Information age optimization method in internet of vehicles under error-prone channel condition

Abstract

The invention relates to the technical field of Internet of Vehicles, and particularly discloses an information age optimization method in the Internet of Vehicles under error-prone channel conditions, which comprises the following steps of: (1) modeling a vehicle data extraction and base station service process as a Poisson point process; (2) constructing a dual-state channel model containing an ideal channel and an error channel, wherein the transition probability of the channel state is dynamically adjusted by the Doppler effect caused by the vehicle speed; (3) modeling information Aol based on a queuing theory; (4) the optimal data extraction rate is determined through simulation, and the vehicle data extraction rate is dynamically adjusted to minimize AoI according to the real-time channel state and the number of vehicles; and (5) designing a dynamic optimization algorithm, and adjusting the data extraction rate of the vehicle on line according to the real-time channel state and the queue load so as to maintain the optimal system AoI.

Description

Technical Field

[0001] The present invention relates to the technical field of Internet of Vehicles, and particularly to an information age optimization method in Internet of Vehicles under error-prone channel conditions. Background Art

[0002] Currently, in-vehicle network technology is evolving towards the development direction of ultra-low latency and ultra-high reliability. Diverse real-time vehicle applications rely on in-vehicle network technology. However, in practical applications, in-vehicle networks still face many challenges. In the context of the Internet of Vehicles (IoV), vehicles often have difficulty independently collecting sufficient data, and their computing resources often cannot meet task requirements. Therefore, they rely on the Internet of Vehicles for data processing. However, frequent data exchange between multiple vehicles and roadside units or base stations can lead to significant resource consumption and even network congestion. Optimizing link scheduling can effectively alleviate this problem. Nevertheless, traditional metrics based on communication rate or reliability cannot comprehensively reflect the communication performance of the system. Therefore, researchers are urgently in need of a new criterion to measure the freshness of information, which has promoted the emergence of the concept of Age of Information (AoI).

[0003] AoI refers to the time interval in a communication system from the generation of a specific piece of information to the reception and processing of that information. This concept is crucial in wireless communication, sensor networks, and real-time data transmission. In a communication system, the update frequency of information usually depends on its freshness, and freshness can be measured by AoI. A lower AoI indicates that the information is more real-time and fresh. Therefore, reducing AoI helps improve the performance of the communication system by reducing transmission latency and enhancing real-time capabilities.

[0004] In addition, the AoI concept plays an important role in optimizing data transmission strategies to ensure the timely delivery of critical information. If the AoI exceeds a certain threshold, it may not provide effective information for tasks with high real-time requirements, potentially leading to incorrect judgments. In addition, AoI can also help estimate the information changes during data transmission. For example, in an autonomous driving scenario, the vehicle position changes over time, and AoI management helps with more precise vehicle control to ensure road safety. By managing and minimizing AoI, the system can be improved to meet the requirements of real-time communication and data transmission.

[0005] In the communication process of the vehicle-to-everything (V2X) network, a large amount of data is exchanged between vehicles and base stations. The base station provides computing power and service support for the vehicles within its coverage area. The resource allocation strategy in the environment significantly affects the waiting time and transmission interval of data packets. Therefore, designing an appropriate resource allocation strategy can help reduce the average age of information (AoI) of the system, and this problem has been widely studied. However, the service rate of each base station is limited; if the request frequency from vehicles is too high, it may increase the number of requests in the queue, thus affecting the freshness of the information received by the base station. Therefore, it is crucial to adjust the request frequency of vehicles to match the service rate of the base station.

[0006] In addition, recent studies often adopt ideal channel models while ignoring the uncertainty of channels in the real world, such as signal attenuation and data packet collisions. In a channel with errors, when the channel is in an unfavorable state, data packets will experience significant interference, resulting in transmission failures and the need for retransmission. This will significantly affect the freshness of information, leading to an increase in AoI. Therefore, considering the impact of the error channel on AoI helps to better characterize the performance of the system in AoI modeling and design. Based on this, the present invention introduces an error channel model to accurately reflect the actual communication environment, optimize AoI, and enhance the real-time response ability of the system. Summary of the Invention

[0007] Aiming at the deficiencies of the prior art, the present invention provides a method for optimizing the age of information in a vehicle-to-everything network under error-prone channel conditions. By introducing an error channel model, it more accurately reflects the actual communication environment, derives an analytical expression of AoI considering the data extraction rate and channel packet loss probability, and proves its convexity, so that convex optimization methods can be used to efficiently optimize parameters to minimize AoI.

