JP7559947B2 - Route control device, communication system, and route control method - Google Patents

Description

The present invention relates to a route control device, a communication system, and a route control method.

For example, in 5G (fifth generation mobile communication system), a system is known that includes MFH (Mobile Front Haul), MMH (Mobile Mid Haul), and MBH (Mobile Back Haul) and is configured to control traffic.

For example, there is a wireless communication system in which communication paths between multiple DUs (Distribution units: remote stations), each of which is connected to multiple RUs (Radio units: radio stations), and multiple CUs (Central units: aggregation base stations) are switched by a SW (switch).

Conventionally, such systems predict (using machine learning) the amount of traffic arriving at a certain CU, and if it is expected that the CU will receive more traffic than it can handle, the SW is switched so that the traffic is sent to a different CU (see, for example, non-patent document 1).

Igor Donevski et al., “Neural Networks for Cellular Base Station Switching” IEEE INFOCOM 2019.

However, in conventional technology, a bandwidth-guaranteed path between CU-DU is set up according to a single algorithm, so when there are transmission delays due to the network configuration or when the network traffic differs, it may happen that the prediction of future traffic volume for a communication path does not keep up with the actual traffic. When the prediction of future traffic volume does not keep up with the actual traffic, bandwidth cannot be guaranteed.

In addition, in conventional technology, CU switching is achieved by turning off the power of unused CUs, making it difficult to switch only traffic with some DUs to a different CU while continuing to accept traffic with other DUs at an operating CU.

The present invention has been made in consideration of the above-mentioned problems, and aims to provide a route control device, a communication system, and a route control method that can determine and switch efficient communication routes even when traffic and propagation delays fluctuate.

A route control device according to one embodiment of the present invention is a route control device that controls a switch that is provided between a plurality of aggregation base stations and a plurality of remote stations and switches communication paths, and is characterized in having a prediction time determination unit that determines a time available for predicting future traffic volume for a communication path based on a transmission time between the switch and the aggregation base station, a transmission time between the switch and the remote station, control information including traffic volume and schedule information between the remote station and the aggregation base station, an operation time of the switch, and a predetermined maximum allowable delay time; a selection unit that selects an algorithm that can determine a communication path within the time available for prediction determined by the prediction time determination unit from among a plurality of algorithms according to a predetermined priority; an execution unit that determines a communication path between the aggregation base station and the remote station by executing the algorithm selected by the selection unit; and a switching control unit that controls switching of the communication path by the switch so that the aggregation base station and the remote station communicate via the communication path determined by the execution unit.

In addition, a communication system according to one embodiment of the present invention is a communication system including a switch provided between a plurality of aggregation base stations and a plurality of remote stations for switching communication paths, and a path control device for controlling the switch, wherein the path control device has a prediction time determination unit for determining a time available for predicting future traffic volume for a communication path based on a transmission time between the switch and the aggregation base station, a transmission time between the switch and the remote station, control information including traffic volume and schedule information between the remote station and the aggregation base station, an operation time of the switch, and a predetermined maximum allowable delay time, a selection unit for selecting an algorithm capable of determining a communication path within the time available for prediction determined by the prediction time determination unit from among a plurality of algorithms according to a predetermined priority, an execution unit for determining a communication path between the aggregation base station and the remote station by executing the algorithm selected by the selection unit, and a switching control unit for controlling the switching of the communication path by the switch so that the aggregation base station and the remote station communicate via the communication path determined by the execution unit.

In addition, a route control method according to one embodiment of the present invention is a route control method for controlling a switch that is provided between a plurality of aggregation base stations and a plurality of remote stations and switches communication routes, the method comprising the steps of: a prediction time determination step for determining a time available for predicting future traffic volume for a communication route based on a transmission time between the switch and the aggregation base station, a transmission time between the switch and the remote station, control information including traffic volume and schedule information between the remote station and the aggregation base station, an operation time of the switch, and a predetermined maximum allowable delay time; a selection step for selecting an algorithm capable of determining a communication route within the determined time available for prediction from among a plurality of algorithms according to a predetermined priority; an execution step for determining a communication route between the aggregation base station and the remote station by executing the selected algorithm; and a switching control step for controlling switching of the communication route by the switch so that the aggregation base station and the remote station communicate via the determined communication route.

According to the present invention, efficient communication paths can be determined and switched even when traffic and propagation delays fluctuate.

