Method for adjusting backoff mechanism and vehicle-mounted communication system
By using quantum neural networks to predict traffic flow information and competition window size, and dynamically adjusting vehicle backoff mechanisms, the problem of frequent data packet collisions in vehicular ad hoc networks is solved, improving the reliability and efficiency of information transmission.
Patent Information
- Application Number
- CN202310662596.0
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Priority Date
- 2023-05-18
- Filing Date
- 2023-06-06
- Publication Date
- 2024-11-22
AI Technical Summary
In vehicular ad hoc networks, data transmission between vehicles is frequently disrupted by constant contention window size, leading to frequent packet collisions, especially in broadcast mode where successful reception cannot be detected, thus affecting the transmission of traffic safety information.
Based on vehicle location and movement information, the server uses a quantum neural network to predict traffic flow information and competition window size for each road segment, and dynamically adjusts the vehicle backoff mechanism to reduce data packet collisions.
It improves the reliability and efficiency of information transmission between vehicles, reduces data packet collisions, and ensures the timely transmission of traffic safety information.
Smart Images

Figure CN119012403A_ABST
Abstract
Description
Technical Field
[0001] The present invention relates to a vehicle-mounted communication technology, and in particular to a method for adjusting a backoff mechanism and a vehicle-mounted communication system. Background Art
[0002] In recent years, in order to improve road safety, smart cities, etc., coupled with the rapid development of 5G, artificial intelligence (AI) and autonomous driving, the Internet of Vehicles has become one of the most demanded and most promising types of IoT in the market.
[0003] Vehicular ad hoc networks (VANET) are a highly anticipated technology that can enable communication between autonomous vehicles and traditional vehicles in future unmanned transportation. VANET applications can be divided into two main categories: traffic safety applications and traffic management applications. There are two common VANET communication methods: (1) 5.9 GHz dedicated short-range communication (DSRC); (2) cellular-based vehicular communication.
[0004] With the rapid development of the fifth-generation communication system (5G) technology for connected and autonomous vehicles, the industry is more inclined to support cellular-based communication technology, which is called cellular vehicle to everything (C-V2X). However, such C-V2X communication still requires roadside units (RSU) or some kind of base station near the road as an infrastructure component to collect and process traffic and vehicle data.
[0005] Research indicates that about 60% of accidents can be controlled if drivers are warned half a second before a vehicle crashes. Therefore, effective provision of information can play a key role in safe and smooth driving. In VANET, the transmission range of vehicles is limited (approximately 300 to 1000m). Therefore, the collective contribution of all vehicles is required to reach the target coverage area. Unfortunately, the decentralized VANET architecture with large and simultaneous transmissions makes reliable and secure broadcasting a discussed research challenge.
[0006] A few years ago, the Federal Communications Commission (FCC) approved 75MHz (5.850-5.925GHz) of spectrum for multi-channel VANET environments. The spectrum is divided into one control channel (CCH) and six service channels (SCH). The DSRC standard specifies a channel switching scheme to allow vehicles to switch between these two classes of applications. In addition, the DSRC standard also recommends that vehicles should access the CCH every 100 milliseconds (this is called the Synchronization Interval (SI)) to send and receive related status messages, and then reliably and within an acceptable delay range to pass these status messages to neighboring vehicles.
[0007] Typically, internal collisions within a single vehicle can be resolved by the scheduler by allowing the highest priority access class (AC) to transmit first and letting lower priority ACs perform (random) backoff. However, external collisions occur when the same AC queue of different vehicles (e.g., AC0 of vehicle 1 and vehicle 2) is granted transmission opportunities (TXOP) at the same time.
[0008] When a collision occurs, (random) backoff is a common solution to contention problems between different vehicles that want to transmit data at the same time. When a vehicle performs a (random) backoff procedure, it waits for the duration of an additional, randomly selected time slot, which must be greater than 0 and less than CW.
[0009] Since vehicles have no priority and have equal access to the channel, packets may collide again when selecting the same backoff time slot. This is a serious problem for broadcasting safety packets on the CCH, because there may be hundreds of vehicles in the affected area, so the probability of any given time slot being selected by only one vehicle is very low.
[0010] Unlike unicast transmission, successful reception is not acknowledged in broadcast mode, so collisions cannot be detected. Furthermore, this may further induce more collisions if the initially selected contention window (CW) size remains unchanged. Summary of the invention
[0011] In view of this, the present invention provides a method for adjusting a backoff mechanism and a vehicle communication system, which can be used to solve the above technical problems.