[0008] To achieve the above objectives, the present invention is realized through the following technical solutions:

[0009] A method for optimizing the age of information in a vehicle-to-everything network under error-prone channel conditions, comprising the following steps:

[0010] (1) Model the process of vehicle data extraction and base station service as a Poisson point process;

[0011] (2) Construct a two-state channel model including an ideal channel and an error channel, where the transition probability of the channel state is dynamically adjusted by the Doppler effect caused by the vehicle speed;;

[0012] (3) Model the age of information Aol based on queuing theory, specifically including:

[0013] Adopt the M / M / 1 model, combine the collision probability and the discard probability, and derive the age-of-information AoI expression;

[0014] Adopt the D / M / 1 model, calculate the expected waiting time using the Lambert W function, and derive the Age of Information (AoI) expression;

[0015] (4) Determine the optimal data extraction rate through simulation, and dynamically adjust the vehicle data extraction rate according to the real-time channel state and the number of vehicles to minimize the AoI;

[0016] (5) Design a dynamic optimization algorithm to online adjust the vehicle data extraction rate according to the real-time channel state and queue load to maintain the optimal system AoI.

[0017] The present invention further defines the technical solution:

[0018] Preferably, the Poisson point process includes a homogeneous Poisson process and a non-homogeneous Poisson process. For the homogeneous Poisson process, the arrival delays are independent and identically distributed exponential random variables, and for the non-homogeneous Poisson process, the arrival rate is dynamically adjusted according to the intensity function;

[0019] For the above homogeneous Poisson process: for {a (i)} i≥1 being a homogeneous Poisson process with rate Λ, the corresponding arrival delays are independent and identically distributed (IID) exponential random variables with an average of 1 / Λ. Therefore,

[0020] P(T i ≤t) = 1 - e -Λt .

[0021] For the non-homogeneous Poisson process: for {A (i)} i≥1 being a non-homogeneous Poisson process with intensity Λ(t), the corresponding arrival rate can be estimated as follows. The time to the first arrival has the following distribution,

[0022]

[0023] After that, given the first arrival delay T1 = A1, the conditional time to the second arrival T2,

[0024]

[0025] Similarly for i = 3, 4,...... For a non-homogeneous Poisson process {A (i)} i≥1 , each point is independently distributed in the interval a ∈ [0, t) with the following distribution,

[0026]

[0027] Preferably, in the above (2):

[0028] The probability that the ideal channel remains ideal is P i , and the probability that the error channel remains in error is P p ;

[0029] The calculation formula for the Doppler frequency is f d = f c v / c (c = 3×10 8 m / s).

[0030] where v is the vehicle speed, f c represents the carrier frequency, and c is the speed of light;

[0031] The channel state transition probability is modeled by the Marcum function, specifically expressed as:

[0032]

[0033] where γ is the received signal-to-noise ratio and γth is the channel state determination threshold.

[0034] Preferably, in the above (3):

[0035] For the M / M / 1 system, the AoI expression is:

[0036]

[0037] where λ is the data extraction rate, μ is the base station service rate, and Pcollision is the collision probability;

[0038] For the D / M / 1 system, the AoI expression is solved by the Lambert W function:

[0039]

[0040] where ρ = λ / μ.

[0041] Preferably, the specific process of the dynamic optimization algorithm in the above (5) includes:

[0042] Real-time monitoring of the channel state (ideal / error), queue length, and collision probability;

[0043] If the channel state or queue load changes, recalculate the optimal data extraction rate λ opt ;

[0044] By adjusting the data request interval of the vehicle, converge the current data extraction rate to λ opt .

[0045] Preferably, it further includes:

[0046] Under the 3GPP LTE Cat.M1 protocol, the AoI is reduced by increasing the collision time slots, and the optimized time slot interval is 3 LTE time slots;

[0047] Analyze the impact of the number of vehicles, channel packet loss rate, and collision time slots on AoI through the Python simulation platform, and verify the performance advantages of the D / M / 1 system compared to the M / M / 1 system.