FIG. 1 is a diagram illustrating an example of a configuration of a communication system according to an embodiment. FIG. 2 is a diagram illustrating an example of the configuration around SW. 1 is a block diagram illustrating functions of a route control device according to an embodiment and its surroundings. FIG. 10A and 10B are diagrams illustrating examples of predicted times determined by a predicted time determination unit; FIG. 11 is a diagram showing a specific example of a plurality of algorithms used by a selection unit to select an algorithm. 1A is a diagram showing an example of the operation of a route control device when traffic volume is used, and FIG. 1B is a diagram showing an example of the operation of a route control device when control information is used; 1A is a diagram showing a schematic diagram of a relationship with traffic when there is sufficient prediction time, and FIG. 1B is a diagram showing a schematic diagram of a relationship with traffic when there is not sufficient prediction time; 1A is a diagram showing the traffic volume of each DU predicted by the path control device, and FIG. 1B is a diagram showing the traffic volume of each CU before and after the path control device allocates the traffic volume; 1A is a diagram showing the traffic volume of each DU predicted by the path control device and the actual traffic volume, and FIG. 1B is a diagram showing the traffic volume of each CU before and after the path control device allocates the actual traffic volume. 1A is a diagram showing the traffic volume of each DU predicted by a route control device according to an embodiment, the traffic volume of each DU predicted by a route control device of a comparative example, and the actual traffic volume, and FIG. 1B is a diagram showing the result of the route control device according to an embodiment predicting and allocating the actual traffic volume, and the result of the route control device of the comparative example predicting and allocating the actual traffic volume.

A communication system according to one embodiment will be described with reference to the drawings as follows. FIG. 1 is a diagram showing an example of the configuration of a communication system 1 according to one embodiment. In the communication system 1, for example, a core network 11 to which an end server 10 is connected is connected to multiple CUs (Central units: aggregation base stations) 12, forming a network such as 5G (fifth generation mobile communication system).

The CU 12 is connected to multiple DUs (Distribution units: remote stations) 13 and a route control device 3 via SWs (switches) 2.

Each DU 13 is connected to multiple RUs (Radio units) 14. The RUs 14 accommodate UEs (User equipment) 15 and enable the UEs 15 to access the end server 10.

SW2 is provided between multiple CUs 12 and multiple DUs 13, and constitutes an MMH (Mobile Mid Haul) that switches the communication path between CUs 12 and DUs 13, for example, according to the control of the route control device 3. Hereinafter, when multiple CUs 12 are to be distinguished from one another, they will be distinguished by adding a number after the #, such as CU#1, CU#2, etc. Also, when multiple DUs 13 are to be distinguished from one another, they will be distinguished by adding a number after the #, such as DU#1 to DU#4, etc.

Figure 2 is a diagram showing an example of the configuration around SW2. SW2 receives information indicating the transmission time (t) and traffic volume (x) for communication from, for example, CU#1 and CU#2. SW2 also receives the transmission time (t) and control information (c) for communication from each of DU#1 to DU#4. In the case of upstream communication, SW2 also receives information indicating the traffic volume (x) from each of DU#1 to DU#4.

The control information is information that UE 15 transmits to DU 13 in order to access, for example, end server 10. The control information that each of DU#1 to DU#4 transmits to SW2 includes, for example, traffic volume and schedule information between DU 13 and CU 12. Specifically, the control information describes the time (t _C ) at which the next traffic will actually be transmitted. Then, the data that has been allocated based on the control information will be transmitted in the next transmission.

Next, the route control device 3 according to one embodiment will be described in detail. Figure 3 is a block diagram illustrating the functions of the route control device 3 according to one embodiment and its surroundings. The route control device 3 has a time analysis unit 30, a predicted time determination unit 31, a memory unit 32, a selection unit 33, an execution unit 34, and a switching control unit 35, and is connected to SW2.

The time analysis unit 30 analyzes the traffic transmission time from the control information (C _DU ) received from the DU 13 , and outputs the traffic transmission time (t _C ) and the control information to the predicted time determination unit 31 and the execution unit 34 .

The time analyzer 30 may receive a time stamp together with the control information, and analyze the time (t _c ) at which traffic will actually be transmitted next time based on a TTI (Transmission Time Interval) period.

The predicted time determination unit 31 determines a time (predicted time: _{tf )} that can be used to predict future traffic volume for a communication path based on the transmission time ( _tCU ) between SW2 and CU12, the transmission time (tDU) between SW2 and DU13, control information including traffic volume and schedule information (tC, etc.) between DU13 and CU12, the operation time of SW2 ( _tSW ), and a predetermined maximum allowable delay time (for example, MMH required delay time: _{tMMH )} , and outputs the determined time to the selection unit 33.