[0012] An embodiment of the present invention provides a method for adjusting a backoff mechanism, which is suitable for an in-vehicle communication system, wherein the in-vehicle communication system includes a server. The method includes: obtaining vehicle information of multiple vehicles in multiple road sections by the server, wherein the vehicle information of each vehicle includes corresponding communication information and movement information, the multiple road sections include a first road section, and the multiple vehicles include multiple first vehicles located in the first road section; the server determines a current traffic flow information of each road section based on the vehicle information of each vehicle; the server determines a plurality of probabilities corresponding to each road section at least based on the current traffic flow information of each road section, wherein the plurality of probabilities corresponding to each road section respectively correspond to a plurality of candidate contention window sizes; the server determines a contention window size corresponding to each road section based on the plurality of probabilities corresponding to each road section; and the server controls an information retransmission mechanism of the plurality of first vehicles based on the contention window size corresponding to the first road section.
[0013] An embodiment of the present invention provides an in-vehicle communication system, including a server. The server is configured to: obtain vehicle information of multiple vehicles in multiple road sections, wherein the vehicle information of each vehicle includes corresponding communication information and movement information, the multiple road sections include a first road section, and the multiple vehicles include multiple first vehicles located in the first road section; determine a current traffic flow information of each road section based on the vehicle information of each vehicle; determine multiple probabilities corresponding to each road section at least based on the current traffic flow information of each road section, wherein the multiple probabilities corresponding to each road section respectively correspond to multiple candidate contention window sizes; determine a contention window size corresponding to each road section based on the multiple probabilities corresponding to each road section; and control an information retransmission mechanism of the multiple first vehicles based on the contention window size corresponding to the first road section. BRIEF DESCRIPTION OF THE DRAWINGS
[0014] Figure 1 is a schematic diagram of a vehicle-mounted communication system according to an embodiment of the present invention;
[0015] Figure 2 is a flow chart of a method for adjusting a backoff mechanism according to an embodiment of the present invention;
[0016] Figure 3 It is a probability diagram corresponding to a road section according to an embodiment of the present invention.
[0017] Explanation of symbols
[0018] 10: Vehicle communication system
[0019] 110: Server
[0020] 120:RSU
[0021] VH: Vehicle
[0022] SG1~SGD:Road section
[0023] R1: Coverage
[0024] S210~S250: Steps
[0025] P 1,1 ,P 1,2 ,…,P 1,S ,P 2,1 ,P 2,2 ,…,P 2,S ,P D,1 ,P D,2 ,…,P D,S :Probability
[0026] P1(t)~P D (t), P(t): vector DETAILED DESCRIPTION
[0027] Please refer to Figure 1 , which is a schematic diagram of a vehicle-mounted communication system according to an embodiment of the present invention. Figure 1 In the embodiment, the in-vehicle communication system 10 includes a server 110, an RSU 120 and a plurality of vehicles VH, wherein the server 110 is, for example, a cloud server and / or an in-vehicle communication server, which may be used to receive various data from one or more RSUs 120, but is not limited thereto.
[0028] In the embodiment of the present invention, the vehicle communication system 10 may form a VANET, for example. For those skilled in the art, VANET is a wireless ad hoc network that provides communication between vehicles and on-board units (OBU) and RSUs, and is intended to provide a wireless environment for different vehicles or passengers to exchange information.
[0029] VANET mainly consists of the following three parts: (1) Onboard Unit, RSU and Trust Authority (TA), and different types of data (such as location information, road condition information and emergency information) can be transmitted in VANET. The purpose of VANET is to provide ubiquitous connectivity with mobile users on the road and provide efficient vehicle-to-vehicle (V2V) communication, making intelligent transportation systems (ITS) possible.
[0030] exist Figure 1 In the example, the RSU 120 has a coverage range R1 (e.g., a communication range within the communication capability of the RSU 120), and the vehicle VH within the coverage range R1 communicates through the RSU 120. In addition, vehicles outside the coverage range R1 (e.g., vehicles outside the dotted line) use V2V communication and vehicle-to-infrastructure (V2I) communication, and the V2V communication and V2I communication must be implemented based on corresponding wireless communication protocols.
[0031] exist Figure 1 In the example, the coverage R1 of the RSU 120 may include one or more roads, and the one or more roads may be divided into a plurality of road segments SG1-SGD (D is a positive integer). Figure 1 The aspects (such as length, width, etc.) of the sections SG1-SGD are only used as examples and are not used to limit possible implementations of the present invention. In different embodiments, the designer may determine the aspects of the sections SG1-SGD according to needs.
[0032] As mentioned above, the vehicles in the coverage area R1 can communicate with the RSU 120 , which means that the vehicles VH on the road segments SG1 ˜SGD can communicate with the RSU 120 .
[0033] In some embodiments, the RSU 120 may obtain corresponding vehicle information from each vehicle VH on each road segment SG1 ˜SGD, and the vehicle information of each vehicle VH may include, for example, corresponding communication information and movement information.
