CN104507169A - Three-dimensional resource dynamic allocation method and device for reducing system uplink transmission time delay - Google Patents

Abstract

The invention discloses a three-dimensional resource dynamic allocation method and device for reducing TD-LTE/TD-LTE-A system uplink transmission delay, and performs reasonable and efficient dynamic allocation of resources from three-dimensional perspectives of time domain, frequency domain and air domain. Including: grouping the user's business according to different delay and rate requirements, calculating the resource occupancy priority parameters of the user's business belonging to the group; calculating the scheduling priority of the time domain and scheduling the user in the time domain accordingly; calculating The scheduling priority of the frequency domain is used to allocate frequency domain resources accordingly; the transmission mode of air domain resources is allocated reasonably according to the channel quality. Finally, after the 3D resource allocation calculation is completed, the system performs a table lookup operation on the generated mapping table, and quickly allocates resources according to the data in the table. The invention can preferentially allocate resources to users with better channel quality, higher time delay requirements and higher rate requirements, and has the characteristics of reducing resource allocation algorithm complexity, enhancing transmission reliability and reducing packet loss rate.

Description

Three-dimensional resource dynamic allocation method and device for reducing system uplink transmission delay

technical field

The invention relates to a three-dimensional resource dynamic allocation method and device for reducing the uplink transmission time delay of a TD-LTE/TD-LTE-A system, and belongs to the research field of wireless resource management in a mobile communication system.

Background technique

At present, LTE has been able to provide a downlink rate of up to 100Mbps. With the development of massive MIMO and UHF transmission technology, a 1Gbps experimental bureau has been successfully built, and the theoretical rate is as high as 10Gbps. On the other hand, with the substantial improvement of mobile phone hardware performance and the explosive growth of applications, user needs have evolved from a single voice service to a series of services such as voice, streaming media, web browsing, and online games. The business also shows the characteristics of various types, high delay requirements, wide range of rate changes, and strong suddenness. Therefore, more efficient and flexible resource management methods are required.

Traditional resource management is mainly realized jointly through methods such as admission control, power control, handover control, radio resource scheduling and allocation. Among them, wireless resource scheduling and allocation algorithms mainly include sequential scheduling and allocation methods, and resource scheduling and allocation algorithms that consider independent factors such as channel quality, transmission packet size, HoL waiting time, and user fairness. These algorithms are optimally designed for the use of limited wireless resources in mobile communication systems, and improve the performance of the system to a certain extent. The future mobile communication system will provide more services, and wireless resources will also present multi-dimensional characteristics of time domain, frequency domain, code domain, and air domain. Therefore, it is no longer sufficient to use a relatively single consideration target when allocating radio resources, and only considering certain aspects of performance will often lead to deterioration of other performances. Therefore, under the condition of carrying multiple services, it should be considered to design a wireless resource allocation algorithm from a multi-dimensional perspective, so as to maximize the reasonable and full use of wireless resources and improve system performance without additional investment.

Contents of the invention

Purpose of the invention: the present invention studies the three-dimensional resource dynamic allocation method for reducing TD-LTE/TD-LTE-A system uplink transmission time delay, faces practical engineering applications, combines uplink time domain, frequency domain and space domain three-dimensional resources, proposes A method and device capable of reducing transmission delay and resource block allocation complexity in a mobile communication system are provided, which can simply and quickly realize uplink three-dimensional resource allocation in a multi-user and multi-service scenario.

Technical scheme: in order to realize the foregoing invention object, the present invention adopts following technical scheme:

A three-dimensional resource dynamic allocation method for reducing TD-LTE/TD-LTE-A system uplink transmission delay, characterized in that the following steps are repeated in each operation cycle:

Step 1: Group the services included in the user set Ω _UE ={UE ₁ ,UE ₂ ,…,UE _i ,…,UE _N }, i=1,…,N according to different performance requirements, and establish a service grouping table ;

Step 2: According to the long-term delay and long-term rate of the service that user i belongs to service group j, as well as the delay threshold of service group j and the rate requirement value of service group j, calculate the resources of the service that user i belongs to service group j occupancy priority parameter PRI _i,j (t);

Step 3: Calculate the scheduling priority T _i (t) in the time domain according to the maximum rate of user i at time t, the long-term rate, the cache size of the data to be sent, and the priority parameters, and perform time-domain scheduling on the user i accordingly. Scheduling (TDPS), which schedules users sequentially in descending order of T _i (t), and stops scheduling when the system has no resources to allocate.