The prediction time determination unit 31 may also include a maximum allowable delay time determination unit 310 that determines the maximum allowable delay time of the MMH as the maximum allowable delay time used to determine the time available for prediction according to, for example, the traffic priority, or determines the maximum allowable delay time used to determine the time available for prediction to be 0. In this case, the prediction time determination unit 31 determines the total time of the time from the reception of the control information to the arrival of the main signal and the maximum allowable delay time determined by the maximum allowable delay time determination unit 310 as the time available for prediction.

For example, the predicted time determination unit 31 grasps the current time (t _now ) from a grandmaster clock or an original clock. The predicted time determination unit 31 may store the time required to calculate the allocation to each CU 12 from the estimated traffic information as being constant. The predicted time determination unit 31 may store the time required by SW2 from the instruction to switch to the actual switching as being constant. The predicted time determination unit 31 may store the required delay time of the MMH as being constant.

4 is a diagram showing an example of a predicted time determined by the predicted time determination unit 31. For example, the transmission time (t _DU ) between SW2 and DU 13 may be as shown in the following formula (1).

t _DU = t _now +1-t _C ...(1)

The memory unit 32 stores the transmitted traffic volume and outputs it to the execution unit 34 as a continuous value together with previously stored past traffic volumes.

The selection unit 33 selects an algorithm capable of determining a communication path within the time available for prediction (prediction time) determined by the prediction time determination unit 31 from among multiple algorithms according to a predetermined priority, and outputs the selection result to the execution unit 34.

For example, the selection unit 33 selects traffic estimation when there is sufficient prediction time based on the prediction time determined by the prediction time determination unit 31, and selects an algorithm such as time series prediction when there is not enough prediction time. The selection unit 33 may be configured to select multiple algorithms.

Figure 5 is a diagram showing a specific example of multiple algorithms used by the selection unit 33 to select an algorithm. For example, the selection unit 33 selects at least one of multiple algorithms numbered 1 to 3, etc. according to priority as an algorithm capable of determining a communication path within a predicted time.

The higher the priority (1>2>3) of the algorithm, the more accurate the prediction will be, but the longer it will take to calculate.

For example, when the time available for prediction is longer than 0.7 ms, the selection unit 33 selects a high priority RB (resource block) allocation and traffic estimation (calculation) by multiplication using an MCS (Modulation Coding Scheme).

At this time, the selection unit 33 acquires the number of resource allocations (RIV: resource indication value) and the MCS from the control information (DCI: Downlink Control Information). For example, the selection unit 33 estimates the line quality from the MCS, obtains the traffic volume of the RB, and multiplies it by the RIV to calculate the traffic volume of each DU13.

In addition, when the time available for prediction is 0.3 to 0.7 ms, the selection unit 33 selects linear prediction, which has the second highest priority, because the top priority algorithm is equal to or longer than the prediction time. In linear prediction, increasing the amount of traffic increases the accuracy, but the time required for prediction increases. Therefore, the selection unit 33 adjusts the amount of traffic used depending on the time required for prediction.

In addition, when the time available for prediction is shorter than 0.3 ms, the selection unit 33 selects time series prediction by machine learning, which has the third priority. As with linear prediction, in time series prediction, increasing the amount of traffic increases the accuracy, but the time required for prediction increases. Therefore, the selection unit 33 adjusts the amount of traffic to be used depending on the time required for prediction.

The execution unit 34 determines a communication path between CU12 and DU13 by executing the algorithm selected by the selection unit 33.

For example, the execution unit 34 performs traffic estimation when there is sufficient prediction time, and performs time series prediction when there is not enough prediction time. Traffic estimation is an algorithm that predicts each traffic volume from control information. Time series prediction is an algorithm that calculates the predicted traffic volume for the next time using past traffic information. In time series prediction, the traffic information used may be changed depending on the time available for prediction.

In addition, when the selection unit 33 selects multiple algorithms, the execution unit 34 may be configured to execute the multiple algorithms and merge the multiple prediction results to output one prediction result. For example, the execution unit 34 weights the accuracy or time of the multiple prediction results, or merges the multiple prediction results by majority vote, etc.

For example, the execution unit 34 includes an aggregation unit 340 that aggregates multiple algorithms, and when the selection unit 33 selects multiple algorithms, the aggregation unit 340 executes the multiple algorithms and merges the multiple prediction results to output one prediction result. At this time, the aggregation unit 340 weights the accuracy and time of the multiple prediction results, or merges the multiple prediction results by majority vote, etc.