[0034] In one embodiment, the communication information of each vehicle VH includes, for example, vehicle-mounted communication information corresponding to each vehicle VH (e.g., various road condition information and / or emergency information, etc.). In addition, the movement information of each vehicle VH is, for example, corresponding global positioning system (GPS) information, which may include the vehicle position and movement speed of each vehicle VH, but is not limited thereto.
[0035] In one embodiment, after the RSU 120 obtains corresponding vehicle information from each vehicle VH on each road segment SG1 ˜SGD, the RSU 120 may provide the collected vehicle information of each vehicle VH to the server 110 for further analysis.
[0036] Similarly, the server 110 may also obtain vehicle information collected by other RSUs managed by the server 110, and the server 110 may perform similar control / configuration on the vehicles corresponding to each RSU based on the information obtained from each RSU.
[0037] To facilitate understanding of the concept of the present invention, the following only takes RSU 120 and its corresponding segments SG1-SGD as an example for explanation, and a person with ordinary knowledge in the art should be able to understand the operations performed by the server 110 based on the information collected by other RSUs.
[0038] In the embodiments of the present invention, the server 110 may be implemented as various intelligent devices and / or computer devices, but is not limited thereto. In some embodiments, the server 110 may include a storage circuit and a processor, wherein the storage circuit may be, for example, any type of fixed or removable random access memory (RAM), read-only memory (ROM), flash memory, hard disk or other similar device or a combination of these devices, and may be used to record multiple program codes or modules.
[0039] In addition, the processor in the server 110 is coupled to the storage circuit and can be a general-purpose processor, a special-purpose processor, a traditional processor, a digital signal processor, multiple microprocessors, one or more microprocessors combined with a digital signal processor core, a controller, a microcontroller, an application-specific integrated circuit (ASIC), a field programmable gate array (FPGA), any other type of integrated circuit, a state machine, an advanced RISC machine (ARM)-based processor, and the like.
[0040] In an embodiment of the present invention, the processor of the server 110 may access the modules and program codes recorded in the storage circuit to implement the method for adjusting the backoff mechanism proposed by the present invention, the details of which are described as follows.
[0041] Please refer to Figure 2 , which is a flow chart of a method for adjusting the backoff mechanism according to an embodiment of the present invention. The method of this embodiment can be Figure 1 The server 110 executes the following Figure 1 Description of components shown Figure 2 Details of each step.
[0042] First, in step S210, the server 110 obtains vehicle information of each of the vehicles VH in the plurality of road segments SG1 to SGD. For ease of explanation, it is assumed that the vehicle information of each vehicle VH obtained by the server 110 includes the vehicle position, moving speed, and road condition information of each vehicle VH, but is not limited thereto.
[0043] In step S220 , the server 110 determines the current traffic flow information of each road segment SG1 ˜SGD based on the vehicle information of each vehicle VH.
[0044] In one embodiment, the current traffic flow information of the dth road segment (d is an index value) of the road segments SG1-SGD can be represented as:
[0045] X d (t) = [x d,1 (t), x d,2 (t), ..., x d,V (t)]
[0046] , where t is the time index value, x d,i (t) represents the vehicle information of the i-th vehicle among multiple vehicles in the d-th road section, 1≤i≤V, i is an integer, and V is the number of multiple vehicles in the d-th road section.
[0047] In one embodiment, x d,i (t) can be characterized as:
[0048]
[0049] in is the moving speed of the i-th vehicle in the d-th road section, is the vehicle position of the i-th vehicle in the d-th road segment (which can be represented by GPS coordinates), M d,i (t) is the vehicle-borne communication information of the i-th vehicle in the d-th road section.
[0050] In one embodiment, for the i-th vehicle in the d-th road segment, it corresponds to the time from the t0-th time point to the t2-th time point. and M d,i (t) may be as shown in the following Table 1, but is not limited thereto.
[0051]
[0052] Table 1
[0053] In one embodiment, the current traffic flow information corresponding to the road segments SG1 to SGD at the tth time point can be represented by X1(t) to X2(t). D (t).
[0054] Then, in step S230, the server 110 determines a plurality of probabilities corresponding to each road segment SG1-SGD based on at least the current traffic flow information of each road segment SG1-SGD, wherein the plurality of probabilities corresponding to each road segment SG1-SGD respectively correspond to a plurality of candidate contention window sizes. Each road segment SG1-SGD may correspond to S candidate contention window sizes, and the S candidate contention window sizes corresponding to the dth road segment may be represented as CW d,1 (t)~CW d,S (t), but is not limited thereto.
[0055] In one embodiment, during the execution of step S230, the server 110 may first obtain at least one piece of historical traffic flow information of each road segment SG1-SGD. Taking the dth road segment as an example, the corresponding at least one piece of historical traffic flow information (the number of which is, for example, M) may be represented as X d (t-1)~X d (tM), but is not limited thereto. In one embodiment, X d (tk) is, for example, the historical traffic flow information corresponding to the tkth time point, and it can also be understood as the current traffic flow information at the tkth time point (k is a positive integer), but is not limited to this.
