JP7746967B2 - Wireless communication device and communication method - Google Patents

Description

This disclosure relates to a wireless communication device and a communication method.

The communications system described in Patent Document 1 has multiple gateways, measures line conditions, and offloads packets to other lines when one line becomes congested. Patent Document 1 assumes an environment in which a fixed communications network is used as the main line and a mobile communications network is used as a secondary line. Furthermore, Ping measurement or congestion control is used to determine the status of the communications lines, but both require end-to-end communication, which takes at least one round-trip communication time.

Japanese Patent Application Laid-Open No. 2020-10215

One aspect of the present disclosure aims to provide a wireless communication device that can detect degradation in communication quality early.

One aspect of the present disclosure is a communication system including: a plurality of wireless communication modules, each having a buffer;
a monitoring module that monitors the state of the buffer and predicts a sign of congestion in each wireless communication module;
a control module that performs control so as to perform wireless communication by preferentially using a wireless communication module having a low congestion sign among the wireless communication modules based on the result of the congestion sign prediction;
The wireless communication device includes:

Another aspect of the present disclosure is a communication method in a wireless communication device including a plurality of wireless communication modules each having a buffer, the method comprising:
a monitoring step of monitoring the state of the buffer and predicting a sign of congestion in each wireless communication module;
a control step of controlling the wireless communication modules so that a wireless communication module having a low congestion sign is preferentially used for wireless communication based on the result of the congestion sign prediction;
A communication method comprising:

According to aspects of the present disclosure, degradation in communication quality can be detected earlier, enabling faster congestion control.

FIG. 2 is a diagram illustrating a functional configuration of a mobile terminal according to an embodiment. FIG. 2 is a diagram illustrating a hardware configuration of a mobile terminal according to an embodiment. 10 is a flowchart showing the flow of a packet transmission control process performed by a control module in an embodiment. 1A and 1B are diagrams illustrating details of a monitoring module and a congestion predictor, respectively, in accordance with an embodiment of the present invention; 10A and 10B are operational sequence diagrams showing communication control during redundant transmission in one embodiment ((A) is a case where there is no sign of congestion, and (B) is a case where there is a sign of congestion). 10A and 10B are operational sequence diagrams showing communication control during parallel transmission in one embodiment ((A) is a case where there is no sign of congestion, and (B) is a case where there is a sign of congestion).

One embodiment of the present disclosure is a wireless communication device including multiple wireless communication modules, a monitoring module, and a control module.

A wireless communication device may be a mobile device (mobile body) or a fixed device. Examples of mobile bodies include devices that can move themselves, such as vehicles equipped with wireless communication and mobility functions, and devices that can be moved by being carried by a user or mounted on a mobile body, such as smartphone terminals (portable computers).

For example, the multiple wireless communication modules are each connected to a different network. Each wireless communication module has a buffer for temporarily storing transmission packets, extracts the packets stored in the buffer, performs analog conversion, modulation, and the like, and transmits them as wireless signals. The wireless communication modules may be based on any communication standard, such as cellular communication (5G, LTE, 3G, etc.), wireless LAN, or LPWA (Low Power Wide Area). Furthermore, the communication standards of the multiple wireless communication modules may be the same or different.

The monitoring module monitors the buffer status and predicts signs of congestion for each wireless communication module. The buffer status can be predicted based on, for example, at least one of the buffer capacity, buffer usage, and packet residence time in the buffer. Prediction of signs of congestion may be performed using rule-based processing, machine learning-based processing, or a combination of these. Furthermore, signs of congestion may be predicted in two stages, with signs present or without signs, or in three or more stages, such as no signs, signs present (low), and signs present (high).