The switching control unit 35 has a judgment unit 350 and an allocation determination unit 352, and controls the switching of the communication path by SW2 so that CU12 and DU13 communicate via the communication path determined by the execution unit 34.

For example, the determination unit 350 determines whether or not switching of SW2 is necessary based on the predicted traffic volume calculated by the execution unit 34, and if switching is not necessary, does not switch the communication path, and if switching is necessary, outputs the predicted traffic volume to the allocation determination unit 352.

As a specific example, the judgment unit 350 determines whether or not switching of SW2 is necessary by determining whether or not the predicted traffic volume exceeds a threshold value (e.g., bandwidth x 0.8).

In addition, the judgment unit 350 judges that switching of SW2 is not necessary if the traffic volume does not change in each communication path, as shown in the following equation (2).

|x _CU#1 -x _CU#2 |<Max(x _CU )×0.1 ...(2)

In other cases, the determination unit 350 determines that switching of the communication path is necessary.

The allocation determination unit 352 determines the allocation of SW2 based on the predicted traffic volume input from the judgment unit 350, and sends an instruction to SW2 to switch the communication path.

Then, SW2 switches the communication path in accordance with the instructions sent by the path control device 3.

Next, each operation example of the route control device 3 will be described. Figure 6 is a diagram that shows a schematic diagram of an operation example according to information used by the route control device 3. Figure 6(a) is a diagram that shows a schematic diagram of an operation example of the route control device 3 when traffic volume (traffic information) is used. Figure 6(b) is a diagram that shows a schematic diagram of an operation example of the route control device 3 when control information is used.

As shown in Figure 6 (a), when the route control device 3 attempts to switch the communication route using actual traffic information, the route control device 3 can only use the actual traffic information to predict traffic for a short period of time between when the data is received and when the data is transmitted, and therefore the prediction cannot keep up with the actual traffic.

Therefore, in order to switch communication paths using actual traffic information, the route control device 3 performs traffic prediction using the previous data and switches communication paths. In this case, since the route control device 3 uses the previous data, the accuracy of the traffic volume prediction may decrease.

As shown in Figure 6 (b), by using the received control information, the route control device 3 can ensure that there is time available for prediction and switching of the communication route between receiving the control information and transmitting the data.

In this case, even if the previous data is not used, the prediction time increases by the time from when the control information is received to when the data is received. Also, by using the control information to make a prediction, prediction accuracy is guaranteed more than if the previous data were used.

If the time from receiving the control information to receiving the data is longer than the time required from predicting the traffic volume to switching the communication path, it is possible to switch the communication path when the data is received (real-time path switching). Therefore, the route control device 3 analyzes the time required for prediction, and if the prediction time is sufficient, it can perform real-time path switching with high accuracy based on the control information.

Figure 7 is a diagram showing a schematic diagram of the relationship between the time available for prediction (prediction time) calculated by the route control device 3 and traffic. Figure 7(a) is a diagram showing a schematic diagram of the relationship with traffic when there is sufficient prediction time. Figure 7(b) is a diagram showing a schematic diagram of the relationship with traffic when there is insufficient prediction time.

Here, the time until data is transmitted calculated by the time analysis unit 30 is denoted as t _C. The predicted time determination unit 31 calculates the time (predicted time: t f ) that can be used to predict the traffic volume using the following formula (3) based on the transmission time t _CU between CU 12 and SW2, the transmission time t _DU between DU 13 and SW2, the time until data transmission t _C , the operation time of SW2 (time until allocation and switching of communication path ₎ t _SW , and the delay requirement t _MMH of MMH.

t _f =t _MMH −(Max(t _CU )+Max(t _DU −t _C )−t _SW )
...(3)

Then, the selection unit 33 selects an algorithm for calculating the traffic volume for switching the communication path based on the predicted time _tf .

For example, as shown in FIG. 7(a), when there is sufficient prediction time (e.g., when there is time to transmit traffic), the selection unit 33 selects an algorithm for estimating (calculating) traffic using control information. The execution unit 34 then executes the traffic estimation algorithm to calculate traffic and outputs it to the switching control unit 35.

On the other hand, as shown in FIG. 7(b), when there is insufficient prediction time (for example, when there is no time to transmit traffic), the selection unit 33 selects an algorithm that uses past traffic data to perform a time series prediction of future traffic volume. At this time, the selection unit 33 may change parameters such as the number of traffic data to be used depending on the time required for prediction.

