An Adaptive Resource Reservation Interval Method Based on SB-SPS

Abstract

The present invention discloses an adaptive resource reservation interval method and system based on SB-SPS, which relates to the technical field of cellular vehicle-to-everything communication, and includes the following steps: S1: The vehicle uses a fuzzy function to respectively perform fuzzy processing on the road vehicle density and the communication resource busy rate as input parameters of the fuzzy inference engine; S2: The fuzzy inference engine calculates the membership degree of RRI and infers the matching RRI value state based on the fuzzy rule base; S3: Defuzzification is performed to select the best RRI value to be applied to the SB-SPS scheduling mechanism to allocate communication resources. The present invention sets the change of the RRI state value as an adaptive interval to reserve resources for transmitting subsequent message packets, reduce the message transmission delay, and improve the packet transmission rate and communication resource efficiency.

Description

Technical Field

[0001] The present invention relates to the technical field of cellular vehicle-to-everything communication, and in particular to an adaptive resource reservation interval method based on SB-SPS. Background Art

[0002] In the application of 5G cellular network communication technology, autonomous driving is one of the specific implementations of ultra-reliable and low-latency applications and is the future development direction in the field of vehicle-to-everything. The NR-V2X standard in the cellular vehicle-to-everything communication specification released by the 3rd Generation Partnership Project supports two communication resource allocation modes, Mode1 and Mode2, to achieve vehicle-to-vehicle communication. Among them, Mode1 performs centralized resource allocation by the base station, and Mode2 realizes distributed resource allocation by the on-vehicle communication terminal. Currently, Mode 2 uses a sensing-based semi-persistent scheduling mechanism to allocate available communication resources. This algorithm includes three processes: sensing, candidate resource selection, and resource selection (or resource reselection), as Figure 5 shown. In the resource selection or reselection process, the UE sets a resource reservation interval to continuously transmit subsequent message packets using the selected communication resources. When RRI≥100ms, the UE randomly selects a value for the resource reselection counter (RC) within the value range [5,15]. When RRI<100ms, the UE randomly selects a value for RC within the value range [5*C,15*C] according to the set value of RRI, where after each message transmission, the UE decrements the RC value by 1. When RC = 0, with probability p k retains the currently selected resource for subsequent communication transmission, and with probability (1 - p k ) reselects a new resource to transmit subsequent message packets. The value of RRI directly affects the performance of message communication transmission and the utilization efficiency of communication resources. Optimizing the RRI value can improve the success rate of message transmission, reduce the message transmission delay, and improve the safety of autonomous driving applications in vehicle-to-everything.

[0003] Currently, in the SB-SPS resource scheduling mechanism, there are multiple pre-configuration methods for the RRI value, but generally it is set to a fixed value and cannot be adaptively adjusted according to the resource occupancy situation in vehicle-to-everything. In fact, when the RRI value is large, the probability that different UEs use the same communication resource will decrease, which can improve the success rate of message transmission, but also increase the message transmission delay. At the same time, once the UE wrongly selects a communication resource, it will cause consecutive loss of multiple message packets, resulting in the vehicle driving state information not being updated in a timely manner. When the RRI value is small, the UE frequently occupies reserved resources, reducing the available candidate resources, increasing the collision probability that the UE selects the same resource during the resource selection or reselection process, and reducing the message packet reception rate and communication resource efficiency. Summary of the Invention

[0004] Objective of the Invention: Aiming at the above-mentioned existing technologies, an adaptive resource reservation interval method based on SB-SPS is proposed. In C-V2X, the on-vehicle UE applies a fuzzy system to infer the RRI value state that matches the network state, so as to obtain the best communication performance.

[0005] Technical Solution: An adaptive resource reservation interval method based on SB-SPS includes the following steps:

[0006] S1: The vehicle applies fuzzy functions to respectively perform fuzzy processing on the road vehicle density and the communication resource busy rate, as the input parameters of the fuzzy inference engine;

[0007] S2: The fuzzy inference engine calculates the RRI membership degree and infers the matching RRI value state based on the fuzzy rule base;

[0008] S3: Defuzzification is performed to select the best RRI value and apply it to the SB-SPS scheduling mechanism to allocate communication resources.

