast-moving environment of VANET, and also adapting the vehicle's state to changes over time.

[0096] To verify the reliability of the ABPR algorithm in a more realistic VANET scenario, the experimental dataset used was the HighD dataset of roads near Cologne, Germany. The dataset was compared with the GPSR and RSAR algorithms in terms of average total energy consumption for vehicle unloading, number of selfish vehicles, average percentage of RSU connection time, route hop count, and link lifetime. The number of vehicles on the road was selected as 10, 20, 30, 40, 50, 60, 70, 80, and 90, respectively. Furthermore, the ABPR algorithm performance was verified under different beacon power levels using ±70%, ±50%, and ±30% of the standard power.

[0097] It should be noted that the longest single road segment in the HighD dataset is 420m, and there are no records of continuous vehicles between multiple road segments. However, this experiment requires recording the continuous connection time of vehicles. Therefore, in order to match the data in the dataset, the interval of RSU was shortened from the usual 2000 intervals to 400m, and the standard beacon power coverage radius of the corresponding vehicles was scaled proportionally from 300m to 60m.

[0098] The specific experimental simulation parameters are shown in Table 3.

[0099] Table 3 Simulation parameter settings

[0100]

[0101] Figure 5 The figure shows the simulation results of the method of the present invention on the HighD dataset.

[0102] Vehicle beacon power value: For the ever-developing new energy vehicles, range has always been a key concern for users. Therefore, while ensuring normal vehicle communication, the lower the energy consumption of unrelated vehicle operation, the better the vehicle's range will be. Figure 5 As shown in (a), when the total number of vehicles on the road is greater than 60, the number of vehicles within the range of a single vehicle beacon continuously increases. When the threshold N for event triggering is reached, event 2, the energy-saving mode, is triggered, reducing the beacon power of the current vehicle to the minimum set value, thereby saving vehicle energy consumption. In contrast, the GPSR and RSAR algorithms used in the comparative experiment maintain a fixed beacon throughout the process, operating in standard power consumption mode.

[0103] On the other hand, communication capabilities are also a crucial foundation for vehicles to achieve intelligent driving. To improve communication capabilities, it's necessary to proactively increase beacon power when there are few neighboring vehicles, and to find enough available vehicles to help enhance data offloading capabilities, for example... Figure 5 When the total number of vehicles in (a) is 10, the three different comparison algorithms of ABPR all become the maximum values ​​within their respective ranges, which is used to improve the communication capabilities of the vehicles.

[0104] A comparison of the number of vehicles exhibiting selfishness: The definition of selfishness is based on practical considerations. For example... Figure 5As shown in (b), both the GPSR and RSAR algorithms use fixed beacon power, causing the number of selfish vehicles to increase continuously with the number of vehicles. However, the ABPR algorithm, after the number of vehicles exceeds 60, reduces the average beacon power value of the vehicles, decreasing energy consumption and consequently reducing the number of selfish vehicles. Furthermore, as the number of vehicles increases, more and more vehicles actively reduce their beacon power, further reducing the number of selfish vehicles. Because the GPSR and RSAR algorithms consistently operate at standard energy consumption values, and since the initial energy values ​​of the vehicles follow a normal distribution, the number of selfish vehicles increases with the number of vehicles.

[0105] When the total number of vehicles is 10, the ABPR algorithm increases beacon power to find available vehicles, but this also increases vehicle energy consumption and the number of selfish vehicles. However, overall, the variable beacon power used by ABPR, ranging from ±30%, ±50%, to ±70%, reduces the number of selfish vehicles by 13.79%, 14.94%, and 17.24% compared to the standard beacon power.

[0106] Comparison of RSU connection time: Reducing beacon power can save vehicle energy to some extent, but it increases the risk of vehicles failing to connect to RSUs compared to standard power. To increase routing connection stability while reducing beacon power, the ABPR algorithm uses an improved Q-learning algorithm, with a dynamic learning rate based on the vehicle's motion characteristics. This allows the vehicle to select a node with similar motion characteristics as the next-hop routing node, thus maintaining a relatively stable connection within the coverage area of ​​the lowest beacon power. The variable beacon power used in ABPR, from ±30% and ±50%, improves by 0.63% and 0.2% respectively compared to the fixed value. However, the ±70% parameter actually decreases by 1.75%, indicating that continuously reducing the minimum beacon power value in an attempt to save energy can have counterproductive effects. Figure 5 As shown in (c), the ABPR algorithm, at ±70% beacon power, has a lower average vehicle connection time to RSU than the standard beacon power, regardless of whether there are few vehicles or congestion.