[0056] Afterwards, the server 110 may determine the plurality of probabilities corresponding to each road segment SG1 ˜SGD based on the current traffic flow information and the historical traffic flow information corresponding to each road segment SG1 ˜SGD.
[0057] In one embodiment, the server 110 may integrate the current traffic flow information and the historical traffic flow information of each road segment SG1-SGD into corresponding statistical traffic flow information. Taking the dth road segment as an example, its corresponding statistical traffic flow information at the tth time point may be represented as:
[0058]
[0059] where h0(k) is the kth coefficient in the impulse response (expressed as h(t)).
[0060] In one embodiment, the coefficients of the impulse response h(t) can be determined by the designer according to the requirements, but are not limited thereto. In one embodiment, the impulse response h(t) describes the response of the system to a unit impulse signal in time, which can be obtained through mathematical tools such as Fourier transform or Laplace transform. The coefficients h in the impulse response in this formula are n (k) describes the response strength and time delay of the input signal at the kth time unit.
[0061] Based on this, in one embodiment, the statistical traffic information corresponding to the road segments SG1-SGD at the tth time point can be represented as But it is not limited to this.
[0062] Afterwards, the server 110 may generate the statistical traffic flow information (eg, ) is fed into a quantum neural network (QNN), wherein the quantum neural network (hereinafter referred to as M1) determines the multiple probabilities corresponding to each road segment SG1~SGD in response to the statistical traffic information corresponding to each road segment SG1~SGD.
[0063] In one embodiment, the multiple probabilities corresponding to the dth road segment at the tth time point can be represented as:
[0064] P d (t) = [P d,1 , P d,2 , ..., P d,S ] T
[0065] , where P d,s is the sth probability among the multiple probabilities corresponding to the dth road segment, S is the number of the multiple probabilities corresponding to the dth road segment. The sth probability corresponds to the sth candidate contention window size among the multiple candidate contention window sizes.
[0066] For example, the multiple probabilities corresponding to the first road segment at the tth time point can be represented as “P1(t)=[P 1,1 , P1,2,...,P1,S] T "; The multiple probabilities corresponding to the second road segment at the tth time point can be represented as "P2(t) = [P 2,1 , P 2,2 , ..., P 2,S ] T "; The multiple probabilities corresponding to the D-th road segment at the t-th time point can be represented as "P D (t) = [P D,1 , P D,2 , ..., P D,S ] T In some embodiments, P1(t)~P D (t) is, for example, a vector of length S, but is not limited thereto.
[0067] Please refer to Figure 3 , which is a probability diagram corresponding to the road sections SG1-SGD according to an embodiment of the present invention. Figure 3 In the example, the multiple probabilities corresponding to the sections SG1 to SGD determined by the quantum neural network M1 can be represented as It is, for example, a vector of length D×S, but is not limited thereto.
[0068] In one embodiment, in order to enable the quantum neural network M1 to have the above capabilities, during the training process of the quantum neural network M1, the designer may feed specially designed training data into the quantum neural network M1 so that the quantum neural network M1 can perform corresponding learning. For example, after obtaining the statistical traffic flow information of a certain road section at a certain time point and the corresponding probabilities of selecting different competition window sizes, the server 110 may feed the statistical traffic flow information and the probabilities corresponding to different competition window sizes as a piece of training data into the quantum neural network M1 in training so that the quantum neural network M1 can perform corresponding learning.
[0069] Based on the similar concept, after different statistical traffic flow information and corresponding competition window sizes are fed into the quantum neural network M1 in training as training data, the quantum neural network M1 can perform corresponding learning. ), the quantum neural network M1 can accordingly predict / determine the probabilities corresponding to different competition window sizes (e.g., P d (t) = [P d,1 , P d,2 , ..., P d,S ] T ), but is not limited thereto.
[0070] Next, in step S240 , the server 110 determines a contention window size corresponding to each of the road segments SG1 ˜SGD based on the plurality of probabilities corresponding to each of the road segments SG1 ˜SGD.
[0071] In one embodiment, the server 110 may determine the contention window size corresponding to each segment SG1-SGD based on a similar method. For ease of description, one of the segments SG1-SGD (hereinafter referred to as the first segment) is used as an example for description.
[0072] In one embodiment, the probability corresponding to the first road segment includes a maximum probability, and the maximum probability corresponds to a specific candidate contention window size of the plurality of candidate contention window sizes. In this case, the server 110 may select the specific candidate contention window size as the contention window size corresponding to the first road segment.
