JP7311457B2 - Control device, communication system, control method - Google Patents

Description

The present invention relates to control technology, and more particularly to a control device, communication system, and control method for controlling relay.

In wireless communication between wireless devices, for example, the transmission rate is determined by selecting a modulation method or coding method according to propagation loss, which is one of the propagation environment factors of the wireless medium. This can be said to be control in the physical layer of the OSI reference model. Also, when an ad-hoc network or a mesh network is formed by a plurality of wireless devices, a relay route including two or more wireless devices is formed, and signals are transferred along the relay route. A shortest route or a random route is selected as the relay route (see Patent Document 1, for example). This can be said to be control in the data link layer or transport layer of the OSI reference model.

Enomoto, M.; , Shibata, N.; , Yasumoto, K.; , Ito, M.; and Higshino, T.; , "A Demand-Oriented Information Retrieval Method on MANET," Proc. of Int'l Workshop on Future Mobile and Ubiquitous Information Technologies (FMUIT '06) (2006)

To improve communication characteristics, the transmission rate is determined according to the propagation environment of the physical layer only. Also, the relay route is selected so as to obtain low delay or high throughput, which are communication characteristics of the data link layer or transport layer. However, these are not controlled so as to obtain high performance for applications actually used by users.

The present invention has been made in view of such circumstances, and its object is to provide a technique for executing control suitable for transmission of applications.

In order to solve the above problems, a control device according to one aspect of the present invention is a control device for controlling the transmission speed and relay path of each of a plurality of wireless devices in a communication system in which relay is performed by a plurality of wireless devices. an acquisition unit that acquires a transmission rate that has already been set and measured communication characteristics for each of a plurality of wireless devices; and QoE (Quality of Experience) estimation from the communication characteristics acquired by the acquisition unit. A derivation unit for deriving a value, and performing multi-objective optimization so that the QoE estimation value derived in the derivation unit and the communication characteristics are high, thereby deriving a relay route candidate including two or more wireless devices. New transmission speed and relay route are determined based on the relay route candidates derived by the optimizer and the optimizer, the QoE estimation value derived by the derivation unit, and the transmission speed and communication characteristics acquired by the acquisition unit. a learning agent to be derived; and an instruction unit for instructing each of the plurality of wireless devices to use the new transmission rate and relay route derived by the learning agent.

Another aspect of the invention is a communication system. This communication system includes a plurality of wireless devices capable of relaying, and a control device that controls the transmission speed and relay path of each of the plurality of wireless devices. The control device includes an acquisition unit that acquires a transmission rate that has already been set and measured communication characteristics for each of the plurality of wireless devices, and a QoE (Quality of Experience) from the communication characteristics acquired by the acquisition unit. A derivation unit for deriving an estimated value, and by executing multi-objective optimization so that the QoE estimation value derived in the derivation unit and the communication characteristics are high, deriving relay route candidates including two or more wireless devices. A new transmission rate and relay path based on the optimization unit, the relay route candidate derived by the optimization unit, the QoE estimated value derived by the derivation unit, and the transmission speed and communication characteristics acquired by the acquisition unit and an instruction unit for instructing each of the plurality of wireless devices to use the new transmission rate and relay route derived by the learning agent.

Yet another aspect of the invention is a control method. This method is a control method for controlling the transmission speed and relay path of each of a plurality of wireless devices in a communication system in which relay is performed by a plurality of wireless devices. a step of acquiring a set transmission rate and a measured communication characteristic; a step of deriving a QoE (Quality of Experience) estimated value from the acquired communication characteristic; deriving a relay route candidate including two or more wireless devices by performing multi-objective optimization as follows; the derived relay route candidate, the derived QoE estimate, and the obtained transmission rate; a step of deriving a new transmission speed and relay route based on the communication characteristics; and a step of instructing each of the plurality of wireless devices to use the derived new transmission speed and relay route.

Any combination of the above constituent elements, and conversion of expressions of the present invention into methods, devices, systems, recording media, computer programs, etc. are also effective as aspects of the present invention.

