CN112399427A - A method for intelligent spectrum sharing in 5G communication system - Google Patents

Abstract

The invention relates to a method for intelligent spectrum sharing in a 5G communication system, and belongs to the field of intelligent spectrum sharing. The purpose is to solve the problem of lack of flexibility in spectrum sharing in the prior art. The method of the invention aims at reducing the conflict between the CUs when the AU receives the interference of the CU, and ensures the normal communication of the 5G communication system. Spectrum sharing can be studied in the frequency, time, and space domains to realize intelligent spectrum sharing methods in 5G communication systems, analyze the competition between CUs in the frequency domain, and determine the number of CUs in the space and the available channels. The number of CU and AU is judged in the time domain, and the total utility of the system is statistically analyzed to improve the performance of the utility target of the simulation analysis system, thereby increasing the total data exchange volume and meeting the data transmission requirements.

Description

A method for intelligent spectrum sharing in 5G communication system

technical field

The invention relates to a method for intelligent spectrum sharing in a 5G communication system, in particular to a method for intelligent spectrum sharing in a 5G communication system, and belongs to the technical field of communication.

Background technique

The arrival of the 5G era is accelerating, and spectrum sharing is the general trend. The existing 5G communication system is short of spectrum resources, and there is an urgent requirement to establish a good wireless communication system. Through refined spectrum management, spectrum sharing can be achieved in the frequency domain, time domain, and spatial domain. The contradiction in using spectrum holes is the conflict between Authorized User (AU) and Cognitive User (CU). In order to ensure the normal communication of AU on its frequency band, when AU returns, CU needs to release its The occupied channel is "returned" to the AU. However, considering that the AU returns to a certain frequency band is a random behavior, and the CU suddenly switches to an idle channel, the performance of the system is affected to varying degrees, so that the CU can fully realize its own communication requirements in the spectrum hole used by it. , it is necessary to analyze the state of the channel and the length of time it maintains.

The idle spectrum is the spectrum hole generated by the AU not using it. A spectrum hole generally has the attributes of spatial domain, frequency domain and time domain, that is, in a certain area (the number of CUs), in a certain section of spectrum (a channel), there is no AU in a certain time range. In the research in this field, the "space-time-frequency domain" characteristics of spectrum resources have not been paid enough attention. amount of data exchange. However, if the variation characteristics of the idle spectrum in the time domain, the number of CUs in the space domain, and the AU idle channels in the frequency domain are not considered, there will be a problem that the shared channels cannot meet the actual needs of the CUs, which affects the feasibility of the system. sex.

In most of the current spectrum sharing research, the system objective cannot be well optimized without studying the spectrum hole situation. Moreover, in the common multiple sharing, although the period can be long or short, it is still regarded as a fixed value, which will limit the flexibility of spectrum resource sharing. Therefore, the present invention proposes a method for intelligent spectrum sharing in a 5G communication system from the perspective of the number of times of multiple sharing.

The key problem of spectrum sharing technology is to avoid interfering with the normal communication of AUs. The main principle of spectrum sharing technology based on greedy algorithm is to randomly select each CU, count the expected usage time and spectrum characteristics of the CU, and solve the CU in turn. The performance on each spectrum is comprehensively considered to share the spectrum with the best performance for the user. At the same time, the spectrum is deleted from the free spectrum pool, and the same method is used to select other users. When all the spectrum is shared, the method ends. The greedy algorithm considers the performance differences of different CUs on different attribute channels, but under the condition that the number of idle channels is less than the number of CUs, it is considered that each CU can only be shared with one or no sub-channels, although it is helpful to study the time of spectrum sharing. Domain characteristics, but not in line with the actual communication situation. Aiming at the time-frequency characteristics of spectrum resource holes, the previously studied stochastic algorithms pointed out the necessity of resource sharing in the time domain. The ON/OFF model is used to analyze the duration distribution of spectrum holes, which can be used to describe the transition between the channel occupied and idle states of the AU. ON is occupied, indicating that data is being sent, and OFF is idle, indicating that no data is being sent. . The previously proposed greedy spectrum sharing algorithm uses the Markov model to analyze the amount of transmitted data and obtains the transformation relationship between the time domain and the frequency domain. Spectral holes are analyzed, but all are considered when a single CU uses spectral holes. In practical applications, CUs of different service types have different communication times, and therefore have different degrees of sensitivity to the time-variation of the same spectrum hole. For a CU with a short communication time, the sensitivity to the time-variation of spectral holes is relatively low, but for a CU with a long communication time, the sensitivity to the time-variation of spectral holes is relatively high. Therefore, to consider the time-varying spectrum holes, not only from the perspective of spectrum holes, but also from the perspective of CU usage.