[0073] For example, assuming that the first road segment under consideration is the d-th road segment, the candidate contention window size corresponding to the d-th road segment is, for example, CW d,1 (t)~CW d,S (t), while CW d,1 (t)~CW d,S (t) corresponds to P d P in (t) d,1 ~P d,S). In one embodiment, it is assumed that P d,1 In P d (t) is the maximum probability, the server 110 can choose the corresponding d,1 The CW d,1 (t) (ie, the specific candidate contention window size corresponding to the maximum probability) is used as the contention window size corresponding to the dth road segment.
[0074] exist Figure 3 In the scenario, it is assumed that the first segment under consideration is segment SG1, wherein the candidate contention window size corresponding to segment SG1 is, for example, CW 1,1 (t)~CW 1,S (t), while CW 1,1 (t)~CW 1,S (t) respectively correspond to P in P1(t) 1,1 ~P 1,S ). In one embodiment, it is assumed that P 1,1 In P1(t), the server 110 can select the corresponding 1,1 The CW 1,1 (t) is the contention window size of the segment SG1.
[0075] For another example, assume that the first segment under consideration is segment SG2, where the candidate contention window size corresponding to segment SG2 is, for example, CW 2,1 (t)~CW 2,S (t), while CW 2,1 (t)~CW 2,S (t) respectively correspond to P in P2(t) 2,1 ~P 2,S ). In one embodiment, assuming P 2,2 The server 110 can select the one corresponding to P2(t) to obtain the maximum probability. 2,2 The CW 2,2 (t) is the contention window size of the segment SG2.
[0076] For another example, assume that the first segment under consideration is segment SGD, wherein the candidate contention window size corresponding to segment SGD is, for example, CW D,1 (t)~CW D,S (t), while CW D,1 (t)~CW D,S (t) corresponds to P D P in (t) D,1 ~P D,S ). In one embodiment, it is assumed that P D,S In P D (t) is the maximum probability, the server 110 can choose the corresponding D,S The CWD,S (t) is the contention window size for segment SGD.
[0077] In other embodiments, the server 110 may also adjust the method of determining the specific candidate contention window size of each road segment SG1 -SGD according to the designer's requirements, and may not be limited to the method described in the above embodiments.
[0078] For example, assume that the first road segment considered is the dth road segment, and P d,1 In P d (t) is the second largest probability, the server 110 may choose to use the corresponding d,1 The CW d,1 (t) as the contention window size corresponding to the dth road segment. That is, the server 110 may use the candidate contention window size corresponding to the second largest probability as the specific candidate contention window size, but is not limited thereto.
[0079] For ease of explanation, the vehicles located in the first road section are referred to as first vehicles below, but are not limited thereto. Based on this, after determining the contention window size corresponding to the first road section, in step S250, the server 110 controls the information retransmission mechanism of the plurality of first vehicles according to the contention window size corresponding to the first road section.
[0080] In one embodiment, the server 110 may send a control command indicating a contention window size corresponding to the first road segment to each first vehicle through the RSU 120 , so that each first vehicle executes an information retransmission mechanism accordingly.
[0081] For ease of understanding, one of the multiple first vehicles (hereinafter referred to as a reference vehicle) is taken as an example for explanation below.
[0082] In one embodiment, the reference vehicle (hereinafter referred to as RH) may receive the contention window size corresponding to the first road segment from the server 110 (via the RSU 120) and perform the information retransmission mechanism accordingly. That is, when the reference vehicle RH performs the random backoff procedure later for some reason, the reference vehicle RH will wait for an additional, randomly selected time slot duration, and the randomly selected duration must be greater than 0 and less than the contention window size corresponding to the first road segment.
[0083] Therefore, in one embodiment, in response to the reference vehicle RH determining that a first data packet cannot be successfully broadcast, the reference vehicle RH performs a random backoff procedure based on the contention window size corresponding to the first road segment to obtain a first backoff time slot (which corresponds to a duration randomly selected in the above manner). For example, when the reference vehicle RH performs a random backoff procedure at the jth time slot and randomly selects a duration (whose length is, for example, p time slots), the reference vehicle RH can use the j+pth time slot as the first backoff time slot (j and p are positive integers), but is not limited thereto.
[0084] Thereafter, the reference vehicle RH may attempt to broadcast the first data packet (which was previously unsuccessfully broadcast) again during the first backoff time slot.
[0085] In some embodiments, the reference vehicle RH may also adaptively adjust the contention window size for executing the random backoff procedure according to the packet sending situation of the reference vehicle RH.
[0086] In one embodiment, in response to determining that the first data packet has expired, the reference vehicle RH may determine a first specific contention window size based on a contention window size corresponding to the first road segment.
[0087] In one embodiment, the reference vehicle RH may obtain a first reference contention window size by dividing the contention window size corresponding to the first road segment by a specific factor (e.g., 2). Afterwards, in response to determining that the first reference contention window size is less than the window size lower limit, this means that the reference vehicle RH is not allowed to use the first reference contention window size that is too low as the contention window size for executing the random backoff procedure. Based on this, the reference vehicle RH may use the contention window size corresponding to the first road segment as the first specific contention window size. That is, the reference vehicle RH may not change the contention window size for executing the random backoff procedure.