[0048] Beneficial effects

[0049] Compared with the prior art, the following beneficial effects are achieved:

[0050] By introducing a dynamic channel model, deriving the AoI expression, and designing an online optimization algorithm, the present invention significantly reduces the average age of information in the vehicle networking system. Simulations show that the D / M / 1 system outperforms the traditional M / M / 1 system under various channel conditions. The optimal data extraction rate strategy can reduce AoI by up to 30%. This method is applicable to high-speed mobile scenarios and provides a reliable solution for real-time communication in vehicle networking. Description of the drawings

[0051] Figure 1 It is a diagram showing the impact of the data extraction rate on the average age of information under different systems in this embodiment;

[0052] Figure 2 It is a diagram showing the impact of the data extraction rate selected by the random strategy and the optimal strategy on the age of information under different systems in this embodiment;

[0053] Figure 3 It is a diagram showing the impact of the number of vehicles on the age of information at different data extraction rates in this embodiment;

[0054] Figure 4 It is a diagram showing the impact of the channel discard probability on the age of information at different data extraction rates in this embodiment;

[0055] Figure 5 It is a diagram showing the impact of different collision time slots on the age of information of the system in this embodiment. Detailed implementation manners

[0056] Next, the technical solutions in the embodiments of the present invention will be clearly and completely described in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative efforts shall fall within the protection scope of the present invention.

[0057] Embodiment 1

[0058] This embodiment considers that the base station and the roadside unit (RSU) provide services for vehicles. The vehicle and the base station communicate through a wireless channel, and they follow the 3GPP LTE Cat.M1 rules. This embodiment assumes that the process of vehicle data extraction follows a homogeneous Poisson point process with a density of λ. The vehicle transmits data packets to the nearest base station through the wireless uplink channel, and all vehicles use the same transmission power P. Similarly, the service process of the base station also follows a homogeneous Poisson point process with a density of μ. Without loss of generality, the present invention considers a base station near a vehicle. And a base station can be related to multiple vehicles.

[0059] In the queue selection problem, the messages sent by the vehicle to the base station follow the First-Come-First-Served (FCFS) principle. Consider two queue models in queuing theory, the D / M / 1 model and the M / M / 1

[0060] model. The age of information for the communication between the vehicle and the base station can be derived theoretically. Specifically, the present invention considers the local data arrival of the vehicle as the starting point. There is a certain time interval between data arrival and transmission. During the data transmission process,

[0061] there is a certain delay. Due to the uncertainty of the channel conditions, errors may occur during the packet transmission, resulting in the need for retransmission.

[0062] Therefore, it is necessary to consider the additional delay caused by the probability of channel errors. The data may need to wait in the queue at the receiving end. When the data is successfully received, the accurate age of information can be obtained according to the difference between the timestamp generated by the data packet and the real time. The present invention analyzes the age of information of the system through theoretical calculations.

[0063] Based on the above, this embodiment provides an age of information optimization method in a vehicle-to-everything (V2X) network under error-prone channel conditions, which specifically includes the following steps:

[0064] (1) Model the processes of vehicle data extraction and base station service as Poisson point processes, specifically:

[0065] Vehicle data extraction process: The local data extraction process of the vehicle can be described as a Poisson point process. This process A is a Poisson point arrival process with the following two properties.

[0066] The process has random points A([a, b)) located in a bounded interval which can be measured by a Poisson random variable with an average of λ([a, b)) as a Radon measure (non-negative);

[0067] The points of process A fall in intervals [a1, b1), [a2, b2),..., [ak , b k ) to form k independent random variables with the desired

[0068] λ([a1, b1)), λ([a2, b2)),..., λ([a k , b k ));

[0069] For a point process {A (i)} i≥1 defined on non - negative real numbers, the points can be arranged in ascending order according to their arrival times, A1 ≤ A2 ≤ A3 ≤..., so the effective distance between two points is a random variable, the time interval,

[0070] T i = A i - A i-1 . (1.1)

[0071] For,

[0072] T1 = A1, (1.2)

[0073] And,

[0074] i = 2, 3, 4,...(1.3)

[0075] is the so - called inter - arrival time or arrival delay;

[0076] In this embodiment, the Poisson point process includes a homogeneous Poisson process and a non - homogeneous Poisson process.

