CN111314938B - An optimization method for cellular network time-frequency domain resource allocation for a single cell - Google Patents

Abstract

The present invention relates to an optimization method for cellular network time-frequency domain resource allocation for a single cell, which transforms a cellular network power consumption optimization model into a single-cell user uplink transmission system energy minimization combined with time-frequency domain The model involves resource allocation and transmit power control. First, according to the idea of Lagrangian dual algorithm, the optimal transmit power of each cellular user in each spectrum is obtained, and then the optimization model is solved by combining cooperative game theory. Compared with the common single-dimensional resource allocation, the energy consumption of the system is greatly reduced by the time-frequency domain resource allocation optimization method of the present invention.

Description

An optimization method for cellular network time-frequency domain resource allocation for a single cell

technical field

The invention relates to the technical field of cellular network resource allocation, in particular to an optimization method for cellular network time-frequency domain resource allocation for a single cellular cell.

Background technique

With the rapid development of mobile multimedia services and related applications, multimedia services such as high-definition video and online live broadcast have experienced explosive growth, and their high-traffic characteristics have brought huge pressure on operators' core networks and spectrum resources; 5G is to deal with the explosive growth of mobile data traffic. The multi-service and multi-technology converged network proposed in response to the growing demand for connection of massive devices is characterized by high speed, ubiquitous network, low power consumption, and low latency. Therefore, efficient and flexible allocation of cellular network resources can improve network performance, reduce system energy consumption, and achieve sustainable development.

However, the existing technology of resource allocation in cellular networks only performs single-dimensional resource allocation in the frequency domain or time domain to improve the performance of the system. The pure time domain resource allocation does not consider the frequency selective fading on the subcarriers, and the pure The frequency domain resource allocation does not take into account the different Quality of Service (QoS) required by users for different applications, which may lead to waste of resources or insufficient resources.

In view of this, the inventor of the present invention has deeply conceived the problem of resource allocation existing in the above-mentioned cellular network, and this case came into being.

SUMMARY OF THE INVENTION

In view of the problems existing in the prior art, the purpose of the present invention is to provide an optimization method for resource allocation in the time-frequency domain of a cellular network for a single cell, which combines resource allocation and transmission power control to minimize system energy consumption .

For achieving the above object, the technical scheme adopted in the present invention is:

A method for optimizing cellular network time-frequency domain resource allocation for a single cell, the method comprising the following steps:

Step 1. Obtain the total energy consumption of all cellular users in the cellular network under their respective resource allocations;

The energy consumption in the entire communication system is expressed by the summation formula of the respective energy consumption of all cellular users, then the total energy consumption in the communication system can be expressed as:

Among them, E _i is the energy consumption of cellular user i, and P _i is the transmit power of cellular user i, expressed as

中的X和P分别指问题所要优化的资源分配变量集

和功率变量集P＝(p_i,k)，

为蜂窝用户i在资源块的第k个子载波上第t个时隙的分配情况，R_i,min表示蜂窝用户i的最小速率要求；s _i ·P _i,cir is the circuit energy consumption generated by user i during the open time of sending data, (Ts _i ) P _i,idle is the idle power of cellular user i in the remaining time; M is the number of cellular users , s _i is the device opening time of the cellular user i, 0≤s _i ≤T, each single-cellular user in the cellular network only turns on its own sending device when it transmits to the base station; T is the time slot when the resource block is transmitted Quantity, Pi _,cir is the power consumed by other circuit blocks when the cellular user transmits resources in the uplink; when the cellular user has no data to send or after all data is sent, it closes all sending circuit blocks to save energy, but At this time, the energy consumption caused by the existence of leakage current is called the idle power of the cellular user, which is expressed as P _i,idle ;

X and P respectively refer to the resource allocation variable set to be optimized by the problem

and the power variable set P=(pi _,k ),

is the allocation of the t-th time slot on the k-th subcarrier of the resource block by the cellular user i, R _i,min represents the minimum rate requirement of the cellular user i;

In the cellular network, there is no interference between multiple cellular users in different subcarriers in the same time slot, and ri _,k is the achievable rate of cellular user i on subcarrier k, which can be expressed as

Among them, pi _,k is the transmit power of cellular user i on subcarrier k, k∈K, i∈M, N ₀ is the power spectral density of white Gaussian noise, g _i,k is the power spectral density of cellular user i on subcarrier k The channel gain on k, W is the bandwidth of the subcarrier, and K is the number of subcarriers in a resource block;

