CN104486767B - Dynamic ABS disturbance restraining methods based on sub-clustering in isomery cellular network - Google Patents

Abstract

The invention relates to a cluster-based dynamic ABS interference suppression method in a heterogeneous cellular network, which belongs to the technical field of wireless communication. This method firstly calculates the interference weight between dense small cells, establishes the interference relationship between small cell networks, that is, whether there is an interference edge between small cells, and then establishes an interference graph about the small cell interference relationship; then based on the established small cell interference graph , allocate small cells to K clusters, ensure the maximum interference weight between clusters during the allocation process, and finally allocate different frequency band resources for different clusters. In order to further reduce intra-cluster interference, a cross-type dynamic ABS allocation method is adopted for the small cells in the cluster. The small cells in the cluster are divided into two groups based on the size of the load. According to the load of the two groups, the average user transmission rate information, dynamic adjustment The ratio of ABS between groups further reduces the interference of small cells in the cluster, improves resource utilization, and thus improves system throughput.

Description

Clustering-Based Dynamic ABS Interference Suppression Method in Heterogeneous Cellular Networks

technical field

The invention belongs to the technical field of wireless communication, and relates to a dynamic ABS interference suppression method based on clustering in a heterogeneous cellular network.

Background technique

With the development of wireless communication networks, the future wireless network is evolving towards network intelligence, broadband, diversification and integration. With the widespread popularity of smart terminals, data services will experience explosive growth. In the future 5G network, data services will be mainly distributed indoors and hot spots, which makes the ultra-dense network one of the main means to realize the 1000 times traffic demand of future 5G. The ultra-dense network can improve network coverage, greatly increase system capacity, and offload services, enabling more flexible network deployment and more efficient frequency reuse. In the future, more dense network solutions will be adopted for high-frequency bands and large bandwidths, and more than 100 small cells/sectors will be deployed.

At the same time, the increasingly dense network deployment also makes the network topology more complex, and inter-cell interference has become the main factor restricting the growth of system capacity, which greatly reduces the network energy efficiency.

The interference management method in the traditional heterogeneous network mainly solves the interference problem from two perspectives: the first is from the perspective of the frequency domain, by allocating different resource blocks to different users at the same time to avoid interference in the same frequency band, or Some derivative methods, such as reducing the power of some network nodes in the same frequency band, thereby reducing the interference between the same frequency, such methods comprehensively consider the division of frequency domain resources, the algorithm complexity is generally high, and only from the perspective of the frequency domain To allocate resources, it is impossible to make an overall plan for system resources in terms of time, and the allocation of resources will be unreasonable in a long period of time; the other is from the perspective of time domain, such as eICIC technology, which is mostly used In the example of reducing the interference of the macro cell to the small cell, the range extension technology is used for the small cell, by reducing the power on the subframe of some time slots of the macro base station, thereby reducing the interference on the subframe of the corresponding small cell, so that in a certain This method improves the communication quality of the users in the small cells and expands the capacity of the small cells, but this method cannot coordinate the interference between the small cells well when the small cells are dense.

Therefore, in the high-density cellular network of 5G in the future, an effective interference management scheme is needed, which can effectively reduce the interference in the dense cellular network, and at the same time properly balance the load in the dense cellular network, thus extremely Greatly improve the capacity of the 5G system, obtain better spectrum utilization, and further improve the quality of service for users.

Contents of the invention

In view of this, the object of the present invention is to provide a cluster-based dynamic ABS interference suppression method in a heterogeneous cellular network, which can well improve the problem of severe interference in a dense cellular network.

To achieve the above object, the present invention provides the following technical solutions:

A method for suppressing dynamic ABS interference based on clustering in a heterogeneous cellular network, comprising the following steps:

Step 1: Calculate the interference weight between dense small cells, establish the interference relationship between the small cell networks, that is, whether there is an interference edge between the small cells, so as to establish an interference graph about the small cell interference relationship;

Step 2: Based on the established small cell interference graph, allocate small cells to K clusters, ensure that the interference weight between clusters is the largest during the allocation process, and finally allocate different frequency band resources for different clusters;

Step 3: The small cells in the cluster are divided into two groups based on the load, the primary group G ₁ and the secondary group G ₂ , where the load of each small cell in the primary group G ₁ is higher than that of each small cell in the secondary group G ₂ According to the load of the small cells, according to the load of the master-slave group, the ABS ratio is assigned to the master group _G1 as θ, and the ABS sub-ratio is assigned to the slave group G2 as ₁ -θ.