According to the present invention, control suitable for application transmission can be executed.

1 is a diagram showing the configuration of a communication system according to an embodiment; FIG. FIG. 2 is a diagram showing the configuration of the wireless device in FIG. 1; FIG. 2 is a diagram showing the configuration of a fourth radio device in FIG. 1; FIG. 2 is a sequence diagram showing a processing procedure by the communication system of FIG. 1; FIG. 3 is a flow chart showing a processing procedure by the wireless device of FIG. 2; FIG. 4 is a flow chart showing a processing procedure by the fourth radio device of FIG. 3; FIG.

Before describing this embodiment in detail, an overview will be given first. The present embodiment relates to a communication system including multiple wireless devices. In a communication system, a relay path is formed by two or more wireless devices, and signals are transmitted along the relay path. As mentioned above, appropriate control is required for the transmission of application signals. Here, as an example, it is assumed that the application is video streaming. A QoE (Quality of Experience) index that objectively indicates a subjective quality of experience as a performance of video streaming is ITU-T Recommendation G.3. 1071. However, QoE metrics have not been used as a control for relay routing in ad-hoc or mesh networks.

On the other hand, in recent years, AI (Artificial Intelligence) and machine learning such as deep learning have been actively researched and developed, and are also being tackled in the wireless communication field. Machine learning is applicable to various fields, and is particularly suitable for image recognition/identification. For example, when machine learning is applied to wireless communication, steep changes in throughput are estimated by learning and recognizing buildings or obstructions along the propagation path from images captured by an imaging device.

However, it is not possible to select a transmission speed or a relay route by using evaluation parameters unique to wireless communication such as propagation error, interference, transmission speed, network routing delay, packet loss, and throughput for machine learning. not done. In this way, the propagation environment, transmission speed and relay path for obtaining high performance are evaluated individually for each layer and used for their selection. However, since the transmission speed and the relay route evaluated across multiple layers are not used to select the transmission speed or the relay route, the maximum number of applications from the transmission speed and relay route considering all layers Performance is desired.

An object of the present embodiment is to improve the communication quality of video streaming in wireless communication. Therefore, in a communication system, each wireless device collects information such as packet loss and delay time in a wireless section, and machine learning is performed on the collected information to calculate transmission speeds and relay routes that increase performance. At that time, a QoE index that obtains the user's subjective quality of experience in video streaming using objective parameters is used as an evaluation function in machine learning.

FIG. 1 shows the configuration of a communication system 1000. As shown in FIG. The communication system 1000 includes first radio equipment 100a to seventh radio equipment 100g, which are collectively referred to as radio equipment 100 . The number of wireless devices 100 included in communication system 1000 is not limited to seven.

One of the plurality of wireless devices 100 is the source wireless device 100, another one is the destination wireless device 100, and the rest are relay wireless devices 100. FIG. For example, when the first wireless device 100a is the source wireless device 100 and the seventh wireless device 100g is the destination wireless device 100, the second wireless device 100b to the sixth wireless device 100f are the relay wireless devices 100. becomes. At that time, the packet signal transmitted from the first radio device 100a is relayed by the relay radio device 100 and received by the seventh radio device 100g. The source wireless device 100 and the destination wireless device 100 may be arbitrary wireless devices 100 . Thus, in the communication system 1000, relaying is performed by a plurality of wireless devices 100. FIG. A well-known technique may be used for relaying the packet signal, so the explanation is omitted here.

A transmission rate and a relay route are set for each wireless device 100 . Here, the transmission rate is indicated by, for example, a combination of an encoding method and a modulation method, and the relay path is the relay radio equipment 100 included on the way from the transmission source radio equipment 100 to the destination radio equipment 100. is indicated by a combination of A control device (not shown) controls the transmission speed and relay path of each wireless device 100 . A control device is included in one of the plurality of wireless devices 100 . For example, the control device is included in the fourth radio device 100d. The configuration of the control device is not limited to this, and each of the plurality of wireless devices 100 may include a control device, and only the control device included in any one of the wireless devices 100 may operate. The control device may be configured separately from radio device 100 .