When using the idle spectrum, it is necessary to analyze the different states of the channel and their duration, so that the CU can meet the communication needs when using the spectrum hole. Since the frequency band used by the AU is random, in order not to affect the AU, when the AU returns to the frequency band, the CU must unconditionally vacate the channel being used and return it to the AU. If the AU does not use the spectrum temporarily, it will bring some spectrum holes. The shared channel may not meet the actual requirements of the CU, and the optimization goal of the system cannot be achieved.

SUMMARY OF THE INVENTION

The present invention is to solve the problem that the sharing competition period is fixed and lacks flexibility in the prior art when spectrum resources are shared, and proposes a method for intelligent spectrum sharing in a 5G communication system. The specific steps of the method are as follows:

Step 1: Define α as the ratio of the number of unoccupied spectrum in the spectrum system to the number of spectrum in the AU, α ₀ is the threshold for statistical analysis by the spectrum sharing agency If α ≥ α ₀ , the spectrum will be re-shared, if α < α ₀ , the spectrum will not be re-shared;

为CU n共享m_n时的速率，其中m_n为CU n能够共享的空闲信道数；Step 2: Spectrum sharing agency receives N CU application requirements

is the rate when CU _n shares mn, where mn is the number of idle channels that can be shared by CU _n ;

单位信道宽度Δf、数据传输需求D_n、开始共享的CU数量q，且q≤N；Step 3: Initial setting of parameters: total data exchange volume TR, rate

Unit channel width Δf, data transmission requirement D _n , number q of CUs to start sharing, and q≤N;

然后将

从大到小排序，根据D_n、Δf和m_n个空闲信道的带宽

求得每个CU在其可使用的信道上的传输数据时间长度t_n：Step 4: Find N CUs

followed by

Sort from largest to smallest, according to the bandwidth of D _n , Δf and m _n idle channels

Find the transmission data time length t _n of each CU on its available channels:

Step 5: Obtain the probability P(t _n ) that the CU successfully shares the use time length according to t _n , and then obtain the total data exchange volume of the system:

为m_q个可使用信道的带宽；where P(t _n ) is the statistical average data exchange volume, m _q represents the number of available channels shared with q CUs,

is the bandwidth of m _q available channels;

Step 6: The newly added CU n' selects the set of shared channels:

JH ₁ ={n|m _n ≥m _n′ ,n,n′∈[0,q-1]},JH ₂ ={n|m _n ≥1,n∈JH ₁ },

代表新增的CU n′在m_q上的速率，JH₁为符合m_n≥m_n′情况下CU的集合，JH₂为符合m_n≥1情况下CU的集合，JH₃为符合

情况下CU的集合，如果JH₃为空，停止频谱共享，执行步骤七，如果JH₃有值，则转至步骤五；where m _n' represents the number of channels shared by the newly added CU n',

Represents the rate of the newly added CU n' on m _q , JH ₁ is the set of CUs that conform to the condition of m _n ≥ m _n' , JH ₂ is the set of CUs that conform to the condition of m _n ≥ 1, and JH ₃ is the set of CUs that conform to the condition of m n ≥ 1

In the case of the set of CUs, if JH ₃ is empty, stop spectrum sharing, go to step 7, if JH ₃ has a value, go to step 5;

Step 7: Pre-share the channel for the newly added CU n';

Step 8: The newly added CU n' uses the minimum time μ of each channel at this time. If μ > t _n' , confirm the trial shared channel, transform parameters: q=q+1, go to step 9; if μ≤t _{n ′} , the sharing fails, then m _q = 0, t _q = 0, m _n′ = 2m _n′ , t _n′ = t _n′ /2, go to step ten;

Step 9: After this sharing is over, update the values of m _q and t _q , and then feed back the results to the CU, each CU accesses the channel used for sharing, update α=0, and go to step 1; start sharing adjustment for other CUs , when the number of user channels for obtaining spectrum is reduced, and when a stable value is reached, perform step 10;

Step 10: Sharing ends.