Then, the execution unit 34 predicts the next traffic volume using a time series prediction algorithm and outputs it to the switching control unit 35.

In other words, when there is sufficient prediction time, the route control device 3 controls the communication route in real time with high accuracy, and when there is insufficient prediction time, it controls the communication route with the same accuracy as switching the communication route using traffic information.

Next, we will explain a specific example of the operation of the route control device 3. First, we will explain the case where the route control device 3 controls real-time path switching.

For example, the time analysis unit 30 receives control information (C _DU ), calculates the traffic transmission time (t _C =4 TTI=4 ms later), and transmits it to the predicted time determination unit 31 .

As shown in the following equations (4) and (5), the prediction time determination unit 31 calculates the time required for traffic transmission (t fRT ) and the time available for prediction (t f ) from the required delay time of MMH (t _MMH =5 ms) using the traffic transmission time (t _C ) at the current time (t _now ), the transmission and reception time between SW2 and CU 12 (t _CU =0.1 ms), the transmission and reception time between SW2 and _DU 13 (t _DU =0.05 ms), the CU allocation decision time (t _path =1 ms), and the communication path switching time (t _SW =0.1 _ms ).

t _fRT = t _C + t _DU - t _now - t _path - t _SW = 2.85 ms (4)
t _f =t _MMH +t _C -t _now -t _CU -t _path -t _SW =7.8ms...(5)

If there is a discrepancy between the time of DU13 and the time of the route control device 3, the predicted time determination unit 31 may synchronize the times, or may send a timestamp from DU13 to correct the discrepancy between _tDU and the time it actually arrives at SW2.

Then, the selection unit 33 selects an algorithm based on the predicted time ( _tf ). Here, since _tf = 7.8 ms, the selection unit 33 selects an estimate (calculation, 1 ms) of the traffic to be transmitted by the control information. After executing the algorithm, the execution unit 34 transmits each traffic volume to the switching control unit 35. If switching of the communication path by SW2 is not required, the execution unit 34 does not transmit.

For example, if the traffic volume for CU#1 is 8 Gbps and the traffic volume for CU#2 is 2 Gbps, the allocation determination unit 352 changes the allocation of DU13 as a switching of the communication path. In other words, the allocation determination unit 352 transmits a switching instruction of the communication path to SW2 in order to change CU12 to which DU13 is connected.

Next, we will explain an example of control performed by the route control device 3 when there is sufficient prediction time.

The time analysis unit 30 receives the control information (C _DU ), calculates the traffic transmission time (t _C =4 TTI=1 ms later), and transmits it to the predicted time determination unit 31 .

The predicted time determination unit 31 obtains the transmission/reception time between the SW2 and the CU 12 (t _CU =0.1 ms) and the transmission/reception time between the SW2 and the DU 13 (t _DU =0.05 ms).

Then, as shown in the following equations (6) and (7), the prediction time determination unit 31 calculates the time required for traffic transmission (t fRT ) and the time available for prediction (t f ) from the required delay time of MMH (t _MMH =1 ms) using the traffic transmission time (t _C ) at the current time (t _now ), the transmission and reception time between SW2 and CU 12 (t _CU =0.1 ms), the transmission and reception time between _SW2 and DU 13 (t _DU =0.05 ms), the CU allocation determination time (t _path =1 ms), and the communication path switching time (t _SW =0.1 _ms ).

t _fRT = t _C + t _DU - t _now - t _path - t _SW = -0.15ms (6)
t _f =t _MMH +t _C -t _now -t _CU -t _path -t _SW =0.8ms...(7)

If there is a discrepancy between the time of DU13 and the time of the route control device 3, the predicted time determination unit 31 may synchronize the times, or may send a timestamp from DU13 to correct the discrepancy between _tDU and the time it actually arrives at SW2.

Then, the selection unit 33 selects an algorithm based on the predicted time ( _tf ). Here, since _tf = 0.8 ms, the selection unit 33 selects an estimate (calculation, 0.7 ms) of the traffic to be transmitted by the control information. After executing the algorithm, the execution unit 34 transmits each traffic volume to the switching control unit 35. If switching of the communication path by SW2 is not required, the execution unit 34 does not transmit.

For example, if the traffic volume for CU#1 is 9 Gbps and the traffic volume for CU#2 is 8 Gbps, the allocation determination unit 352 changes the allocation of DU13 as a switch of the communication path. In other words, the allocation determination unit 352 transmits a communication path switch instruction to SW2 in order to change CU12 to which DU13 is connected.

Next, we will explain an example of control performed by the route control device 3 when the prediction time is insufficient.