[0077] Among them, for the homogeneous Poisson process: for {A (i)} i≥1 being a homogeneous Poisson process with rate Λ, the corresponding arrival delays

[0078] are independent and identically - distributed (IID) exponential random variables, with an average of 1 / Λ. Therefore,

[0079] P(T

[0080] ≤ t) = 1 - e i .(1.4) -Λt

[0081] For the non - homogeneous Poisson process: for {A (i)} i≥1 being a non - homogeneous Poisson process with intensity Λ(t), the corresponding arrival rate

[0082] ​Can be estimated as follows. The time of first arrival has the following distribution,

[0083]

[0084] After that, given the first arrival delay T1 = A1, the second conditional arrival time T2,

[0085]

[0086] Similarly for i = 3, 4,...... For a non-homogeneous Poisson process {A (i)} i≥1 , each point is independently distributed in the interval a ∈ [0, t) with the following distribution,

[0087] Build a two-state channel model that includes an ideal channel and an error channel, specifically:

[0088]

[0089] The channel state is divided into two states: ideal and poor. Frames transmitted in the ideal channel will not fail, while transmissions in the error channel will fail with a certain probability, and the states of the ideal channel and the error channel will change. The probability that the ideal channel remains ideal is P

[0090] , and the probability that the error channel remains in error is P i , then the probability that the ideal channel transitions to the error channel is 1 - P p , and the probability that the error channel transitions to the ideal is 1 - P i , and the probability that the error channel transitions to the ideal is 1 - P p . It should be particularly noted that the mobility of the vehicle itself will affect the communication performance. In the section on 0 model mathematical modeling, the present invention considers the impact of vehicle speed on communication as the impact of vehicle speed on the channel, because the Doppler effect caused by vehicle speed affects the retention and transition probabilities of the error channel and the ideal channel.

[0091] Model mathematical modeling

[0092] Let θ represent the maximum rate at which the channel can transmit link layer frames to the vehicle (θ) = bit rate / frame bit rate. Let v represent the vehicle speed, and f c represent the carrier frequency. The Doppler frequency is

[0093] f d = f c v / c (c = 3 × 10 8 m / s).(1.9)

[0094] Considering F as the fading margin and given the physical modulation and coding scheme. If the received Signal Noise Ratio (SNR) is below the threshold,

[0095] SNR threshold = E[SNR] / F.(1.10)

[0096] then the channel is in a poor state. Otherwise, the channel is in an ideal state. The average probability of frame transmission failure due to errors in the channel is,

[0097]

[0098] There is also a medium coefficient,

[0099]

[0100] Also,

[0101] ρ = J0(2πf d / θ).(1.13)

[0102] where ρ is the Gaussian correlation coefficient of two samples of the amplitude of a fading channel with frequency f d . 1 / θ is the frame transmission time over the channel, and J0(.) is the zero-order Bessel function. Finally, the static transition probability of the system is expressed using the Marcum function as follows,

[0103]

[0104] where the probability that the ideal channel remains ideal is P i , and the probability that the error channel remains in error is P p .

[0105] According to the work of Pokhrel et al.

[49] , the probability that an update in the channel is deprecated once,

[0106]

[0107] where v L is the probability of at least one frame transmission failure in the poor channel, and l L is the probability of at least one frame transmission failure in the ideal channel

[0108] .

[0109] (3) Age of Information calculation: For an M / M / 1 system, where the vehicle extraction data rate is λ and the base station service rate is μ, that is, update packets are generated as a Poisson process with rate λ and submitted to the system, so the time between state updates

[0110] is generated and submitted to the system, so the time between state updates

[0111] Arrival time X i are independent and identically distributed exponential random variables. Then,

[0112] E[X] = 1 / λ. (1.16)

[0113] In addition,

[0114] E[X 2 = 2 / λ 2 . (1.17)

[0115] Since a base station is connected to M vehicles, there is a probability of collision in the data transmission of these M vehicles. In the specification of LTE CAT M1, it should be spaced 2 to 3 time slots to avoid collision. The present invention adopts 3 time slots,

[0116] τ c = 3τ t , (1.18)τ t is the time slot interval. Then there is a collision probability,

[0117]