[0009] Furthermore, in step S1, the road vehicle density is fuzzified through the fuzzy membership function set of the road vehicle density state; the construction method of the fuzzy membership function set of the vehicle density state includes the following steps:

[0010] Construct the state parameter set of the road vehicle density. The states of the vehicle density at least include the traffic density levels of {lower, medium, higher} in three road areas, and the corresponding state parameter set D = {d L , d M , d H}; Based on the state parameter set D, construct the fuzzy membership function set of the road vehicle density state μ(d) = {μ L (d), μ M (d), μ H (d)}:

[0011]

[0012]

[0013]

[0014] Among them, D L is the threshold for the lower state of the road vehicle density, D H is the threshold for the higher state of the road vehicle density, and the threshold for the medium state of the road vehicle density

[0015] Further, in step S1, the communication resource busy rate is fuzzified by using the fuzzy membership function set of the communication resource busy rate; the construction method of the fuzzy membership function set of the communication resource busy rate includes the following steps:

[0016] Construct the communication resource busy rate on the vehicle driving section. The communication resource busy rate includes at least three channel load states of {light load, normal, heavy load}, and the corresponding parameter set B = {b L , b M , b H}; Based on the parameter set B, construct the fuzzy membership function set of the communication resource busy rate μ(b) = {μ L (b), μ M (b), μ H (b)}:

[0017]

[0018]

[0019]

[0020] Among them, B L is the threshold for the communication resource busy rate in the light load state, B H is the threshold for the communication resource busy rate in the heavy load state, and the threshold for the communication resource busy rate in the normal state

[0021] Further, in step S2, the establishment method of the fuzzy rule base includes the following steps:

[0022] Suppose the RRI value-taking state includes at least three interval size value-taking states of {smaller, medium, larger}, and construct the corresponding state set RRI = {r L , r M , r H}; Based on the state set RRI, construct the fuzzy membership function set of the RRI value-taking state μ(r) = {μ L (r), μ M (r), μ H (r)}:

[0023]

[0024]

[0025]

[0026] Among them, RRI L is the threshold for RRI to take a smaller interval value, RRI M is the threshold for RRI to take a medium interval value, and RRI HThe threshold for taking a larger interval value for RRI;

[0027] According to the fuzzy membership function set D of the road vehicle density state and the fuzzy membership function set B of the communication resource busy rate, calculate the RRI value state according to the following formula:

[0028]

[0029] Wherein, Represents the fuzzy product operation;

[0030] Based on the RRI value state calculated by inference and the fuzzy membership function set of the RRI value state, establish a fuzzy rule base for the RRI value state of the adaptive vehicle networking state.

[0031] Furthermore, in step S2, the fuzzy inference engine calculates the membership degrees μ(r) = {μ L (r), μ M (r), μ H (r)} of the three RRI value states according to the formula μ(r) = {μ(d) × μ(b)};

[0032] In step S3, based on the RRI value state obtained in step S2, defuzzification is performed according to the following rules:

[0033] When the RRI value state is r L , with probability p L Take a random integer value within [0:49], and with probability (1 - p M ) take a random integer value within [50:99];

[0034] When the RRI value state is r M , with probability p H Take a random integer value within [50:99], and with probability (1 - p H ) randomly take a value that is an integer multiple of 100 within [100, 500];

[0035] When the RRI value state is r H , with probability (1 - p H ) randomly take a value that is an integer multiple of 100 within [100, 500], and with p H Probability randomly take a value that is an integer multiple of 100 within [600, 1000];

[0036] Wherein, the probability p L = μ L (r), p M = μ M (r), P H = μ H (r).