[0088] On the other hand, in response to determining that the reference contention window size is not less than the window size lower limit, it means that the reference vehicle RH is allowed to use a lower first reference contention window size as the contention window size for executing the random backoff procedure. Based on this, the reference vehicle RH can use the first reference specific contention window size as the first specific contention window size.
[0089] Thereafter, in one embodiment, in response to the reference vehicle RH determining that a second data packet has not been successfully broadcast, the reference vehicle RH may perform a random backoff procedure based on the first specific contention window size to obtain a second backoff slot, and attempt to broadcast the second data packet again in the second backoff slot.
[0090] In short, if the reference vehicle RH fails to successfully broadcast a data packet (such as the first data packet mentioned above) within the time limit, resulting in the data packet being overdue, the reference vehicle RH can adaptively reduce the contention window size used to execute the random backoff procedure so that subsequent data packets can be sent out earlier with fewer backoff time slots.
[0091] In some embodiments, the reference vehicle RH may repeatedly execute the above mechanism.
[0092] For example, in one embodiment, in response to determining that the second data packet has expired, the reference vehicle RH may determine the second specific contention window size based on the first specific contention window size.
[0093] In one embodiment, the reference vehicle RH may obtain a second reference contention window size by dividing the first specific contention window size by a specific factor (e.g., 2). Afterwards, in response to determining that the second reference contention window size is less than the window size lower limit, this means that the reference vehicle RH is not allowed to use the second reference contention window size that is too low as the contention window size for executing the random backoff procedure. Based on this, the reference vehicle RH may use the first specific contention window size as the second specific contention window size. That is, the reference vehicle RH may not change the contention window size for executing the random backoff procedure.
[0094] On the other hand, in response to determining that the second reference contention window size is not less than the window size lower limit, this means that the reference vehicle RH is allowed to use the lower second reference contention window size as the contention window size for executing the random backoff procedure. Based on this, the reference vehicle RH can use the second reference specific contention window size as the second specific contention window size.
[0095] Afterwards, in one embodiment, in response to the reference vehicle RH determining that the third data packet cannot be successfully broadcast, the reference vehicle RH may perform a random backoff procedure based on the second specific contention window size to obtain a third backoff time slot, and attempt to broadcast the third data packet again in the third backoff time slot. For related details, please refer to the description in the previous embodiment, which will not be repeated here.
[0096] In summary, the method of the embodiment of the present invention can determine multiple probabilities corresponding to each road section after obtaining the current traffic flow information of each road section, and these probabilities correspond to multiple candidate contention window sizes. Afterwards, the method of the embodiment of the present invention can select the contention window size corresponding to each road section according to the corresponding multiple probabilities, and control the information retransmission mechanism of each vehicle in each road section accordingly. In this way, a more appropriate contention window size can be selected for the vehicles in each road section, so that the vehicles in each road section can better execute information retransmission mechanisms such as random backoff procedures.
[0097] From another point of view, common traffic prediction models and prediction methods mostly use linear time series models, such as autoregressive models, moving average models, autoregressive moving average models, and autoregressive integrated moving average models.
[0098] Since network traffic time series is a nonlinear process, traditional linear models cannot capture the uncertainty and time-varying nature of network traffic. How to capture abnormal / individual / outstanding traffic characteristics, such as self-similarity on a large time scale, short-range dependence, and multiple fractals on a small time scale. Considering the impact of individual heterogeneity on traffic volume, the embodiment of the present invention uses probability-time as a theory to capture individual characteristics or abnormal events, and then uses quantum neural networks to propose a retreat mechanism that can comprehensively consider the relationship between the dynamically developing traffic volume, vehicle speed, and uncertainty of the competition window.
[0099] Specifically, the embodiment of the present invention infers that the vehicle's movement information and vehicle communication information are time-varying, uncertain, and have a trade-off relationship (which is similar to the form of Heisenberg's uncertainty principle), and estimates the probabilities corresponding to different contention window sizes accordingly.
[0100] In theory, neural networks are able to provide better performance in handling nonlinear relationships between output and input. Neural networks have been successfully used to model complex nonlinear systems and predict signals for a wide range of engineering applications. However, to make artificial intelligence more realistic, the key bottleneck is computing speed, and one of the main differences between classical data and quantum data is that quantum data cannot be copied.
[0101] Since the vehicle's mobility information and on-board communication information are affected by different factors every second (such as weather, traffic accidents, the speed of neighboring vehicles, etc.), in principle, if you want to obtain a suitable contention window for each time slot in different traffic flows, you need an infinite number of samples, which will lead to an exponential demand for classical computing resources.