Step 2. Define the matrix

为K×T的矩阵，表示蜂窝用户i在一个资源块中的资源分配情况，其中，i＝1,2,…,M；k＝1,2,…,K；t＝1,2,…T；根据

可以令C_i,k＝A_i ^T·E统计用户i在各个子载波上传输的时间，将公式(1)的优化模型转写为：

is a K×T matrix, representing the resource allocation of cellular user i in a resource block, where i=1,2,...,M; k=1,2,...,K; t=1,2,... T; according to

C _i,k =A _i ^T ·E can be used to count the transmission time of user i on each subcarrier, and the optimization model of formula (1) can be transcribed as:

Step 3. Through the Lagrangian dual algorithm, the optimal transmit power P _opt =(pi _,k ) of each cellular user on each subcarrier can be obtained to meet the respective rate requirements of the cellular users;

Decompose the problem corresponding to formula (4) into M independent sub-problems, and write the corresponding Lagrangian function for the optimization model of each sub-problem, and set the current iteration count=0 in the optimization model of each sub-problem , the maximum number of iterations is countmax, and the convergence error e1=e- ¹⁰ ; in each iteration, the partial derivative of p _i,k is 0, and the equation can be obtained

为当前第i个子问题对应的优化模型的解，即用户i在子信道k上的发送功率，λ_i是指拉格朗日函数中的对偶变量；in,

is the solution of the optimization model corresponding to the current i-th sub-problem, that is, the transmit power of user i on sub-channel k, and λ _i refers to the dual variable in the Lagrangian function;

Step 4. Use the subgradient method to update the Lagrangian dual variables:

An extremely small step size μ _n is defined to ensure the convergence of the optimal value of the subgradient method, and the update of the Lagrangian dual variables is performed according to the following formula:

λ _i (n+1)=[λ _i (n)+μ _n d(λ _i (n))] ⁺ , i=1,2,...,M (7);

In each iteration, the updated dual variable is used to update the formula (5), and the solution of the optimization model corresponding to the final sub-problem will converge to the unique optimal solution;

Step 5. If the relative dual gap ||λ _i (n)-λ _i (n-1)||≤e1 or the current number of iterations exceeds countmax, the iteration is stopped, and the current cellular user is allocated in the resource block. The optimal transmit power feasible solution P _opt =(pi _,k ) is applied to the communication system, and the total energy consumption of all cellular users in the current system is calculated by formula (1); otherwise, proceed to step 3-4;

Step 6. Adjust and optimize the resource allocation of cellular users through cooperative game theory;

Define a maximum number of failures fail_max, and randomly extract a resource block to obtain the subcarrier and time slot corresponding to the resource block, as well as the cellular user User_ori currently assigned to it, and reassign this resource block randomly to all users except the original user. For any user User_new, update the resource occupancy list of the two users, and use steps 3-5 to obtain the corresponding optimal transmission power, compare the total energy consumption E_new of the system after redistribution and the total energy consumption E_ori of the original system, and compare There are three situations:

If E_new<E_ori, remove the resource block from the original user User_ori resource occupation list, add it to the resource occupation list of user User_new, update the total energy consumption of the system, and set the current number of failures fail=0;

If E_new=E_ori, update the current number of failures;

If E_new>E_ori, remove the resource block from the resource possession list of the user User_new and add it to the resource possession list of the user User_ori;

If the current number of failures reaches the maximum number of failures, the game theory is stopped, and the final user allocation and their respective optimal transmission powers are obtained to minimize the total energy consumption of system users.

After adopting the above scheme, the present invention transforms the power consumption optimization model of the cellular network into the energy consumption minimization model of the single-cell user uplink transmission system combined with the time-frequency domain, which involves resource allocation and transmission power control. The idea of the dual algorithm obtains the optimal transmit power of each cellular user in each frequency spectrum, and then combines the cooperative game theory to solve the optimization model. Compared with the common single-dimensional resource allocation, the energy consumption of the system is greatly reduced by the time-frequency domain resource allocation optimization method of the present invention.

Description of drawings

1 is a schematic structural diagram of a single cellular user communication system;

FIG. 2 is an example diagram of resource allocation for a single cellular user;

Figure 3 is a flow chart of a single-cell user controlling power.