For example, the following method can be considered for rule-based congestion prediction. First, the four states of the wireless communication module are considered: "no congestion signs," "congestion in progress," "congestion signs present," and "congestion subsiding." Furthermore, the indicators used for prediction can be values based on at least one of the buffer usage, buffer usage percentage, and packet residence time in the buffer. The buffer usage, buffer usage percentage, and packet residence time on which the indicators are based may be current values obtained from the wireless communication module or future predicted values. The predicted values can be predicted using, for example, a machine learning model. Then, when there are no congestion signs, the monitoring module determines that the system has transitioned to a congestion in progress state if the moving average of the indicator is on an increasing trend and the indicator exceeds a first threshold. Furthermore, the monitoring module determines that, in a congestion-in-progression state, the state has transitioned to a state with signs of congestion when the indicator exceeds a second threshold value that is greater than the first threshold value, and determines that the state has transitioned to a state with signs of congestion decreasing when the moving average of the indicator is on a decreasing trend. Furthermore, the monitoring module determines that, in a congestion-inprogression state, the state has transitioned to a state with signs of congestion decreasing when the moving average of the indicator is on an increasing trend, and determines that the state has transitioned to a state with no signs of congestion when the moving average of the indicator is on a decreasing trend and the indicator is less than the first threshold. Note that "congestion in progress" and "congestion decreasing" may be treated as intermediate states to provide hysteresis, or may be treated as states with a low level of signs of congestion between "signs present" and "no signs present."

Based on the result of congestion prediction, the control module controls wireless communication so that wireless communication modules with the lowest congestion signs are used preferentially among the multiple wireless communication modules. Note that "using preferentially" means using wireless communication modules so that the amount of packets transmitted by wireless communication modules with the lowest congestion signs is equal to or greater than the amount of packets transmitted by wireless communication modules with the highest congestion signs. For example, in the case of redundant transmission in which a single packet is transmitted by multiple wireless communication modules, the packet can be transmitted in duplicate using only wireless communication modules with congestion signs lower than a threshold. Furthermore, in the case of parallel transmission in which multiple packets are distributed to the multiple wireless communication modules for transmission, the packets can be distributed so that the number of packets transmitted by wireless communication modules with the lowest congestion signs is increased.

According to the above embodiment, signs of congestion can be detected simply by monitoring the buffer of the wireless communication module, and end-to-end communication (between the wireless communication device and the communication destination) is not required to detect signs of congestion. Therefore, early detection of signs of congestion is possible, and by switching the communication line to be used based on early signs of congestion, deterioration of communication quality throughout the system can be suppressed.

Embodiments of the present disclosure will be described below with reference to the drawings. The configurations of the following embodiments are examples, and the present disclosure is not limited to the configurations of the embodiments.

(System configuration)
FIG. 1 illustrates the overall configuration of a communication system including a mobile terminal 100 according to one embodiment. The mobile terminal 100 is a device mounted on a mobile object such as a vehicle, and includes a function for executing an application program (hereinafter, "app") 110 and a communication device 200. The communication device 200 includes three wireless communication modules 210a, 210b, and 210c, a monitoring module 220, and a control module 230. The communication methods of the three wireless communication modules may be the same or different. In one embodiment, the wireless communication modules 210a, 210b, and 210c communicate via cellular communication (LTE/5G), Wi-Fi (IEEE 802.11ac/ax), and LPWA (IEEE 802.11ah), respectively. The wireless communication module 210a includes a baseband unit 211a, a buffer 212a, and an RF unit 213a. The wireless communication modules 210b and 210c are similarly configured. In the following description, when the wireless communication module is generally described, the subscript will be omitted and the module will be referred to as the wireless communication module 210. The same applies to the base station.

Figure 2 is a diagram showing the hardware configuration of a mobile terminal 100 according to one embodiment. The mobile terminal 100 includes a processor 11 such as a CPU, a main memory device 12 such as RAM, an auxiliary memory device 13 such as ROM or an SSD (Solid-State Disk), an input/output interface 14, and a communication interface 15, all connected via a bus. The processor 11 loads a computer program stored in the auxiliary memory device 13 into the main memory device 12 and executes it, thereby executing an app 110. A communication device 200 is also connected via the communication interface 15. The communication device 200 includes a microcontroller 21, memory 22, RF circuitry 23, and antenna 24. The microcontroller 21 executes a computer program stored in the memory 22, thereby performing the functions of the baseband unit 211, buffer 212, monitoring module 220, and control module 230, and the RF circuitry 23 and antenna 24 implement the functions of the RF unit 213. Note that each function in this embodiment may be implemented by a combination of a general-purpose processor and a program, or by a dedicated circuit such as an ASIC.