[0102] A typical data-driven model treats traffic volume as a random process, extracts or learns structural patterns from massive historical traffic volume data, and then simulates traffic volume. Therefore, data-driven models require representative, high-quality training data to achieve accurate simulation. However, data-driven models also have other problems. For example, the impact of individual heterogeneity on traffic volume is not considered, making it difficult to respond / present abnormal events, which also limits the applicability of data-driven models to simulate traffic volume.
[0103] To improve the above problems, many models add individual accurate state data (e.g., speed, acceleration, and distance between vehicles) to improve granularity. However, it is difficult to collect such accurate state data in high-speed or high-vehicle traffic.
[0104] Even if these accurate state data are available, complex calculations are still required to construct individual parameter sets. Some studies also use empirical parameters or simple statistical methods to estimate average state parameters, but they still cannot reflect the impact of individual heterogeneity.
[0105] The instantaneous processing of massive amounts of data using quantum technology has become one of the most promising tools in the field of machine learning. Quantum computing is a new computing method based on the laws of quantum mechanics, using quantum properties such as superposition, entanglement, and interference of quantum states. The combination of quantum computing and neural networks produces quantum neural networks. Therefore, in response to the above needs, the application of quantum computing combined with the simulation of traffic flow based on individual heterogeneity is one of the spirits of the embodiments of the present invention.
[0106] Due to individual heterogeneity, the position of vehicles in high-speed traffic changes irregularly, making it difficult to describe the traffic flow. Since it is difficult to determine the exact position of the vehicle, the general method is to use the average position (the ideal position at the average speed) to approximate the description. In addition, the position of the vehicle fluctuates around the corresponding average position in each time slot, forming a "probability-time" distribution. In other words, the irregular evolution of the vehicle position is difficult to describe using a classical deterministic framework. Therefore, from the perspective of uncertainty, expressing it in the form of probability is a feasible and more accurate method, which is the starting point of the embodiment of the present invention.
[0107] Although the present invention is disclosed in conjunction with the above embodiments, they are not intended to limit the present invention. Any ordinary technician in the technical field may make slight changes and modifications without departing from the spirit and scope of the present invention. Therefore, the scope of protection of the present invention shall be based on the definition of the attached claims.
Claims
1. A method for adjusting a backoff mechanism, suitable for an in-vehicle communication system, wherein the in-vehicle communication system comprises a server, include: Obtaining, from the server, vehicle information of a plurality of vehicles in a plurality of road sections, wherein the vehicle information of each of the vehicles includes corresponding communication information and movement information, the road sections include a first road section, and the vehicles include a plurality of first vehicles located in the first road section; The server determines the current traffic flow information of each road section based on the vehicle information of each vehicle; The server determines, at least according to the current traffic flow information of each road section, a plurality of probabilities corresponding to each road section, wherein the probabilities corresponding to each road section respectively correspond to a plurality of candidate contention window sizes; The server determines the contention window size corresponding to each road section based on the probabilities corresponding to each road section; as well as The server controls the information retransmission mechanism of the first vehicles according to the contention window size corresponding to the first road section. 2 . The method as claimed in claim 1 , wherein the server obtains the vehicle information of each of the vehicles from a roadside unit that manages the road sections. 3 . The method as claimed in claim 1 , wherein the communication information of each of the vehicles comprises vehicle-mounted communication information corresponding to each of the vehicles, and the movement information of each of the vehicles comprises a corresponding vehicle position and movement speed. The method as claimed in claim 3 , wherein the in-vehicle communication information corresponding to each of the vehicles includes road condition information.
5. The method of claim 3, wherein the current traffic flow information of the dth road segment of the road segments is characterized as: X d (t)=[x d,1 (t),x d,2 (t),…,x d,V (t)] Where t is the time index value, x d,i (t) vehicle information representing the i-th vehicle in the d-th road section, 1≤i≤V, i is an integer, and V is the number of the vehicles in the d-th road section.
6. The method according to claim 5, in: in is the moving speed of the i-th vehicle in the d-th road section, is the vehicle position of the i-th vehicle in the d-th road section, M d,i (t) is the vehicle-mounted communication information of the i-th vehicle in the d-th road section.
7. The method of claim 1, wherein the step of determining the probabilities corresponding to each of the road sections at least based on the current traffic flow information of each of the road sections include: Obtaining at least one historical traffic flow information of each road section from the server; The server determines the probabilities corresponding to each of the road sections based on the current traffic flow information corresponding to each of the road sections and the at least one historical traffic flow information.
8. The method of claim 7, wherein the step of determining the probabilities corresponding to each of the road sections based on the current traffic flow information of each of the road sections and the at least one historical traffic flow information is as follows: include: Integrate the current traffic flow information and the at least one historical traffic flow information of each road section into corresponding statistical traffic flow information; The statistical traffic flow information corresponding to each of the road sections is fed into a quantum neural network, wherein the quantum neural network determines the probabilities corresponding to each of the road sections in response to the statistical traffic flow information corresponding to each of the road sections.