Step 4: Calculate the scheduling priority F _i,m in the frequency domain according to the maximum rate that user i can achieve on resource block m, the long-term rate and T _i (t), and accordingly determine the scheduling priority in TDPS The user allocates resource blocks (RBs), selects the RB-UE pair {RB _x , UE _y } with the largest F _i,m in the unallocated RB set and the unallocated user set, allocates RB _x to UE _y and uses RB _x is centered and extended to both sides, and adjacent unallocated RBs are allocated to UE _y until the resource blocks on the left and right sides cannot be expanded or the number of allocated RBs is greater than or equal to the number of RBs required by the user;

Step 5: Select the transmission method of airspace resources according to the channel quality. When the signal-to-noise ratio SINR>SINR ₀ , choose to use spatial multiplexing to transmit data, otherwise, choose to use space diversity to transmit data, where SINR ₀ is the selected transmission method threshold value;

Step 6: After the relevant calculation of three-dimensional resource allocation in the time domain, frequency domain and air domain is completed, a three-dimensional resource mapping table is generated, which includes a list of users who have obtained scheduling qualifications, a list of correspondence between users and resource blocks, and a transmission form of MIMO; The system performs table lookup operations on the mapping table, and quickly allocates resources according to the data in the table;

The process of each scheduling and RB allocation is to repeat the above steps 1 to 6 at multiples of the transmission time interval.

According to the first step

$G_i\∈[d_j^{\mathrm{budget}},r_j^{\mathrm{budget}}],j=\mathrm{1,2},\·\&Center\; Dot;\&Center\; Dot;,J$ [Rule 1]

Group the services included in the user, where, is the delay threshold of service group j, The required value for the rate of business packet j.

In the second step according to

[Rule 2]

$\left\{\begin{array}{c}{\mathrm{D.}}_{i,j}\left(t\right)=(1-\mathrm{\α}){\mathrm{D.}}_{i,j}(t-1)+\mathrm{\α}{\mathrm{D.}}_{i,j}\left(t\right)\\ R_{i,j}\left(t\right)=(1-\mathrm{\α})R_{i,j}(t-1)+\mathrm{\α}{R}_{i,j}\left(t\right)\end{array}\right.$ [Rule 3]

Calculate the resource occupancy priority parameter PRI _i,j (t) of the business that user i belongs to business group j, where D _i,j (t) and R _i,j (t) are respectively The long-term delay and long-term rate of the business, α is a sliding window constant, 0<α<1, It is a correction value introduced to avoid packet loss caused by services that have not been allocated resources when the maximum delay is reached.

In the third step according to

$T_i\left(t\right)=\frac{r_i\left(t\right)}{R_i\left(t\right)}\&Center\; Dot;Q_i\left(t\right)\&Center\; Dot;\underset{j}{\mathrm{\&Sigma;}}{\mathrm{PRI}}_{i,j}\left(t\right)$ [Rule 5]

Calculate the scheduling priority T _i (t) in the time domain, where r _i (t) is the maximum rate that user i can achieve at time t, and R _i (t) is the long-term rate of user i at time t. The calculation method is the same as [Rule 3], Q _i (t) is the cache size of the amount of data to be sent reported by user i through the buffer status report, Resource occupation priority parameter for user i.

In the fourth step according to

$f_{i,m}=\frac{r_{i,m}\left(t\right)}{R_{i,m\left(t\right)}}{T}_i\left(t\right)$ [Rule 6]

Calculate the scheduling priority F _i,m in the frequency domain, where r _i,m (t) and R _i,m (t) are the maximum data that can be achieved on RB m when RB m is allocated to user i sending rate and long-term rate; according to

[Rule 7] Calculate the number of RBs required by the user, where Q _i (t) is the cache size of the data to be sent reported by user i through the buffer status report, and M is the total number of all resource blocks in the system.