[0107] The reason why the average time spent connecting to the RSU for all vehicles did not reach 100% is that when a vehicle enters a new environment, there is a certain lag in the propagation of beacon information and the establishment of a link connection. Therefore, during the entire driving cycle of a vehicle, there is a period of time spent waiting to establish a connection. When there are only 10 vehicles in total, the standard beacon power cannot cover enough vehicles due to the small number of vehicles, resulting in a smaller percentage of time spent connecting to the RSU.

[0108] Average Route Hop Count Comparison: The average route hop count refers to the number of forwarding vehicles a data packet passes through from the offloading vehicle to the RSU. A higher average route hop count indicates greater instability in the dynamic environment of VANET. For example... Figure 5 As shown in (d), the GPSR routing algorithm, due to its greedy approach, selects only the nearest vehicle as the next-hop route, resulting in an increasing average number of hops per vehicle as the number of vehicles increases. The RSAR algorithm, on the other hand, uses the traditional Q-learning algorithm to select the next route based on the maximum value in the Q-table and continuously learns from neighboring nodes using a fixed learning rate, thus reducing the average number of hops compared to the GPSR algorithm.

[0109] The GPSR algorithm shows a continuous increase in the number of routes as the number of vehicles increases, while the RSAR algorithm shows an initial increase followed by a decrease. This is because initially, there are fewer vehicles and fewer routing nodes. As the total number of vehicles increases, the number of route hops also increases, reaching a stable state when the number of vehicles reaches 40. As the number of vehicles continues to increase, traffic congestion on the road worsens, and vehicle speeds decrease. Most vehicles within the beacon range begin to stabilize. When selecting the next-hop route, the RSAR algorithm excludes some vehicles traveling in the opposite direction from consideration, thus reducing unnecessary hops, and the final average number of route hops begins to decrease.

[0110] However, compared to ABPR, RSAR's fixed learning rate does not consider the dynamic changes in VANET. In experimental environments, vehicles are constantly moving, and the RSUs they choose to unload vary. Therefore, ABPR introduces a dynamic learning rate and discount rate, taking into account vehicle motion characteristics and route hop counts. The learning rate and discount rate are dynamically calculated each time the Q-table is updated, ensuring that data packets select vehicles with more similar motion characteristics and shorter route hop counts to reach the RSU as the next node, thus reducing the average route hop count overall. On the other hand, when beacon power decreases, the number of selectable vehicles further reduces the route hop count. The ABPR algorithm uses three different beacon power levels, from ±30%, ±50%, to ±70%, and the average route hop count decreases as the beacon power decreases. Overall, compared to GPSR and RSAR, the average route hop count is reduced by 55.34%, 56.59%, and 61.10%, and by 21.82%, 24.01%, and 31.91%, respectively.

[0111] Route lifetime comparison: Route lifetime refers to the time a link remains active; the longer the link remains active, the more stable the communication. For example... Figure 5As shown in (e), the average link hold-up time of the GPSR algorithm is only 24.6%, while that of the RSAR algorithm is 33.1%. The average hold-up time of the three different power levels of ABPR is 38.8%, 37.8%, and 35.7% respectively, ranging from ±30%, ±50% to ±70%.

[0112] The GPSR algorithm has a low link hold time because it only selects the next route according to a greedy forwarding strategy. The nearest selected vehicle might be traveling in the opposite direction to the current vehicle or it might be accelerating. Therefore, relying solely on the nearest node to the person selecting the next hop results in a low hold time.

[0113] The RSAR algorithm uses the traditional Q-learning algorithm to select the next-hop route. Since the traditional Q-learning algorithm finds the next hop based on the maximum value in the Q-table, and each vehicle's Q-table is stored within itself, according to the Q-table maintenance rules, the longer a neighboring vehicle has been in the neighbor list, the more stable the vehicle is, the larger its Q-table value, and the more stable the link to the selected next-hop vehicle, resulting in a longer link hold time. The ABPR algorithm improves link hold time by ±30%, ±50%, and ±70% compared to GPSR and RSAR, with overall improvements of 57.65%, 53.71%, and 45.1%, and 17.18%, 14.25%, and 7.85%, respectively.

