Time-frequency resource scheduling method for vehicle-to-vehicle communication

Abstract

The invention relates to a vehicle-to-vehicle communication-oriented time-frequency resource scheduling method, which comprises the following steps of: setting a plurality of cache queues for each vehicle, and generating a cache matrix, an end-to-end maximum time delay matrix, a priority matrix and a reliability matrix based on the cache queues; a transfer rule is established to transfer the data packets in the cache matrix to a sending queue; calculating the number of single-time scheduling data packets to determine a single-time scheduling data set; determining a candidate time-frequency resource block set and the total number of resource blocks; determining and revising the maximum number of candidate time-frequency resource blocks which can be occupied by a single data packet; and allocating at least one time-frequency resource block with good quality to each data packet under the condition that the maximum available candidate time-frequency resource block is allocated to all the data packets. According to the method, the service quality of services with different priorities is classified and guaranteed, the performance of a vehicle-to-vehicle communication network is optimized, the reliability of key data transmission is improved, and the time delay of data transmission is reduced.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of vehicle-to-vehicle communication, and particularly relates to a time-frequency resource scheduling method for vehicle-to-vehicle communication. BACKGROUND

[0002] The C-V2X (Cellular Vehicle-to-Everything) system needs to solve the problem of limited time-frequency resources when performing vehicle-to-vehicle communication, so as to meet the communication needs of different types of services under the condition of limited time-frequency resources. The quality of service of different services is classified and guaranteed according to priority, end-to-end delay and reliability, so as to improve the transmission reliability of key data and reduce the transmission delay. SUMMARY

[0003] In view of the above analysis, the present application aims to provide a time-frequency resource scheduling method for vehicle-to-vehicle communication, which realizes efficient management and allocation of limited time-frequency resources, optimizes the performance of vehicle-to-vehicle communication network, and improves the efficiency and reliability of data transmission.

[0004] The purpose of the present application is mainly realized through the following technical scheme:

[0005] The present application discloses a time-frequency resource scheduling method for vehicle-to-vehicle communication, comprising:

[0006] A plurality of cache queues are set for each vehicle according to service requirements, and a cache matrix, an end-to-end maximum delay matrix, a priority matrix and a reliability matrix are generated based on the cache queues;

[0007] According to the priority matrix, the end-to-end maximum delay matrix and the reliability matrix, a transfer rule is established to transfer data packets in the cache matrix to the sending queue; the number of data packets in a single scheduling is calculated to determine a single scheduling data set;

[0008] The candidate time-frequency resource block set and the total number of resource blocks for vehicle-to-vehicle communication are determined;

[0009] The maximum number of candidate time-frequency resource blocks that can be occupied by a single data packet is determined based on the reliability matrix; the maximum number of candidate time-frequency resource blocks that can be occupied by a single data packet is revised based on the relationship between the total number of data packets in a single scheduling and the total number of candidate time-frequency resource blocks;

[0010] The data packets in the single scheduling set are allocated to candidate resources; under the condition that all data packets are allocated to the maximum number of candidate time-frequency resource blocks, each data packet is allocated to at least one good time-frequency resource block.

[0011] Further, three cache queues Q buffer1 , Q buffer2 , Q buffer3 are set for each vehicle according to service requirements.

[0012] Buffer queue Q buffer1 Safety emergency services, such as vehicle emergency braking, vehicle out-of-control warning, etc., which have the highest priority;

[0013] Buffer queue Q buffer2 Traffic assistance non-emergency services, such as speed limit warning, congestion ahead warning, dangerous road warning, etc.

[0014] Buffer queue Q buffer3 Large bandwidth non-emergency services, such as navigation planning, audio and video entertainment, etc.

[0015] Further, the transfer rules include:

[0016] 1) According to the end-to-end maximum delay matrix M delay of each data packet T n,l from small to large, n = 1, 2, 3, l = 1, 2, …, L, L is the maximum number of data packets that can be stored in the buffer queue;

[0017] 2) When the end-to-end maximum delay of the data packet is the same, according to the priority matrix M priority of each data packet P n,l from small to large;

[0018] 3) When the end-to-end delay and priority of the data packet are consistent, according to the reliability matrix M reliability of each data packet R n,l from large to small;

[0019] 4) When the end-to-end maximum delay, priority and reliability of the data packet are consistent, the data packets are transferred in order according to the time of generation.

