An energy-saving optimization method and device for realizing opportunistic routing in vehicle-to-everything (V2X) networks

Abstract

This invention discloses an energy-saving optimization method for implementing opportunistic routing in vehicle-to-everything (V2X) networks, comprising: S1, formalizing the vehicle-to-vehicle scenario, link model, and energy consumption model of vehicle relay node sensors; S2, starting data forwarding from the highest priority relay node and estimating the actual number of data packets forwarded by the link; S3, defining the forwarding set of transmitting vehicles, optimizing the expected travel distance, expected energy consumption, and optimized expected energy consumption and expected distance for each packet; S4, normalizing the distance from the source node to the relay node, the expected number of transmissions, and the energy consumption of vehicle relay node sensors to obtain a communication efficiency cost index, and obtaining a path that meets a specified metric based on the communication efficiency cost index; S5, using the HEOR algorithm to obtain the optimal path selection. This invention also discloses an energy-saving optimization device for implementing opportunistic routing in V2X networks. This invention surpasses existing energy efficiency solutions in terms of throughput, data packet transmission rate, and energy efficiency, and can be widely applied in the field of wireless communication.

Description

Technical Field

[0001] This invention relates to the field of wireless communication, and in particular to an energy-saving optimization method and apparatus for realizing opportunistic routing in vehicle-to-everything (V2X) networks. Background Technology

[0002] In recent years, with the continuous expansion of the electric vehicle market and the rising sales of electric vehicles, the scarcity of charging stations often leads to longer waiting times and difficulty in obtaining available charging stations when needed. Therefore, energy efficiency is one of the important issues to consider in vehicle-to-everything (V2X) routing design. However, due to the limited number of charging facilities, electric vehicles often have to wait a long time to charge or cannot find available charging stations. The situation is even worse during extreme weather or urban traffic congestion. In V2X networks equipped with on-board units (OBUs), electric vehicles require additional energy to transmit data.

[0003] With the popularization and development of wireless communication technology and portable wireless communication smart devices, vehicle-to-everything (V2X) technology, as an important component of future intelligent transportation technology, has become one of the important research areas. V2X has attracted much attention due to its diverse traffic control and passenger entertainment applications. However, the high mobility of vehicles in V2X leads to frequent link interruptions and network topology fluctuations. Therefore, routing design in such networks is a challenging task.

[0004] Opportunistic Routing (OR) is an emerging routing strategy that enhances the performance of multi-hop wireless networks by leveraging the broadcast characteristics of wireless networks. It has attracted widespread attention in recent years. Compared to traditional routing mechanisms, OR does not specify a fixed next-hop node for forwarding packets. Instead, it flexibly selects multiple available neighboring nodes for data transmission at each stage and assigns priorities to them. This strategy reduces dependence on a single forwarding node and mitigates the negative impact of wireless link instability on data transmission. Therefore, OR can significantly improve wireless network throughput, enhance data transmission reliability, and improve end-to-end latency performance. However, traditional energy-efficient routing methods in vehicular networks typically only consider the utility of line-of-sight (LOS) links. When LOS links are blocked or unavailable, designing an energy-efficient routing scheme to achieve green communication goals is crucial in vehicular networks. Summary of the Invention

[0005] The purpose of this invention is to overcome the shortcomings of the above-mentioned background technology and provide an energy-saving optimization method and device for realizing vehicular network opportunistic routing, which surpasses existing energy efficiency solutions in terms of throughput, data packet transmission rate and energy efficiency by effectively utilizing NLOS links.

[0006] This invention provides an energy-saving optimization method for opportunistic routing in vehicle-to-everything (V2X) networks, comprising the following steps: S1, acquiring vehicle-to-vehicle scenarios, establishing a link model and a vehicle relay node sensor energy consumption model, and formalizing the vehicle-to-vehicle scenario, the link model, and the vehicle relay node sensor energy consumption model, respectively; S2, the highest priority relay node forwards data first; if the highest priority relay node fails to forward the data, the lower priority relay nodes continue to forward the data, based on contribution probability. S3. Estimate the actual number of data packets forwarded on the link from the source node to the relay node; S4. Define the forwarding set of the transmission vehicles according to the horizontal distance between the transmission vehicle and the target vehicle, optimize the expected travel distance and expected energy consumption of each packet by the expected packet progress, and combine distance and energy to optimize the expected energy consumption and expected distance when considering the packet transmission rate; S5. Contribution probability from the source node to the relay node. Obtain the expected number of transmissions from the source node to the relay node. Normalize the distance from the source node to the relay node, the expected number of transmissions from the source node to the relay node, and the energy consumption of the vehicle relay node sensor to obtain the probability values ​​of the distance from the source node to the relay node, the probability value of the expected number of transmissions from the source node to the relay node, and the probability value of the energy consumption of the vehicle relay node sensor. The average of the sums of the probability values ​​of the distance from the source node to the relay node, the probability value of the expected number of transmissions from the source node to the relay node, and the probability value of the energy consumption of the vehicle relay node sensor is used to obtain the communication efficiency cost index. The index values ​​of adjacent vehicles within the threshold range are excluded from the communication efficiency cost index to obtain all paths that meet the specified metric. S5. The HEOR algorithm is obtained by optimizing the Dinkelbach algorithm. The HEOR algorithm is applied to any path to obtain the optimal path selection.