[0118] where, is the time when any two vehicles send data. And because of Equation (1.45), so is a Poisson process

[0119] Adding a constant to the time delay still results in a Poisson process. Then there is,

[0120]

[0121] In addition, the service time is also independent and identically distributed exponentially, and the average service time is,

[0122] E[T ser = 1 / μ. (1.21)

[0123] The optimization objective can be generated, calculate the age of the system, and then find the server utilization rate that minimizes the average age Δ,

[0124] ρ = λ / μ. (1.22)

[0125] Referring to the work of Kual et al.,

[0126] T i = W i + S i . (1.23)

[0127] where W i is the waiting time, the time from generation to the end of queuing. Si The service time is the time for transmission to the server. Mistransmission in the channel will lead to a longer waiting time.

[0128] The final average age-of-information time is calculated as

[0129]

[0130]

[0131] where

[0132] f T (t) = μ(1 - ρ)e -μ(1-ρ)t , t ≥ 0. (1.26)

[0133] Considering the collision probability p c and the discard probability p d , Equation (1.24) is updated to

[0134]

[0135] Referring to Equation (1.20), we have

[0136]

[0137] It can be seen that the latter term is just a coefficient and is irrelevant to the independent variable. So we have

[0138]

[0139] Therefore, Equation (1.24) is updated to

[0140]

[0141] The extreme value in Equation (1.30) cannot be easily obtained through mathematical derivation, so numerical solutions need to be obtained through simulation.

[0143] Similarly, in the D / M / 1 system, the average age of information is calculated as

[0144]

[0145] where β is the Laplace transform of the inter-arrival time distribution

[0146] β = L X (μ(1 - β)). (1.32)

[0147] It can be calculated using the Lambert W function W(.)

[0148] β = e -μ(1-β)D = -ρW(-ρ-1 e (-1 / ρ) ).(1.33)

[0149] Considering the collision probability and the discard probability, Equation (1.24) is updated to

[0150]

[0151] With the analytical expressions of M / M / 1 and D / M / 1 in the error channel, the age of information derived above can be analyzed through simulation

[0152]

[0153] (4) Age of information optimization:

[0154] According to Equation (1.30) and Equation (1.34), the extreme point of the function can be obtained through simulation. Replacing the data extraction rate corresponding to the extreme point with the vehicle's own data extraction rate can minimize the age of information of the system

[0155] Such as Figure 1 The present invention compares the average age of information of the D / M / 1 system and the M / M / 1 system under different system utilization rates ρ. In fact, since the value of the system service rate μ is fixed, adjusting the value of ρ is equivalent to adjusting the value of the vehicle data extraction rate λ. It can be seen that the age of information is different in the M / M / 1 and D / M / 1 systems. In the M / M / 1 system, the optimal value of ρ is approximately 0.53, and the corresponding average age of information is approximately 3.49. Between 0.2 and 0.53, a larger ρ will result in a smaller age of information. Between 0.53 and 0.8, a larger ρ will result in an increase in the age of information. In the D / M / 1 system, the optimal value of ρ is approximately 0.515, and the corresponding average age of information is approximately 2.26. Between 0.2 and 0.515, a larger ρ results in a smaller age of information. Between 0.515 and 0.8, a larger ρ results in an increase in the age of information. This is determined by the age of information calculation formula

[0156] Figure 2 Shows the average age of information caused by using different strategies to determine the value of ρ in the D / M / 1 system and the M / M / 1 system. In the D / M / 1 system, the average age of information caused by the random strategy is approximately 15% higher than that of the optimal strategy, while in the M / M / 1 system, the average age of information caused by the random strategy is approximately 32% higher than that of the optimal strategy. It can be seen that in the same system, the optimal strategy is significantly better than the random strategy. This is because choosing an appropriate optimal data extraction rate according to the environmental changes can better adapt to the environmental changes than randomly choosing the data extraction rate. This verifies the effectiveness of the online optimization method of vehicle data extraction rate proposed by the present invention