[0020] Further, the calculation formula of the number of data packets N once-data in a single scheduling data set S once-data is:

[0021]

[0022] represents the floor function; P is the average priority of all data packets to be transmitted in the sending queue, which is solved according to the priority value P n,l of each data packet in the sending queue and the number of data packets N send ; the value range of N once-data is [1, 2, 3, 4, 5]; the number of data packets in a single transmission scheduling is between 1 and 5;

[0023] The corresponding relationship is as follows:

[0024] N sendWhen N = 0, the number of data packets N in a single scheduling operation once-data =0;

[0025] N send ≥1.5L At that time, the number of data packets N in a single scheduling operation once-data It is 5;

[0026] N send ≥1.5L At that time, the number of data packets N in a single scheduling operation once-data It is 4;

[0027] N send ≥1.5L At that time, the number of data packets N in a single scheduling operation once-data It is 3;

[0028] N send ≥1.5L At that time, the number of data packets N in a single scheduling operation once-data It is 2;

[0029] In other cases, the number of data packets N in a single scheduling operation once-data The value is 1.

[0030] Furthermore, a perception-based semi-persistent scheduling algorithm is used to determine the candidate time-frequency resource block set and the total number of resource blocks.

[0031] Furthermore, the process of determining the candidate time-frequency resource block set and the total number of resource blocks includes:

[0032] 1) Determine the resource awareness window and the resource selection window; the awareness window is a time-frequency resource block consisting of the 1000 subframes prior to the resource reselection time; the resource selection window is a time-frequency resource block within a set time window after the resource reselection time.

[0033] 2) In the listening and sensing window, there are 1000 historical subframes. For subframes that can be correctly received and decoded by SCI, some candidate resources reserved for other services in the selection window are excluded based on the SCI information.

[0034] 3) Calculate the average received power (RSRP) of the frequency domain position reference signal at historical N*100ms times for the remaining candidate resources; then compare the average RSRP with the threshold Th. rsrp In comparison, those exceeding the threshold Th rsrp Candidate resources are excluded if they are less than the threshold Th. rsrp , continue to be considered as candidate resources;

[0035] 4) For the remaining candidate resources, if the number is less than the total number N of resource blocks in the resource selection window... selcet-window 20% of that, then the RSRP threshold Th rsrpIncrease 3db, repeat the judgment of step 3) until the number of remaining candidate resources is greater than the total number of resource blocks N in the resource selection window selcet-window .

[0036] 5) Calculate the average received signal strength indication of the corresponding frequency domain position at the time of N*100ms for the remaining candidate resource blocks

[0037] 6) Sort from low to high according to the value of the average received signal strength indication, and select the 20% of the resource blocks with the lowest value as the candidate time-frequency resource block set S candidate , the total number of which is N total-candidate .

[0038] Further, the expression of the average received power RSRP is:

[0039]

[0040] wherein, RSRP(x,y) represents the average RSRP value of the candidate resource corresponding to the xth subframe in the time domain and the yth subchannel in the frequency domain in the remaining candidate resources at the historical N*100ms time; RSRP (x-i*100,y) RSRP(x-i*100,y) represents the value of RSRP corresponding to the x-i*100th subframe in the time domain and the yth subchannel in the frequency domain in the remaining candidate resources.

[0041] The expression of the average received signal strength indication is:

[0042]

[0043] wherein, RSSI(x,y) represents the average RSSI value of the candidate resource corresponding to the xth subframe in the time domain and the yth subchannel in the frequency domain in the remaining candidate resources at the historical N*100ms time; RSSI (x-i*100,y) RSSI(x-i*100,y) represents the RSSI value corresponding to the x-i*100th subframe in the time domain and the yth subchannel in the frequency domain in the remaining candidate resources.

[0044] Further, based on the reliability matrix, the calculation formula of the maximum number of candidate time-frequency resource blocks N data-select that a single data packet can occupy is:

[0045]

[0046] wherein, represents the floor function, and the reliability R n,l of various types of data packets is obtained from the reliability matrix M reliability .

[0047] Further, the formula for revising the maximum candidate time-frequency resource block that a single data packet can occupy based on the relationship between the total number of single scheduling data packets and the total number of candidate time-frequency resource blocks is:

[0048]

[0049] wherein, represents rounding down; N total-once is the total number of candidate time-frequency resource blocks expected to be occupied by single scheduling; N total-candidate is the total number of candidate time-frequency resources;

[0050] After revision, it is continuously determined whether the ratio of the total number of candidate resources expected to be occupied by single scheduling N total-once to the total number of candidate time-frequency resources N total-candidate is greater than 1 / 2; when it is greater than 1 / 2, revision is continuously performed through the above formula until the ratio of the total number of candidate resources expected to be occupied by single scheduling N total-once to the total number of candidate time-frequency resources N total-candidate falls below 1 / 2.

[0051] Further, the rule for allocating candidate resources to data packets in the single scheduling set is:

[0052] 1) N candidate candidate resource blocks with the lowest RSSI value are selected from the candidate time-frequency resource block set S once-data , and are sequentially allocated to N once-data data packets in the single scheduling set S once-data according to the data packet arrangement order;

[0053] 2) the remaining N data-select -1 resource blocks required by each data packet are randomly selected from the remaining N total-candidate -N once-data candidate resources.