[0007] In the above technical solution, the specific process of step S1 is as follows: S11, Formalization of the vehicle-to-vehicle scenario: Establish a vehicle network set V consisting of N vehicles, labeled i = {1, 2, ..., N}. All vehicles in set V move in the same direction, and their density and speed are distributed according to a Gaussian distribution; S12, Formalization of the link model: The communication of each node in set V is divided into intra-layer links R. intra and inter-layer links R inter And R inter =δR intra Where δ is the degradation ratio of the transmission range, 0 < δ ≤ 1, and the vehicle source node v i With vehicle relay node v j Packet delivery probability of intra-layer links for: Vehicle source node v i With vehicle relay node v j inter-layer link packet transmission probability for: Where, d ij Indicates the vehicle source node v i With vehicle relay node v j Distance; S13, Formalization of the energy consumption model of vehicle relay node sensors: Calculate the energy consumption of vehicle source node v i Transmit the k bits of data from the data packet to a location d. ij Vehicle relay node v outside the meter j The energy consumed by the sensor Among them, E elec It is the energy consumption required to operate each onboard unit, ε fs These are the transmission parameters of the free space model.

[0008] In the above technical solution, in step S2, the contribution probability The definition process is as follows: S21, Define the vehicle topology graph Among them, vehicle vector Sort by horizontal distance to the destination node, where v′1 is the source vehicle, v′... N The destination vehicle is E, and both E and S are N×N matrices with elements... Indicates whether the link crosses layers; if the link is a cross-layer link, then set e. ij =1; otherwise, let e ij =0; element Indicates whether the link is available; if the link is available, s ij Set to 1; otherwise, s ij Setting it to 0, we can obtain s ij for: Derivation of vehicle source node v i Successfully transmitted to vehicle relay node v j Contribution probability

[0009] S22, When the transmitting vehicle is the source vehicle v′ i When selecting the vehicle to forward, v′ j Contribution probability of (i≤j≤N) The calculation is as follows: S23. As a symmetric matrix, the probability matrix Transform into a probability matrix

[0010] In the above technical solution, the specific process of step S3 is as follows: S31, Define the forwarding vehicle v′ according to the horizontal distance between the forwarding vehicle and the target vehicle. i forwarding set π i Size ki , The i-th line: The EPA defines the expected grouping progress as follows: Similarly, the source vehicle v′ i The expected energy consumption for the expected distance traveled is defined as follows: S32. When considering packet transmission rate, combine distance and energy to optimize expected energy consumption and expected distance: Where Ω represents the set of all forwarding node choices, β ij To select the binary variable of adjacent vehicles, E h It is the energy threshold between adjacent forwarding vehicles; η EE C1 is the ratio of the expected forwarding distance of the forwarding node to the expected total energy consumption, C2 represents the expected data packet transmission rate constraint of all neighboring vehicles to the forwarding vehicle, and C3 represents the expected energy consumption constraint between adjacent forwarding vehicles.

[0011] In the above technical solution, the normalization process in step S4 is as follows: The distances between all adjacent vehicles and the destination vehicle are normalized within the range [0, 1]. For the normalized vehicle source node v i With vehicle destination node v d The probability value of the distance is calculated as follows: Vehicle source node v i With vehicle relay node v j The formula for calculating the expected number of transfers between ETX is as follows: Using the distance d from the source node to the relay node jd The same method is used to normalize the expected number of transmissions from the source node to the relay node (ETX). ij Energy consumption E of vehicle relay node sensors ij .

[0012] In the above technical solution, the calculation process for all paths satisfying the specified metric in step S4 is as follows: The adjacent vehicle source node v is obtained by calculating the transmission distance, the expected number of transmissions, and the required energy. i With relay node v j CECM (Communication Efficiency Cost Indicator):

[0013] By excluding specific thresholds CECM h For adjacent vehicles within a certain range, obtain all paths ψ that satisfy the specified metric; the proof of all paths ψ is as follows: Represent all paths that satisfy the specified conditions as Ψ={ψ1,ψ2,...,ψ m}, where element ψm ∈Ψ is a binary vector; if in path ψ m Select the mode of transportation v′ i Let λ i ∈ψ m =1; Ψ and ψ m The sizes are M and N respectively; assume Indicates in ψ m The set of indicators in which all values ​​are 1, there are k in total. m For each path, the expected grouping schedule and energy consumption are as follows: Where, φ i,i+1 =φ i φ i+1 Derivation of EPA from path selection vector Ω′ tot and EC tot :

[0014]

[0015] In the formula, EC0 represents the initial energy consumption of the network; the path selection problem is transformed as follows: Among them, E c It is the energy constraint on each link, β m It is a binary variable for path selection.

[0016] In the above technical solution, in step S4, η EE It has superlinear convergence, and the proof is as follows: Lemma 1: Let G(Ω′)=η EE Let C be a convex set such that EC tot ≠0, then G(Ω′) is both pseudo-convex and pseudo-concave on C; Proof: Formula (17) represents a linear fractional function, then G(Ω′) is both pseudo-convex and pseudo-concave on C; Lemma 2: G(Ω′) is strictly quasi-concave, and is strictly quasi-concave on C; Proof: Based on the fact that G(Ω′) is both pseudo-convex and pseudo-concave in Lemma 1, then G(Ω′) is also strictly quasi-concave, and is also strictly quasi-concave on C; Lemma 3: Every local optimum of G(Ω′) on C is also its global solution; Proof: Based on Lemma 2, the properties of strictly quasi-convex and strictly quasi-concave functions lead to Lemma 3; According to Lemma 3, use an arbitrary nonlinear programming NLP solver to solve the linear fractional programming (LFP) problem. get Theorem 1: The global maximum value is found if and only if The optimal solution Ω is obtained * Proof: Let Ω * If ∈C is the optimal solution of (18), then we have

[0017] Due to ECtot (Ω′)>0, therefore we know therefore, It is the maximum value of (18), Ω * It is the optimal solution of (17); Then Ω * It is the optimal solution of (18); Theorem 2: For problem (17), the Dinkelbach algorithm is superlinearly convergent; Proof: The Dinkelbach algorithm will Update to the previous value G(Ω′), that is Where Ω′ i For F(Ω) i The optimal solution is found, and the algorithm converges at a speed of . If it is non-increasing, then each Sequences obtained by Dinkelbach algorithm Superlinear convergence to

[0018] In the above technical solution, the specific process of step S5 is as follows: According to Theorem 1, problem (17) is transformed into a parameter subtraction form, that is...