Detailed ways

The present invention discloses an optimization method for resource allocation in the time-frequency domain of a cellular network in a single cell. In the cellular network of multiple cellular users in a single cell, the aim is to optimize the energy consumption of the system and consider any frequency spectrum. At any time, it can only be used for transmission by a single cellular user, and a user can occupy multiple time periods on multiple sub-channels for transmission. The total energy consumption of all users in the system is not only related to the allocation of cellular users in this resource block, but also depends on Based on their respective transmit powers, and considering these two factors at the same time, under the premise of ensuring the communication quality of each user, optimize the transmit power of each user in the system on each spectrum, and adjust the resource allocation of users to minimize the total system. energy consumption.

As shown in Figure 1, a cellular network consists of one base station and M cellular users. M cellular users share a resource block. Here, a resource block has K orthogonal subcarriers (Subcarriers) and T time slots. The T time slots constitute the data transmission time of Ts, and the bandwidth of each subcarrier is W, each user will have different channel gain and power on different subcarriers. As shown in Figure 2, any time period on each subcarrier can be allocated to any user, and the remaining time of the subcarrier can be shared among users, that is, one user can occupy multiple time periods on multiple subcarriers. All channel information is known. In a total resource block (Resource Block, RB) consisting of 12 consecutive subcarriers in a frequency domain and a time slot in the time domain, the base station can allocate suitable frequency bands to different users, and determine The transmit power of each user in this frequency band is the corresponding time, so that each user can complete their own transmission within one RB.

The optimization method of the present invention specifically comprises the following steps:

Step 1. Obtain the total energy consumption of all cellular users in the cellular network under their respective resource allocations.

In the uplink transmission link of the cellular network, the energy consumption of each user _i can be divided into three parts, one is the transmission power consumption P _i ; P _i,cir , and the third is the idle power consumption (Ts _i ) P _i,idle in the remaining time. Among them, s _i , 0≤s _i ≤T is the device turn-on time of the cellular user i, and each single-cell user in the cellular network turns on its own sending device only when it transmits to the base station.

The energy consumption in the entire communication system is expressed by the summation formula of the respective energy consumption of all cellular users, then the total energy consumption in the communication system can be expressed as:

中的X和P分别指问题所要优化的资源分配变量集

和功率变量集P＝(p_i,k)，

为蜂窝用户i在资源块的第k个子载波上第t个时隙的分配情况，R_i,min表示蜂窝用户i的最小速率要求。In formula (1), s _i ·P _i,cir is the circuit energy consumption generated by user i in the open time of sending data, (Ts _i ) P _i,idle is the idle power of cellular user i in the remaining time; M is the number of cellular users, s _i is the device opening time of cellular user i, 0≤s _i ≤T, each single-cellular user in the cellular network only turns on its own sending device when it transmits to the base station; T is the resource The number of time slots during block transmission, P _i,cir is the power consumed by other circuit blocks when the cellular user sends resources in the uplink; when the cellular user has no data to send or all data is sent, it closes all sending circuits block to save energy, but the energy consumption caused by the existence of leakage current is called the idle power of cellular users, expressed as P _i,idle ;

X and P respectively refer to the resource allocation variable set to be optimized by the problem

and the power variable set P=(pi _,k ),

is the allocation of the t-th time slot on the k-th subcarrier of the resource block for the cellular user i, R _i,min represents the minimum rate requirement of the cellular user i.

P _i can be expressed as:

为K×T的矩阵，表示蜂窝用户i在一个资源块中的资源分配情况，其中，i＝1,2,…,M；k＝1,2,…,K；t＝1,2,…T；；蜂窝网络中同一个时隙上的处于不同子载波的多个蜂窝用户之间没有干扰现象，r_i,k为用户i在子载波上k上的可达速率，可表示为definition

is a K×T matrix, representing the resource allocation of cellular user i in a resource block, where i=1,2,...,M; k=1,2,...,K; t=1,2,... T;; There is no interference between multiple cellular users in different subcarriers on the same time slot in the cellular network, ri _,k are the achievable rates of user i on subcarrier k, which can be expressed as

where pi _,k is the transmit power of user i on subcarrier k, k∈K, i∈M, N ₀ is the power spectral density of white Gaussian noise, g _i,k is the transmit power of user i on subcarrier k channel gain.