Further, the calculation of the interference weight and the establishment criteria and method of the interference graph in step 1 specifically include:

1) Count the number of users within the center of the small cell as N, the number of edge users as M, and M+N≠0;

2) The average channel quality RASQ _ij of the small cell area served by cell i and interfered by cell j is expressed as:

in Indicates the SINR of the Kth user at the edge of the cell, Indicates the SINR of the Kth user in the center of the cell;

3) where the interference weight If W _ij ≥ W _th , there is an interference edge between the two small cells, where W _th is a preset value, and MAX(RASQ _ij , RASQ _ji ) represents the maximum value among RASQ _ij and RASQ _ji ;

4) The interference graph G(V, E) can be determined by sequentially calculating the interference weights between every two small cells, where V represents a set of small cells, and E is a set of interference weights W _ij ;

5) Periodically update the interference map.

Further, from step 1, the interference weight W _ij between every two small cells in the interference graph G(V,E) can be known, and the system spectrum resource is divided into K sub-sections, R={R ₁ ,R ₂ ,..., R _k }, assign the point set V to K clusters, so that the weight between clusters is the largest, that is:

Step two specifically includes:

1) Initialization: Ω _k represents the degree of node k, W _i is the weight of cluster R _i , for cluster R _i there is when the algorithm starts There is W _i = 0, and V' represents the remaining small cells, which are sorted in descending order according to the degree of the small cell point set, and the first to K nodes are selected from the set of V' in order of Ω _k from large to small and assigned to In K clusters;

2) Continue to assign nodes to K clusters according to the order of nodes Ω _k from large to small. If node m is assigned to cluster R _i , calculate the intra-cluster interference weight

3) Select the cluster R _j with the smallest interference weight in the cluster, and assign node m to the cluster R _j , at this time V'=V'-m;

4) Repeat the above steps 2), 3) until the remaining small cell set V' is empty;

5) Re-cluster when the interference graph changes.

Further, the dynamic ABS subframe configuration scheme of small cells in the cluster based on load and user average service rate in step 3 specifically includes:

1) The entire system adopts a proportional fair scheduling algorithm, and the small cells in the cluster are divided into two groups G ₁ and G ₂ , and the load of any small cell in group G ₁ is greater than the load of any small cell in group G ₂ ;

2) The two groups use a cross subframe allocation scheme. When the main group G ₁ uses normal subframes, the slave group G ₂ uses the corresponding ABS subframes until the number of ABS subframes of the two groups meets its preset value;

3) Assuming that the long-term average service rate of user j in small cell C _i is R _ij , and the allocation ratio of ABS subframes in G ₁ is θ, then the allocation ratio of ABS subframes in G ₂ is 1-θ, 0.4≤θ≤0.6 ;

4) Calculation method for full business model θ:

_The load in group _G1 is determined by the number of users in group _{G1, Num G1 ,} and the load in group G2 is determined by the number of users in group G2, Num _{G2 ;}

The maximum throughput metric is:

The optimal ABS subframe allocation can be seen from this:

5) Calculation method for non-full business model θ:

Partial business load is determined by the total amount of data to be transmitted by the base station;

Assuming that the amount of data that base station C _i prepares to transmit for user j is B _ij , the time for user j in the main group G ₁ to transmit these data Time required to transmit data from user j in group _G2

The total transit time is T _Σ :

make Make have Therefore, T _Σ has a minimum value, and the optimal ratio of θ ^opt is:

6) Periodically update the ABS ratio θ.

The beneficial effect of the present invention is that: the method provided by the present invention uses the small cell interference map establishment and the clustering scheme to rationally allocate the frequency spectrum, and suppresses the interference of small cells between clusters; through the load-based dynamic ABS allocation method, the cluster Interference of small cells within the cluster, and balance the load within the cluster to improve spectrum utilization.

Description of drawings

In order to make the purpose, technical scheme and beneficial effect of the present invention clearer, the present invention provides the following drawings for illustration:

FIG. 1 is a schematic diagram of a system architecture in which control channels and data channels are separated;

Fig. 2 is the overall schematic diagram of the method of the present invention;

Fig. 3 is a schematic diagram of establishing an interference graph;

Figure 4 is a schematic diagram of the clustering process;

FIG. 5 is a schematic diagram of spectrum allocation after clustering;

Figure 6 is a schematic diagram of a dynamic ABS allocation process;

Fig. 7 is a schematic diagram of ABS configuration in a cluster.

detailed description

The preferred embodiments of the present invention will be described in detail below with reference to the accompanying drawings.