9. The method as claimed in claim 8, wherein the statistical traffic flow information of the dth road segment among the road segments is characterized as: Where t is the time index value, h 0 (k) is the kth coefficient in an impulse response, X d (tk) is one of the at least one historical traffic flow information corresponding to the tk-th time point, and M is the number of the at least one historical traffic flow information.
10. The method of claim 8, wherein the probabilities corresponding to the d-th road segment are characterized as: P d (t)=[P d,1 ,P d,2 ,…,P d,S ] T Where P d,s is the s-th probability among the probabilities corresponding to the d-th road segment, S is the number of the probabilities corresponding to the d-th road segment, and the s-th probability corresponds to the s-th candidate contention window size among the candidate contention window sizes.
11. The method of claim 10, wherein the probabilities corresponding to each of the road sections determined by the quantum neural network are characterized as: Where D is the number of these road sections.
12. The method of claim 1, wherein the probabilities corresponding to the first road segment include a maximum probability, the maximum probability corresponds to a specific candidate contention window size among the candidate contention window sizes, and the contention window size corresponding to the first road segment is the specific candidate contention window size.
13. The method of claim 1, wherein the vehicle communication system further comprises a reference vehicle among the first vehicles, and the method further comprises: include: Receiving, by the reference vehicle, the contention window size corresponding to the first road segment from the server; In response to the reference vehicle determining that the first data packet cannot be successfully broadcast, the reference vehicle performs a random backoff procedure based on the contention window size corresponding to the first road section to obtain a first backoff time slot; The reference vehicle attempts to broadcast the first data packet again during the first backoff time slot.
14. The method of claim 13, further comprising: include: In response to determining that the first data packet has expired, determining, by the reference vehicle, a first specific contention window size based on the contention window size corresponding to the first road segment; In response to the reference vehicle determining that the second data packet cannot be successfully broadcasted, the reference vehicle executes the random backoff procedure based on the first specific contention window size to obtain a second backoff time slot; The reference vehicle attempts to broadcast the second data packet again during the second backoff time slot.
15. The method of claim 14, wherein the step of determining the first specific contention window size by the reference vehicle based on the contention window size corresponding to the first road segment include: Obtaining a first reference contention window size by dividing the contention window size corresponding to the first link segment by a specific factor; In response to determining that the first reference contention window size is smaller than a window size lower limit, using the contention window size corresponding to the first road segment as the first specific contention window size; In response to determining that the first reference contention window size is not less than the window size lower limit, the first reference specific contention window size is used as the first specific contention window size.
16. The method of claim 14, further comprising: include: In response to determining that the second data packet has expired, determining, by the reference vehicle, a second specific contention window size based on the first specific contention window size; In response to the reference vehicle determining that the third data packet cannot be successfully broadcasted, the reference vehicle executes the random backoff procedure based on the second specific contention window size to obtain a third backoff time slot; The reference vehicle attempts to broadcast the third data packet again during the third backoff time slot.
17. The method of claim 16, wherein the step of determining the second specific contention window size by the reference vehicle based on the first specific contention window size include: Obtaining a second reference contention window size by dividing the first specific contention window size by a specific factor; In response to determining that the second reference contention window size is smaller than a window size lower limit, using the first specific contention window size as the second specific contention window size; In response to determining that the second reference contention window size is not less than the window size lower limit, the second reference specific contention window size is used as the second specific contention window size.
18. A vehicle communication system, include: A server configured to: Obtaining vehicle information of a plurality of vehicles in a plurality of road sections, wherein the vehicle information of each vehicle includes corresponding communication information and movement information, the road sections include a first road section, and the vehicles include a plurality of first vehicles located in the first road section; Determine the current traffic flow information of each road section based on the vehicle information of each vehicle; Determining a plurality of probabilities corresponding to each of the road sections at least according to the current traffic flow information of each of the road sections, wherein the probabilities corresponding to each of the road sections respectively correspond to a plurality of candidate contention window sizes; Determining a contention window size corresponding to each road segment based on the probabilities corresponding to each road segment; as well as The information retransmission mechanism of the first vehicles is controlled according to the contention window size corresponding to the first road section.
19. The in-vehicle communication system as claimed in claim 18, further comprising a reference vehicle among the first vehicles, wherein the reference vehicle is configured to: receiving the contention window size corresponding to the first link segment from the server; In response to the reference vehicle determining that the first data packet cannot be successfully broadcast, performing a random backoff procedure based on the contention window size corresponding to the first road segment to obtain a first backoff time slot; Attempt to broadcast the first data packet again in the first backoff time slot.