In the fourth step, the center of RB _x is extended to both sides, and the following rules are used when allocating adjacent unallocated RBs to UE _y : when [rule 8] is satisfied, RB _x+1 is allocated to UE _y , assigning RB _x-1 to UE _y when [Rule 9] is satisfied,

$\left\{\begin{array}{c}{f}_{\mathrm{the\; y},x-1}\&le;f_{\mathrm{the\; y},x+1}\\ \max \left(f_{i,x+1}\right)=f_{\mathrm{the\; y},x+1}\end{array}\right.$ [Rule 8]

$\left\{\begin{array}{c}{f}_{\mathrm{the\; y},x-1}>f_{\mathrm{the\; y},x+1}\\ \max \left(f_{i,x-1}\right)=f_{\mathrm{the\; y},x-1}\end{array}\right.$ [Rule 9]

Wherein, F _y,x- 1 is the frequency domain scheduling priority value when UE _y occupies resource block x-1, and F _y,x+1 is the frequency domain scheduling priority value when UE _y occupies resource block x+1.

A three-dimensional resource dynamic allocation device for reducing the uplink transmission delay of TD-LTE/TD-LTE-A system, including:

User queue module to be scheduled;

The parameter input module is used to receive the user queue to be scheduled output by the user queue module to be scheduled, set the performance requirement parameters of each user to be scheduled, and output the user queue to be scheduled with the performance requirement parameter, and the performance requirement parameter includes the user's maximum time Delay and speed requirements;

The scheduler module is used to sequentially perform time-domain resource allocation, frequency-domain resource allocation and air-space resource allocation on the user queue to be scheduled with performance requirement parameters;

And, a three-dimensional resource mapping table module, which is used to generate a three-dimensional resource mapping table in combination with the user queue to be scheduled and the allocation result of the scheduler module, and the mapping table includes the list of users, users and resource blocks that have obtained scheduling qualifications in time-domain scheduling Correspondence list and transmission form of MIMO;

The scheduler module includes:

The time-domain resource allocator unit is used to group the user's business according to different delay and rate requirements, calculate the resource occupation priority parameter of the user's business and the scheduling priority of the time domain, and perform time-domain scheduling for the user accordingly. Scheduling, output a list of users with scheduling qualifications;

The frequency domain resource allocator unit is used to calculate the scheduling priority F _i,m in the frequency domain, and accordingly allocate the frequency domain resource RB to the users with scheduling qualifications output by the time domain resource allocator;

And, the airspace resource allocator unit is used to select the transmission mode of the airspace resource according to the channel quality, when the signal-to-noise ratio is greater than the set threshold value, select to use space multiplexing to transmit data, otherwise, select to use space diversity to transmit data .

Beneficial effects: A three-dimensional resource dynamic allocation method and device for reducing TD-LTE/TD-LTE-A system uplink transmission delay according to the present invention, by grouping services with different performance requirements such as delay, including delay-sensitive Users with services that require rate guarantees can obtain higher priority. In the time domain, they can get priority scheduling, and in the frequency domain, they can get more resource blocks, thereby enhancing the effectiveness and effectiveness of user communication. reliability. Finally, by switching diversity and multiplexing according to channel quality in the air domain, the utilization rate of resources is further improved. The main work of the algorithm is to calculate the user's resource occupation priority in the time domain. In the frequency domain, the allocation method of extending resource blocks to both sides is used and the transmission form of air domain resources is considered, so that the complexity is not high in the actual system operation process. The allocation of system resources has higher efficiency.

Description of drawings

FIG. 1 is a flowchart of a three-dimensional resource dynamic allocation method for reducing uplink transmission delay in a TD-LTE/TD-LTE-A system.

Fig. 2 is an implementation diagram of a three-dimensional resource dynamic allocation method for reducing uplink transmission delay in a TD-LTE/TD-LTE-A system.

FIG. 3 is a schematic diagram of resource block allocation in a three-dimensional resource dynamic allocation method for reducing uplink transmission delay in a TD-LTE/TD-LTE-A system.

Detailed ways

Below in conjunction with specific embodiment, further set forth the present invention, should be understood that these embodiments are only used to illustrate the present invention and are not intended to limit the scope of the present invention, after having read the present invention, those skilled in the art will understand various equivalent forms of the present invention All modifications fall within the scope defined by the appended claims of the present application.

In this embodiment, a system bandwidth of 20M is used, and the number of available resource blocks is 100. Each TTI includes two 0.5ms time slots, each time slot includes 7 OFDM symbols, and 10 TTIs form a radio frame length. The specific parameters of the system are shown in Table 1.