Further, in the described step 1, α is specifically:

表示此刻频谱池中的可以共享频谱总量，

表示M个AU占有的频谱总数。

Indicates the total amount of spectrum that can be shared in the spectrum pool at this moment,

Indicates the total number of spectrums occupied by M AUs.

Further, in the step 4, the data transmission requirement D _n is specifically:

When the amount of data required by CU n is known, the data transmission requirements are:

为m_n个可以共享使用信道的带宽；单位信道带宽记为Δf；f_s为采样点频率，

为噪声功率；K为无编码QAM调制方式和香农容量的信噪比关系，若信道为瑞利的，K可以表示为：where p′ _n,m represents the transmission power of CU n on channel m, t _n represents the time length of data transmission, |H _n,m | ² represents the power gain of CU n on channel m, assuming that all CUs on all channels have The gain is the same, namely: |H _n,m | ² =|H _m | ² ,

is the bandwidth of m _n channels that can be shared; the unit channel bandwidth is recorded as Δf; f _s is the sampling point frequency,

is the noise power; K is the signal-to-noise ratio relationship between the uncoded QAM modulation mode and the Shannon capacity. If the channel is Rayleigh, K can be expressed as:

where BER represents the bit error rate.

Further, the specific process of obtaining the total data exchange volume in the step 5 is:

时，B_n-y和T_n-y分别代表目前贪婪频谱共享策略中的CU被分配的信道带宽和期望使用的时间长度，B_n-h和T_n-h分别代表智能频谱共享算法中CU被分配的信道带宽和期望使用的时间长度，若满足下式，则5G通信系统能够使用智能频谱共享方法：Assuming that the utility target is the total data exchange volume at this time, when the data transmission demand D _n and transmission rate of CU n are known

, B _ny and T _ny represent the allocated channel bandwidth and expected use time of the CU in the current greedy spectrum sharing strategy, respectively, and B _nh and T _nh represent the allocated channel bandwidth and expected use of the CU in the smart spectrum sharing algorithm, respectively. If the following formula is satisfied, the 5G communication system can use the intelligent spectrum sharing method:

表明CU n在m_n上的效率，其中

记为频谱m_n的带宽B，设共享的空闲信道的初始状态都是空闲的，开始使用是0时刻，则获得概率为：Among them, N ₁ is the number of CUs that have shared channels; N ₂ is the number of CUs obtained in the intelligent spectrum sharing method; P(T _nh ) is the probability that CU n can fully use the expected occupancy time using the intelligent spectrum sharing algorithm, P(T _ny ) is the probability of using the current greedy spectrum sharing method for CU n to use the duration;

show the efficiency of CU n on m _n , where

It is denoted as the bandwidth B of the spectrum m _n . Assuming that the initial state of the shared idle channel is all idle, and the start of use is at time 0, the acquisition probability is:

Among them, λ _n is the case from unshareable to shareable, μ _n is from shareable to unshareable, t is T _n-new or T _n-old , and x is the integral variable of time.

Further, the added CU n' pre-shared channel in the step 7 is:

m _n′ =m _n′ /2,t _n′ =2t _n′ ;

where D _q represents the data transmission requirement of CU q, and t _n' represents the expected holding time of the increased CU n' on its shared channel.

The beneficial effects of the present invention are embodied in:

From the perspective of the times of multiple sharing, an intelligent spectrum sharing method is proposed. An effective spectrum sharing algorithm is proposed to improve system performance, ensure CU communication quality, and share suitable spectrum to CU.

In the usual multiple sharing, the sharing interval, although it can be fast or slow, is still a constant value, which will restrict the flexibility of spectrum sharing. Therefore, the present invention proposes an intelligent spectrum sharing method considering how many times to share. The state of mutual influence between AU and CU is studied from the perspective of space, the probabilistic influence degree of AU and CU from the perspective of time, and the competition relationship between CUs from the perspective of frequency. When analyzing the total utility of the system, study the probability of mutual influence, which is in line with the realistic scenario of the system.

The advantages of spectrum sharing are: dynamic response can be made according to the real-time changes of the system, spectrum sharing can be started more intensively during the period when the system idle resources are abundant, and the number of sharing times can be intelligently reduced during the period when the system idle resources are relatively tight;

It can be seen from the above start-up conditions that the existence of α ₀ and the difference in the time requirements of each CU leave an appropriate amount of idle resources in the system. This part of idle resources can ensure that the spectrum switching requirements of CUs are met at necessary moments, thereby reducing the need for The call drop rate of the small system enhances the robustness of the system.