The time analysis unit 30 receives the control information (C _DU ), calculates the traffic transmission time (t _C =4 TTI=0.5 ms later), and transmits it to the predicted time determination unit 31 .

The predicted time determination unit 31 obtains the transmission/reception time between the SW2 and the CU 12 (t _CU =0.1 ms) and the transmission/reception time between the SW2 and the DU 13 (t _DU =0.05 ms).

Then, as shown in the following equations (8) and (9), the prediction time determination unit 31 calculates the time required for traffic transmission (t fRT ) and the time available for prediction (t f ) from the required delay time of MMH (t _MMH =1 ms) using the traffic transmission time (t _C ) at the current time (t _now ), the transmission and reception time between SW2 and CU 12 (t _CU =0.1 ms), the transmission and reception time between _SW2 and DU 13 (t _DU =0.05 ms), the CU allocation decision time (t _path =1 ms), and the communication path switching time (t _SW =0.1 _ms ).

t _fRT = t _C + t _DU - t _now - t _path - t _SW = -0.55 ms (8)
t _f =t _MMH +t _C -t _now -t _CU -t _path -t _SW =0.3ms...(9)

If there is a discrepancy between the time of DU13 and the time of the route control device 3, the predicted time determination unit 31 may synchronize the times, or may send a timestamp from DU13 to correct the discrepancy between _tDU and the time it actually arrives at SW2.

In addition, instead of the memory unit 32, the route control device 3 may be provided with a traffic prediction unit that predicts each traffic volume using the previous traffic volume, and may predict each traffic volume using the previous traffic volume before determining the prediction time.

Then, the selection unit 33 selects an algorithm based on the predicted time ( _tf ). Here, since _tf = 0.3 ms, the selection unit 33 selects an algorithm that predicts future traffic information from past traffic information (determines the amount of data used so as to meet the predicted time). After executing the algorithm, the execution unit 34 transmits each traffic volume to the switching control unit 35. If switching of the communication path by SW2 is not required, the execution unit 34 does not transmit.

For example, if the traffic volume for CU#1 is 8 Gbps and the traffic volume for CU#2 is 2 Gbps, the allocation determination unit 352 changes the allocation of DU13 as a switching of the communication path. In other words, the allocation determination unit 352 transmits a switching instruction of the communication path to SW2 in order to change CU12 to which DU13 is connected.

Next, we will explain an example of control performed by the route control device 3 without using control information.

The predicted time determination unit 31 obtains the transmission/reception time between the SW2 and the CU 12 (t _CU =0.1 ms) and the transmission/reception time between the SW2 and the DU 13 (t _DU =0.05 ms).

Then, as shown in the following equation (10), the predicted time determination unit 31 can use the transmission and reception time between SW2 and CU 12 (t _CU = 0.1 ms), the transmission and reception time between SW2 and DU 13 (t _DU = 0.05 ms), the CU allocation determination time (t _path = 1 ms), and the communication path switching time (t _SW = 0.1 ms) to determine whether or not there is time available for switching (t determination) from the time required for traffic transmission (t _fRT ) and the required delay time of MMH (t _MMH = 1 ms).

t judgment = t _MMH - t _DU - t _CU - t _path - t _SW = -0.25ms...(10)

At this time, the judgment time (tjudge) is a negative value. In other words, even if the communication path is switched due to the traffic volume, it will not be in time.

Therefore, the route control device 3 switches the communication route based on the prediction of the traffic volume of the next section. In this case, the time available for prediction increases by 4 TTI periods (1 ms), so _tf = 0.75 ms. Then, the selection unit 33 selects an algorithm based on the prediction time (tf).

Here, since _tf = 0.75 ms, an algorithm is selected to predict future traffic information from past traffic information (determine the amount of data used so as to meet the prediction time). However, since this is a time-series prediction, the prediction accuracy is lower than when control information is used.

After executing the algorithm, the execution unit 34 transmits each traffic volume to the switching control unit 35. The execution unit 34 does not transmit if switching of the communication path by SW2 is not required.

For example, if the traffic volume for CU #1 is 8 Gbps and the traffic volume for CU #2 is 7.5 Gbps, the allocation determination unit 352 does not allocate DU13 as a switching of the communication path.

Next, a specific example of when the route control device 3 switches the communication route will be described. The route control device 3 controls the switching of the communication route by changing the allocation of the traffic volume of each DU 13 to each CU 12. This allows the route control device 3 to distribute the load on the communication route to each CU 12.