[0019] In the formula, Let represent the optimal solution in the k-th iteration. Indicates the EPA at the k-th iteration. tot With EC tot The ratio; substituting formulas (15) and (16) into formula (22), we get:

[0020]

[0021] Transform formula (23) into: Therefore, Theorem 3 is given: For any given path ψ i Optimal path selection The formula is as follows: in, This represents the path that satisfies the constraints in the k-th iteration. Formula (23) is equivalent to: For φ m For paths ≤ 0, let... Maximize the objective function; for φ m If the value is >0, select the path that satisfies the given conditions; combine equations (17) and (23) to determine the optimal path selection.

[0022] The above technical solution also includes step S6, which involves simulation and comparison of throughput, packet transmission rate, and energy efficiency.

[0023] The present invention also provides an energy-saving optimization device for realizing opportunistic routing in vehicle-to-everything (V2X) networks, which has a computer program that can execute an energy-saving optimization method for realizing opportunistic routing in V2X networks.

[0024] The present invention provides an energy-saving optimization method and apparatus for opportunistic routing in vehicle-to-everything (V2X) networks, which has the following beneficial effects:

[0025] (1) Impact of traffic density on performance: Throughput increases with increasing λ at low traffic densities, but decreases after λ reaches 0.21 due to bandwidth consumption. The HEOR protocol outperforms TSOR and SDORP in congestion mitigation and energy efficiency through multi-path selection and traffic optimization. PDR increases initially with increasing vehicle density, but decreases after exceeding 0.21 due to network congestion. TSOR and SDORP are significantly affected by channel contention and interference, while HEOR maintains stable and efficient routing through algorithm optimization. Energy efficiency index η EE As the number of connected vehicles increases, HEOR performs better in maintaining high energy efficiency through its proactive waste disposal strategy.

[0026] (2) Impact of End-to-End Distance on Performance: In the impact of end-to-end distance on routing performance, throughput increases as the distance increases from 600m to 750m, subsequently decreasing due to increased latency and retransmission rate. The HEOR protocol utilizes the Dinkelbach algorithm to optimize the path, providing higher throughput and PDR, especially peaking at an end-to-end distance of 750m. Energy efficiency index η EE The efficiency of HEOR increases with distance, and it maintains the highest energy efficiency through a globally optimal solution algorithm, even after 800m. EE Even as energy consumption begins to decline, high energy efficiency can be maintained through link assessment and optimization, reducing long-term energy consumption. Attached Figure Description

[0027] Figure 1 This is a schematic diagram illustrating the overall principle of the energy-saving optimization method for opportunistic routing in vehicle-to-everything (V2X) according to the present invention.

[0028] Figure 2 This is a schematic diagram of the overall process of the energy-saving optimization method for opportunistic routing in vehicle-to-everything (V2X) according to the present invention.

[0029] Figure 3 This is a schematic diagram of the vehicle network set V in step S1 of the energy-saving optimization method for vehicle network opportunistic routing in the present invention.

[0030] Figure 4 This is a schematic diagram of the logic code for the HEOR algorithm in step S5 of the energy-saving optimization method for opportunistic routing in the Internet of Vehicles of the present invention.

[0031] Figure 5This is a schematic diagram of the architecture of the energy-saving optimization device for realizing vehicle-to-everything (V2X) opportunistic routing according to the present invention;

[0032] Figure 6 The diagram shows a comparison of the energy-saving optimization method and apparatus for vehicle-to-everything (V2X) opportunistic routing of the present invention with existing technologies in terms of throughput, packet transmission rate and energy efficiency during congested traffic.

[0033] Figure 7 The diagram shows a comparison of the energy-saving optimization method and apparatus for vehicle-to-everything (V2X) opportunistic routing in this invention with existing technologies in terms of throughput, packet transmission rate, and energy efficiency under end-to-end distance conditions. Detailed Implementation

[0034] The present invention will now be described in further detail with reference to the accompanying drawings and embodiments, but these embodiments should not be construed as limiting the present invention.

[0035] Technical Principles

[0036] Traditional high-efficiency routing methods in vehicular networks typically only consider the utility of line-of-sight (LOS) links while neglecting the utility of non-linear line-of-sight (NLOS) links, thus failing to utilize NLOS links for data transmission within the vehicular network. When LOS links are blocked or unavailable, NLOS links can provide an alternative communication method, expanding communication coverage and improving communication reliability.

[0037] In energy-efficient vehicular networking routing, non-line-of-sight (NLOS) links hold great potential for opportunistic routing, enhancing communication capabilities and saving energy. Although elevated bridge surfaces attenuate NLOS links, choosing NLOS links can effectively improve energy efficiency and communication reliability when vehicle visibility is limited or when distances from other vehicles are significant.