Table 1 System parameters

parameter parameter value bandwidth 20MHz Number of subcarriers 100 slot length 0.5ms eNodeB radius 1km Transmitting Antenna×Receiving Antenna 2×2 mobile model Random Walk multipath model Jakes model shadow fading model logarithmic model path loss model L＝128.1+37.6log _10d user movement speed 3km/h

As shown in Figure 1, the specific implementation steps of a three-dimensional resource dynamic allocation method for reducing the uplink transmission delay of the TD-LTE/TD-LTE-A system provided by the embodiment of the present invention are as follows:

Step 1: Group the services included in the user set Ω _UE ={UE ₁ ,UE ₂ ,…,UE _i ,…,UE _N }, i=1,…,N according to different performance requirements, and establish a service grouping table . The grouping criteria are:

$G_i\&Element;[d_j^{\mathrm{budget}},r_j^{\mathrm{budget}}],j=\mathrm{1,2},\&CenterDot;\&CenterDot;\&Center\; Dot;,J$ [Rule 1]

in, is the delay threshold of service group j, The required value for the rate of business packet j. The grouping situation is shown in Table 2.

Table 2 The rate budget value and maximum transmission delay value of the service

business type Rate Budget (kbps) Maximum transmission delay (ms) VoIP 10 (G.729 encoding scheme) 50 video 200 100 HTTP 100 300 FTP 50 300

Step two: according to

[Rule 2]

The resource occupancy priority parameter PRI _i,j (t) of the service belonging to the service group j of the user i is calculated. Wherein, D _i,j (t) and R _i,j (t) are the long-term delay and long-term rate of the services of user i belonging to service group j, respectively. D _i,j (t) and R _i,j (t) according to

$\left\{\begin{array}{c}{\mathrm{D.}}_{i,j}\left(t\right)=(1-\mathrm{\&alpha;}){\mathrm{D.}}_{i,j}(t-1)+\mathrm{\&alpha;}{\mathrm{D.}}_{i,j}\left(t\right)\\ R_{i,j}\left(t\right)=(1-\mathrm{\&alpha;})R_{i,j}(t-1)+\mathrm{\&alpha;}{R}_{i,j}\left(t\right)\end{array}\right.$ [Rule 3]

inferred. This is an exponentially decreasing weighted moving average, where the weighted influence of each value decreases exponentially over time. The weighting degree is determined by the constant sliding window α (0<α<1), and α=0.5 is generally used. It is a correction value introduced to avoid packet loss caused by services that have not been allocated resources when the maximum delay is reached. according to

[Rule 4]

inferred. in, is a threshold value of delay.

Step 3: Perform time-domain scheduling (TDPS) on users according to resource occupation priority, and the scheduling priority T _i (t) in the time domain is based on

$T_i\left(t\right)=\frac{r_i\left(t\right)}{R_i\left(t\right)}\&Center\; Dot;Q_i\left(t\right)\&CenterDot;\underset{j}{\mathrm{\&Sigma;}}{\mathrm{PRI}}_{i,j}\left(t\right)$ [Rule 5]

Calculated. Among them, r _i (t) is the maximum rate that user i can achieve at time t, which can be obtained according to Shannon’s formula, R _i (t) is the long-term rate of user i at time t, the calculation method is the same as [rule 3], Q _i (t) is the cache size of the amount of data to be sent reported by user i through the buffer status report (BSR), Resource occupation priority parameter for user i. The users are scheduled sequentially in descending order of T _i (t), until the system stops scheduling when there is no resource to allocate.

Step 4: The frequency domain scheduling (FDPS) process allocates resource blocks (RBs) to the users to be scheduled determined in the TDPS. according to

$f_{i,m}=\frac{r_{i,m}\left(t\right)}{R_{i,m\left(t\right)}}{T}_i\left(t\right)$ [Rule 6]

Calculate the scheduling priority F _i,m in the frequency domain. Among them, r _i,m (t) and R _i,m (t) are respectively the maximum data transmission rate and long-term rate that can be achieved on RB m when RB m is allocated to user i, and the calculation method is the same as r _i (t ) and R _i (t). according to

[Rule 7]