FIG. 2 shows the configuration of the wireless device 100. As shown in FIG. Wireless device 100 includes communication section 10 , processing section 20 , measurement section 30 and setting section 40 . The communication unit 10 performs wireless communication via an antenna. The processing unit 20 is connected to the communication unit 10 and executes various processes in wireless communication. If the wireless device 100 in FIG. 2 is the source wireless device 100 , the processing unit 20 outputs data to be transmitted to the communication unit 10 . After generating a packet signal including data from the processing unit 20, the communication unit 10 performs encoding/modulation on the packet signal, and transmits the resultant packet signal from an antenna.

If the wireless device 100 in FIG. 2 is a relay wireless device 100, the communication unit 10 receives packet signals via an antenna. The communication unit 10 obtains data by demodulating and decoding the packet signal. After generating a packet signal including the acquired data, the communication unit 10 performs encoding, modulation, and the like on the packet signal, and transmits the resulting packet signal from an antenna.

2 is the destination wireless device 100, the communication unit 10 receives the packet signal via the antenna. The communication unit 10 obtains data by demodulating and decoding the packet signal. The communication unit 10 outputs the acquired data to the processing unit 20 . The processing unit 20 executes predetermined processing on data received from the communication unit 10 .

The measurement unit 30 measures propagation error, packet loss, and interference for the packet signal received by the communication unit 10 . These measurements correspond to measurements of communication characteristics in the physical layer. The measurement unit 30 also measures network routing delay and throughput for the packet signal received by the communication unit 10 . These measurements correspond to measurements of communication characteristics in the data link layer or transport layer. Since well-known techniques may be used for these measurements, descriptions thereof are omitted here. The processing unit 20 outputs the communication characteristics measured by the measuring unit 30 to the communication unit 10, and the communication unit 10 sends a packet containing the identification information (hereinafter referred to as "ID") of the wireless device 100 and the communication characteristics. Send a signal. The destination of the packet signal is the fourth wireless device 100d. When this radio device 100 is not adjacent to the fourth radio device 100d, the packet signal is transferred by another radio device 100 to the fourth radio device 100d.

The setting unit 40 receives information on the transmission speed and the relay route from the fourth wireless device 100d via the communication unit 10 . Here, the transmission rate information indicates a modulation method and an encoding method (encoding rate). Also, in the relay route information, the wireless devices 100 along the relay route are indicated in order. The setting unit 40 sets the modulation method and the coding method (coding rate) to the communication unit 10, so that the communication unit 10 uses the modulation method and the coding method (coding rate) in communication. In addition, the setting unit 40 specifies two wireless devices 100 arranged before and after this wireless device 100 in the relay route, and causes the communication unit 10 to receive a packet signal from one of the wireless devices 100. , causes the other wireless device 100 to transmit a packet signal. Furthermore, when the antenna of the communication unit 10 is a directional antenna whose direction can be controlled, the setting unit 40 controls the directivity of the antenna so that the antenna is directed toward the specified two wireless devices 100 .

FIG. 3 shows the configuration of the fourth radio device 100d. In the fourth radio device 100d, the control unit 50 is included in the processing unit 20 of the radio device 100 in FIG. The control unit 50 includes an acquisition unit 200 , a derivation unit 210 , an optimization unit 220 , a learning agent 230 and an instruction unit 240 . The control unit 50 corresponds to the aforementioned control device. When the fourth radio device 100d is the source radio device 100, the relay radio device 100, or the destination radio device 100, the processing is the same as before, so here, Description is omitted. The communication unit 10 receives packet signals from each of the plurality of wireless devices 100 . The packet signal includes the ID and the communication characteristics measured by wireless device 100 that transmitted the packet signal. The communication unit 10 outputs communication characteristics to the processing unit 20 .