[0038] By studying the potential of NLOS links to increase OR communication opportunities in elevated bridge structures, this paper designs a method to transform the NP-hard (non-deterministic polynomial-time hardness) multi-hop OR problem in vehicular networks into a solvable form. Furthermore, a Hybrid-link Energy Optimization Routing (HEOR) algorithm is proposed to maximize energy efficiency, particularly for elevated bridge structures in the Internet of Things (IoT). Extensive simulations of the algorithm's performance are conducted, comparing it with the Thompson Sampled Opportunity Routing (TSOR) algorithm and the Software-Defined Networking Asynchronous Duty Cycle Wireless Sensor Network Opportunity Routing (SDORP) algorithm. The results show that HEOR, by effectively utilizing NLOS links, surpasses existing energy-efficient solutions in terms of throughput, packet delivery rate, and energy efficiency.

[0039] Detailed technical solution

[0040] Example: See Figure 1 This invention provides an energy-saving optimization method for opportunistic routing in vehicle-to-everything (V2X) communication, comprising the following steps, detailed below. Figure 2 :

[0041] S1. Formalization of vehicle-to-vehicle scenarios, link models, and energy consumption models, the specific process of which is as follows:

[0042] S11. Formalization of Car-to-Car Scenarios

[0043] As attached Figure 3 The diagram shows a vehicle network consisting of a set V (N vehicles), labeled i = {1, 2, ..., N}. The vehicles in set V travel on a straight, two-level highway. On the upper level, vehicles travel on an elevated highway that passes through a busy section of the city center. On the lower level, vehicles travel on a straight road segment below the elevated highway, without intersections. It is assumed that all vehicles move in the same direction, and that their density and speed follow a Gaussian distribution.

[0044] S12, Formalization of the Link Model

[0045] Each node in set V is equipped with an on-board unit, enabling adjacent nodes to communicate via the IEEE 802.11p protocol. Furthermore, each node can obtain its position, velocity, and direction via a Global Positioning System (GPS) device, or via BeiDou, Galileo, or GLONASS systems. The radio transmission range is assumed to be a circle, although this is not always the case in reality. Communication between nodes in set V is divided into intra-layer links and inter-layer links; the transmission range of an intra-layer link is denoted by R. intra R represents the transmission range of the inter-layer link. inter It indicates that, and R inter =δR intra Where 0 < δ ≤ 1 represents the degradation ratio within the transmission range. Vehicle source node v i With vehicle relay node v j Packet delivery probability of intra-layer links for:

[0046]

[0047] Vehicle source node v i With vehicle relay node v j inter-layer link packet transmission probability for:

[0048]

[0049] Where, d ij Indicates the vehicle source node v i With vehicle relay node v j The distance.

[0050] S13, Formalization of Energy Consumption Model

[0051] This invention uses free-space channels. Vehicle source node v i Transmit the k bits of data from the data packet to a location d. ij Vehicle relay node v outside the meter j The energy consumed by the sensor E ij The energy E ij The calculation formula is as follows:

[0052]

[0053] Among them, E elec It is the energy consumption required to operate each onboard unit, ε fs These are the transmission parameters of the free space model.

[0054] S2. The contribution probability model is formalized, and the specific process is as follows:

[0055] In OR (Relay Relay), the sender has multiple candidate relays, any of which can receive and forward data packets. Since only one node can transmit at a time, the highest-priority relay node forwards the message first. If the highest-priority relay node fails to forward, lower-priority relay nodes can continue to participate in data forwarding. In OR, the relay denoted by j does not need to forward all data packets received from the source denoted by i. In fact, only a small portion of the received data packets are responsible for transmission and contribute to data relay. Contribution probability is used. To estimate how many data packets should actually be forwarded on the link from source node i to relay node j, in one or more embodiments, the link may be a LOS link or an NLOS link.

[0056] S21. Define the vehicle topology graph. Vehicle vector Sort by horizontal distance to the destination node. Therefore, v′1 is the source vehicle, v′ N The destination vehicle is E. Both E and S are N×N matrices. Elements Indicates whether the link crosses layers; if the link is a cross-layer link, then set e. ij =1; otherwise, set it to 0. Similarly, element This indicates whether the link is available; if the link is available, it is set to 1; otherwise, it is set to 0. Therefore, s... ij for:

[0057]

[0058] Therefore, the vehicle source node v can be derived. i Successfully transmitted to vehicle relay node v j Contribution probability

[0059]

[0060] S22, When the transmitting vehicle is the source vehicle v′ i When selecting the vehicle to forward, v′ j Contribution probability of (i≤j≤N) The calculation is as follows:

[0061]

[0062] S23, Probability Matrix It is a symmetric matrix. For ease of calculation, the probability matrix is... By transforming it into an upper triangular matrix form using formula (6), a new probability matrix can be obtained.

[0063]

[0064] S3: Problem formulation, the specific process is as follows:

[0065] S31, Define π i For forwarding vehicle v′ i The forwarding set, whose size is k i Since forwarding vehicles are sorted according to their horizontal distance from the target vehicle, non-zero elements in the same row are consecutive. This indicates that... The i-th line:

[0066] Expected Packet Advance (EPA) is a crucial optimization objective in Operational Relationship Management (ORM). It represents the expected progress of each packet during transmission. The EPA can be derived as follows:

[0067]

[0068] Similarly, the source vehicle v′ can be derived. i Expected energy consumption:

[0069]

[0070] S32. Considering packet transmission rate, and combining distance and energy, the energy-saving transmission problem in vehicular networks is studied. It is expressed as the following optimization plan:

[0071]

[0072] Where Ω represents the set of all forwarding node choices, β ij To select the binary variable of an adjacent vehicle, if the vehicle source node v i Select adjacent vehicle relay node v j , then β ij If β is 1, then β is 1. ij It is 0. E h It is the energy threshold between adjacent forwarding vehicles.