[0157] Such asFigure 3 It can be observed that as the maximum number of connected vehicles increases, the age of information shows an increasing trend in both the M / M / 1 system and the D / M / 1 system. Moreover, as the value of ρ increases, this increasing trend becomes more evident. In the M / M / 1 system, when the number of vehicles increases from 2 to 10, the average age of information of the curve with ρ = 0.53 increases by 0.04, the average age of information of the curve with ρ = 0.70 increases by 0.16, and the average age of information of the curve with ρ = 0.80 increases by 0.37. In the D / M / 1 system, when the number of vehicles increases from 2 to 10, the average age of information of the curve with ρ = 0.515 increases by 0.02, the average age of information of the curve with ρ = 0.70 increases by 0.7, and the average age of information of the curve with ρ = 0.80 increases by 0.14. This increase can be attributed to the change in the collision probability. As the maximum number of vehicles connected to the base station increases, the adjacent time points for vehicles to send information become closer, resulting in an increase in the collision probability. When the value of ρ is larger, the data arrival frequency of the system is higher, the vehicle requests are more frequent, and collisions are more likely to occur. Moreover, this effect is greater when the number of vehicles is larger. The increase in the collision probability leads to an increase in the waiting time for information to be transmitted in the system, and thus the age of information also increases accordingly.

[0158] This relationship can be further explained as a progressive process. First, the increase in the maximum number of connected vehicles leads to an increase in the collision probability, which in turn leads to an increase in the waiting time for information to be transmitted in the system. As the waiting time increases, the age of information also increases. Therefore, the final result is that the performance in terms of the age of information is affected.

[0159] It is worth noting that the D / M / 1 system performs better than the M / M / 1 system. Regardless of how the maximum number of connected vehicles changes on the coordinate axis, the age of information of the D / M / 1 system is less than that of the M / M / 1 system. This indicates that the D / M / 1 system is superior to the M / M / 1 system. This may be because the D / M / 1 system can handle collisions more effectively and has a lower collision probability, thereby reducing the waiting time of information in the system and ultimately reducing the age of information.

[0160] From Figure 4 The observed curve changes clearly show that as the packet loss probability increases, the age of information also shows an increasing trend. This phenomenon is common in both the M / M / 1 and D / M / 1 systems. This is because the packet loss probability, as an indicator reflecting the channel state, an increase in it means an exacerbation of channel congestion, thus increasing the probability of errors occurring.

[0161] The increase in the packet loss probability leads to an increase in the number of information retransmissions because once a loss occurs, the information needs to be resent. This prolongs the waiting time for information to be transmitted in the system, thereby increasing the age of information. Therefore, the increase in the packet loss probability directly leads to an increase in the age of information.

[0162] By further comparing the M / M / 1 system and the D / M / 1 system, it can be clearly seen that regardless of how the packet loss probability changes on the coordinate axis, the age of information of the M / M / 1 system is always greater than that of the D / M / 1 system. This indicates that the M / M / 1 system is not as good as the D / M / 1 system.

[0163] This phenomenon can be attributed to the difference in system complexity. In the M / M / 1 system, the data retrieval process is not a deterministic Poisson process but a more complex process, resulting in more obvious instability in the performance of the M / M / 1 system compared to the D / M / 1 system. As shown in the figure, as the collision slot increases, the age of information of both the M / M / 1 and D / M / 1 systems shows a decreasing trend. This is because the increase in collision slots reduces the collision probability at adjacent time points when different vehicles send data.

[0164] From Figure 5 it can be seen that when the collision slot increases, the time interval between data transmissions between vehicles becomes larger, so the possibility of collision between adjacent vehicles decreases. This reduces the collision probability of the system and the number of information retransmissions. Since the number of retransmissions decreases, the waiting time for transmission in the system also decreases. Therefore, the age of information of the system will also decrease accordingly. Generally speaking, the increase in collision slots enables the system to more effectively avoid collisions, reduce the number of information retransmissions, thereby reducing the waiting time of the system, and further reducing the age of information. This shows that to a certain extent, increasing the collision slot can improve the performance of the system and reduce the transmission delay of information. Therefore, the D / M / 1 system is more efficient in handling losses and retransmissions, thus showing a lower age of information.

[0165] The present invention models the process of vehicle data extraction and base station service as a Poisson point process, considers the channel model with errors, and derives an analytical expression for the relationship between the vehicle data extraction rate and the AoI.