Calculate the number of RBs required by user i. All resource blocks RB in the system constitute the RB set Ω _RB ={RB ₁ ,RB ₂ ,...,RB _m ,...,RB _M }, m=1,...,M, in the initial case, the unallocated RB set unassigned user set According to [Rule 6], as shown in Figure 3, in and Select the RB-UE pair with the largest F _i,m {RB _x , UE _y }, 1≤x≤M, 1≤y≤N, allocate RB _x to UE _y and update and Centered on the resource block RB _x allocated to UE _y in Expand the resource blocks inward to both sides, consider the left resource block RB _x-1 and the right resource block RB _x+1 , satisfy the following rules

$\left\{\begin{array}{c}{f}_{\mathrm{the\; y},x-1}\&le;f_{\mathrm{the\; y},x+1}\\ \max \left(f_{i,x+1}\right)=f_{\mathrm{the\; y},x+1}\end{array}\right.$ [Rule 8]

And RB _x+1 is not allocated, then allocate RB _x+1 to UE _y and update The case of F _y,x-1 >F _y,x+1 is handled in the same way. In [Rule 8], F _y,x- 1 is the frequency domain scheduling priority value when UE _y occupies resource block x-1, and F _y,x+1 is the frequency domain scheduling priority value when UE _y occupies resource block x+1 level value, max(F _i,x+1 )=F _y,x+1 means that on RB _x+1 , for all UE _i , UE _y has the largest frequency-domain scheduling priority value. Repeat this process until the resource blocks at RB _x-9 and RB _x+7 in Figure 3 are not scalable, or the number of allocated RBs is greater than or equal to _ni,RB , then repeat the fourth step for the next user For resource allocation, as shown in Figure 3, the resource block allocation of UE _z is performed at RB _x+12 , and the resource block allocated to UE _y is finally [RB _x-8 , RB _x+6 ]. if If it is empty, the resource block allocation process ends.

Step 5: After the RBs are allocated, the user uses the allocated RBs to transmit on the 2×2 MIMO channel. When the channel quality is good, that is, SINR>SINR ₀ , use spatial multiplexing to transmit data to increase the data transmission rate; otherwise, use space diversity to transmit data to improve the reliability of data transmission. Among them, SINR ₀ is the threshold value for selecting the transmission mode, which can be set to 10dB.

Step 6: After the relevant calculations of three-dimensional resource allocation in the time domain, frequency domain and air domain are completed, a three-dimensional resource mapping table is generated. The table includes a list of users eligible for scheduling (time domain), a list of correspondence between users and resource blocks (frequency domain), and a transmission form of MIMO (air domain). The system performs table lookup operations on the mapping table, and performs fast resource allocation based on the data in the table.

The process of each scheduling and RB allocation is to repeat the above first step to Step six.

As shown in Figure 2, a three-dimensional resource dynamic allocation device for reducing TD-LTE/TD-LTE-A system uplink transmission delay provided by an embodiment of the present invention includes a user queue module to be scheduled, a parameter input module, and a scheduler module and a three-dimensional resource mapping table module, wherein the scheduler module includes a time-domain resource allocator unit, a frequency-domain resource allocator unit and an air-domain resource allocator unit.

The parameter input module receives the user queue to be scheduled output by the user queue module to be scheduled, sets the performance requirement parameters of each user to be scheduled (including user maximum delay and rate requirements, etc.), and outputs the user queue to be scheduled with the performance requirement parameters to the dispatcher module; the scheduler module sequentially performs time-domain resource allocation, frequency-domain resource allocation and air-domain resource allocation on the user queue to be scheduled with performance requirements parameters, and outputs the scheduling result to the three-dimensional resource mapping table module to generate a three-dimensional resource mapping table, The table includes a list of users qualified for scheduling in time-domain scheduling, a list of correspondence between users and resource blocks, and a transmission form of MIMO. The system performs table lookup operations on the generated mapping table, and performs rapid resource allocation based on the data in the table.

The time-domain resource allocator unit of the scheduler module is used to group the user's business according to different delay and rate requirements, calculate the resource occupation priority parameters of the user's business and the scheduling priority of the time domain, and accordingly The user performs time-domain scheduling, and outputs a list of users with scheduling qualifications; the frequency-domain resource allocator unit is used to calculate the scheduling priority in the frequency domain, and accordingly performs frequency-domain Allocation of resource RB; the airspace resource allocator unit is used to select the transmission mode of airspace resources according to the channel quality. When the signal-to-noise ratio is greater than the set threshold value, choose to use space multiplexing to transmit data, otherwise, choose to use space Diversity transmission data.