[0073] In the optimization problem of (10), η EE C1 represents the ratio of the total expected forward distance of the forwarding nodes to the total expected energy consumption, and C2 represents the packet delivery rate constraint on each link. C3 represents the expected packet delivery rate constraint from all neighboring vehicles to the forwarding vehicle. C4 represents the expected energy consumption constraint between adjacent forwarding vehicles. The expected energy consumption is described using link probabilities and binary choice variables because the goal is to optimize the expected value.

[0074] In opportunistic routing, each node has many candidate nodes at each hop, so the matrix representation of all available paths grows exponentially. Therefore, each node has multiple paths to the receiver. In the worst case, each vehicle can communicate with every other vehicle, with a time complexity of O(log n). Therefore, this problem is NP-hard.

[0075] S4: Problem transformation, the specific process is as follows:

[0076] Since the goal of (10) is to obtain a 0-1 matrix rather than a 0-1 vector, we cannot directly use the new Dinkelbach algorithm developed for solving linear programming problems. To solve this problem, we transform it from link selection to path selection. This invention uses a new metric, namely the Communication Efficiency Cost Metric (CECM), to exclude adjacent vehicles that exceed a threshold. Only vehicles within the transmission range are considered. After the sending vehicle obtains the distances between all adjacent vehicles and the destination vehicle, it determines the distances based on the distances between the sending vehicle and the destination vehicle.

[0077]

[0078] The distances between all adjacent vehicles and the destination vehicle are normalized in the range [0, 1]. For the normalized vehicle source node v i With vehicle destination node v d The probability value is the distance. The purpose of this is to assign higher priority to vehicles that are closer to the destination.

[0079] Expected Transmission Count (ETX) is an important metric for measuring link quality in OR (Robotic Link), where the vehicle source node v... i With vehicle relay node v j The formula for calculating the expected number of transfers between ETX is as follows: Given that links with lower energy consumption have higher priority, the link is determined by the distance d from the source node to the relay node. jd The same method is used to normalize the expected number of transmissions from the source node to the relay node (ETX). ij Energy consumption E of vehicle relay node sensors ij .

[0080] The source vehicle is determined by the following formula:

[0081]

[0082] Calculate the source node v of each adjacent vehicle i With vehicle relay node v jThe Communication Efficiency Cost Index (CECM) is a measure of the communication efficiency cost, where the attribute in CECM is the average of three probability values ​​in formula (12). These probability values ​​are derived from the transmission distance, the expected number of transmissions, and the required energy, respectively.

[0083] The higher the link quality, the smaller the CECM. Subsequently, it will exclude adjacent vehicles that do not exceed a certain threshold, which, in one or more embodiments, is denoted as the CECM. h Finally, all paths ψ that satisfy the specified metric can be obtained.

[0084] The proof for all paths ψ is as follows:

[0085] All paths that satisfy the specified conditions are represented as Ψ={ψ1,ψ2,...,ψ m}, where element ψ m ∈Ψ is a binary vector. If in path ψ m Select the mode of transportation v′ i Let λ i ∈ψ m Ψ and ψ are 1. m The sizes are M and N, respectively. Assume... Indicates in ψ m The set of indicators in which all values ​​are 1, there are k in total. m Therefore, the expected grouping schedule and energy consumption for each route are as follows:

[0086]

[0087] Where φ i,i+1 =φ i φ i+1

[0088] Therefore, the EPA can be derived from the path selection vector Ω′. tot and EC tot .

[0089]

[0090] In the formula, EC0 represents the initial energy consumption of the network.

[0091] Therefore, the path selection problem can be transformed as follows:

[0092]

[0093] Among them, E c It is the energy constraint on each link, β m It is a binary variable for path selection; if path ψ is selected... m Then β m If β is 1, then β is 1. mis 0.

[0094] So, how can we prove that the sequence obtained by the Dinkelbach algorithm... Superlinear convergence to The specific proof process is as follows:

[0095] Lemma 1: Let G(Ω′) = η EE Let C be a convex set such that EC tot If ≠0, then G(Ω′) is both a pseudo-convexity and a pseudo-concaveness on C.

[0096] Proof: Formula (17) represents a linear fractional function, so G(Ω′) is both pseudo-convex and pseudo-concave on C.

[0097] Lemma 2: G(Ω′) is strictly quasi-concave and strictly quasi-concave on C.

[0098] Proof: Based on the fact that G(Ω′) is both pseudo-convex and pseudo-concave in Lemma 1, we can conclude that G(Ω′) is also strictly quasi-concave, and also strictly quasi-concave on C.

[0099] Lemma 3: Every local optimum of G(Ω′) on C is also its global solution.

[0100] Proof: Based on Lemma 2, the properties of strictly quasi-convex and strictly quasi-concave functions lead to Lemma 3.

[0101] According to Lemma 3, the linear fractional programming (LFP) problem can be solved using an arbitrary nonlinear programming (NLP) solver. It can be obtained The global maximum value.

[0102] Theorem 1: If and only if

[0103]

[0104] The optimal solution Ω can be obtained. * .