[0054] The present application has the following beneficial effects:

[0055] The time-frequency resource scheduling method for vehicle-to-vehicle communication disclosed in the present application efficiently manages and allocates limited time-frequency resources to meet the communication needs of different types of services. The quality of service of different services is classified and guaranteed according to priority, end-to-end delay and reliability, which improves the transmission reliability of critical data and reduces the transmission delay. BRIEF DESCRIPTION OF DRAWINGS

[0056] The accompanying drawings are included to provide a further understanding of the present application and are incorporated in and constitute a part of this specification, illustrate embodiments of the present application and together with the description serve to explain the principles of the present application, and, together with the description, to specify the characteristics of the working procedures of the embodiments; and in the drawings:

[0057] Figure 1A flow chart of the time-frequency resource scheduling method for vehicle-to-vehicle communication of the embodiment of the present application is shown in the figure.

[0058] Figure 2 A transfer flow schematic diagram of the data packets in the cache matrix being transferred to the sending queue of the embodiment of the present application is shown in the figure.

[0059] Figure 3 A flow schematic diagram of determining the single scheduling data set of the embodiment of the present application is shown in the figure.

[0060] Figure 4 A flow schematic diagram of determining the candidate time-frequency resource block set and the total number thereof of the embodiment of the present application is shown in the figure.

[0061] Figure 5 A time-frequency resource scheduling instance schematic diagram of the embodiment of the present application is shown in the figure. DETAILED DESCRIPTION

[0062] The preferred embodiments of the present application will be described in detail below with reference to the accompanying drawings, wherein the accompanying drawings form a part of the present application and are used to explain the principles of the embodiments of the present application.

[0063] The embodiment discloses a time-frequency resource scheduling method for vehicle-to-vehicle communication, as shown in the figure, comprising: Figure 1

[0064] Step S1, a plurality of cache queues are set for each vehicle according to the service requirement, and a cache matrix, an end-to-end maximum time delay matrix, a priority matrix and a reliability matrix are generated based on the cache queue;

[0065] Step S2, a transfer rule is made according to the priority matrix, the end-to-end maximum time delay matrix and the reliability matrix, the data packets in the cache matrix are transferred to the sending queue, the number of single scheduling data packets is calculated to determine the single scheduling data set;

[0066] Step S3, the candidate time-frequency resource block set for vehicle-to-vehicle communication and the total number of resource blocks are determined;

[0067] Step S4, the maximum number of candidate time-frequency resource blocks that can be occupied by a single data packet is determined based on the reliability matrix, and the maximum number of candidate time-frequency resource blocks that can be occupied by a single data packet is revised based on the relationship between the total number of single scheduling data packets and the total number of candidate time-frequency resource blocks.

[0068] Step S5, the candidate resource allocation is performed on the data packets in the single scheduling set, and each data packet is allocated to at least one good time-frequency resource block under the condition that all data packets are allocated to the maximum number of candidate time-frequency resource blocks.

[0069] ​In the vehicle-to-everything system, vehicles need to interact with each other through wireless communication devices. Vehicle-to-everything services can be roughly divided into three types: safety emergency services, traffic non-emergency services, and information entertainment large-bandwidth non-emergency services. Each type of service has different performance parameters such as end-to-end maximum delay and reliability.

[0070] In the 3GPP protocol standard, according to the different requirements of vehicle-to-everything services for delay and reliability, the ProSe Per-Packet Priority (PPPP) is proposed. The vehicle-to-everything application layer data is divided into eight priority levels as the basis for determining the priority of the data packet and constraining the maximum transmission delay and ensuring reliability. As shown in the following table, it is the constraint condition of part of the data service reliability, delay and other performance indicators.

[0071]

[0072] Specifically, in step S1, three buffer queues Q buffer1 , Q buffer2 , Q buffer3 are set for each vehicle according to the service requirements; each buffer queue contains a storage area with a length of L, which is used to store up to L data packets that have arrived from the upper layer but have not yet been sent out. The buffer queue is emptied and initialized when the communication device starts.

[0073] The three buffer queues are represented as follows:

[0074] Q buffer1 = [D 1,1 D 1,2 ... D 1,l ... D 1,L ];

[0075] Q buffer2 = [D 2,1 D 2,2 ... D 2,l ... D 2,L ];

[0076] Q buffer3 = [D 3,1 D 3,2 ... D 3,l ... D 3,L ];

[0077] wherein D n,l , n = 1, 2, 3 represents the data packet in the nth buffer queue in the lth storage area.

[0078] According to the priority of the vehicle communication data packet, each buffer queue stores different types of data services; wherein,

[0079] Buffer queue Q buffer1 Safety class emergency services, including vehicle emergency braking, vehicle out-of-control warning, etc. These services have the highest priority, and low latency and high reliable transmission of these services should be guaranteed first.