[0037] Beneficial effects: The adaptive RRI implementation method of the present invention can, according to the current road vehicle density and communication resource busy rate parameter status, use a fuzzy inference machine to obtain an optimized RRI status value and apply it to the SB-SPS resource scheduling mechanism for vehicle-to-vehicle communication in C-V2X, so as to reduce the vehicle-to-vehicle communication delay in the state of sparse road vehicles or light communication resource load, and reduce the vehicle-to-vehicle communication access collision probability in the state of dense road vehicles or heavy communication resource load, improving the reliability and resource efficiency of vehicle-to-vehicle communication. Description of the drawings

[0038] Figure 1 is the working flow chart of the present invention;

[0039] Figure 2 is the membership function of the road vehicle density of the present invention;

[0040] Figure 3 is the membership function of the communication resource busy rate of the present invention;

[0041] Figure 4 is the membership function of the RRI status value of the present invention;

[0042] Figure 5 is the schematic diagram of the SB-SPS resource scheduling principle of the present invention. Detailed implementation manners

[0043] The following further explains the present invention with reference to the drawings.

[0044] As Figure 1 shown, an adaptive resource reservation interval method based on SB-SPS has the following specific steps:

[0045] S1: The vehicle-mounted computing unit uses fuzzy functions to perform fuzzy processing on the road vehicle density and communication resource busy rate respectively as the input parameters of the fuzzy inference machine.

[0046] S2: The fuzzy inference machine calculates the RRI membership degree and infers the matching RRI value state based on a pre-established fuzzy rule base.

[0047] S3: Defuzzify according to the RRI membership degree calculated in S2, and select the best RRI value to apply to the SB-SPS scheduling mechanism to allocate communication resources.

[0048] In step S1, the process of fuzzy processing the road vehicle density is as follows:

[0049] Let the road vehicle density state parameter set be D = {d L , d M , d H} represent the traffic density of {low, medium, high} three road areas respectively, and can also be further set based on actual needs to increase or decrease the number of states. The three states here are three common road states.

[0050] If the threshold for the low state of road vehicle density is D L , and the threshold for the congested state of road vehicle density is D H , then the fuzzy membership function μ(d) of the road vehicle density state = {μ L (d), μ M (d), μ H (d)}, and the formula is as follows:

[0051]

[0052]

[0053]

[0054] Among them, in this implementation, set the threshold for the medium state of road vehicle density or make other value restrictions on the threshold D for the medium state based on actual applications M .

[0055] The process of fuzzy processing the communication resource busy rate is as follows:

[0056] Let the parameter set of the communication resource busy rate on the vehicle driving section be B = {b L , b M , b H} represent {light load, normal, heavy load} three channel load states respectively. If the light load threshold for the communication resource busy rate is B L , and the heavy load threshold for the communication resource busy rate is B H , then the fuzzy membership function μ(b) of the road communication resource busy rate = {μ L (b), μ M (b), μ H (b)}, and the expression is as follows:

[0057]

[0058]

[0059]

[0060] Among them, set the threshold for the normal state of the communication resource busy rate In this embodiment, the three channel load states of the communication resource busy rate are set for the above technical solution.

[0061] In step S2, the method for establishing the fuzzy rule base includes the following steps:

[0062] Let the RRI value state set be RRI = {r L , r M , r H}, which respectively represent three interval size value states of {smaller, medium, larger}. If the RRI takes the smaller interval threshold as RRI L , the medium interval threshold as RRI M , and the larger interval threshold as RRI H , then the fuzzy membership function μ(r) of the RRI value state = {μ L (r), μ M (r), μ H (r)}, and the expression is as follows:

[0063]

[0064]

[0065]

[0066] According to the fuzzy membership function set D of the road vehicle density state and the fuzzy membership function set B of the communication resource busy rate, calculate the RRI value state according to the following formula:

[0067]

[0068] where, represents the fuzzy product operation.