The wireless communication module 210 transmits packets requested for transmission by the control module 230 to the base station 300, and also receives packets transmitted from the base station and passes them to the control module 230. The following mainly describes transmission. Transmission packets passed from the control module 230 are processed by the baseband unit 211a, and the processed packets are stored in the buffer 212a. The packets are temporarily accumulated in the buffer 212a, and when the timing for transmission arrives, the packets are retrieved from the buffer 212a and subjected to analog conversion, modulation, amplification, etc. by the RF unit 213a and transmitted as wireless signals.

The monitoring module 220 monitors the status of the buffers 212a-212c of each wireless communication module 210a-210c and predicts signs of congestion on the lines used by each wireless communication module 210a-210c. For example, the monitoring module 220 references a Buffer Status Report or UL Scheduling Grant for cellular communication, or a priority transmission queue within a device or driver for Wi-Fi, to acquire at least one of the buffer capacity, buffer usage, and packet residence time in the buffer, and calculates an index (metric) based on these values. The buffer usage and packet residence time may be current values or future estimated values. For example, future values can be predicted by inputting the current usage and residence time into a time-series prediction machine learning model such as a recurrent neural network (RNN), a transformer, or a decision tree. Such machine learning models can be generated by using the history (time-series data) of buffer usage or packet residence time as training data to predict future buffer usage or packet residence time from time-series data over a certain period. The prediction may be performed by other methods such as a Kalman filter. The congestion sign estimation process will be described later with reference to FIG.

The control module 230 determines which of the wireless communication modules 210a-c to use to transmit data requested by the application 110. Transmission methods include single transmission, in which data is transmitted using one of the wireless communication modules; redundant transmission, in which the same packet is transmitted using multiple wireless communication modules; and parallel transmission, in which multiple packets are distributed to multiple wireless communication modules and transmitted simultaneously. There are no particular limitations on how the control module 230 determines the transmission method. For example, the application 110 may specify the transmission method, or the control module 230 may determine the method based on the type of data to be transmitted. As an example, data related to driving control requires immediacy and reliability of communication, so it may be transmitted using redundant transmission, while audio data and video data may be transmitted using parallel transmission, since some loss of data is acceptable. The control module 230 also determines which wireless communication module to use for communication based on the predicted results of line congestion predictions. Here, the control module 230 controls wireless communication so that wireless communication is performed with priority given to the wireless communication module with the lowest level of congestion predictions among the multiple wireless communication modules. Determining which wireless communication module to use will be explained later with reference to Figures 5 and 6.

Figure 3 is a flowchart showing the flow of the packet transmission control process performed by the control module 230. The process shown in Figure 3 is executed periodically.

In step S101, the control module 230 obtains a list of available wireless communication modules.

In step S102, the control module 230 initializes the in-use line list and the excluded line list. The in-use line list is a list of lines (wireless communication modules) to be used for communication, and the excluded line list is a list of lines (wireless communication modules) not to be used for communication, both of which are stored in memory 22.

In step S103, the result of line congestion sign estimation is obtained from the monitoring module 220. Here, the congestion sign estimation process by the monitoring module 220 will be described.

Figure 4(A) shows the detailed configuration of the monitoring module 220, and Figure 4(B) is a state transition diagram showing the line state. As shown in Figure 4(A), the monitoring module 220 has a buffer status acquisition unit 221, a buffer status prediction unit 222, and a congestion sign estimation unit 223. The buffer status acquisition unit 221 periodically acquires current values of at least one of the buffer capacity, usage, and packet residence time from the buffers 212a to 212c of the wireless communication modules 210a to 210c. The buffer status prediction unit 222 has a time-series prediction machine learning model and predicts future values from the time-series data of buffer usage and packet residence time up to the present. The congestion sign estimation unit 223 uses the future values predicted by the buffer status prediction unit 222 as an index to estimate whether there are signs of congestion on the line.