[0080] Buffer queue Q buffer2 Traffic assistance class non-emergency services, including speed limit warning, congestion ahead warning, dangerous road warning, etc.

[0081] Buffer queue Q buffer3 Large bandwidth non-emergency services, including navigation planning, audio and video entertainment, etc.

[0082] Specifically, the buffer queue of each vehicle generated according to the buffer queue is used as a buffer matrix M buffer :

[0083]

[0084] The vehicle performs queue polling every ΔT time interval to monitor the queue buffer condition and read the number of data packets in each buffer queue, the end-to-end maximum delay of each data packet, the transmission priority of each data packet, and the reliability.

[0085] The maximum number of data packets that can be stored in each buffer queue is L, but in actual work, it is not necessarily full, and the actual number of data packets stored in each queue is represented by N buffer1 N buffer2 N buffer3 . The total number of data packets to be transmitted in the buffer queue is represented by N buffer-total , N buffer-total =N buffer1 +N buffer2 +N buffer3 .

[0086] The buffer queue is polled to obtain the end-to-end maximum delay of each data packet, which is represented by an end-to-end maximum delay matrix M delay ; where T n,l , n=1, 2, 3 represent the end-to-end maximum delay of the data packet in the Lth storage area in the nth buffer queue.

[0087]

[0088] The buffer queue is polled to obtain the transmission priority of each data packet, which is represented by a priority matrix M priority , where P n,l , n=1, 2, 3 represent the priority of the data packet in the Lth storage area in the nth buffer queue.

[0089]

[0090] Polling the buffer queue, obtaining the reliability of each data packet, using the reliability matrix M reliability , wherein R n,l , n = 1, 2, 3 represents the reliability of the data packet in the nth buffer queue.

[0091]

[0092] In step S2, according to the priority matrix M priority , the end-to-end maximum delay matrix M delay , and the reliability matrix M reliability , a transfer rule is established to transfer the data packets in the buffer matrix M buffer to the sending queue Q send ; the length of the sending queue Q send is set to be 3 times the length of the buffer queue L, i.e. 3L.

[0093] In order to ensure the low-delay service to be transmitted preferentially, the established transfer rule includes:

[0094] 1) The data packets are transferred from small to large according to the end-to-end maximum delay T delay of each data packet in the end-to-end maximum delay matrix M n,l ;

[0095] 2) When the end-to-end maximum delays of the data packets are the same, the data packets are transferred from small to large according to the priority P priority of each data packet in the priority matrix M n,l ;

[0096] 3) When the end-to-end delays and the priorities of the data packets are the same, the data packets are transferred from large to small according to the reliability R reliability of each data packet in the reliability matrix M n,l ;

[0097] 4) When the end-to-end maximum delays, the priorities, and the reliabilities of the data packets are the same, the data packets are transferred in order according to the time of generation of the data packets.

[0098] The transfer flowchart is shown in Figure 2 .

[0099] In step S2,

[0100] If one data packet is selected from the sending queue Q send for transmission according to the order each time, when the number of data packets to be transmitted is large, or the proportion of high-priority data packets is large, the delay will be increased and the transmission efficiency will be reduced because only one data packet is dispatched each time. Therefore, a plurality of data packets can be selected for simultaneous transmission in single dispatch, and N once-data data packets are selected to form a single dispatch data set Sonce-data .

[0101] Single scheduling dataset S once-data It is the send queue Q send A subset containing the data packets for each scheduling session. The dataset S for a single scheduling session needs to be determined. once-data Number of data packets N once-data .

[0102] Priority matrix M priority Priority P of each data packet n,l Values ​​range from 0 to 7, and there are 8 priority types, P n,l The smaller the value, the higher the priority of the data packet. Therefore, the priority value P of each data packet in the sending queue can be used as a reference. n,l Number of data packets N send Solve for the average priority of all pending data packets in the sending queue.

[0103] N send Represents the sending queue Q send The number of data packets N to be transmitted. After each scheduling cycle, the number of data packets N to be transmitted is... send Updated once.

[0104]

[0105] Then, the single scheduling dataset S once-data Number of data packets N once-data The process of determination includes:

[0106] 1) When the total number of data packets N in the sending queue send When the value is 0, it indicates that there is no data to be transmitted, and the number of data packets N in a single scheduling operation is [number missing]. once-data The value is 0, and the single scheduling dataset S is... once-data It is an empty set;

[0107] 2) When the average priority of all pending data packets in the sending queue is... The data packet is less than the priority threshold and the number of packets N waiting to be transmitted in the sending queue is... send Greater than the sending queue Q send When the length is half, it indicates that the sending queue Q is... send There are many data packets to be transmitted, and the data packets have high priority. At this time, it is necessary to schedule multiple data packets at once. The system priority threshold is set to 4.

[0108] 3) Other cases indicate that there is no need to increase the number of data packets per scheduling, so scheduling is carried out according to 1 data packet per scheduling.