The present invention is suitable for dense small cell network, in the dense cell network, there is not only interference between the small cells, but also strong interference between the macro base station and the small cells.

As shown in Figure 1, in order to avoid the interference of the macro base station, the system architecture shown in Figure 1 is adopted. In this architecture, the macro cell is dedicated to the broadcast of control information, while the small cell is responsible for the transmission of data information. First, the UE accesses a macro cell, and then with the assistance of the macro cell, the UE accesses the small cell with the highest average receiving power.

Fig. 2 is the overall flowchart of this method, can find out from the figure that this method is mainly divided into three main parts:

Step 1: Calculate the average regional channel quality RASQ _ij between small cells, obtain the interference weight W _ij between small cells, and establish the interference graph G(V,E);

Step 2: cluster the small cells according to the interference graph G(V,E), and use different spectrum resources for the small cells in different clusters;

Step 3: Based on the load and user average transmission rate information, the dynamic ABS ratio allocation is performed on the small cells in the cluster.

Considering user mobility and load changes, the above steps need to be repeated for periodic T. In a short period of time, users can be considered stationary and user load changes can be considered to be approximately constant, so T is set to a short period of time. And considering that the short time will increase the system overhead and the calculation is too frequent, T is set to 1s, which can be adjusted according to the actual situation, such as the user's moving speed.

Step 1 is detailed as follows:

Step 1.1: First count the number of users in the center of the small cell as N, and the number of edge users as M: if the user measures RSRP > RSRP _th , it is counted as a central user; if the user measures RSRP < RSRP _th , it is counted as an edge user; RSRP _th is the default value.

Step 1.2: The regional average channel quality RASQ _ij of small cell i interfered by small cell j is expressed as:

in represents the SINR of the edge user, Indicates the user SINR of the cell center;

Step 1.3: Calculate the interference weight between two adjacent small cells:

Step 1.4: If W _ij ≥ W _th , the interference edge exists, establish an interference graph G(V,E) with the small cell as the vertex V, and the interference weight W _ij as the edge, and update the interference graph at a period T.

Figure 3 is a schematic diagram of the establishment of interference graph

As shown in Figure 3, M _i represents the number of users at the edge of small cell i, and N _i represents the number of users at the center of small cell i, assuming that the distances between small cells 1, 2, and 3 are equidistant, that is, d ₁₂ =d ₁₃ =d ₂₃ , and the transmission power of the three small cells P ₁ =P ₂ =P ₃ ;

The relationship between the total number of three small cells is: M ₁ +N ₁ =M ₂ +N ₂ =M ₃ +N ₃ ;

The relationship between the number of edge users is: M ₁ <M ₂ <M ₃ ;

Since the edge users in the small cell are greatly interfered, when there are many small cell edges, the interfered users of the small cell are concentrated in the edge area; assuming that the SINR of all the edge users is the same, and the SINR of the central user is the same, according to the above process It is not difficult to introduce:

For small cells 1 and 2: RASQ ₁₂ > RASQ ₂₁ , then

For small cells No. 2 and No. 3: RASQ ₂₃ > RASQ ₃₂ , then

For small cells 1 and 3: RASQ ₁₂ > RASQ ₃₁ , then

And W ₁₃ =W ₂₃ >W ₁₂ ;

It can be seen that when many users are distributed at the edge of the cell, we will select the cell with more users at the edge of the cell as an important basis for judging the interference weight. Frequently received strong interference, resulting in poor performance of edge users.

Step 2 is detailed as follows:

Figure 4 is the detailed process of clustering the interference graph, as shown in the figure:

Step 2.1: Initialization: Ω _k represents the degree of node k, W _i is the weight of cluster R _i , for cluster R _i has when the algorithm starts have W _i =0;

Step 2.2: V' represents the remaining small cells, sorted in descending order according to the degree of the small cell point set, and i=0;

Step 2.3: Select the element m with the largest Ω _k in the set V' for distribution;

Step 2.4: Determine whether there is an empty set in the K clusters, if yes, continue to step 2.5, otherwise skip to step 2.6;

Step 2.5: assign element m to R _i , i=i+1, return to step 2.4;

Step 2.6: Calculate the interference weight of element m with each cluster Select the cluster with the smallest interference weight, and assign element m to this cluster;

Step 2.7: Repeat the above steps 2.3 to 2.5 until the remaining small cell set V' is empty.