可以令C_i,k＝A_i ^T·E统计用户i在各个子载波上传输的时间，将三维资源分配矩阵转化为二维资源分配矩阵，优化问题可以转写为：Step 2. According to the possible resource allocation of cellular users in a resource block

C _i,k =A _i ^T ·E can be used to count the transmission time of user i on each subcarrier, and the three-dimensional resource allocation matrix can be transformed into a two-dimensional resource allocation matrix. The optimization problem can be transcribed as:

Step 3. Through the Lagrangian dual algorithm, the optimal transmit power P _opt =(pi _,k ) of each cellular user on each sub-carrier can be obtained to meet the respective rate requirements of the cellular users,

Decompose the problem corresponding to formula (4) into M independent sub-problems, and write the corresponding Lagrangian function for the optimization model of each sub-problem, and set the current iteration count=0 in the optimization model of each sub-problem , the maximum number of iterations is countmax, and the convergence error e1=e- ¹⁰ ; in each iteration, the partial derivative of p _i,k is 0, and the equation can be obtained

为当前第i个子问题对应的优化模型的解，即用户i在子信道k上的发送功率，λ_i是指拉格朗日函数中的对偶变量。in,

is the solution of the optimization model corresponding to the current i-th sub-problem, that is, the transmit power of user i on sub-channel k, and λ _i refers to the dual variable in the Lagrangian function.

Step 4. At the same time, use the subgradient method to update the Lagrangian dual variable:

An extremely small step size μ _n is defined to ensure the convergence of the optimal value of the subgradient method, and the update of the Lagrangian dual variables is performed according to the following formula:

λ _i (n+1)=[λ _i (n)+μ _n d(λ _i (n))] ⁺ ,i=1,2,...,M (7)

Step 5. If the relative dual gap ||λ _i (n)-λ _i (n-1)||≤e1 or the current number of iterations exceeds countmax, the iteration is stopped, and the current cellular user is allocated in the resource block. The feasible solution of transmit power P _opt =(pi _,k ) is applied to the communication system, and the total energy consumption of all cellular users in the current system is calculated by formula (1); otherwise, proceed to steps 3 and 4.

Steps 2 to 5 belong to the known resource allocation, optimizing power control part, and the flowchart of this part is shown in FIG. 3 .

Step 6: Adjust and optimize the resource allocation of cellular users through cooperative game theory.

Define a maximum number of failures fail_max=50, randomly select a small resource block from the resource block, you can know the subcarrier and time slot it corresponds to, and which user User_ori it is currently assigned to, and reassign the resource block randomly For any user User_new except the original user, update the resource occupancy list of these two users, and use steps 3, 4, and 5 to obtain the corresponding optimal transmission power, and compare the total energy consumption E_new and The total energy consumption E_ori of the original system is divided into three comparisons:

If E_new<E_ori, remove the resource block from the original user User_ori resource occupation list, add it to the resource occupation list of user User_new, update the total energy consumption of the system, and set the current number of failures fail=0;

If E_new=E_ori, update the current number of failures;

If E_new>E_ori, the resource block is removed from the resource occupation list of the user User_new and added to the resource occupation list of the user User_ori.

If the current number of failures has reached the maximum number of failures, the game theory is stopped, and the final user allocation and their respective optimal transmission power are obtained, so as to minimize the total energy consumption of system users.

The key point of the present invention is that the present invention converts the power consumption optimization model of the cellular network into the energy consumption minimization model of the single-cell user uplink transmission system combined with the time-frequency domain, which involves resource allocation and transmission power control. The idea of the sun-dual algorithm obtains the optimal transmit power of each cellular user in each frequency spectrum, and then combines the cooperative game theory to solve the optimization model. And it is proved by experiments that, compared with the common single-dimensional resource allocation, the time-frequency domain resource allocation optimization method of the present invention greatly reduces the energy consumption of the system by 8-25%.

The above are only the embodiments of the present invention and do not limit the technical scope of the present invention. Therefore, any minor modifications, equivalent changes and modifications made to the above embodiments according to the technical essence of the present invention still belong to the present invention. within the scope of the technical solution.

Claims (1)

1. A method for optimizing time-frequency domain resource allocation for a cellular network of a single cell, characterized by: the optimization method comprises the following steps:

step 1, solving the sum of energy consumption of all cellular users in a cellular network under respective resource allocation;

by using the energy consumption in the whole communication system as the sum formula of the energy consumption of each cellular user, the total energy consumption in the communication system can be expressed as:

wherein E is_iFor cellular user i energy consumption, P_iThe transmit power for cellular user i, denoted as

s_i·P_i,cirFor the circuit energy consumption generated by user i during the on-time of the transmitted data, (T-s)_i)P_i,idleIdle power for cellular user i for the remaining time; m is the number of cellular users, s_iFor the equipment on time of cellular user i, s is more than or equal to 0_iT is less than or equal to T, each single cellular user in the cellular network opens own sending equipment only when transmitting to the base station; t is time of resource block transmissionNumber of slots, P_i,cirIs the power consumed by other circuit blocks when a cellular user transmits resources in the uplink; when the cellular user has no data to transmit or has all data transmitted, it turns off all transmitting circuit blocks to save energy, but the energy consumption caused by the leakage current is called idle power of cellular user, and is denoted as P_i,idle；