[0109] That is, the single scheduling dataset S once-data Number of data packets N once-dataThe calculation formula is:

[0110]

[0111] Indicates rounding down; N once-data The value range is [1, 2, 3, 4, 5]; the number of data packets scheduled in a single transmission is between 1 and 5;

[0112] The correspondence is as follows:

[0113] N send When N = 0, the number of data packets N in a single scheduling operation once-data =0;

[0114] N send ≥1.5L At that time, the number of data packets N in a single scheduling operation once-data It is 5;

[0115] N send ≥1.5L At that time, the number of data packets N in a single scheduling operation once-data It is 4;

[0116] N send ≥1.5L At that time, the number of data packets N in a single scheduling operation once-data It is 3;

[0117] N send ≥1.5L At that time, the number of data packets N in a single scheduling operation once-data It is 2;

[0118] In other cases, the number of data packets N in a single scheduling operation once-data The value is 1.

[0119] Because the sending queue Q send The data packet scheduling order is already determined, and the number of data packets N scheduled in a single operation is [number]. once-data Once determined, the single scheduling dataset S once-data It can then be obtained automatically. Assume N once-data =3, S once-data like Figure 3 As shown.

[0120] Specifically, in step S3, a perceptual-based semi-persistent scheduling algorithm is used to determine the candidate time-frequency resource block set S. candidate and the total number of resource blocks N total-candidate .

[0121] like Figure 4 As shown, the process of determining the candidate time-frequency resource block set and the total number of resource blocks includes:

[0122] 1) Determine the resource sensing window and the resource selection window; the sensing window is the time-frequency resource block composed of the 1000 subframes before the historical resource reselection time; the resource selection window is the time-frequency resource block in the set time window after the resource reselection time, and the candidate resource exclusion and resource selection are performed in the resource selection window;

[0123] Assume that the resource reselection is triggered at T time, and the sensing window range is [T-1000, T-1]. The selection window range is [T+T1, T+T2], wherein T1 takes the value range [1, 4], and T2 takes the value range [20, 100]. The total number of time-frequency resource blocks in the resource selection window is N selcet-window .

[0124] 2) Listen to the historical 1000 subframes in the sensing window, and for the subframes that can correctly receive and decode the SCI (sidelink control information), exclude part of the candidate resources reserved for other services in the selection window according to the SCI information;

[0125] 3) Calculate the average received power RSRP of the reference signal corresponding to the frequency domain position at the historical N*100ms time for the remaining candidate resources; compare the average RSRP with the threshold Th rsrp , exclude the candidate resources greater than the threshold Th rsrp , and continue to use the candidate resources less than the threshold Th rsrp as candidate resources;

[0126] The expression of the average received power RSRP is:

[0127]

[0128] , wherein, represents the average RSRP value of the candidate resource corresponding to the xth subframe in the time domain and the yth subchannel in the frequency domain in the remaining candidate resources at the historical N*100ms time; RSRP (x-i*100,y) represents the value of RSRP corresponding to the x-i*100th subframe in the time domain and the yth subchannel in the frequency domain in the remaining candidate resources. The average RSRP is actually the linear average of RSRP with a time interval of 100ms, and N is generally 10; the value range of the 3GPP defined RSRP threshold Th rsrp is [-128dBm, 0dBm].

[0129] 4) For the remaining candidate resources, if the number is less than 20% of the total number N selcet-window of resource blocks in the resource selection window, the RSRP threshold Th rsrp is increased by 3db, and the judgment of step 3) is repeated until the number of remaining candidate resources is greater than 20% of the total number N selcet-window of resource blocks in the resource selection window;

[0130] 5) Calculate the average RSSI (Received Signal Strength Indication) at the frequency domain position corresponding to the remaining candidate resource blocks at time N*100ms. The calculation formula is as follows:

[0131]

[0132] in, RSSI represents the average RSSI value of the candidate resource corresponding to the x-th sub-frame in the time domain and the y-th sub-channel in the frequency domain at historical N*100ms intervals. (x-i*100,y) This represents the RSSI value corresponding to the xi*100th subframe in the time domain and the yth subchannel in the frequency domain among the remaining candidate resources. Calculating the average RSSI is actually the same as calculating the linear mean of the RSSI over 100ms intervals, where N is typically 10.

[0133] 6) According to the average received signal strength indication Sort the values ​​from low to high, and select one of them. The 20% of resource blocks with the lowest values ​​are selected as the candidate time-frequency resource block set S. candidate The total number is N total-candidate .