Figure 5 is a schematic diagram of spectrum allocation after clustering:

After the clustering is over, the inter-cluster interference is the largest, and the intra-cluster interference weight is small, so we use different frequency bands 1, 2, 3, and 4 for different clusters R ₁ , R ₂ , R ₃ , and R ₄ ; The larger interference between clusters can be eliminated, but there is still interference within the cluster at this time, and we will suppress the interference within the cluster in the subsequent steps.

Step 3 is detailed as follows:

Referring to Figure 6, the intra-cluster dynamic ABS allocation method:

Step 3.1: Statistically measure the SINR of all users in the cluster, the user acceptance power is P, the total interference power is I, and the noise power is N ₀ , then the SINR can be expressed as:

Step 3.2: Considering that the ABS subframe θ adjustment period is short, its instantaneous service rate is approximately equal to the average service rate within the period, so the calculation method of the average rate can be simplified as follows:

Carry out CQI mapping on the SINR measured by the user, and obtain its modulation and coding method. According to the modulation and coding method, calculate the size of its transmission block, and then obtain its service rate R; or it can be simply calculated by the Shannon formula: R=log(1+ SINR), the average service rate R can be obtained;

Step 3.3: The small cells in the cluster are divided into two groups based on the load, the primary group G ₁ and the secondary group G ₂ , where the load of each small cell in the primary group G ₁ is higher than that of each small cell in the secondary group G ₂ small cell load;

In-cluster load metrics:

For full business models:

Since the business volume that all users in the cluster need to transmit is approximately infinite, we use the number of users as the standard to measure the load, and the load in the _G1 group is determined by the number of users in the _G1 group Determine the load in _G2 group by the number of users in _G2 group Decide;

For non-full business models:

The load in the cluster is determined by the total traffic that needs to be transmitted by small cells, and this load is given by the specific business model;

Step 3.4: According to the load of the master-slave group, assign the ABS subframe ratio to the master group G ₁ as θ, and assign the ABS subframe ratio to the slave group G ₂ as 1-θ, 0.4≤θ≤0.6;

Calculation method for full business model θ:

Under the full service model, the total amount of data to be transmitted by users is unlimited, so the transmission volume per unit time, that is, the throughput C _Σ is an important metric to measure system performance, so the metric C _Σ is established as follows, and C _Σ Take the maximum, and θ is the optimal ABS subframe ratio;

The maximum throughput metric is:

The optimal ABS subframe allocation can be seen from this:

Calculation method for non-full business model θ:

Partial business load is determined by the total amount of data to be transmitted by the base station;

Assuming that the amount of data that base station C _i prepares to transmit for user j is B _ij , the time for user j in the main group G ₁ to transmit these data Time required to transmit data from user j in group _G2 For the non-full business model, the total amount of business is changing dynamically, and the average service rate of users is also changing, but in general, the smaller the total transmission time, the better, so the construction of the total transmission time function T _Σ is established, and T _Σ is the smallest, and the calculated θ is the most ABS subframe ratio;

The total transit time is T _Σ :

make Make have Therefore, T _Σ has a minimum value, and the optimal ratio of θ ^opt is:

Step 3.6: Configure the ABS subframe obtained from the above solution:

Referring to Figure 7, the ABS subframe configuration scheme:

It can be seen from the above that the ABS subframe optimization model obtains the allocation ratio θ of ABS subframes in the cluster. In order to further reduce the interference in the cluster, a cross-subframe allocation scheme is used for the two groups:

In a frame structure, assuming that there are 8 subframes used for normal data transmission, after the above solution, G ₁ group θ=0.4, then G ₂ group θ=0.6, and the number of ABS subframes in G ₁ group can be approximately 3, The number of ABS subframes in group G2 can be approximately 4. From Figure ₇ , we can see that when the main group _G1 uses normal subframes, the slave group G2 uses the corresponding ABS subframes until the _two groups are satisfied. The numbers of ABS subframes satisfy their preset values.

Finally, it should be noted that the above preferred embodiments are only used to illustrate the technical solutions of the present invention and not to limit them. Although the present invention has been described in detail through the above preferred embodiments, those skilled in the art should understand that it can be described in terms of form and Various changes may be made in the details without departing from the scope of the invention defined by the claims.