[0069] According to the RRI value state obtained by the inference calculation and the fuzzy membership function set of the RRI value state, use real data to establish a fuzzy rule base of the RRI value state of the adaptive vehicle network state after a large number of trainings, as shown in Table 1:

[0070] Table 1

[0071] D B RRI <![CDATA[d L > <![CDATA[b L > <![CDATA[r L > <![CDATA[d L > <![CDATA[b M > <![CDATA[r M > <![CDATA[d L > <![CDATA[b H > <![CDATA[r M > <![CDATA[d M > <![CDATA[b L > <![CDATA[r M > <![CDATA[d M > <![CDATA[b M > <![CDATA[r M > <![CDATA[d M > <![CDATA[b H > <![CDATA[r H > <![CDATA[d H > <![CDATA[b L > <![CDATA[r M <!-- 5 -->]]> <![CDATA[d H > <![CDATA[b M > <![CDATA[r M > <![CDATA[d H > <![CDATA[b H > <![CDATA[r H >

[0072] In step S2, the fuzzy inference engine calculates the membership degrees μ(r) = {μ L (r), μ M (r), μ H (r)} of the three RRI value states according to the input information and the formula μ(r) = {μ(d) × μ(b)}. To avoid the differences between the value ranges of vehicle state parameters, all input parameters are normalized to the range of [0, 1] during the operation.

[0073] When the fuzzy inference engine infers and outputs a certain value state of RRI based on the fuzzy rule base, the defuzzification module outputs the current optimal RRI value according to the decision table shown in Table 2 and applies it to the SB-SPS scheduling mechanism for communication resource allocation.

[0074] Table 2 Defuzzification Decision Table

[0075]

[0076] Among them, the probability p L = μ L (r), p M = μ M (r), p H = μ H (r).

[0077] In the invention, by estimating the dynamically changing road vehicle density and communication resource load status in the vehicle network, the determination of the RRI state value is implemented as an adaptive interval to reserve resources for transmitting subsequent message packets, reduce the message transmission delay, and improve the packet transmission rate and communication resource efficiency.

[0078] To better illustrate the principle and application of the present invention, this embodiment describes the technical solution in combination with specific cases and corresponding parameters.

[0079] For example, an on-vehicle communication terminal is equipped on the vehicle to support the C-V2X communication specification. Suppose in a cellular vehicle network on a 2-km long, two-way six-lane highway section, at time t, a vehicle v triggers a communication resource selection or reselection mechanism to select available communication resources to transmit message packets. According to the principle of the SB-SPS resource scheduling mechanism, the UE will retain the currently selected communication resources and set an RRI value, and use the retained communication resources to transmit subsequent message packets according to the RRI interval period.

[0080] The on-vehicle UE performs fuzzy processing on the road vehicle density and communication resource busy rate parameters to estimate the road traffic state and channel usage state. Subsequently, the fuzzy values of the road traffic state and channel usage state parameters are used as input parameters for the fuzzy inference engine, and the fuzzy inference engine infers the RRI value state to be set according to the fuzzy rule base. Finally, the defuzzification decision outputs the RRI value for the SB-SPS resource scheduling mechanism. The specific content includes:

[0081] 1. Fuzzify the state parameters:

[0082] For the vehicle v traveling on the above-mentioned section, the on-vehicle UE fuzzifies the road vehicle density parameter value d, where The total number of vehicles is counted as the number of vehicles sending CAM (Cooperative Awareness Message). On the premise of ensuring high-speed traffic safety, the vehicle spacing in a high-density traffic state is generally not less than 50m. At this time, the vehicle density D of the road section set H = 120 vehicles / km. To ensure the reliability of vehicle-to-vehicle communication transmission, the maximum vehicle spacing is generally 200m. The road section set is in a sparse traffic state, and the vehicle density D of the road L = 30 vehicles / km. The threshold calculation for the normal state of road traffic is Therefore, the membership degrees μ(d) = {μ L , μ M , μ H} of the road vehicle density state set D = {d L , d M , d H} are realized by formulas (1) to (3), and are shown as follows respectively.