Description of drawings

1 is a schematic flowchart of an intelligent spectrum sharing system and method;

Fig. 2 is the relation diagram of CU time requirement and channel statistics duration;

Fig. 3 is the time length diagram of increasing the shared channel and the less channel;

Figure 4 is a system model diagram; b1--bN represents the information from CU1 to N used for SMD judgment and analysis, X[1]-X[N] represents the spectrum information available from CU1 to CUN; English meaning, AU-i represents the spectrum i of the authorized user, AU-i+1 represents the spectrum i+1 of the authorized user, CU1 represents the cognitive user 1, CU6 represents the cognitive user 6, AU1 represents the authorized user 1, and AU2 represents the authorized user. User 2, AU3 represents authorized user 3, and AU4 represents authorized user 4;

Fig. 5 is the relation diagram of data traffic and CU number;

FIG. 6 shows the relationship between the data exchange amount and the number of CUs.

Detailed ways

The specific embodiments of the invention are described in conjunction with the accompanying drawings, and the specific implementation process is as follows:

Embodiment 1: If X _n represents the state at time n, and the set of random variables X _i constitutes the state space, then the conditional probability distribution of X _n+1 to the past state is only the expression of X _n , that is,

P(Xn _{+1 =} m| _X0 ,X1,..., _Xn )=P(Xn ₊₁ =m| _Xn ) (1)

This theory is an important model. When the state of the random process at time t ₀ is determined, the state at time t (t ₀ <t) has nothing to do with the state before time t ₀ .

Suppose the state space S={1,2,…,N} of a homogeneous Markov process {X′(t),t≥0} with continuous state discrete parameters, the transition probability matrix is P(t), and the density matrix is Q, and the initial distribution is P={p ₁ ,p ₂ ,...,p _N }. It is assumed that when the system is in state 1,2,…,K, the system can work normally; and when it is in state K+1,…,N, the system fails to operate normally and needs to be repaired.

The expected occupation time requirement t _n of the CUn is greater than the idle time of the channel.

Since the regression of the AU has random characteristics, the actual idle duration of the channel may be greater than or less than this value. In order to meet data transmission requirements and avoid interference, if the amount and rate of transmission data are known, the system can reasonably share the expected time length and the number of channels required by the CU.

This embodiment adopts a basic model. In this model, the spectrum information randomly released by the AU and the demand for spectrum resources of each CU are uniformly collected and managed by the system management department SMD set by the system. The spectrum system is used to collect the current usage of each spectrum by each AU. If a certain segment of spectrum is not used by the corresponding AU, the spectrum in this segment is marked as available, otherwise it is unavailable. In addition, when an AU returns to a certain spectrum. When it is on the spectrum of the segment, the system will make corresponding marker rewriting.

In the model of the embodiment, the SMD adopts the intelligent spectrum sharing method, and the system performs the sharing in an intelligent manner each time, and re-shares when the number of available spectrums in the current system is greater than the threshold value of the statistical analysis of the spectrum sharing agency. For _CUi , whenever it has a spectrum sharing requirement, it submits its application requirement bi to SMD, and SMD judges the requirement and feeds back the result to CU when re-sharing. The requirement _bi mainly includes the information used for _{SMD judgment} and _analysis , which is often different in different goal-oriented systems. t _i ), j _i , t _i and d _i represent the channel unit price, required duration and required quantity that the CU is willing to pay respectively; in the system aiming at the system data exchange volume, _bi = (γ _i ,d _i , t _i ), γ _i is the channel utilization rate of the CU (reflecting the transmission capability of the CU).

The assumptions of the intelligent spectrum sharing method in spectrum holes are as follows:

1. The spectrum system can count the average value T _aver-j of the idle duration of different channels. The system channel has only two states, namely the occupied state (Y(t)) and the idle state (X(t));

2. The target of CU access spectrum is to transmit data, and the data exchange volume D _i and transmission rate R _n are determined;

3. The time lengths of the idle and working states of the channel all obey the negative exponential distribution, the constant λ represents the average statistical time in the idle state, and the constant μ represents the average statistical time from the working state, then the idle time probability is P{X≤t}= 1-e- ^μt , t≥0, μ>0, where X represents the idle time. P{ ^Y≤t }=1-e-λt, where t≥0, λ>0, Y represents the occupied time; the state can be changed at any time, but cannot be changed more than twice within a very small time Δt; on different channels users will not interfere with each other, that is, the user's transmit power value is not restricted.