[0105] Proof: Let Ω * If ∈C is the optimal solution of (18), then we have

[0106]

[0107] Due to EC tot (Ω′)>0, therefore we know

[0108]

[0109] therefore, It is the maximum value of (18), Ω * It is the optimal solution of (17).

[0110] because,

[0111]

[0112] This means Ω * It is the optimal solution of (18).

[0113] Theorem 2: For problem (17), the Dinkelbach algorithm is superlinearly convergent.

[0114] Proof: Dinkelbach's algorithm will Update to the previous value G(Ω′), that is Where Ω′ i For F(Ω) i The optimal solution is found. The algorithm converges at a speed of... It is non-increasing. Therefore, for each Sequences obtained by Dinkelbach algorithm Superlinear convergence to (These proofs are to demonstrate that the Dinkelbach algorithm is superlinearly convergent.)

[0115] S5: For any given path, find the optimal path selection. The specific process is as follows:

[0116] This invention will develop a new Dinkelbach algorithm, named the HEOR algorithm, to improve energy efficiency. According to Theorem 1, problem (17) can be transformed into a parameter subtraction form, i.e.

[0117]

[0118] In the formula, Let represent the optimal solution in the k-th iteration. Indicates the EPA at the k-th iteration. tot With EC tot The ratio of .

[0119] Substituting equations (15) and (16) into equation (22), we get:

[0120]

[0121] For simplicity, let:

[0122]

[0123] Then, the following theorem is given:

[0124] Theorem 3: For any given path ψi Optimal path selection The formula is as follows:

[0125]

[0126] in, This represents the path that satisfies the constraints in the k-th iteration.

[0127] Proof: To prove the conclusion following Theorem 2, we first rewrite formula (23) equivalently as follows:

[0128]

[0129] For φ m Paths with a value ≤ 0 need to be set To maximize the objective function. For φ m If the value is >0, select the path that satisfies the given conditions.

[0130] Combining formulas (17) and (23), this invention proposes the HybridLink Energy Optimized Routing (HEOR) algorithm to optimally determine path selection.

[0131] In the HEOR algorithm, local points are obtained in the i-1th iteration by solving problem (22). This process is repeated for each path. Then, the source vehicles merge them into The total EPA and EC are obtained. Finally, η is obtained. EE The new value of F. When the value of F is less than ∈, the iteration stops, and the optimal solution Ω′ is obtained. Let ∈ be 10. -5 The detailed implementation of the HEOR algorithm is attached. Figure 4 .

[0132] S6: Formalizing data transmission network performance metrics under two factors, the specific process is as follows:

[0133] Experimental simulations were conducted using MATLAB R2019B and the traffic simulator SUMO. A client-server mechanism was enabled between MATLAB (client) and SUMO (server) using TaCI4Matlab. The simulation ran on a 3D urban viaduct scene measuring 3500m × 5000m × 25m. The viaduct was 15 meters high. All vehicles traveled in the same direction. Vehicles traveled at 60-80 km / h on the viaduct, while their speed decreased to 40-60 km / h on the ground-level roads. (The last sentence appears to be incomplete and possibly refers to a different simulation, "same-layer link R..."). LOS and cross-layer link R NLOSThe transmission distances are 250m and 200m respectively. There are three lanes on each of the upper and lower layers at the edge. The number of vehicles in each lane follows an exponential distribution with parameter λ. The upper layer has fewer vehicles, and the lower layer has more, facilitating the study of the communication probability brought by the NLOS link. The data rate and packet size are set to 2Mbps and 512 bytes respectively. Each experiment in this invention runs for 200 seconds, repeated 500 times with different seeds, and the average value is reported. Two sets of simulations were performed to evaluate the impact of different vehicle densities and end-to-end distances on routing performance. Three performance parameters (throughput, packet transmission rate, and energy efficiency η) were considered. EE The simulation was compared. λ was set between 0.15 and 0.25, with the end-to-end distance fixed at 850m. For the other group, the end-to-end distance was set between 600m and 900m.

[0134] Throughput is defined as the number of bits successfully transmitted per second in a network. PDR (packet delivery ratio) is the ratio of the number of packets successfully received at the destination to the total number of packets generated at the source.

[0135] Simulation results show that the HEOR algorithm outperforms existing TSOR and SDORP energy-efficient schemes in terms of throughput, packet transmission rate, and energy efficiency by effectively utilizing NLOS links.

[0136] See Figure 5 The present invention provides an energy-saving optimization device for realizing opportunistic routing in vehicle-to-everything (V2X) communication, comprising the following parts:

[0137] Formalization module: Obtain vehicle-to-vehicle scenarios, establish link models and vehicle relay node sensor energy consumption models, and formalize vehicle-to-vehicle scenarios, link models, and vehicle relay node sensor energy consumption models, respectively.

[0138] Contribution probability model module: The highest priority relay node forwards data first. If the highest priority relay node fails to forward the data, the lower priority relay nodes continue to forward the data, based on the contribution probability. Estimate the number of data packets actually forwarded on the link from the source node to the relay node;

[0139] Problem Formulation Module: Define the forwarding set of transmission vehicles according to the horizontal distance between the transmission vehicle and the target vehicle, optimize the expected forward distance and expected energy consumption of each packet by the expected packet progress, and combine distance and energy to optimize expected energy consumption and expected distance when considering packet transmission rate;

[0140] Problem transformation module: Contribution probability from source node to relay node Obtain the expected number of transmissions from the source node to the relay node. Normalize the distance from the source node to the relay node, the expected number of transmissions from the source node to the relay node, and the energy consumption of the vehicle relay node sensor to obtain the probability values ​​of the distance from the source node to the relay node, the probability value of the expected number of transmissions from the source node to the relay node, and the probability value of the energy consumption of the vehicle relay node sensor. The average of the sums of the probability values ​​of the distance from the source node to the relay node, the probability value of the expected number of transmissions from the source node to the relay node, and the probability value of the energy consumption of the vehicle relay node sensor is used to obtain the communication efficiency cost index. The index values ​​of adjacent vehicles within the threshold range are excluded from the communication efficiency cost index to obtain all paths that meet the specified metric.