[0166] Using Python software, the present invention simulates the derived expression (a simple convex function), can obtain the optimal solution, and designs an online optimization algorithm to optimize the average age of information of the system by adjusting the vehicle data extraction rate. Finally, the present invention analyzes the difference in AoI between the D / M / 1 system and the M / M / 1 system.

[0167] The results show that the proposed method effectively reduces the average information timeliness under various channel conditions and enhances the real-time response ability of the system.

[0168] It should be noted that, in this article, relational terms such as first and second are only used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any actual relationship or order between these entities or operations.

Claims

1. A method for optimizing information age in an Internet of Vehicles under error-prone channel conditions, characterized in that: The following steps are involved: (1) The process of vehicle data extraction and base station service is modeled as a Poisson point process; (2) Construct a dual-state channel model consisting of an ideal channel and an error channel, in which the transition probability of the channel state is dynamically adjusted by the Doppler effect caused by the vehicle speed; (3) Modeling information age Aol based on queuing theory, including: The M / M / 1 model is used to combine the collision probability and the discard probability to derive the information age AoI expression; The D / M / 1 model is adopted, the Lambert W function is used to calculate the expected waiting time, and the expression of information age AoI is derived; (4) Determine the optimal data extraction rate through simulation, and dynamically adjust the vehicle data extraction rate to minimize AoI based on the real-time channel status and the number of vehicles; (5) Design a dynamic optimization algorithm to adjust the vehicle data extraction rate online according to the real-time channel status and queue load to maintain the optimal system AoI.

2. The information age optimization method in the Internet of Vehicles under error-prone channel conditions according to claim 1 is characterized in that: The Poisson point process includes a homogeneous Poisson process and a non-homogeneous Poisson process, wherein the arrival delay of the homogeneous Poisson process is an independent and identically distributed exponential random variable, and the arrival rate of the non-homogeneous Poisson process is dynamically adjusted according to the intensity function; The above homogeneous Poisson process: In {A ( i ) } ≥1 For a homogeneous Poisson process with rate Λ, the corresponding arrival delay are independent and identically distributed (IID) exponential random variables with mean 1 / Λ, so, P(T i ≤t)=1-e -Λt . Non-homogeneous Poisson process: For {A ( i ) } ≥1 For a nonhomogeneous Poisson process of strength Λ(t), the corresponding arrival rate can be estimated as follows. First arrival time Has the following distribution, After that, given the first arrival delay T1=A1 and the second condition arrival time T2, The same is true for i = 3, 4, ... For a non-homogeneous Poisson process {A (i) } i≥1 , each point is independently distributed in the interval a∈[0,t) with the following distribution, 3. The information age optimization method in the Internet of Vehicles under error-prone channel conditions according to claim 1 is characterized in that: In (2) described above: The probability that an ideal channel remains ideal is P i , the probability that the error channel remains error-free is P p ; The calculation formula of Doppler frequency is f d =f c v / c(c=3×10 8 m / s). Where v is the vehicle speed, f c represents the carrier frequency, c is the speed of light; The channel state transition probability is modeled by the Malkum function, which is specifically expressed as: Where γ is the received signal-to-noise ratio, and γth is the channel state determination threshold.

4. The method for optimizing information age in the Internet of Vehicles under error-prone channel conditions according to claim 1, characterized in that: In (3): For the M / M / 1 system, the AoI expression is: Where λ is the data extraction rate, μ is the base station service rate, and Pcollision is the collision probability; For the D / M / 1 system, the AoI expression is solved by the Lambert W function: Where ρ = λ / μ.

5. The method for optimizing information age in the Internet of Vehicles under error-prone channel conditions according to claim 1, characterized in that: The specific process of the dynamic optimization algorithm in (5) includes: Real-time monitoring of channel status (ideal / error), queue length and collision probability; If the channel status or queue load changes, recalculate the optimal data extraction rate λ opt ; By adjusting the data request interval of the vehicle, the current data extraction rate converges to λ opt .

6. The method for optimizing information age in a vehicle network under error-prone channel conditions according to any one of claim 1, characterized in that: Also includes: Under the 3GPP LTE Cat.M1 protocol, AoI is reduced by increasing collision time slots; The Python simulation platform is used to analyze the impact of the number of vehicles, channel packet loss rate, and collision time slot on AoI, verifying the performance advantage of the D / M / 1 system over the M / M / 1 system.

Images