Claims (4)

1. a dynamic ABS interference suppression method based on clustering in a heterogeneous cellular network, is characterized in that: comprise the following steps: Step 1: Calculate the interference weight between dense small cells, establish the interference relationship between the small cell networks, that is, whether there is an interference edge between the small cells, so as to establish an interference graph about the small cell interference relationship; Step 2: Based on the established small cell interference graph, allocate small cells to K clusters, ensure that the interference weight between clusters is the largest during the allocation process, and finally allocate different frequency band resources for different clusters; Step 3: The small cells in the cluster are divided into two groups based on the load, the primary group G ₁ and the secondary group G ₂ , where the load of each small cell in the primary group G ₁ is higher than that of each small cell in the secondary group G ₂ According to the load of the small cells, according to the load of the master-slave group, the ABS ratio is assigned to the master group _G1 as θ, and the ABS sub-ratio is assigned to the slave group G2 as ₁ -θ. 2. the dynamic ABS interference suppression method based on clustering in a kind of heterogeneous cellular network according to claim 1, it is characterized in that: the calculation of interference weight in step 1 and the establishment criterion and method of interference graph specifically comprise: 1) Count the number of users within the center of the small cell as N, the number of edge users as M, and M+N≠0; 2) The average channel quality RASQ _ij of the small cell area served by cell i and interfered by cell j is expressed as: <mrow><msub><mi>RASQ</mi><mrow><mi>i</mi><mi>j</mi></mrow></msub><mo>=</mo><mfrac><mrow><mi>M</mi><mo>*</mo><munderover><mo>&amp;Sigma;</mo><mrow><mi>k</mi><mo>=</mo><mn>1</mn></mrow><mi>M</mi></munderover><mrow><msubsup><mi>SINR</mi><mi>k</mi><mi>O</mi></msubsup></mrow><mo>+</mo><mi>N</mi><mo>*</mo><munderover><mo>&amp;Sigma;</mo><mrow><mi>k</mi><mo>=</mo><mn>1</mn></mrow><mi>N</mi></munderover><mrow><msubsup><mi>SINR</mi><mi>k</mi><mi>I</mi></msubsup></mrow></mrow><msup><mrow><mo>(</mo><mi>M</mi><mo>+</mo><mi>N</mi><mo>)</mo></mrow><mn>2</mn></msup></mfrac><mo>-</mo><mo>-</mo><mo>-</mo><mrow><mo>(</mo><mn>1</mn><mo>)</mo></mrow></mrow> in Indicates the SINR of the Kth user at the edge of the cell, Indicates the SINR of the Kth user in the center of the cell; 3) where the interference weight The meaning is that the interference weight between two small cells is inversely proportional to the maximum value of the channel average quality between them, that is, the better the channel quality between two small cells, the smaller the interference weight between them; if W _ij ≥ W _th , there is an interference edge between the two small cells, where W _th is a preset value, and MAX(RASQ _ij , RASQ _ji ) represents the maximum value among RASQ _ij and RASQ _ji ; 4) The interference graph G(V, E) can be determined by sequentially calculating the interference weights between every two small cells, where V represents a set of small cells, and E is a set of interference weights W _ij ; 5) Periodically update the interference map. 3. the dynamic ABS interference suppression method based on clustering in a kind of heterogeneous cellular network according to claim 2, it is characterized in that: by step 1, every two small cells in the known interference graph G (V, E) Interference weight W _ij , divide the system spectrum resource into K sub-segments, R={R ₁ , R ₂ ,..., R _k }, assign the point set V to K clusters, so that the inter-cluster weight is the largest, that is for: <mrow><mi>m</mi><mi>a</mi><mi>x</mi><munderover><mo>&amp;Sigma;</mo><mrow><mi>i</mi><mo>=</mo><mn>1</mn></mrow><mrow><mi>K</mi><mo>-</mo><mn>1</mn></mrow></munderover><munderover><mo>&amp;Sigma;</mo><mrow><mi>j</mi><mo>=</mo><mi>i</mi><mo>+</mo><mn>1</mn></mrow><mi>K</mi></munderover><munder><mo>&amp;Sigma;</mo><mrow><msub><mi>v</mi><mn>1</mn></msub><mo>&amp;Element;</mo><msub><mi>R</mi><mi>i</mi></msub><mo>,</mo><msub><mi>v</mi><mn>2</mn></msub><mo>&amp;Element;</mo><msub><mi>R</mi><mi>j</mi></msub></mrow></munder><mi>w</mi><mrow><mo>(</mo><msub><mi>v</mi><mn>1</mn></msub><mo>,</mo><msub><mi>v</mi><mn>2</mn></msub><mo>)</mo></mrow><mo>-</mo><mo>-</mo><mo>-</mo><mrow><mo>(</mo><mn>2</mn><mo>)</mo></mrow></mrow> Where v ₁ and v ₂ are elements in the small cell set V; Step two specifically includes: 1) Initialization: Ω _k represents the degree of node k, W _i is the weight of cluster R _i , for cluster R _i there is At the beginning of the algorithm for There is W _{i =} 0; use the set V' to represent the small base stations that have not been allocated to the cluster, sort them in descending order according to the degree of the small cell point set, and select the first one from the set of V' to The Kth node is assigned to K clusters in turn; w _{u, v} is the interference weight between two nodes u and v in the cluster R _i ; 2) Continue to assign nodes to K clusters according to the order of nodes Ω _k from large to small. If node m is assigned to cluster R _i , calculate the change value of the interference weight in the cluster w _{m, u} is the interference weight between the new node m and the existing node u in the cluster; 3) Select the cluster R _j with the smallest interference weight in the cluster, and assign node m to the cluster R _j , at this time V'=V'-m; 4) Repeat the above steps 2), 3) until the remaining small cell set V' is empty; 5) Re-cluster when the interference graph changes. 4. the dynamic ABS interference suppression method based on clustering in a kind of heterogeneous cellular network according to claim 1, it is characterized in that: the small cell in the cluster in the step 3 is based on the dynamic ABS subframe of load and user average service rate The configuration scheme specifically includes: 1) The entire system adopts a proportional fair scheduling algorithm, and the small cells in the cluster are divided into two groups G ₁ and G ₂ , and the load of any small cell in group G ₁ is greater than the load of any small cell in group G ₂ ; 2) The two groups use a cross subframe allocation scheme. When the main group G ₁ uses normal subframes, the slave group G ₂ uses the corresponding ABS subframes until the number of ABS subframes of the two groups meets its preset value; 3) Assuming that the long-term average service rate of user j in small cell C _i is R _ij , and the allocation ratio of ABS subframes in G ₁ is θ, then the allocation ratio of ABS subframes in G ₂ is 1-θ, 0.4≤θ≤0.6 ; 4) Calculation method for full business model θ: The load in G ₁ group is used for the number of users in G ₁ group Determine the load in group _G2 by the number of users in group _G2 Decide; The maximum throughput metric is: <mrow><msub><mi>C</mi><mi>&amp;Sigma;</mi></msub><mo>=</mo><mrow><mo>(</mo><mn>1</mn><mo>-</mo><mi>&amp;theta;</mi><mo>)</mo></mrow><mo>*</mo><munder><mo>&amp;Sigma;</mo><mrow><mi>i</mi><mo>&amp;Element;</mo><msub><mi>G</mi><mn>1</mn></msub></mrow></munder><munder><mo>&amp;Sigma;</mo><mrow><mi>j</mi><mo>&amp;Element;</mo><msub><mi>C</mi><mi>i</mi></msub></mrow></munder><msub><mi>R</mi><mrow><mi>i</mi><mi>j</mi></mrow></msub><mo>+</mo><mi>&amp;theta;</mi><mo>*</mo><munder><mo>&amp;Sigma;</mo><mrow><mi>i</mi><mo>&amp;Element;</mo><msub><mi>G</mi><mn>2</mn></msub></mrow></munder><munder><mo>&amp;Sigma;</mo><mrow><mi>j</mi><mo>&amp;Element;</mo><msub><mi>C</mi><mi>i</mi></msub></mrow></munder><msub><mi>R</mi><mrow><mi>i</mi><mi>j</mi></mrow></msub><mo>-</mo><mo>-</mo><mo>-</mo><mrow><mo>(</mo><mn>3</mn><mo>)</mo></mrow></mrow> The optimal ABS subframe allocation can be seen from this: <mrow><msubsup><mi>&amp;theta;</mi><mi>&amp;Sigma;</mi><mrow><mi>o</mi><mi>p</mi><mi>t</mi></mrow></msubsup><mo>=</mo><mfenced open = "{" close = ""><mtable><mtr><mtd><mrow><msub><mi>&amp;theta;</mi><mi>max</mi></msub><mo>,</mo><munder><mo>&amp;Sigma;</mo><mrow><mi>i</mi><mo>&amp;Element;</mo><msub><mi>G</mi><mn>2</mn></msub></mrow></munder><munder><mo>&amp;Sigma;</mo><mrow><mi>j</mi><mo>&amp;Element;</mo><msub><mi>C</mi><mi>i</mi></msub></mrow></munder><msub><mi>R</mi><mrow><mi>i</mi><mi>j</mi></mrow></msub><mo>&gt;</mo><munder><mo>&amp;Sigma;</mo><mrow><mi>i</mi><mo>&amp;Element;</mo><msub><mi>G</mi><mn>1</mn></msub></mrow></munder><munder><mo>&amp;Sigma;</mo><mrow><mi>j</mi><mo>&amp;Element;</mo><msub><mi>C</mi><mi>i</mi></msub></mrow></munder><msub><mi>R</mi><mrow><mi>i</mi><mi>j</mi></mrow></msub></mrow></mtd></mtr><mtr><mtd><mrow><msub><mi>&amp;theta;</mi><mi>min</mi></msub><mo>,</mo><mi>o</mi><mi>t</mi><mi>h</mi><mi>e</mi><mi>r</mi><mi>w</mi><mi>i</mi>><mi>s</mi><mi>e</mi></mrow></mtd></mtr></mtable></mfenced><mo>-</mo><mo>-</mo><mo>-</mo><mrow><mo>(</mo><mn>4</mn><mo>)</mo></mrow></mrow> 5) Calculation method for non-full business model θ: Partial business load is determined by the total amount of data to be transmitted by the base station; Assuming that the amount of data that base station C _i prepares to transmit for user j is B _ij , the time for user j in the main group G ₁ to transmit these data Time required to transmit data from user j in group _G2 The total transit time is T _Σ : <mrow><msub><mi>T</mi><mi>&amp;Sigma;</mi></msub><mo>=</mo><munder><mo>&amp;Sigma;</mo><mrow><mi>i</mi><mo>&amp;Element;</mo><msub><mi>G</mi><mn>1</mn></msub></mrow></munder><munder><mo>&amp;Sigma;</mo><mrow><mi>j</mi><mo>&amp;Element;</mo><msub><mi>C</mi><mi>i</mi></msub></mrow></munder><mfrac><msub><mi>B</mi><mrow><mi>i</mi><mi>j</mi></mrow></msub><mrow><mo>(</mo><mn>1</mn><mo>-</mo><mi>&amp;theta;</mi><mo>)</mo><msub><mi>R</mi><mrow><mi>i</mi><mi>j</mi></mrow></msub></mrow></mfrac><mo>+</mo><munder><mo>&amp;Sigma;</mo><mrow><mi>i</mi><mo>&amp;Element;</mo><msub><mi>G</mi><mn>2</mn></msub></mrow></munder><munder><mo>&amp;Sigma;</mo><mrow><mi>j</mi><mo>&amp;Element;</mo><msub><mi>C</mi><mi>i</mi></msub></mrow></munder><mfrac><msub><mi>B</mi><mrow><mi>i</mi><mi>j</mi></mrow></msub><mrow><mi>&amp;theta;</mi><mo>*</mo><msub><mi>R</mi><mrow><mi>i</mi><mi>j</mi></mrow></msub></mrow></mfrac><mo>-</mo><mo>-</mo><mo>-</mo><mrow><mo>(</mo><mn>5</mn><mo>)</mo></mrow></mrow> make Make have Therefore, T _Σ has a minimum value, and the optimal ratio of θ ^opt is: <mrow><msup><mi>&amp;theta;</mi><mrow><mi>o</mi><mi>p</mi><mi>t</mi></mrow></msup><mo>=</mo><mfrac><msqrt><msub><mi>T</mi><msub><mi>G</mi><mn>2</mn></msub></msub></msqrt><mrow><msqrt><msub><mi>T</mi><msub><mi>G</mi><mn>2</mn></msub></msub></msub>msub></msqrt><mo>+</mo><msqrt><msub><mi>T</mi><msub><mi>G</mi><mn>1</mn></msub></msub></msqrt></mrow></mfrac><mo>-</mo><mo>-</mo><mo>-</mo><mrow><mo>(</mo><mn>6</mn><mo>)</mo></mrow></mrow> 6) Periodically update the ABS ratio θ.