The congestion sign prediction unit 223 predicts the line state based on the state transition diagram shown in Figure 4 (B). First, the line transitions from the initial state S0 to the no congestion sign state S1. The no congestion sign state S1 transitions to the congestion in progress state S2 when the moving average of the index is on an increasing trend and the index exceeds a lower threshold (first threshold). The congestion in progress state S2 transitions to the congestion sign state S3 when the index exceeds a higher threshold (second threshold), and transitions to the congestion decreasing state S4 when the moving average of the index is on a decreasing trend. The congestion sign state S3 transitions to the congestion decreasing state S4 when the index falls below the higher threshold. The congestion decreasing state S4 transitions to the congestion in progress state S2 when the moving average of the index is on an increasing trend, and transitions to the no congestion state S1 when the moving average of the index is on a decreasing trend and the index falls below the lower threshold. In this embodiment, the congestion in progress state S2 and congestion decreasing state S4 are employed to provide hysteresis to the state transitions and suppress jittering of the state transitions. Therefore, even if the congestion-free state S1 transitions to the congestion in progress state S2 or the congestion decreasing state S4, the line is determined to have no signs of congestion, and it is not until it transitions to the congestion-predicted state S3 that it is determined to have signs of congestion. Similarly, even if the congestion-predicted state S3 transitions to the congestion in progress state S2 or the congestion decreasing state S4, it is not until it transitions to the congestion-free state S1 that it is determined to have no signs of congestion.

Returning to the explanation of the flowchart in Figure 3, after obtaining the estimated demodulation prediction results for each line from the monitoring module 220 in step S103, the control module executes loop processing L1, including steps S104 to S106, for each line. Note that parallel processing may be used instead of loop processing.

In step S104, the estimated congestion signs for the line in question are determined to be either present or absent. If there are no signs (S104-NO), processing proceeds to step S105, where the control module 230 adds the line to the list of available lines. If there are signs of congestion (S104-YES), processing proceeds to step S106, where the control module 230 adds the line to the list of excluded lines.

When loop processing L1 is completed for all lines, processing proceeds to S107, and control module 230 determines to send the packet using a line (wireless communication module) stored in the list of lines in use.

Figures 5(A) and 5(B) are operational sequence diagrams showing communication control during redundant transmission.

First, a redundant transmission operation when there are no signs of congestion on any of the lines will be described with reference to FIG. 5A. In step S11, application 110 sends a communication request for packet 1 to control module 230. In step S12, control module 230 determines a transmission policy. Here, it is assumed that redundant transmission using all available lines has been selected. By referencing the list of available lines, control module 230 determines that all of wireless communication modules 210a to 210c are available, and therefore determines to perform redundant transmission using all of wireless communication modules 210a to 210c. In steps S13 to S15, control module 230 sends packet 1 to each of wireless communication modules 210a to 210c. In steps S16 to S18, each of wireless communication modules 210a to 210c transmits packet 1 to server 400. In step S19, server 400 receives packet 1 after eliminating duplicates. In this way, the transmission of packet 1 from application 110 to server 400 is completed.

Next, the redundant transmission operation when there are signs of congestion on the line of wireless communication module 210b will be described using Figure 5 (B). In step S21, application 110 sends a communication request for packet 1 to control module 230. In step S22, control module 230 determines a transmission policy. Here, it is assumed that redundant transmission using all available lines has been selected. By referring to the list of available lines, control module 230 determines that wireless communication modules 210a and 210c are available, and therefore determines to perform redundant transmission using wireless communication modules 210a and 210b. In steps S23 and S24, control module 230 sends packet 1 to each of wireless communication modules 210a and 210b. In steps S25 and S26, wireless communication modules 210a and 210b each transmit packet 1 to server 400. In step S19, server 400 receives packet 1 after eliminating duplicates. In this way, transmission of packet 1 from app 110 to server 400 is completed.

Figures 6(A) and 6(B) are operational sequence diagrams showing communication control during parallel transmission.