Based on the above analysis and assumptions, the steps, feasibility and effectiveness of the intelligent spectrum sharing method for 5G communication systems are described below.

The working and idle state change probability of the system can be expressed by Equation (2):

where o(Δt) represents an order of magnitude smaller than Δ(t), which can be ignored.

Assuming that the probabilities of the system being in the idle state and the occupied state are P ₀ =P{X(t)=0} and P ₁ =P{Y(t)=1}, respectively, according to the definition of the differential equation, equation (3) can be obtained :

The Laplace transform of formula (3) can be obtained:

Wherein P ₀ ^* (s) represents the Laplace transform of P ₀ '(t), and P ₁ ^* (s) represents the Laplace transform of P ₁ '(t).

For the system in the idle state at the initial moment, (P ₀ (0),P ₁ (0))=(0,1), the above equation can be solved to get:

For the system in the occupied state at the initial moment, (P ₀ (0), P ₁ (0))=(1,0) can be solved:

存在，记为Γ，则称Γ为系统的稳态有效度，它表示运行时间足够长时系统正常的概率。No matter what state the system is in at the initial moment, P ₀ (0) is called the validity (credibility) of the system, and is represented by Γ(t), if

exists, denoted as Γ, then Γ is called the steady-state effectiveness of the system, which represents the probability that the system is normal when the running time is long enough.

The longer the time t, the lower the effectiveness of a CU in using a channel Γ(t)=0. This embodiment refers to the definition of "task effectiveness" in reliability theory to describe the probability that a CU occupies an idle channel within a period of time T. In the range of time [t ₁ , t ₂ ], the probability that the system is in the normal use state is called the task effectiveness, expressed as Γ(t ₁ , t ₂ ), then:

Step 1: If α≥α ₀ , re-share the spectrum, otherwise SMD continues to serve and wait; where SMD is the spectrum detection department, α represents the percentage of the total spectrum that can be shared in the current system to the total spectrum owned by the AU, α ₀ The minimum input value set for the spectrum monitoring department, SMD is the spectrum management department;

Step 2: SMD receives the requirements of N CUs

Step 3: The system sets variables, data exchange amount TR, data transmission demand D _n , unit channel width Δf, initially shared CU q, q≤N, the transmission rate of CU _n on the spectrum mn

从大到小排列；根据确定的数据传输需求量D_n，计算各个CU在其信道上的使用时长t_n：Step 4: Individual CU Compliance

Arrange from large to small; according to the determined data transmission demand D _n , calculate the usage duration t _n of each CU on its channel:

为m_n个空闲信道的带宽。where m _n represents the number of available channels obtained by CU n, Δf is the unit channel bandwidth,

is the bandwidth of m _n idle channels.

Step 5: Obtain the probability P(t _n ) that the CU can manipulate the duration according to t _n , this probability indicates the system data exchange amount, and the system data exchange amount obtained by formula (9) is:

为m_q个空闲信道的带宽；where m _q represents the number of channels obtained by q CUs,

is the bandwidth of m _q idle channels;

Step 6: For CU n' newly added to the system:

代表新加入的CU n′在频谱m_q上的传输速率，JH₁为合乎m_n≥m_n′时CU的集合，JH₂为合乎m_n≥1时CU的集合，JH₃为合乎

时CU的集合，推断JH₃是不是空集，假使其为空集，频谱共享停止，实施步骤7，假使不是空集，就实施步骤5；where m _n' represents the number of channels obtained by the newly added CU n',

Represents the transmission rate of the newly added CU n' on the spectrum m _q , JH ₁ is the set of CUs when m _n ≥ m _n' , JH ₂ is the set of CUs when m _n ≥ 1, JH ₃ is the set of CUs

When the set of CUs, it is inferred whether JH ₃ is an empty set. If it is an empty set, the spectrum sharing is stopped, and step 7 is implemented. If it is not an empty set, step 5 is implemented;

Step 7: Initial sharing for the newly added CU n';

Step 8: The newly added CUn′ uses the minimum value μ in each channel. If μ>t _n′ , confirm the initial sharing, replace the parameters: q=q+1, and implement step 9; if μ≤t _n′ , the sharing fails , then m _q = 0, t _q = 0, m _n′ = 2m _n′ , t _n′ = t _n′ /2, and step 10 is performed;

Step 9: After this sharing ends, replace the m _q and t _q values, and feed back the results to each CU, each CU accesses the channel obtained by each CU, replace α=0, and implement step 1; start to share with other CUs, after obtaining When the CU of the spectrum can no longer reduce the number of its channels, step 10 is performed;

Step 10: Sharing ends.