[0083]

[0084]

[0085]

[0086] In this embodiment, it is assumed that the channel busy rate is used to represent the communication resource busy rate. Then, the parameter of the vehicle-mounted UE for fuzzifying the communication resource busy rate is b = N busy / N C , where N busy is the number of all sub-channel time slots in the inner link received signal strength index greater than the predefined threshold (generally -90dB) in the time slot [t - 100, t - 1]; N C is the number of all sub-channel time slots in the time slot [t - 100, t - 1]. It is assumed that the threshold value for the lightly loaded state of CBR is B L = 0.3, and the threshold value for the heavily loaded state is B H = 0.7. Then, when CBR is in the normal state, its threshold value is Therefore, the membership degrees μ(b) = {μ L , μ M , μ H} of the CBR state set are realized by formulas (4) to (6), and are shown as follows respectively:

[0087]

[0088]

[0089]

[0090] 2. Establish a fuzzy rule base

[0091] Currently, in the NR-V2X standard Release 17, the RRI values are divided into two intervals, that is, a random value is taken within the range of [0, 99] ms, or a random value within the range of [100, 1000] ms is an integer multiple of 100. Let RRI M = 100 ms be the threshold for the medium-sized interval state, then the threshold for the smaller interval state is The threshold for the larger interval state is the median of the integer multiples of 100 in [100, 1000], that is, RRI H = median[100, 1000] = 600 ms. The membership degrees μ(r) = {μ L (r), μ M (r), μ H (r)} of the RRI value states are respectively shown as follows:

[0092]

[0093]

[0094]

[0095] According to formula (10), perform fuzzy inference to calculate the RRI value state, and establish a fuzzy rule base for the RRI value state as shown in Table 1 according to the above formula.

[0096] 3. Output the RRI value

[0097] The vehicle density parameter value d = 60 vehicles / km is calculated by vehicle v. The membership degrees of the three state parameter sets D = {d L , d M , d H} of the road vehicle density are respectively μ(d) = {μ L (d), μ M (d), μ H (d)} = {0.33, 0.67, 0}. At this time, the vehicle density value state is D = {d M}. If the vehicle v calculates that the current road channel resource busy rate b = 0.4, the membership degrees of the three states of the communication resource busy rate can be calculated by formulas (14) - (16) as μ(b) = {μ L (b), μ M (b), μ H (b)} = {0.5, 0.5, 0}. At this time, the channel resource busy rate value state is B = {b L , bM}。From Table 1, it can be seen that at this time, the resource reservation interval value state is RRI = {r M}。

[0098] Using the formula μ(r) = {μ(d) × μ(b)}, the membership degrees of the three value states of RRI are calculated as μ(r) = {μ L (r), μ M (r), μ H (r)} = {0.4, 0.6, 0}. At this time, the probability that the RRI value state is r M is μ M (r) = 0.6. From Table 2, it can be seen that the defuzzification module takes a random integer value within the range of [50:99] with a probability p M = μ M (r) = 0.6, or takes a random integer value within the range of [100:500] with a probability of 0.4, and outputs it as the RRI value for the SB-SPS scheduling mechanism.

[0099] Among them, since the fuzzy inference machine has inferred that the current RRI value state of medium size is the most appropriate according to Table 1, the RRI value decision probability is directly related to the possibility (fuzzy membership degree) of different states of RRI here, that is, directly take p M = μ M (r) = 0.6.

[0100] The process of calculating the membership degrees μ(r) = {μ L (r), μ M (r), μ H (r)} of the three value states of RRI is as follows:

[0101] When performing fuzzy product operation, the relative value of the normalized vehicle density parameter value is Then

[0102]

[0103]

[0104] μ H (b) = 0

[0105] The present invention adjusts the resource reservation interval size by adapting to the vehicle networking state. When the road vehicles are sparse or the communication resources are lightly loaded, the average delay of vehicle-to-vehicle communication is reduced by reducing the access interval delay of subsequent message packets; when the road vehicles are dense or the communication resources are heavily loaded, by increasing the access interval time of subsequent message packets and increasing the available resource pool, the access collision probability of vehicle-to-vehicle communication is reduced, and the reliability and resource efficiency of vehicle-to-vehicle communication are improved.

[0106] The above are only the preferred embodiments of the present invention. It should be noted that for those of ordinary skill in the art, without departing from the principle of the present invention, several improvements and refinements can be made, and these improvements and refinements should also be regarded as the protection scope of the present invention.