The acquisition unit 200 of the processing unit 20 acquires communication characteristics of each of the plurality of wireless devices 100 . In addition, the processing unit 20 holds the transmission speed and relay route already output to each of the plurality of wireless devices 100, and the acquisition unit 200 obtains the transmission speed and relay route of each of the plurality of wireless devices 100. get.

The derivation unit 210 derives a QoE estimated value based on the communication characteristics acquired by the acquisition unit 200, such as the communication characteristics of each of the plurality of wireless devices 100, such as propagation error, packet loss, and interference. The deriving unit 210 also derives the difference between the derived QoE estimated value and the previously derived QoE estimated value.

The optimization unit 220 performs multi-objective optimization so that the derived QoE estimation value and the communication characteristics of each of the plurality of wireless devices 100 are improved. In multi-objective optimization, for example, Pareto optimal solutions are derived. A Pareto optimal solution indicates a relay route candidate including two or more wireless devices 100 . Generally, multiple relay route candidates are derived. That is, the QoE index is introduced as an objective function of an optimization method in an objective quality evaluation method estimated from communication characteristics of the communication system 1000 such as propagation error, packet loss, and interference.

The learning agent 230 uses the difference between the QoE estimated values derived by the derivation unit 210 as a reward, and based on the relay route candidates derived by the optimization unit 220 and the transmission speed and communication characteristics acquired by the acquisition unit 200, New transmission speeds and relay routes are derived by performing machine learning. Here, a new transmission rate is derived for each wireless device 100 . Also, in the relay route, as described above, the wireless devices 100 along the relay route are shown in order.

Instructing unit 240 transmits information on the new transmission speed and relay route to communication unit 10 in order to instruct each of the plurality of wireless devices 100 to use the new transmission speed and relay route derived by learning agent 230. Output. The communication unit 10 transmits to each of the plurality of wireless devices 100 a packet signal containing information on the new transmission speed and relay route.

In terms of hardware, this configuration can be realized by any computer's CPU, memory, and other LSIs, and in terms of software, it is realized by programs loaded in the memory, etc., but here it is realized by linking them. It depicts the function blocks to be used. Therefore, those skilled in the art will understand that these functional blocks can be realized in various forms by using only hardware or a combination of hardware and software.

The operation of the communication system 1000 configured as above will be described. FIG. 4 is a sequence diagram showing a processing procedure by the communication system 1000. As shown in FIG. Here, for clarity of explanation, only the second radio device 100b to the fourth radio device 100d of the communication system 1000 are shown as an example. The second radio apparatus 100b measures communication characteristics (S10), and the third radio apparatus 100c measures communication characteristics (S12). The second radio device 100b transmits communication characteristics (S14), and the third radio device 100c transmits communication characteristics (S16). The fourth radio device 100d derives the transmission rate and the relay route (S18). The fourth radio device 100d transmits the transmission speed and the relay route to the second radio device 100b (S20), and transmits the transmission speed and the relay route to the third radio device 100c (S22). The second radio device 100b sets the transmission speed and the relay route (S24), and the third radio device 100c sets the transmission speed and the relay route (S26).

FIG. 5 is a flow chart showing a processing procedure by the wireless device 100. As shown in FIG. The measurement unit 30 measures communication characteristics (S100). The communication unit 10 transmits communication characteristics (S102). If the communication unit 10 does not receive information on the transmission speed and the relay route (N in S104), it waits. When the communication unit 10 receives information on the transmission speed and the relay route (Y of S104), the setting unit 40 sets the transmission speed and the relay route (S106).

FIG. 6 is a flowchart showing a processing procedure by the fourth radio device 100d. The acquisition unit 200 acquires the transmission speed and communication characteristics (S150). The derivation unit 210 derives the QoE estimated value (S152). The optimization unit 220 derives relay route candidates as Pareto optimal solutions (S154). The learning agent 230 derives the transmission speed and the relay route through learning (S156). The instructing unit 240 instructs the use of the transmission speed and relay route (S158).