[0114] Simulation experiments on the Cologne dataset in Germany show that, compared with GPSR and RSAR routing algorithms, the proposed ABPR can not only adaptively adjust beacon power according to different situations to save vehicle energy consumption, but also outperforms the comparative algorithms in terms of vehicle selfishness, average vehicle connection RSU time, routing hop count, and link hold-up time when the beacon power variation range is ±30% of the standard power.

[0115] It should be noted that the specific embodiments described above are exemplary, and those skilled in the art can devise various solutions inspired by the disclosure of this invention. These solutions all fall within the scope of this invention and its protection. Those skilled in the art should understand that this specification and its accompanying drawings are illustrative and not intended to limit the scope of the claims. The scope of protection of this invention is defined by the claims and their equivalents.

Claims

1. A vehicle-to-everything (V2V2I) routing method based on adaptive variable beacon power, characterized in that, Includes the following steps: Step 1: Collect beacon information of neighboring vehicles within one hop of signal range; Step 2: Analyze whether there are any usable vehicles among the neighboring vehicles; Step 3: If there is one or more available vehicles, proceed to step 5; otherwise, continue to step 4. Step 4: Increase beacon power according to the adaptive formula and collect available vehicles; If the vehicle's beacon power is at its maximum and the vehicle search timeout occurs, event 1 is triggered and the current phase of unloading work ends early. Event 1: In sparsely populated roads, if a vehicle is at its maximum beacon power value for a long time and cannot find any available vehicles, that is, when a vehicle detects that the number of neighbors in its list is less than the threshold for a long time, event 1 will be automatically triggered to reduce the vehicle's beacon power value to the minimum value and reduce unnecessary energy consumption. Step 5: Analyze the states of available vehicles and the current vehicle; if the states of available vehicles and the current vehicle are not in state S4, they need to be dynamically adjusted to state S4, where state S4 is the normal communication state; Step 6: Collect beacon information of available vehicles, and determine the next-hop route based on vehicle selfishness, vehicle mobility, and link stability, using improved Q-learning, where improved Q-learning includes: The link dynamic evaluation function is used as the learning rate. The link dynamic evaluation function between vehicle i and vehicle m is defined as follows: in, It is a weighting factor. It is the influencing factor of vehicle motion characteristics between vehicle i and vehicle m. The more similar the motion characteristics between vehicles, the greater the possibility of maintaining the link for a long time. It is the duration that vehicle m exists in the neighbor list of vehicle i. The longer the duration, the stronger the reliability. It is the remaining battery power factor of vehicle m; when the battery power is more sufficient, the vehicle's selfish behavior is weaker, and the greater the possibility of continuing to forward routes in the future. The discount rate is set as a dynamic parameter of the routing node length, and the discount factor is defined as shown in equation (10): in, This indicates the total number of routing nodes that vehicle i will reach after selecting vehicle m as its next-hop routing node. It is the sum of all path nodes leading to the RSU from the list of neighboring vehicles of vehicle i; Step 7: Offload data according to the selected route.

2. The method as described in claim 1, characterized in that, The vehicle's status includes: Status S1: Vehicles not connected to the RSU can receive beacon information from vehicles V1 that are connected to the RSU, but vehicle V1 cannot receive beacons from V2; State S2: Unconnected RSU vehicle V2 does not receive any beacons from neighboring vehicles or RSUs, while connected RSU vehicle V1 can receive beacons from vehicle V2; Status S3: Vehicle beacon signals do not overlap with each other.

3. The method as described in claim 2, characterized in that, The formula for the range of beacon power variation is defined as follows: P adap =P base *Δad Δad∈{N,1,2-N},(1>N>0) (1) Where P adap P represents the range of varying power. base The value represents the base beacon power, and Δad represents the increment factor. Δad can only be one of three numbers: the maximum value of 2-N, the minimum value of N, and 1. The value of N needs to be determined based on the hardware conditions of actual communication. When the increment is 1, it means that the standard beacon power is maintained.

4. The method as described in claim 3, characterized in that, When a vehicle is continuously in an environment with few vehicles or traffic jams, it can trigger power saving events 1 and 2, among which, Event 2: When vehicles are in congested areas, a large number of vehicle broadcasts will cause wireless channel congestion; when a vehicle detects that the number of neighbors in the list exceeds the threshold for a long time, Event 2 will be automatically triggered, that is, the beacon power will be reduced while ensuring normal connection.

Images