Below, we will explain an example where the number of CU12 is 2 (CU#1, CU#2) and the number of DU13 is 8 (DU#1 to DU#8).

Figure 8 is a schematic diagram showing the traffic volume of each DU13 predicted by the route control device 3 and the results of allocating the traffic volume of each DU13 to each CU12 using the predicted traffic volume. Figure 8(a) is a diagram showing the traffic volume of each DU13 predicted by the route control device 3. Figure 8(b) is a diagram showing the traffic volume of each CU12 before and after the route control device 3 allocates the traffic volume (before and after load balancing). Here, it is assumed that the traffic volume predicted by the route control device 3 matches the actual traffic volume.

If the traffic volume of each DU13 after the path control device 3 executes the above-mentioned algorithm is the traffic volume shown in Figure 8 (a), the communication system 1 will have a traffic volume of 8 Gbps in CU#1 and a traffic volume of 2 Gbps in CU#2 (before load balancing) if SW2 does not switch the communication path.

At this time, the traffic volume accounts for 80% of CU#1's bandwidth of 10 Gbps, and there is sufficient free bandwidth in CU#2 (judgment condition: |8-2|>8 x 0.1), so the route control device 3 switches the communication path using SW2.

For example, the allocation determination unit 352 changes the allocation of the traffic volume of each DU13 to each CU12 so as to distribute the traffic volume of each DU13, and sends an instruction to that effect to SW2.

As a result, as shown after load balancing in Figure 8 (b), even if an error occurs in the traffic volume prediction by the route control device 3, there is a margin of error so that the bandwidth of each CU 12 is not exceeded.

Figure 9 is a schematic diagram showing the results of allocating the traffic volume of each DU13 to each CU12 using the traffic volume of each DU13 predicted by the route control device 3 and the actual traffic volume. Figure 9(a) is a diagram showing the traffic volume of each DU13 predicted by the route control device 3 and the actual traffic volume. Figure 9(b) is a diagram showing the traffic volume of each CU12 before and after the route control device 3 allocates the actual traffic volume (before and after load balancing).

If the traffic volume of each DU13 after the route control device 3 executes the above-mentioned algorithm is the traffic volume shown in the upper part of Figure 9 (a), the communication system 1 will have a traffic volume of 8 Gbps in CU#1 and a traffic volume of 2 Gbps in CU#2 if SW2 does not switch the communication path.

However, when the actual traffic volume is the traffic volume shown in the lower part of Figure 9(a), the traffic volume becomes excessive for the bandwidth of 10 Gbps of CU #1, as shown before load balancing in Figure 9(b).

At this time, there is sufficient available bandwidth for CU#2 (judgment condition: |8-2|>8x0.1), so the route control device 3 switches the communication path using SW2.

For example, the allocation determination unit 352 changes the allocation of the traffic volume of each DU13 to each CU12 so as to distribute the traffic volume of each DU13, and sends an instruction to that effect to SW2.

As a result, as shown after load balancing in Figure 9 (b), even if an error occurs in the traffic volume prediction by the route control device 3, there is a margin of error so that the bandwidth of each CU 12 is not exceeded.

Figure 10 is a schematic diagram showing the result of allocating traffic volume for each DU13 based on the traffic volume for each DU13 predicted by a route control device 3 in one embodiment, and the result of allocating traffic volume for each DU13 based on the traffic volume for each DU13 predicted by a route control device in a comparative example.

Figure 10(a) is a diagram showing the traffic volume of each DU13 predicted by a route control device 3 according to one embodiment, the traffic volume of each DU13 predicted by a route control device of a comparative example, and the actual traffic volume. Figure 10(b) is a diagram showing the results of the route control device 3 according to one embodiment predicting and allocating the actual traffic volume, and the results of the route control device of the comparative example predicting and allocating the actual traffic volume.

As shown in Figure 10, the result of allocating the traffic volume of each DU13 based on the traffic volume of each DU13 predicted by the route control device 3 in one embodiment shows that the communication path was switched appropriately even if the actual traffic volume was greater than the predicted traffic volume.

On the other hand, in the comparative example, when the route control device allocated the traffic volume of each DU13 based on the traffic volume of each DU13 predicted by the route control device, when the actual traffic volume was greater than the predicted traffic volume, the traffic volume became excessive for the 10 Gbps bandwidth of CU#1.

In other words, because the route control device 3 has high traffic volume prediction accuracy, even if an error occurs in the traffic volume prediction, there is enough margin so that the bandwidth of each CU12 is not exceeded.