The present invention can achieve priority resource allocation for users with better channel quality, higher time delay requirements and higher rate requirements. On the basis of ensuring the service rate requirements, the uplink transmission delay of the system is reduced. In order to meet the requirement that uplink resource blocks must be continuous, a simplified method of extending from the best quality RB to both sides is adopted in RB allocation. The invention has the characteristics of reducing the complexity of the resource allocation algorithm, enhancing the reliability of wireless signal transmission and reducing the packet loss rate.

Claims (7)

1. A three-dimensional resource dynamic allocation method for reducing uplink transmission time delay of a TD-LTE/TD-LTE-A system is characterized by comprising the following steps in each operation period:

the first step is as follows: set users to omega_UE＝{UE₁,UE₂,…,UE_i,…,UE_NGrouping services contained in N according to different performance requirements, and establishing a service grouping table;

the second step is that: according to the long-term time delay and the long-term rate of the service group j belonging to the user i and the service group jCalculating resource occupation priority parameter PRI of service of user i belonging to service packet j_i,j(t)；

The third step: calculating the scheduling priority T of the time domain according to the maximum rate and the long-term rate of the user i at the time T, the cache size of the data volume to be sent and the resource occupation priority parameter_i(T) and performing time domain scheduling (TDPS) for the users according to the T_iAnd (t) scheduling the users in sequence in a descending order until the system stops scheduling when no resources are available. (ii) a

The fourth step: according to the maximum rate, long-term rate and T which can be achieved by the user i on the resource block m_i(t) calculating the scheduling priority F of the frequency domain_i,mAnd allocating Resource Blocks (RBs) to the users to be scheduled determined in the TDPS according to the resource blocks, and selecting F in the unallocated RB set and the unallocated user set_i,mMaximum RB-UE Pair { RB_x,UE_y}, RB will be_xTo a UE_yAnd with RB_xExpanding towards two sides for the center, and allocating adjacent unallocated RBs to UE_yUntil the resource blocks on the left side and the right side can not be expanded or the number of the distributed RBs is larger than or equal to the number of the RBs required by the user;

the fifth step: selecting a transmission mode of space domain resources according to the channel quality when the signal to noise ratio SINR is larger than the SINR₀When selecting to use space multiplexing to transmit data, otherwise, selecting to use space diversity to transmit data, wherein, SINR₀A threshold value for selecting a transmission mode;

and a sixth step: after the relevant calculation of time domain, frequency domain and space domain three-dimensional resource allocation is finished, generating a three-dimensional resource mapping table, wherein the three-dimensional resource mapping table comprises a user list for acquiring scheduling qualification, a corresponding relation list of users and resource blocks and a transmission form of MIMO; the system carries out table look-up operation on the mapping table and carries out rapid resource allocation according to data in the table; the process of scheduling and RB allocation is to repeat the above first to sixth steps at multiples of the transmission time interval.

2. The dynamic three-dimensional resource allocation method for reducing uplink transmission delay of TD-LTE/TD-LTE-A system according to claim 1, wherein said first step is based on

<math> <mrow> <msub> <mi>G</mi> <mi>i</mi> </msub> <mo>&Element;</mo> <mo>[</mo> <msubsup> <mi>d</mi> <mi>j</mi> <mi>budget</mi> </msubsup> <mo>,</mo> <msubsup> <mi>r</mi> <mi>j</mi> <mi>budget</mi> </msubsup> <mo>]</mo> <mi>j</mi> <mo>=</mo> <mn>1,2</mn> <mo>,</mo> <mo>.</mo> <mo>.</mo> <mo>.</mo> <mo>,</mo> <mi>J</mi> </mrow> </math> [ rule 1]

The services comprised by the users are grouped, wherein,is the delay threshold for the traffic packet j,the rate requirement value for traffic packet j.