[0134] Given the send queue Q send Single scheduling set S once-data Number of data packets N in a single scheduling operation once-data Candidate time-frequency resource block set S candidate and the total number N of candidate time-frequency resource blocks total-candidate In this embodiment, step S4 is the single scheduling set S. once-data N in once-data The data packets are in the candidate time-frequency resource block set S candidate Select appropriate time-frequency resources for wireless resource transmission;

[0135] Because the reliability requirements of data packets are different, according to the reliability matrix M priority Reliability of various data packets in China n,l The difference is that data packets with high reliability requirements can be transmitted using multiple candidate resource blocks, that is, multiple candidate resources transmit the same data packet, so as to achieve the effect of signal diversity transmission and improve transmission reliability.

[0136] Calculate the maximum number N of candidate time-frequency resources that a single data packet can occupy. data-select The rules include:

[0137] 1) The reliability range for different types of services is 90%-99.999%. The median of 95% is taken as the system reliability threshold.

[0138] 2) When the reliability R of the data packet n,l When the data packet is less than 95% of the system's reliability threshold, one data packet occupies one candidate time-frequency resource block;

[0139] 3) When the data packet reliability R n,l When the reliability threshold set by the system is greater than 95%, multiple candidate resource blocks can transmit the same data packet, that is, one data packet can occupy multiple candidate time-frequency resource blocks.

[0140] The maximum number N of candidate time-frequency resource blocks that a single data packet can occupy is determined based on the reliability matrix. data-select The calculation formula is:

[0141]

[0142] in, Indicates rounding down, and the reliability R of various data packets. n,l From the reliability matrix M reliability Obtained from [the source].

[0143] When multiple candidate time-frequency resource blocks are used to transmit the same data packet, if the proportion of high-reliability data packets is large and no restrictions are imposed, it will lead to an overload of candidate time-frequency resource block set S. candidate The rapidly decreasing availability of candidate resources is highly detrimental to subsequent packet resource selection. Therefore, it is necessary to consider not only packet transmission reliability but also the overall utilization of candidate resources. Consequently, it is required to determine the maximum number N of candidate time-frequency resource blocks that a single packet can occupy. data-select Make corrections.

[0144] In this embodiment, the maximum number of candidate time-frequency resource blocks that a single data packet can occupy is revised based on the relationship between the total number of data packets in a single scheduling and the total number of candidate time-frequency resource blocks;

[0145] The number of data packets in a single scheduling is N once-data The maximum number of candidate time-frequency resource blocks that each data packet can occupy is N. data-select Therefore, the total number of candidate time-frequency resource blocks expected to be occupied in a single scheduling operation is:

[0146] Based on the total number N of candidate time-frequency resources total-candidate And the total number of candidate resources N expected to be used in a single scheduling operation. total-once The proportional relationship between them, and the number N of candidate time-frequency resources that each data packet can occupy. data-select The rules to be amended include:

[0147] 1) When a single scheduling operation is expected to occupy a total of N candidate resources total-once Total number of candidate time-frequency resources N total-candidate When the ratio is greater than 1 / 2, for the number N of candidate time-frequency resources occupied...data-select the number of candidate time-frequency resources occupied by a data packet greater than 1 data-select performing halving processing;

[0148] 2) the number of candidate time-frequency resources that a single data packet can occupy in other cases data-select unchanged.

[0149] That is, the formula for revising the maximum number of candidate time-frequency resource blocks that a single data packet can occupy based on the relationship between the total number of single-scheduling data packets and the total number of candidate time-frequency resource blocks is:

[0150]

[0151] denotes rounding down;

[0152] After the revision, it is continued to judge whether the ratio of the expected total number of candidate resources N total-once occupied by a single scheduling to the total number of candidate time-frequency resources N total-candidate is greater than 1 / 2. When it is greater than 1 / 2, the revision is continued through the above formula until the ratio of the expected total number of candidate resources N total-once occupied by a single scheduling to the total number of candidate time-frequency resources N total-candidate falls below 1 / 2.

[0153] In step S5, when candidate resources are allocated to data packets in the single-scheduling set, the single-scheduling set S once-data , the number of single-scheduling data packets N once-data , the candidate time-frequency resource block set S candidate , the total number of candidate time-frequency resource blocks N total-candidate , and the maximum number of candidate time-frequency resource blocks N data-select that a single data packet can occupy are known. The following needs to select N once-data resource blocks from the candidate time-frequency resource block set S candidate for all data packets in the single-scheduling set S data-select according to certain rules. The allocation rule is:

[0154] 1) select the N candidate candidate resources with the lowest RSSI value from the candidate time-frequency resource block set S once-data and allocate them to the N once-data data packets in the single-scheduling set S once-data in order according to the data packet arrangement order;

[0155] 2) the remaining N data-select -1 resource blocks required by each data packet are randomly selected from the remaining N total-candidate -N once-data candidate resources.

[0156] In this way, it can be ensured that each data packet has a good quality resource block for transmission, and multiple candidate resources can be selected for hierarchical transmission of the same data packet according to the priority, reliability and the like of the data packet, so as to effectively improve the transmission success rate of high-priority and emergency services.