According to this embodiment, since the QoE estimated value is used for machine learning to derive the transmission rate and the relay route, it is possible to derive the transmission rate and the relay route that increase the QoE index. In addition, since information of all layers is used for machine learning, it is possible to derive transmission speeds and relay routes that increase the QoE index. In addition, since the transmission speed and the relay route that increase the QoE index are derived, even if the propagation environment is bad or the beat rate is low in applications such as video, the current network maintains a high communication quality. can. In addition, since machine learning continues to be executed, the transmission rate and relay path can be automatically adjusted even when the network configuration or propagation environment fluctuates. Also, since the difference in the QoE estimate is derived, the QoE estimate can be reflected in the reward. Also, since the control device is included in any one of the plurality of wireless devices, the configuration can be facilitated.

The above has been described based on the embodiments of the present invention. It should be understood by those skilled in the art that this embodiment is an example, and that various modifications can be made to the combination of each component, and such modifications are within the scope of the present invention.

10 communication unit 20 processing unit 30 measurement unit 40 setting unit 50 control unit 100 wireless device 200 acquisition unit 210 derivation unit 220 optimization unit 230 learning agent 240 instruction unit 1000 communication system.

Claims (5)

In a communication system in which relay is performed by a plurality of wireless devices, a control device for controlling the transmission speed and relay path of each of the plurality of wireless devices,
an acquisition unit that acquires a transmission rate that has already been set and a communication characteristic that has been measured for each of the plurality of wireless devices;
a derivation unit that derives a QoE (Quality of Experience) estimated value from the communication characteristics acquired by the acquisition unit;
an optimization unit that derives relay route candidates including two or more wireless devices by executing multi-objective optimization so that the QoE estimation value and communication characteristics derived by the derivation unit are high;
Deriving a new transmission speed and relay route based on the relay route candidate derived by the optimization unit, the QoE estimation value derived by the derivation unit, and the transmission speed and communication characteristics acquired by the acquisition unit a learning agent that
an instruction unit that instructs each of the plurality of wireless devices to use the new transmission rate and relay route derived by the learning agent;
A control device comprising: The derivation unit derives a difference between the derived QoE estimated value and the previously derived QoE estimated value, outputs the derived difference to the learning agent,
2. The learning agent uses the relay route candidate derived by the optimization unit, the difference derived by the derivation unit, and the transmission speed and communication characteristics acquired by the acquisition unit. The control device according to . 3. The control device according to claim 1, wherein the control device is included in one of the plurality of wireless devices. a plurality of wireless devices capable of relaying;
a control device for controlling transmission speeds and relay paths of each of the plurality of wireless devices;
The control device is
an acquisition unit that acquires a transmission rate that has already been set and a communication characteristic that has been measured for each of the plurality of wireless devices;
a derivation unit that derives a QoE (Quality of Experience) estimated value from the communication characteristics acquired by the acquisition unit;
an optimization unit that derives relay route candidates including two or more wireless devices by executing multi-objective optimization so that the QoE estimation value and communication characteristics derived by the derivation unit are high;
Deriving a new transmission speed and relay route based on the relay route candidate derived by the optimization unit, the QoE estimation value derived by the derivation unit, and the transmission speed and communication characteristics acquired by the acquisition unit a learning agent that
an instruction unit that instructs each of the plurality of wireless devices to use the new transmission rate and relay route derived by the learning agent;
A communication system comprising: A control method for controlling a transmission rate and a relay path of each of a plurality of wireless devices in a communication system in which relay is performed by a plurality of wireless devices, comprising:
acquiring a transmission rate already set and a measured communication characteristic for each of the plurality of wireless devices;
deriving a QoE (Quality of Experience) estimate from the obtained communication characteristics;
deriving candidate relay paths that include more than one wireless device by performing multi-objective optimization to increase the derived QoE estimates and communication characteristics;
a step of deriving a new transmission speed and a relay route based on the derived relay route candidate, the derived QoE estimated value, and the acquired transmission speed and communication characteristics;
instructing each of the plurality of wireless devices to use the derived new transmission rate and relay path;
A control method comprising:

Images