3. The dynamic three-dimensional resource allocation method for reducing uplink transmission delay of TD-LTE/TD-LTE-A system as claimed in claim 1, wherein said second step is based on

[ rule 2]

<math> <mfenced open='{' close=''> <mtable> <mtr> <mtd> <msub> <mi>D</mi> <mrow> <mi>i</mi> <mo>,</mo> <mi>j</mi> </mrow> </msub> <mrow> <mo>(</mo> <mi>t</mi> <mo>)</mo> </mrow> <mo>=</mo> <mrow> <mo>(</mo> <mn>1</mn> <mo>-</mo> <mi>&alpha;</mi> <mo>)</mo> </mrow> <msub> <mi>D</mi> <mrow> <mi>i</mi> <mo>,</mo> <mi>j</mi> </mrow> </msub> <mrow> <mo>(</mo> <mi>t</mi> <mo>-</mo> <mn>1</mn> <mo>)</mo> </mrow> <mo>+</mo> <mi>&alpha;</mi> <msub> <mi>D</mi> <mrow> <mi>i</mi> <mo>,</mo> <mi>j</mi> </mrow> </msub> <mrow> <mo>(</mo> <mi>t</mi> <mo>)</mo> </mrow> </mtd> </mtr> <mtr> <mtd> <msub> <mi>R</mi> <mrow> <mi>i</mi> <mo>,</mo> <mi>j</mi> </mrow> </msub> <mrow> <mo>(</mo> <mi>t</mi> <mo>)</mo> </mrow> <mo>=</mo> <mrow> <mo>(</mo> <mn>1</mn> <mo>-</mo> <mi>&alpha;</mi> <mo>)</mo> </mrow> <msub> <mi>R</mi> <mrow> <mi>i</mi> <mo>,</mo> <mi>j</mi> </mrow> </msub> <mrow> <mo>(</mo> <mi>t</mi> <mo>-</mo> <mn>1</mn> <mo>)</mo> </mrow> <mo>+</mo> <mi>&alpha;</mi> <msub> <mi>R</mi> <mrow> <mi>i</mi> <mo>,</mo> <mi>j</mi> </mrow> </msub> <mrow> <mo>(</mo> <mi>t</mi> <mo>)</mo> </mrow> </mtd> </mtr> </mtable> </mfenced> </math> [ gauge ]Then 3]

Calculating resource occupation priority parameter PRI of service of user i belonging to service group j_i,j(t) wherein D_i,j(t) and R_i,j(t) respectively the long-term delay and the long-term rate of the service packet j belonging to the user i, alpha is a sliding window constant, alpha is more than 0 and less than 1,the method is a correction value introduced for avoiding packet loss caused by that resources are not distributed when the maximum time delay of the service is reached.

4. The three-dimensional resource dynamic allocation method for reducing uplink transmission delay of TD-LTE/TD-LTE-A system according to claim 3, wherein said third step is based on

<math> <mrow> <msub> <mi>T</mi> <mi>i</mi> </msub> <mrow> <mo>(</mo> <mi>t</mi> <mo>)</mo> </mrow> <mo>=</mo> <mfrac> <mrow> <msub> <mi>r</mi> <mi>i</mi> </msub> <mrow> <mo>(</mo> <mi>t</mi> <mo>)</mo> </mrow> </mrow> <mrow> <msub> <mi>R</mi> <mi>i</mi> </msub> <mrow> <mo>(</mo> <mi>t</mi> <mo>)</mo> </mrow> </mrow> </mfrac> <mo>&CenterDot;</mo> <msub> <mi>Q</mi> <mi>i</mi> </msub> <mrow> <mo>(</mo> <mi>t</mi> <mo>)</mo> </mrow> <mo>&CenterDot;</mo> <munder> <mi>&Sigma;</mi> <mi>j</mi> </munder> <msub> <mi>PRI</mi> <mrow> <mi>i</mi> <mo>,</mo> <mi>j</mi> </mrow> </msub> <mrow> <mo>(</mo> <mi>t</mi> <mo>)</mo> </mrow> </mrow> </math> [ rule 5]

Calculating scheduling priority T of time domain_i(t) in which，r_i(t) is the maximum rate that user i can reach at time t, R_i(t) is the long-term rate of user i at time t, and the calculation method is the same as [ rule 3 ]]，Q_i(t) the buffer size of the amount of data to be sent reported by user i through the buffer status report,the resource occupancy priority parameter for user i.