Claims

1. An adaptive resource reservation interval method based on SB-SPS, characterized in that, It includes the following steps: S1: The vehicle applies the fuzzy function to perform fuzzy processing on the road vehicle density and the communication resource busy rate respectively, as the input parameters of the fuzzy inference engine; S2: The fuzzy inference engine calculates the membership degree of RRI and infers the matching RRI value state based on the fuzzy rule base; S3: Defuzzification, selects the optimal RRI value to be applied to the SB-SPS scheduling mechanism to allocate communication resources; In step S2, the method for establishing the fuzzy rule base includes the following steps: Let the RRI value-taking states include at least three interval size value-taking states of {smaller, medium, larger}, and construct the corresponding state set RRI = {r L , r M , r H}; Based on the state set RRI, construct the fuzzy membership function set μ(r) = {μ L (r), μ M (r), μ H (r)}: Among them, RRI L is the threshold for the smaller interval value of RRI, and RRI L = 50 ms, and RRI M is the threshold for the medium interval value of RRI, and RRI M = 100 ms, and RRI H is the threshold for the larger interval value of RRI, and RRI H = 600 ms; According to the fuzzy membership function set D of the road vehicle density state and the fuzzy membership function set B of the communication resource busy rate, calculate the RRI value state according to the following formula: Among them, represents the fuzzy product operation; Based on the RRI value state calculated by inference and the fuzzy membership function set of the RRI value state, establish a fuzzy rule base of the RRI value state for the adaptive vehicle networking state; In step S2, the fuzzy inference engine calculates the membership degrees μ(r) = {μ L (r), μ M (r), μ H (r)} of the three value states of RRI according to the formula μ(r) = {μ(d) × μ(b)}; μ(d) is the fuzzy membership function set of the road vehicle density state, and μ(b) is the fuzzy membership function set of the communication resource busy rate; In step S3, based on the RRI value state obtained in step S2, defuzzification is performed according to the following rules: When the RRI value-taking state is r L , with probability p L take a random integer value within [0:49], and with probability (1 - p M ) take a random integer value within [50:99]; When the RRI value-taking state is r M , with probability p M take a random integer value within [50:99], and with probability (1 - p H ) take a value that is an integer multiple of 100 randomly within [100, 500]; When the RRI value-taking state is r H At this time, with probability (1 - p H ), randomly take an integer multiple of 100 within [100, 500], and with probability p H randomly take an integer multiple of 100 within [600, 1000]; Among them, the probability p L = μ L (r), p M = μ M (r), p H = μ H (r).

2. The adaptive resource reservation interval method based on SB-SPS according to claim 1, characterized in that, In step S1, the road vehicle density is fuzzified through the fuzzy membership function set of the road vehicle density state; the construction method of the fuzzy membership function set of the vehicle density state includes the following steps: Construct a set of state parameters for road vehicle density. The states of vehicle density at least include the traffic intensity levels of {low, medium, high} in three road areas, and the corresponding set of state parameters D = {d L , d M , d H}; Based on the set of state parameters D, construct a set of fuzzy membership functions for the road vehicle density state μ(d) = {μ L (d), μ M (d), μ H (d)}: Among them, D L is the threshold for the low state of road vehicle density, and D H is the threshold for the high state of road vehicle density, and the threshold for the medium state of road vehicle density 3. The adaptive resource reservation interval method based on SB-SPS according to claim 2, characterized in that, In step S1, the communication resource busy rate is fuzzified through the fuzzy membership function set of the communication resource busy rate; the construction method of the fuzzy membership function set of the communication resource busy rate includes the following steps: Construct the communication resource busy rate on the vehicle driving section. The communication resource busy rate includes at least three channel load states: {light load, normal, heavy load}, and the corresponding parameter set B = {b L , b M , b H}; Based on the parameter set B, construct the fuzzy membership function set of the communication resource busy rate μ(b) = {μ L (b), μ M (b), μ H (b)}: Among them, B L is the threshold for the communication resource busy rate in the light load state, B H is the threshold for the communication resource busy rate in the heavy load state, and the threshold for the communication resource busy rate in the normal state

Images