First, the parallel transmission operation when there are no signs of congestion on any of the lines will be described using Figure 6 (A). In step S31, application 110 sends a communication request for packets 1 to 3 to control module 230. In step S32, control module 230 determines a transmission policy. Here, we will assume that parallel transmission has been selected, in which packets 1 to 3 are distributed to all available lines and transmitted simultaneously. By referring to the list of available lines, control module 230 determines that all of wireless communication modules 210a to 210c are available, and therefore determines to perform parallel transmission using all of wireless communication modules 210a to 210c. In steps S33 to S35, control module 230 sends packets 1 to 3 to wireless communication modules 210a to 210c, respectively. In steps S36 to S38, wireless communication modules 210a to 210c transmit packets 1 to 3 to server 400, respectively. In step S39, the server 400 controls the order in which packets 1 to 3 are received. In this way, transmission of packets 1 to 3 from the app 110 to the server 400 is completed.

Next, a parallel transmission operation when there is a sign of congestion on the line of wireless communication module 210b will be described with reference to FIG. 6B. In step S41, application 110 sends a communication request for packets 1 to 3 to control module 230. In step S42, control module 230 determines a transmission policy. This description assumes that parallel transmission, in which packets 1 to 3 are distributed to all available lines and transmitted simultaneously, has been selected. By referencing the list of available lines, control module 230 determines that wireless communication modules 210a and 210c are available, and therefore determines to perform parallel transmission using wireless communication modules 210a and 210c. The allocation of packets to wireless communication modules can be determined, for example, in a predetermined order, in order from the wireless communication module with the lowest index, or randomly. In step S43, control module 230 sends packets 1 and 2 to wireless communication module 210a. In step S44, the control module sends packet 3 to wireless communication module 210c. In step S45, wireless communication module 210a transmits packets 1 and 2 to server 400. In step S46, wireless communication module 210c transmits packet 3 to server 400. In step S47, server 400 controls the order of packets 1 to 3 and receives them. In this way, transmission of packets 1 to 3 from application 110 to server 400 is completed.

Also, if single transmission is selected, the control module 230 refers to the list of lines in use, selects one of the lines (wireless communication modules) that shows no signs of congestion, and transmits the packet using that line (wireless communication module). Methods for selecting the line to use include, for example, sequential selection, selection of the line with the lowest probability of congestion, and random selection.

(Advantageous effects of this embodiment)
According to this embodiment, the communication device 200 predicts signs of congestion on the line from the state of the buffer 212 of the wireless communication module 210, and is able to detect signs of congestion without communicating with the server 400. Therefore, signs of congestion can be detected earlier than in the past. Furthermore, by switching communication lines based on early signs of congestion, it is possible to reduce the communication load before congestion actually occurs, and to suppress deterioration of communication quality throughout the system.

In addition, by detecting signs of congestion and avoiding the imposition of communication load on lines with degraded quality, it is expected that the recovery time from congestion will be improved.

<Other Modifications>
The above-described embodiment is merely an example, and the present disclosure can be modified and implemented as appropriate within the scope that does not deviate from the gist of the disclosure.

For example, congestion signs may be detected using methods other than those described above. In the above embodiment, a future predicted value of the buffer state (either the buffer capacity, buffer usage, or packet residence time in the buffer) is used as an index, and congestion signs are predicted based on the magnitude of the index value and the increase/decrease trend of the moving average. As a variation, congestion signs may be predicted based on, for example, the magnitude of the predicted value and the increase/decrease trend of the moving average of the values up to now, or based on the magnitude of the current value and the increase/decrease trend of the moving average of the predicted value, or based on the magnitude of the current value and the increase/decrease trend of the moving average of the values up to now. Furthermore, predictions may be made using only either the magnitude of the index or the increase/decrease trend of the moving average, or a different rule (algorithm) from that described above may be used. Furthermore, congestion signs may be detected from the current value of the buffer state using a machine learning model. Such a machine learning model can be generated by machine learning using the history of the index (time series data) and the presence or absence of future congestion as learning data.

In the above embodiment, only two states, "with signs of congestion" and "without signs of congestion," are used, but signs of congestion may be classified into three or more levels. For example, in the above embodiment, two thresholds, a low threshold and a high threshold, are used, but by using an additional threshold, it is possible to classify into three or more levels. Furthermore, when selecting a line (wireless communication module) for transmission, more detailed control can be achieved by determining to use lines with lower levels of signs of congestion first.

The processes and means described in this disclosure can be freely combined and implemented as long as no technical contradiction occurs.