X and P in (1) respectively refer to resource allocation variable sets to be optimized by the problem

And power variable set P ═ P (P)_i,k)，

For the allocation situation of the t time slot of the cellular user i on the k subcarrier of the resource block, R_i,minRepresents the minimum rate requirement of cellular user i;

there is no interference phenomenon between multiple cellular users on different sub-carriers in the same time slot in cellular network, r_i,kThe achievable rate for cellular user i on subcarrier k can be expressed as

Wherein p is_i,kTransmitting power of cellular user i on subcarrier K, K belongs to K, i belongs to M and N₀Power spectral density, g, of white gaussian noise_i,kChannel gain of a cellular user i on a subcarrier K is used, W is the bandwidth of the subcarrier, and K is the number of subcarriers of one resource block;

step 2, defining a matrix

Is a K × T matrix, indicating that cellular user i is at oneResource allocation case in resource block, where i ═ 1,2, …, M; k is 1,2, …, K; t is 1,2, … T; according to

Can order C_i,k＝A_i ^TE statistics of the time of transmission of user i on each subcarrier, transcribing the optimization model of equation (1) as:

step 3, obtaining the optimal transmitting power P of each cellular user on each subcarrier through a Lagrange dual algorithm_opt＝(p_i,k) To meet the respective rate requirements of the cellular users;

decomposing the problem corresponding to the formula (4) into M independent sub-problems, writing an optimization model for each sub-problem into a corresponding lagrangian function, setting the current iteration number count to be 0, the maximum iteration number to be count max, and the convergence error e1 to be e in the optimization model for each sub-problem^-10(ii) a For p in each iteration_i,kThe deviation derivative equation is 0, then

Wherein,

the solution of the optimization model corresponding to the current ith sub-problem, i.e. the transmission power of the user i on the sub-channel k, lambda_iIs a dual variable in a Lagrangian function;

and 4, updating Lagrangian dual variables by using a sub-gradient method:

defining a very small step size mu_nIn order to ensure the convergence of the optimal value of the sub-gradient method, the Lagrange dual variable is updated according to the following formula:

λ_i(n+1)＝[λ_i(n)+μ_nd(λ_i(n))]⁺,i＝1,2,...,M (7)；

in each iteration, the formula (5) is updated by using the updated dual variable, and finally, the solution of the optimization model corresponding to the subproblem is converged into a unique optimal solution;

step 5, if the relative dual gap | | | lambda_i(n)-λ_i(n-1) | < e1 or the current iteration times exceed countmax, stopping iteration, and obtaining the feasible solution P of the optimal sending power under the condition that the current cellular user is distributed in the resource block_opt＝(p_i,k) The method is applied to a communication system, and the total energy consumption of all cellular users of the current system is calculated by the formula (1); otherwise, continuing the step 3-4;

step 6, adjusting and optimizing the resource allocation of the cellular users through a cooperative game theory;

defining a maximum failure frequency fail _ max, randomly extracting a resource block, acquiring a subcarrier and a time slot corresponding to the resource block and a cellular User _ ori allocated to the resource block at present, randomly allocating the resource block to any User _ new except an original User again, updating resource occupation lists of the two users, obtaining corresponding optimal transmitting power by using the steps 3-5, and comparing the total system energy consumption E _ new after reallocation with the total system energy consumption E _ ori in the original system, wherein the comparison conditions are divided into three types:

if E _ new < E _ ori, removing the resource block from the original User _ ori resource occupation list, adding the resource block into the User _ new resource occupation list, updating the total energy consumption of the system, and setting the current failure frequency fail to be 0;

if E _ new is equal to E _ ori, updating the current failure times;

if E _ new > E _ ori, removing the resource block from the User _ new resource occupation list, and adding the resource block into the resource occupation list of the User _ ori;

and if the current failure times reach the maximum failure times, stopping the game theory, and obtaining the final user allocation condition and the respective optimal sending power of the users, so that the total energy consumption of the system users is minimum.

Images