[0141] Optimal Path Selection Module: The HEOR algorithm is obtained by optimizing the Dinkelbach algorithm. The HEOR algorithm is then applied to any path to obtain the optimal path selection.

[0142] Simulation comparison module: Performs simulations and comparisons in terms of throughput, packet transmission rate, and energy efficiency.

[0143] Technical effect

[0144] (1) The impact of traffic density on performance

[0145] From the appendix Figure 6 It can be seen that throughput increases with increasing λ at low traffic density, but decreases after λ reaches 0.21 due to bandwidth consumption. The HEOR protocol outperforms TSOR and SDORP in congestion mitigation and energy efficiency through multi-path selection and traffic optimization. PDR increases initially with increasing vehicle density, but decreases after exceeding 0.21 due to network congestion. TSOR and SDORP are significantly affected by channel contention and interference, while HEOR maintains stable and efficient routing through algorithm optimization. Energy efficiency index η EE As the number of connected vehicles increases, HEOR performs better in maintaining high energy efficiency through its proactive waste disposal strategy.

[0146] (2) The impact of end-to-end distance on performance

[0147] Regarding the impact of end-to-end distance on routing performance, see the attached table. Figure 7 The throughput shown increases as the distance increases from 600m to 750m, then decreases due to increased latency and retransmission rate. The HEOR protocol utilizes the Dinkelbach algorithm to optimize the path, providing higher throughput and PDR, especially peaking at an end-to-end distance of 750m. Energy efficiency index η EE The efficiency of HEOR increases with distance, and it maintains the highest energy efficiency through a globally optimal solution algorithm, even after 800m. EE Even as energy consumption begins to decline, high energy efficiency can be maintained through link assessment and optimization, reducing long-term energy consumption.

[0148] Obviously, those skilled in the art can make various modifications and variations to this invention without departing from its spirit and scope. Therefore, if these modifications and variations fall within the scope of the claims of this invention and their equivalents, this invention also intends to include these modifications and variations.

[0149] The contents not described in detail in this specification are existing technologies known to those skilled in the art.

Claims

1. An energy-saving optimization method for realizing opportunistic routing in vehicle-to-everything (V2X) networks, characterized in that: Includes the following steps: S1. Obtain the vehicle-to-vehicle scenario, establish the link model and the energy consumption model of the vehicle relay node sensor, and formalize the vehicle-to-vehicle scenario, the link model and the energy consumption model of the vehicle relay node sensor respectively. S2. The highest priority relay node forwards the data first. If the highest priority relay node fails to forward the data, the lower priority relay nodes continue to forward the data, based on contribution probability. Estimate the number of data packets actually forwarded on the link from the source node to the relay node; S3. Define the forwarding set of the transmission vehicles according to the horizontal distance between the transmission vehicle and the target vehicle, optimize the expected forward distance and expected energy consumption of each packet by the expected packet progress, and combine distance and energy to optimize the expected energy consumption and expected distance when considering the packet transmission rate. S4. Contribution probability from source node to relay node The expected number of transmissions from the source node to the relay node is obtained. The distance from the source node to the relay node, the expected number of transmissions from the source node to the relay node, and the energy consumption of the vehicle relay node's sensors are normalized to obtain probability values ​​for the distance from the source node to the relay node, the expected number of transmissions from the source node to the relay node, and the energy consumption of the vehicle relay node's sensors. The average of these sums is used to obtain the communication efficiency cost index. The index values ​​of adjacent vehicles within a threshold range are excluded from the communication efficiency cost index to obtain all paths that satisfy the specified metric. The calculation process for all paths satisfying the specified metric is as follows: The adjacent vehicle source nodes are obtained by calculating the transmission distance, the expected number of transmissions, and the required energy. With relay node CECM (Communication Efficiency Cost Indicator): By excluding specific thresholds For adjacent vehicles within a given range, obtain all paths that satisfy the specified metric. ; All paths The proof is as follows: All paths that satisfy the specified conditions are represented as , among which, element It is a binary vector; If in the path Choose a mode of transportation ,set up =1; and The sizes are M and N, respectively; Assumption Indicates in The set of indicators in which all values ​​are 1, totaling [number]. For each path, the expected grouping schedule and energy consumption are as follows: in, , From path selection vector Derivation and : In the formula, This represents the initial power consumption of the network. The path selection problem can be transformed as follows: in, It is the energy constraint on each link. It is a binary variable for path selection; S5. The HEOR algorithm is obtained by optimizing the Dinkelbach algorithm. The HEOR algorithm is then applied to any path to obtain the optimal path selection. The specific process is as follows: Transform problem (17) into parameter subtraction form, i.e. In the formula, Let represent the optimal solution in the k-th iteration. Indicates the k-th iteration and The ratio of Substituting equations (15) and (16) into equation (22), we get: Transform formula (23) into: For any given path Optimal path selection The formula is as follows: in, This represents the path that satisfies the constraints in the k-th iteration. ; Formula (23) is equivalent to: for The path, let Maximize the objective function; for Choose the path that meets the given conditions; By combining equations (17) and (23), the optimal path selection is determined.