In this way, in one embodiment of the communication system 1, the route control device 3 selects an algorithm from among multiple algorithms according to a predetermined priority that can determine a communication route within the time available for predicting traffic volume, so that an efficient communication route can be determined and switched even if traffic or propagation delay fluctuates.

In addition, each function of the above-mentioned route control device 3 may be configured in part or in whole by hardware such as a PLD (Programmable Logic Device) or an FPGA (Field Programmable Gate Array), or may be configured as a program executed by a processor such as a CPU.

For example, the route control device 3 in one embodiment can be realized using a computer and a program, and the program can be recorded on a storage medium or provided via a network.

1: Communication system, 2: SW, 3: Route control device, 10: End server, 11: Core network, 12: CU, 13: DU, 14: RU, 15: UE, 30: Time analysis unit, 31: Prediction time determination unit, 32: Memory unit, 33: Selection unit, 34: Execution unit, 35: Switching control unit, 310: Maximum allowable delay time determination unit, 340: Aggregation unit, 350: Determination unit, 352: Allocation determination unit

Claims (6)

A path control device that controls a switch that is provided between a plurality of aggregation base stations and a plurality of remote stations and that switches communication paths,
a prediction time determination unit that determines a time available for predicting a future traffic volume for a communication path based on a transmission time between the switch and the aggregation base station, a transmission time between the switch and the remote station, control information including traffic volume and schedule information between the remote station and the aggregation base station, an operation time of the switch, and a preset maximum allowable delay time;
a selection unit that selects an algorithm capable of determining a communication path within the time available for prediction determined by the prediction time determination unit from among a plurality of algorithms in accordance with a predetermined priority;
an execution unit that executes the algorithm selected by the selection unit to determine a communication path between the aggregation base station and the remote station;
a switching control unit that controls switching of the communication path by the switch so that the aggregation base station and the remote station communicate via the communication path determined by the execution unit. The selection unit is
2. The route control device according to claim 1, characterized in that, based on the time determined by the prediction time determination unit, an algorithm for estimating traffic is selected when there is time to transmit traffic, and an algorithm for performing time series prediction of traffic volume is selected when there is no time to transmit traffic. The predicted time determination unit
3. The route control device according to claim 1 or 2, further comprising a maximum allowable delay time determination unit that determines the maximum allowable delay time of the MMH as the maximum allowable delay time used to determine the time available for prediction in accordance with the priority of the traffic, or determines the maximum allowable delay time used to determine the time available for prediction to be 0, and determines the sum of the time from when the control information is received to when the main signal arrives and the maximum allowable delay time determined by the maximum allowable delay time determination unit as the time available for prediction. The selection unit is
Select multiple algorithms,
The execution unit is
The route control device according to any one of claims 1 to 3, further comprising an aggregation unit that executes a plurality of algorithms selected by the selection unit, weights the accuracy or time of the plurality of prediction results, and outputs a single prediction result obtained by merging the prediction results. A communication system including a switch provided between a plurality of aggregation base stations and a plurality of remote stations for switching communication paths, and a path control device for controlling the switch,
The route control device includes:
a prediction time determination unit that determines a time available for predicting a future traffic volume for a communication path based on a transmission time between the switch and the aggregation base station, a transmission time between the switch and the remote station, control information including traffic volume and schedule information between the remote station and the aggregation base station, an operation time of the switch, and a preset maximum allowable delay time;
a selection unit that selects an algorithm capable of determining a communication path within the time available for prediction determined by the prediction time determination unit from among a plurality of algorithms in accordance with a predetermined priority;
an execution unit that executes the algorithm selected by the selection unit to determine a communication path between the aggregation base station and the remote station;
a switching control unit that controls switching of the communication path by the switch so that the aggregation base station and the remote station communicate with each other via the communication path determined by the execution unit. A route control method for controlling a switch that is provided between a plurality of aggregation base stations and a plurality of remote stations and that switches communication routes, comprising:
a prediction time determination step of determining a time available for predicting a future traffic volume for a communication path based on a transmission time between the switch and the aggregation base station, a transmission time between the switch and the remote station, control information including traffic volume and schedule information between the remote station and the aggregation base station, an operation time of the switch, and a preset maximum allowable delay time;
a selection step of selecting an algorithm capable of determining a communication path within a time available for the determined prediction from among a plurality of algorithms according to a predetermined priority;
determining a communication path between the aggregation base station and the remote station by executing a selected algorithm;
a switching control step of controlling switching of the communication path by the switch so that the aggregation base station and the remote station communicate via the determined communication path.

Images