5. The three-dimensional resource dynamic allocation method for reducing uplink transmission delay of TD-LTE/TD-LTE-A system according to claim 3, wherein said fourth step is based on

$F_{i,m}=\frac{r_{i,m}\left(t\right)}{R_{i,m}\left(t\right)}{T}_i\left(t\right)$ [ rule 6]

Calculating the scheduling priority F of the frequency domain_i,mWherein r is_i,m(t) and R_i,m(t) the maximum data transmission rate and long-term rate that can be achieved on RB m when RB m is allocated to user i, respectively; according to

[ rule 7]

Calculating the number of RBs required by the user, wherein Q_iAnd (t) the buffer size of the data volume to be sent reported by the user i through the buffer status report, and M is the total number of all resource blocks in the system.

6. The method as claimed in claim 1, wherein the RB is used in the fourth step_xExpanding towards two sides for the center, and allocating adjacent unallocated RBs to UE_yThe following rules are adopted: when [ rule 8 ] is satisfied]Time to RB_x+1To a UE_yWhen [ rule 9 ] is satisfied]Time to RB_x-1To a UE_y，

<math> <mfenced open='{' close=''> <mtable> <mtr> <mtd> <msub> <mi>F</mi> <mrow> <mi>y</mi> <mo>,</mo> <mi>x</mi> <mo>-</mo> <mn>1</mn> </mrow> </msub> <mo>&le;</mo> <msub> <mi>F</mi> <mrow> <mi>y</mi> <mo>,</mo> <mi>x</mi> <mo>+</mo> <mn>1</mn> </mrow> </msub> </mtd> </mtr> <mtr> <mtd> <mi>max</mi> <mrow> <mo>(</mo> <msub> <mi>F</mi> <mrow> <mi>i</mi> <mo>,</mo> <mi>x</mi> <mo>+</mo> <mn>1</mn> </mrow> </msub> <mo>)</mo> </mrow> <mo>=</mo> <msub> <mi>R</mi> <mrow> <mi>y</mi> <mo>,</mo> <mi>x</mi> <mo>+</mo> <mn>1</mn> </mrow> </msub> </mtd> </mtr> </mtable> </mfenced> </math> [ rule 8]

$\left\{\begin{array}{c}{F}_{y,x-1}>F_{y,x+1}\\ \max \left(F_{i,x-1}\right)=F_{y,x-1}\end{array}\right.$ [ rule 9]

Wherein, F_y,x-1For the UE_yFrequency domain scheduling priority value when occupying resource block x-1, F_y,x+1For the UE_yAnd the frequency domain scheduling priority value when the resource block x +1 is occupied.

7. The three-dimensional resource dynamic allocation device for reducing the uplink transmission time delay of the TD-LTE/TD-LTE-A system is characterized by comprising the following components:

a queue module of the users to be scheduled,

the parameter input module is used for receiving the user queue to be scheduled output by the user queue to be scheduled module, setting the performance requirement parameters of each user to be scheduled, and outputting the user queue to be scheduled with the performance requirement parameters, wherein the performance requirement parameters comprise the maximum time delay and rate requirements of the users;

the scheduler module is used for sequentially performing time domain resource allocation, frequency domain resource allocation and space domain resource allocation on the user queue to be scheduled with the performance requirement parameters;

the three-dimensional resource mapping table module is used for generating a three-dimensional resource mapping table by combining the user queue to be scheduled and the distribution result of the scheduler module, and the mapping table comprises a user list for obtaining scheduling qualification in time domain scheduling, a corresponding relation list of users and resource blocks and a transmission form of MIMO;

the scheduler module comprises:

the time domain resource distributor unit is used for grouping the services of the users according to different time delay and rate requirements, calculating resource occupation priority parameters and time domain scheduling priorities of the services of the users, performing time domain scheduling on the users according to the resource occupation priority parameters and the time domain scheduling priorities, and outputting a user list with scheduling qualification;

the frequency domain resource distributor unit is used for calculating the scheduling priority of the frequency domain and distributing the frequency domain resource RB to the users with scheduling qualification output by the time domain resource distributor;

and the space domain resource distributor unit is used for selecting a transmission mode of space domain resources according to the channel quality, selecting to use space multiplexing to transmit data when the signal-to-noise ratio is greater than a set threshold value, and otherwise, selecting to use space diversity to transmit data.