2. The energy-saving optimization method for realizing opportunistic routing in vehicle-to-everything (V2X) networks according to claim 1, characterized in that: The specific process of step S1 is as follows: S11. Formalization of vehicle-to-vehicle scenarios: Establish a vehicle network set V consisting of N vehicles, labeled i = {1, 2, ..., N}. All vehicles in set V move in the same direction, and their density and speed are distributed according to a Gaussian distribution. S12. Formalization of the link model: Communication between each node in set V is divided into intra-layer links. and inter-layer links ,and ,in, The degradation ratio for the transmission range, , Vehicle source node With vehicle relay node Packet delivery probability of intra-layer links for: Vehicle source node With vehicle relay node inter-layer link packet transmission probability for: in, Indicates the vehicle source node With vehicle relay node The distance; S13. Formalization of the energy consumption model for vehicle relay node sensors: Calculate vehicle source node In the data packet Bit data is transmitted to a distance of... Vehicle relay node outside meters The energy consumed by the sensor : in, This refers to the energy consumption required to operate each onboard unit. These are the transmission parameters of the free space model.

3. The energy-saving optimization method for realizing opportunistic routing in vehicle-to-everything (V2X) networks according to claim 2, characterized in that: In step S2, the contribution probability The definition process is as follows: S21. Define the vehicle topology graph. Among them, vehicle vector Sort by horizontal distance to the destination node, where It is the source vehicle. It is the destination vehicle. and Both are N × N matrices, with elements Indicates whether the link crosses layers; if the link is a cross-layer link, then set... Otherwise, assume 0; element Indicates whether the link is available; if the link is available, Set to 1; otherwise, Set to 0, and thus we can obtain for: Derivation of vehicle source node Successfully transmitted to the vehicle relay node : S22, When the transmission vehicle is When choosing to forward a vehicle Contribution probability The calculation is as follows: S23. As a symmetric matrix, the probability matrix Transform into a probability matrix : 。 4. The energy-saving optimization method for realizing opportunistic routing in vehicle-to-everything (V2X) networks according to claim 3, characterized in that: The specific process of step S3 is as follows: S31. Define forwarding vehicles according to their horizontal distance from the target vehicle. forwarding set , the size is , The Behavior: , The EPA defines the expected grouping progress as follows: Similarly, The expected energy consumption for the expected distance traveled is defined as follows: S32. When considering packet transmission rate, combine distance and energy to optimize expected energy consumption and expected distance: Where Ω represents the set of all forwarding node choices. Binary variables for selecting adjacent vehicles. It is the energy threshold between adjacent forwarding vehicles; C1 is the ratio of the expected forwarding distance of the forwarding node to the expected total energy consumption, C2 represents the expected data packet transmission rate constraint of all neighboring vehicles to the forwarding vehicle, and C3 represents the expected energy consumption constraint between adjacent forwarding vehicles.

5. The energy-saving optimization method for realizing opportunistic routing in vehicle-to-everything (V2X) networks according to claim 4, characterized in that: In step S4, the normalization process is as follows: The distances between all adjacent vehicles and the destination vehicle are normalized in the range [0, 1]. For normalized vehicle source nodes With vehicle destination node The probability value of the distance is calculated as follows: Vehicle source node With vehicle relay node The formula for calculating the expected number of transfers between ETX is as follows: The distance from the source node to the relay node Normalize the expected number of transmissions from the source node to the relay node in the same way. Energy consumption of vehicle relay node sensors .

6. The energy-saving optimization method for realizing opportunistic routing in vehicle-to-everything (V2X) communication according to claim 5, characterized in that: In step S4 It exhibits superlinear convergence, and the proof is as follows: Lemma 1: Let ,set up It is a convex set, such that ,but exist The upper part is both a false convexity and a false concaveness; Proof: Formula (17) represents a linear fractional function, then exist The upper part is both a false convexity and a false concaveness; Lemma 2: It is strictly concave, in The upper part is strictly concave; Proof: Based on Lemma 1 If it is both a false convexity and a false concaveness, then... It is also strictly concave, and The upper part is also strictly concave; Lemma 3: exist Every local optimum on the surface is also its global solution; Proof: Based on Lemma 2, the properties of strictly quasi-convex and strictly quasi-concave functions lead to Lemma 3; According to Lemma 3, the linear fractional programming (LFP) problem can be solved using an arbitrary nonlinear programming (NLP) solver. ,get The global maximum; Theorem 1: If and only if To obtain the optimal solution ; Proof: Let If it is the optimal solution of (18), then we have because It can be known that therefore, It is the maximum value of (18). It is the optimal solution of (17); ， but It is the optimal solution of (18); Theorem 2: For problem (17), the Dinkelbach algorithm is superlinearly convergent; Proof: Dinkelbach's algorithm will Update to the previous value ,Right now ,in for The optimal solution, the algorithm convergence speed is If it is non-increasing, then each The sequence obtained by Dinkelbach algorithm Superlinear convergence to .

7. The energy-saving optimization method for realizing opportunistic routing in vehicle-to-everything (V2X) networks according to claim 6, characterized in that: It also includes step S6, which involves simulation and comparison of throughput, packet transmission rate, and energy efficiency.

8. An energy-saving optimization device for realizing vehicle-to-everything (V2X) opportunistic routing, comprising a computer program, characterized in that: The computer program is capable of executing the energy-saving optimization method for implementing vehicle-to-everything (V2X) opportunistic routing as described in any one of claims 1 to 7.

Images