[0157] In the given algorithm example,

[0158] Suppose that the reliabilities of the D14, D16 and D11 data packets in a single scheduling set are 99.99%, 99.9% and 95% respectively; the buffer queue length L=10; and the total number of candidate time-frequency resource blocks N total-candidate is 20.

[0159] The number of candidate time-frequency resources that can be occupied by the D14 data packet is

[0160] The number of candidate time-frequency resources that can be occupied by the D16 data packet is

[0161] The number of candidate time-frequency resources that can be occupied by the D11 data packet is

[0162] Therefore, the total number of candidate time-frequency resource blocks expected to be occupied in a single scheduling period N total-once =4+3+1=8.

[0163] Since after correction, the number of candidate time-frequency resources that can be occupied by a single data packet remains unchanged, that is, the D14 still occupies 4, the D16 still occupies 3, and the D11 still occupies 1.

[0164] According to the rules, the resource allocation is as follows: first, allocate one candidate resource with the minimum RSSI value to each of the D14, D16 and D11 data packets. Then, randomly allocate 3 candidate resources to the D14 and 2 candidate resources to the D16 from the remaining 17 candidate resource blocks. In this way, the single scheduling is ended. As shown in Figure 5 , the resource scheduling of each data packet is realized.

[0165] In summary, the time-frequency resource scheduling method for vehicle-to-vehicle communication in the embodiment of the application efficiently manages and allocates limited time-frequency resources to meet the communication needs of different types of services. The quality of service of different services is classified and guaranteed according to the priority, end-to-end delay and reliability, so as to improve the transmission reliability of critical data and reduce the transmission delay.

[0166] The above merely describes a preferred specific embodiment of the application, but the protection scope of the application is not limited thereto, and any changes or replacements within the technical range disclosed by the application can be easily thought of by those skilled in the art, which should be covered within the protection scope of the application.

Claims

1. A time-frequency resource scheduling method for vehicle-to-vehicle communication, characterized in that, Comprise: According to the business needs for each vehicle to set a plurality of cache queue, and based on the cache queue to generate cache matrix, end-to-end maximum delay matrix, priority matrix and reliability matrix; According to the priority matrix, end-to-end maximum delay matrix and reliability matrix to make transfer rules, and transfer the data packet in the cache matrix to the sending queue; Calculate the number of single scheduling data packets to determine the single scheduling data set; Determine the candidate time-frequency resource block set and the total number of resource blocks for vehicle-to-vehicle communication; Based on the reliability matrix, determine the maximum number of candidate time-frequency resource blocks that a single data packet can occupy; Based on the relationship between the total number of single scheduling data packets and the total number of candidate time-frequency resource blocks, revise the maximum number of candidate time-frequency resource blocks that a single data packet can occupy; For the candidate resource allocation of the data packet in the single scheduling set, under the condition of meeting the allocation of the maximum number of candidate time-frequency resource blocks for all data packets, allocate at least one good time-frequency resource block to each data packet.

2. The time-frequency resource scheduling method for vehicle-to-vehicle communication according to claim 1, wherein According to the business requirements, set 3 cache queues Q for each vehicle buffer1 , Q buffer2 , Q buffer3 ; Buffer queue Q buffer1 Safety emergency services, including vehicle emergency braking, vehicle out-of-control warning, etc., which have the highest priority. Buffer queue Q buffer2 Non-urgent traffic assistance services such as speed limit warning, congestion ahead warning, and dangerous road warning. Buffer queue Q buffer3 Store navigation planning, audio and video entertainment, and other non-urgent services with large bandwidth.

3. The time-frequency resource scheduling method for vehicle-to-vehicle communication according to claim 2, wherein The transfer rules include: 1) According to the end-to-end maximum delay matrix M delay The end-to-end maximum delay T of each data packet in the matrix M n,l Transfer from small to large; n = 1, 2, 3, l = 1, 2, …, L, L is the maximum number of data packets that can be stored in the buffer queue; 2) When the maximum end-to-end delay of the data packets is the same, the priority matrix M is followed priority the priority P of each data packet in the matrix n,l is transferred from small to large 3) When the data packet end-to-end delay and priority are consistent, according to the reliability matrix M reliability The reliability R of each data packet in the matrix M n,l Transfer from large to small; 4) When the end-to-end maximum delay, priority and reliability of the data packet are consistent, transfer in order according to the data packet generation time.

4. The time-frequency resource scheduling method for vehicle-to-vehicle communication according to claim 3, wherein Single scheduling data set S once-data Number of middle data packets N once-data The calculation formula is: denotes rounding down; is the priority value of each data packet in the sending queue n,l and the number of data packets N send the average priority of all data packets in the sending queue solved; N once-data The value range of N is [1, 2, 3, 4, 5]; the number of data packets in single transmission scheduling is between 1 and 5. The correspondence is as follows: N send = 0, the number of single scheduling data packets N once-data is 0; N send ≥1.5L when the single scheduling data packet number N once-data is 5; N send ≥1.5L when the single scheduling data packet number N once-data is 4; N send ≥1.5L when the single scheduling data packet number N once-data is 3; N send ≥1.5L when the single scheduling data packet number N once-data is 2; Other cases single scheduling packet number N once-data is 1.