Furthermore, processing described as being performed by one device may be shared and executed by multiple devices. Alternatively, processing described as being performed by different devices may be executed by a single device. In a computer system, the hardware configuration (server configuration) used to realize each function can be flexibly changed.

The present disclosure can also be realized by supplying a computer program that implements the functions described in the above embodiments to a computer, and having one or more processors in the computer read and execute the program. Such a computer program may be provided to the computer via a non-transitory computer-readable storage medium connectable to the computer's system bus, or via a network. Non-transitory computer-readable storage media include, for example, any type of disk, such as a magnetic disk (e.g., a floppy disk, a hard disk drive (HDD), etc.), an optical disk (e.g., a CD-ROM, a DVD disk, a Blu-ray disk), a read-only memory (ROM), a random access memory (RAM), an EPROM, an EEPROM, a magnetic card, a flash memory, an optical card, or any type of medium suitable for storing electronic instructions.

100: Mobile terminal, 110: Application 200: Communication devices 210a, 210b, 210c: Wireless communication modules 211a, 211b, 211c: Baseband units 212a, 212b, 212c: Buffers 213a, 213b, 213c: RF units 300a, 300b, 300c: Base stations 400: Internet 500: Server

Claims (4)

a plurality of wireless communication modules each having a buffer;
a monitoring module that monitors the state of the buffer and predicts a sign of congestion in each wireless communication module;
a control module that performs control so as to perform wireless communication by preferentially using a wireless communication module having a low congestion sign among the plurality of wireless communication modules based on the result of the congestion sign prediction;
Equipped with
The control module
When transmitting one packet by the plurality of wireless communication modules in a redundant transmission, the packet is transmitted using only the wireless communication module having a congestion warning level lower than a threshold value;
When a plurality of packets are distributed to the plurality of wireless communication modules for transmission in parallel, the packets are distributed such that a wireless communication module having a lower likelihood of congestion transmits a larger number of packets.
Wireless communication device. a plurality of wireless communication modules each having a buffer;
a monitoring module that monitors the state of the buffer and predicts a sign of congestion in each wireless communication module;
a control module that performs control so as to perform wireless communication by preferentially using a wireless communication module having a low congestion sign among the plurality of wireless communication modules based on the result of the congestion sign prediction;
Equipped with
The monitoring module
an acquisition unit that acquires a current value of at least one of the buffer usage amount, the buffer usage rate, and the packet residence time in the buffer;
a prediction unit that predicts a future value of at least one of the buffer usage amount, the buffer usage rate, and the packet residence time in the buffer based on the current value;
a prediction unit that predicts a sign of congestion of the wireless communication module using the future value as an index;
Equipped with
Wireless communication device. 1. A communication method in a wireless communication device having a plurality of wireless communication modules each having a buffer, comprising:
a monitoring step of monitoring the state of the buffer and predicting a sign of congestion in each wireless communication module;
a control step of controlling the wireless communication module having a low congestion predictive value among the plurality of wireless communication modules to perform wireless communication preferentially based on the result of the congestion predictive value;
Including ,
In the control step,
When a single packet is transmitted by the plurality of wireless communication modules in redundant transmission, the packet is transmitted using only the wireless communication module having a congestion warning level lower than a threshold value;
In the case of parallel transmission in which a plurality of packets are distributed to the plurality of wireless communication modules and transmitted, the packets are distributed such that a wireless communication module having a lower likelihood of congestion receives a larger number of packets.
Communication method. 1. A communication method in a wireless communication device having a plurality of wireless communication modules each having a buffer, comprising:
a monitoring step of monitoring the state of the buffer and predicting a sign of congestion in each wireless communication module;
a control step of controlling the wireless communication module having a low congestion predictive value among the plurality of wireless communication modules to perform wireless communication preferentially based on the result of the congestion predictive value;
Including,
In the monitoring step,
a current value of at least one of the buffer usage amount, the buffer usage rate, and the packet residence time in the buffer is acquired;
a future value of at least one of the buffer usage amount, the buffer usage rate, and the packet residence time in the buffer is predicted based on the current value;
The future value is used as an indicator to predict a sign of congestion of the wireless communication module.
Communication method.