Embodiment 2: The difference between this embodiment and Embodiment 1 is that α in the step 1 is specifically:

表示当前系统中可以共享的频谱总量，

表示AU拥有的频谱总量。

represents the total amount of spectrum that can be shared in the current system,

Indicates the total amount of spectrum owned by the AU.

In common spectrum sharing, although the sharing interval period may be long or short, it is still a fixed value, which will limit the flexibility of spectrum sharing, and is not in line with the idea of this embodiment. Therefore, considering the time interval of spectrum sharing, an intelligent spectrum sharing method is proposed, and its specific definition is as follows:

Definition: Intelligent spectrum sharing, no fixed period is set between two adjacent spectrum sharing actions, and whether to start a new spectrum sharing is determined by the system.

则启动新一次竞争。其中，

表示当前系统中可以共享的频谱总量，

表示M个AU拥有的频谱总量。The conditions for intelligently starting a new spectrum sharing are: if

Then start a new competition. in,

represents the total amount of spectrum that can be shared in the current system,

Indicates the total amount of spectrum owned by M AUs.

Each CU is sorted according to the transmission rate from largest to smallest. When the newly added CU is shared with a certain spectrum, calculate whether its data exchange volume can increase. If it increases, the corresponding spectrum is given to the newly added CU, and the same goes on. Judge other CUs, on the contrary, if it does not increase, the corresponding spectrum will not be given to CUs.

Other steps and parameters are the same as in the first embodiment.

Embodiment 3: The difference between this embodiment and Embodiment 1 or 2 is that the data traffic D _n in the step 4 is specifically:

If the three dimensions of space, time and frequency are studied together, it can be obtained that the spatial domain (number of CUs), frequency (bandwidth) and time (transmission duration) are related to each other. When the amount of data required by the CU is known, the data transmission can be obtained. The demand is:

为m_n个空闲信道的带宽；单位信道带宽记为Δf；t_n表示传输时长；f_s为采样点频率，σ²为噪声功率where p′ _n,m represents the transmission power of CU n on channel m, and |H _n,m | ² represents the power gain of CU n on channel m, if the gain of each CU on each channel is the same, that is, : |H _n,m | ² = |H _m | ² ,

is the bandwidth of m _n idle channels; the unit channel bandwidth is denoted as Δf; t _n represents the transmission duration; f _s is the sampling point frequency, and σ ² is the noise power

K is the signal-to-noise ratio relationship between the uncoded QAM modulation mode and the Shannon capacity. If it is a Rayleigh channel, K is:

where BER represents the bit error rate.

Other steps and parameters are the same as in the first or second embodiment.

Embodiment 4: The difference between this embodiment and one of Embodiments 1 to 3 is that: the specific process of obtaining the system data exchange amount in the step 5 is as follows:

In view of the relationship between the space domain, time domain and frequency domain explained above, the probability that the CU has been using an idle channel for a period of time can be expressed by referring to the definition of task validity. When the demand and transmission rate are known, the available channels are analyzed. Parameters such as the number of channels and the number of CUs affect the system utility.

时，B_n-y和T_n-y分别代表目前贪婪频谱共享策略中的CU被分配的信道带宽和期望使用的时间长度，B_n-h和T_n-h分别代表智能频谱共享方法中CU被分配的信道带宽和期望使用的时间长度，若满足下式，则5G通信系统能够使用智能频谱共享方法：Assuming that the utility target of the current system is the system data exchange volume, when the data transmission demand D _n and transmission rate of CU n are known

, B _ny and T _ny represent the allocated channel bandwidth and expected time length of the CU in the current greedy spectrum sharing strategy, respectively, and B _nh and T _nh represent the allocated channel bandwidth and expected use of the CU in the smart spectrum sharing method, respectively. If the following formula is satisfied, the 5G communication system can use the intelligent spectrum sharing method:

表示CU为n在频谱m_n上的效率，其中

代表频谱m_n的带宽B，设共享的空闲信道的初始状态都是空闲的，开始使用是0时刻，那么根据式(6)、(7)能够获得此概率的表达式为：where N ₁ is the number of CUs obtained in the existing method; N ₂ is the number of CUs obtained in the intelligent spectrum sharing method; P(T _nh ) is the probability that CU n can fully use the expected occupancy time using the intelligent spectrum sharing method, P(T _ny ) is the probability that CU n can completely use the expected occupancy duration by adopting the current greedy spectrum method;

represents the efficiency of CU for n on the spectrum m _n , where

Representing the bandwidth B of the spectrum m _n , assuming that the initial state of the shared idle channel is all idle, and the start of use is at time 0, then the expression of this probability can be obtained according to equations (6) and (7):

Among them, λ _n represents from working state to idle state, μ _n represents from idle state to working state, t is T _nh or T _ny , and x represents the integral variable of time.

The utility goal of the system of formula (13) is to maximize the amount of system data exchange. It is necessary to comprehensively consider whether to increase the length of time, reduce the number of channels, or increase the number of channels and reduce the length of time, and judge and decide the rate reduction caused by the reduction of the number of channels and the expected use time. The pros and cons of the two situations arising from the conflict situation achieve utility goals.

Other steps and parameters are the same as one of the first to third embodiments.

Embodiment 5: The difference between this embodiment and one of Embodiments 1 to 4 is that the newly added CU n' pre-sharing in step 7 is specifically:

m _n′ =m _n′ /2,t _n′ =2t _n′ (16)

D _q represents the traffic required for data transmission of CU q, and t _n' represents the expected occupied duration of the newly added CU n' on the channel obtained by it.

Other steps and parameters are the same as one of the first to fourth embodiments.

取值为[4,6,8,10,12]bps；D_n的取值为[200,400,600,1000,1200]kbit；μ_n取值为[2,4,6,8,10]；λ_n取值为[4,6,8,10,12]；单位子信道的带宽为15KHz；CU n在信道m上的传输功率

为0.02W；σ2＝10-11W，BER＝10-5。Combined with the description of the above embodiment, compared with the existing method and verified, the system sets the variables:

The value is [4,6,8,10,12]bps; the value of D _n is [200,400,600,1000,1200]kbit; the value of μ _n is [2,4,6,8,10]; _λn The value is [4, 6, 8, 10, 12]; the bandwidth of the unit sub-channel is 15KHz; the transmission power of CU n on channel m

is 0.02W; σ2=10-11W, BER=10-5.

In view of the changing characteristics of spectrum holes, when the number of AUs is set to 28 and the number of CUs is set from 5 to 25, without loss of generality, the system data traffic of the intelligent spectrum sharing method, the two-dimensional method and the greedy method are compared. The number of CUs increases, so the smart spectrum sharing method performs best, outperforming the 2D method and the greedy method. Both the two-dimensional method and the greedy method randomly determine a channel priority and then perform spectrum sharing, but the idea of the intelligent spectrum sharing method is to measure the transmission speed priority to obtain the optimal match.

To evaluate the performance of the three methods from the perspective of system data exchange, when the number of AU idle channels is 28 and the number of CUs is from 5 to 25, without loss of generality, the performance curve is not smooth, because the The value range is randomly selected data, but this does not affect the trend and regular characteristics of the performance parameters.

The above are only preferred embodiments of the present invention and are not intended to limit the present invention. Any modifications, equivalent replacements, improvements, etc. made within the spirit and principles of the present invention shall be included in the protection of the present invention. within the range.

Claims (5)

1. an intelligent spectrum sharing method in a 5G communication system, is characterized in that: the concrete method steps are as follows: Step 1: Define α as the ratio of the number of unoccupied spectrums in the spectrum system to the number of spectrums in the AU at the moment, α ₀ is the threshold for statistical analysis by the spectrum sharing agency, if α ≥ α ₀ , the spectrum will be re-shared, if α <α ₀ , the spectrum will not be re-shared; Step 2: Spectrum sharing agency receives N CU application requirements

is the rate when CU _n shares mn, where mn is the number of idle channels that can be shared by CU _n ; Step 3: Initial setting of parameters: total data exchange volume TR, rate

Unit channel width Δf, data transmission requirement D _n , number q of CUs to start sharing, and q≤N; Step 4: Find N CUs

followed by

Sort from largest to smallest, according to the bandwidth of D _n , Δf and m _n idle channels

Find the transmission data time length t _n of each CU on its available channels:

Step 5: Obtain the probability P(t _n ) that the CU successfully shares the use time length according to t _n , and then obtain the total data exchange volume of the system:

where P(t _n ) is the statistical average data exchange volume, m _q represents the number of available channels shared with q CUs,

is the bandwidth of m _q available channels; Step 6: The newly added CU n' selects the set of shared channels: JH ₁ ={n|m _n ≥m _n′ ,n,n′∈[0,q-1]},JH ₂ ={n|m _n ≥1,n∈JH ₁ },

where m _n' represents the number of channels shared by the newly added CU n',

Represents the rate of the newly added CU n' on m _q , JH ₁ is the set of CUs that conform to the condition of m _n ≥ m _n' , JH ₂ is the set of CUs that conform to the condition of m _n ≥ 1, and JH ₃ is the set of CUs that conform to the condition of m n ≥ 1

In the case of the set of CUs, if JH ₃ is empty, stop spectrum sharing, go to step 7, if JH ₃ has a value, go to step 5; Step 7: Pre-share the channel for the newly added CU n'; Step 8: The newly added CU n' uses the minimum time μ of each channel at this time. If μ > t _n' , confirm the trial shared channel, transform parameters: q=q+1, go to step 9; if μ≤t _{n ′} , the sharing fails, then m _q = 0, t _q = 0, m _n′ = 2m _n′ , t _n′ = t _n′ /2, go to step ten; Step 9: After this sharing is over, update the values of m _q and t _q , and then feed back the results to the CU, each CU accesses the channel used for sharing, update α=0, and go to step 1; start sharing adjustment for other CUs , when the number of user channels for obtaining spectrum is reduced, and when a stable value is reached, perform step 10; Step 10: Sharing ends. 2. The intelligent spectrum sharing method in a 5G communication system according to claim 1, characterized in that: in the step 1, α is specifically:

represents the total amount of spectrum that can be shared in the spectrum system at this moment,

Indicates the total number of spectrums occupied by M AUs. 3. The intelligent spectrum sharing method in a 5G communication system according to claim 2, wherein the data transmission requirement D _n in the step 4 is specifically: When the amount of data required by CU n is known, the data transmission requirements are:

where p′ _n,m represents the transmission power of CU n on channel m, t _n represents the time length of data transmission, |H _n,m | ² represents the power gain of CU n on channel m, assuming that all CUs on all channels have The gain is the same, namely: |H _n,m | ² =|H _m | ² ,

is the bandwidth of m _n channels that can be shared; the unit channel bandwidth is recorded as Δf; f _s is the sampling point frequency,

is the noise power; K is the signal-to-noise ratio relationship between the uncoded QAM modulation mode and the Shannon capacity. If the channel is Rayleigh, K can be expressed as:

where BER represents the bit error rate. 4. a kind of intelligent spectrum sharing method in a 5G communication system according to claim 3, is characterized in that: in the described step 5, the specific process of obtaining total data exchange volume is: Assuming that the utility target is the total data exchange volume at this time, when the data transmission demand D _n and transmission rate of CU n are known

, B _ny and T _ny represent the allocated channel bandwidth and expected use time of the CU in the current greedy spectrum sharing strategy, respectively, and B _nh and T _nh represent the allocated channel bandwidth and expected use of the CU in the smart spectrum sharing algorithm, respectively. If the following formula is satisfied, the 5G communication system can use the intelligent spectrum sharing method:

where N ₁ is the number of CUs that have shared the channel; N ₂ is the number of CUs of the channel obtained in the intelligent spectrum sharing method; P(T _nh ) is the probability that the CUn can fully use the expected occupancy time using the intelligent spectrum sharing algorithm, P(T _ny ) is the probability of using the current greedy spectrum sharing method for CU n to use the duration;

show the efficiency of CU n on m _n , where

It is denoted as the bandwidth B of the spectrum m _n . Assuming that the initial state of the shared idle channel is all idle, and the start of use is at time 0, the acquisition probability is:

Among them, λ _n is the case from unshareable to shareable, μ _n is from shareable to unshareable, t is T _nh or T _ny , and x is the integral variable of time. 5. The method for intelligent spectrum sharing in a 5G communication system according to claim 4, wherein the added CU n' pre-shared channel in the step 7 is:

m _n′ =m _n′ /2,t _n′ =2t _n′ ; where D _q represents the data transmission requirement of CU q, and t _n' represents the expected holding time of the increased CU n' on its shared channel.

Images