5. The time-frequency resource scheduling method for vehicle-to-vehicle communication according to any one of claims 1-4, wherein The candidate time-frequency resource block set and the total number of resource blocks are determined by using a sensing-based semi-persistent scheduling algorithm.

6. The time-frequency resource scheduling method for vehicle-to-vehicle communication according to claim 5, wherein The process of determining the candidate time-frequency resource block set and the total number of resource blocks includes: 1) Determine the resource sensing window and the resource selection window; the sensing window is the time-frequency resource block composed of the last 1000 subframes at the resource reselection time; the resource selection window is the time-frequency resource block in the set time window after the resource reselection time; 2) Listen to the last 1000 subframes in the sensing window, and exclude the part of the candidate resources reserved for other services in the selection window according to the SCI information for the subframes that can correctly receive and decode SCI; 3) calculate the average of the received power RSRP of the reference signal corresponding to the frequency domain position at the historical N*100ms moment for the remaining candidate resources; compare the average of the RSRP with the threshold value Th rsrp , and exclude the candidate resources greater than the threshold value Th rsrp , and continue to use the candidate resources less than the threshold value Th rsrp as the candidate resources; 4) for the remaining candidate resources, if the number is less than 20% of the total number of resource blocks N selcet-window in the resource selection window, the RSRP threshold Th rsrp is increased by 3db, and the judgment of step 3) is repeated until the number of the remaining candidate resources is greater than 20% of the total number of resource blocks N selcet-window in the resource selection window; 5) Calculate the average received signal strength indication of the frequency domain position corresponding to N*100ms; 6) Sort according to the value of average received signal strength indication from low to high, select the 20% of resource blocks with the lowest value as candidate time-frequency resource block set S candidate The total number of which is N total-candidate .

7. The time-frequency resource scheduling method for vehicle-to-vehicle communication according to claim 6, wherein The expression of the average received signal strength indication is: wherein, RSRP x,y (i) represents the average RSRP value of the candidate resource corresponding to the xth subframe in the time domain and the yth subchannel in the frequency domain among the remaining candidate resources at the historical N*100 ms time point; RSRP (x-i*100,y) RSRP x,y (i) represents the value of the RSRP corresponding to the x-i*100th subframe in the time domain and the yth subchannel in the frequency domain among the remaining candidate resources.

8. The time-frequency resource scheduling method for vehicle-to-vehicle communication according to claim 7, wherein wherein, RSSI(x, y) represents the average RSSI value of the candidate resource corresponding to the xth subframe in the time domain and the yth subchannel in the frequency domain among the remaining candidate resources at the historical N*100 ms moment; RSSI (x-i*100,y) RSSI(x, y) represents the average RSSI value of the candidate resource corresponding to the xth subframe in the time domain and the yth subchannel in the frequency domain among the remaining candidate resources at the historical N*100 ms moment; RSSI 9. The time-frequency resource scheduling method for vehicle-to-vehicle communication according to claim 8, wherein The maximum number N of candidate time-frequency resource blocks that can be occupied by a single data packet is determined based on a reliability matrix data-select The calculation formula is: wherein, represents a floor function, and the reliability R of each type of data packet is n,l obtained from the reliability matrix M reliability . The formula for revising the maximum number of candidate time-frequency resource blocks that a single data packet can occupy based on the relationship between the total number of single scheduling data packets and the total number of candidate time-frequency resource blocks is:

10. The time-frequency resource scheduling method for vehicle-to-vehicle communication according to claim 9, wherein wherein, denotes rounding down; N total-once is the total number of candidate time-frequency resource blocks expected to be occupied by a single scheduling; N total-candidate is the total number of candidate time-frequency resources; After the correction, continue to judge whether the ratio of the total number of candidate resources N total-once occupied by the single scheduling expectation to the total number of candidate time-frequency resources N total-candidate is greater than 1 / 2; when it is greater than 1 / 2, continue to correct by the above formula until the ratio of the total number of candidate resources N total-once occupied by the single scheduling expectation to the total number of candidate time-frequency resources N total-candidate falls below 1 / 2. ​ The rule for candidate resource allocation for data packets in a single scheduling set is: 1) from the candidate time-frequency resource block set S candidate select N once-data candidate resource blocks with the lowest RSSI values in the order of data packet arrangement and assign them to N once-data data packets in the single-scheduling set S once-data one by one. 2) the remaining N data-select -1 resource blocks in the remaining N total-candidate -N once-data candidates are randomly selected.

Images