Wireless communication system, connection destination learning device, connection destination learning program, and connection destination learning method

Abstract

To provide a radio communications system, or the like capable of optimizing a communication path among a plurality of communication terminals.SOLUTION: In a radio communications system comprising a plurality of communication terminals containing a master unit and slave units, mutual connection destinations for transmitting and receiving signals among the plurality of communication terminals are defined. The slave unit measures intensity of a received signal from another communication terminal of the plurality of communication terminals. The master unit comprises a learning part that calculates success rates of communications between itself and the respective slave units, leans so as to enhance the success rates of the communications, and outputs output information on the connection destination upon inputted with input information on a communication state containing the intensity of received signal and the success rate of communication. In addition, the master unit updates the connection destination on the basis of the output information.SELECTED DRAWING: Figure 10

Description

[Technical Field]

[0001] The present invention relates to a wireless communication system in which connection destinations for transmitting and receiving signals between a plurality of communication terminals including a master unit and a slave unit are defined. [Background technology]

[0002] Conventionally, wireless communication systems equipped with a plurality of communication terminals such as sensors and repeaters have been known. The wireless communication system includes a plurality of communication terminals connected to each other by wireless communication. The connection destinations between these communication terminals are defined. In such wireless communication systems, the connection destinations between the communication terminals are determined based on, for example, radio wave strength (received signal strength) (see, for example, Patent Document 1). [Prior art documents] [Patent documents]

[0003] [Patent Document 1] Japanese Patent Application Laid-Open No. 2009-218913 Summary of the Invention [Problem to be solved by the invention]

[0004] In a configuration in which the connection destination is determined based on the signal strength as described above, connections may be concentrated on repeaters with high signal strength, reducing the success rate of communication and potentially making the network unstable.

[0005] An object of the present invention is to provide a wireless communication system and the like that can optimize communication paths between a plurality of communication terminals. [Means for solving the problem]

[0006] A first aspect of the present invention provides a wireless communication system comprising a plurality of communication terminals including a parent device and a child device, in which destinations for transmitting and receiving signals between the plurality of communication terminals are defined, wherein the child device measures the strength of a received signal from another communication terminal among the plurality of communication terminals, the parent device includes a learning unit that calculates a communication success rate with each of the child devices, learns to improve the communication success rate, and outputs output information regarding the destinations when input information regarding a communication state including the received signal strength and the communication success rate is received, and the parent device updates the destinations based on the output information.

[0007] The input information includes the number of paths, which are the number of signal propagation paths in wireless communication between each of the plurality of communication terminals and the other communication terminals, and the corresponding propagation and the communication signal strength of each of the paths.

[0008] The learning unit may learn by reinforcement learning using a communication success rate as a reward.

[0009] The parent unit may communicate with a child unit assigned to each of a plurality of time slots in accordance with a communication schedule including a plurality of time slots, and each of the child units may measure the received signal strength based on a signal transmitted by the parent unit or another child unit in a time slot of the plurality of time slots to which the parent unit or another child unit is assigned.

[0010] Each of the parent unit and the child unit may be configured to transmit a synchronization signal in an assigned time slot among a plurality of time slots, and each of the child units may measure the received signal strength based on the synchronization signal transmitted by the parent unit or the other child unit in the assigned time slot among the plurality of time slots.

[0011] The communication success rate is the rate at which replies are received in response to a reply request sent from a master unit to a slave unit in wireless communication, and the master unit may calculate the communication success rate for each of the slave units.

[0012] A connection destination learning device provided by a second aspect of the present invention is a connection destination learning device for application to a wireless communication system equipped with a plurality of communication terminals including a parent device and a child device, in which the connection destinations of the plurality of communication terminals for transmitting and receiving signals between each other are specified, and is equipped with an acquisition means for acquiring input information regarding the communication state including the received signal strength from other communication terminals measured by the child device and the communication success rate between each of the parent device and the child device, a learning means for learning to improve the communication success rate, and an output means for outputting output information regarding the connection destination when input information regarding the communication state is input.

[0013] The connection destination learning program provided by the third aspect of the present invention causes a connection destination learning device for application to a wireless communication system equipped with a plurality of communication terminals including a parent unit and a child unit, in which the connection destinations of the plurality of communication terminals that transmit and receive signals to and from each other are specified, to function as an acquisition means for acquiring input information regarding the communication state including the received signal strength from other communication terminals measured by the child unit and the communication success rate between the parent unit and each of the child units, a learning means for learning to improve the communication success rate, and an output means for outputting output information regarding the connection destination when input information regarding the communication state is input.

[0014] A connection destination learning method provided by a fourth aspect of the present invention is a connection destination learning method for application to a wireless communication system equipped with a plurality of communication terminals including a parent unit and a child unit, in which the connection destinations of the plurality of communication terminals for transmitting and receiving signals to and from each other are specified, and includes an acquisition process for acquiring input information regarding the communication state including the received signal strength from other communication terminals measured by the child unit and the communication success rate between each of the parent unit and the child unit, a learning process for learning to improve the communication success rate, and an output process for outputting output information regarding the connection destination when input information regarding the communication state is input. [Effects of the Invention]

[0015] According to this invention, learning is performed to improve the communication success rate, and information regarding each other's connection destinations between multiple communication terminals is automatically updated through learning, thereby reducing communication failures and enabling the optimization of communication routes between multiple communication terminals. [Brief explanation of the drawings]

[0016] [Figure 1] 1 is a schematic diagram of a wireless communication system according to an embodiment of the present invention. [Figure 2] FIG. 1 is a block diagram of a data station according to an embodiment of the present invention. [Figure 3] FIG. 2 is a block diagram of a repeater according to an embodiment of the present invention. [Figure 4] FIG. 1 is a block diagram of a sensor according to an embodiment of the present invention. [Figure 5] FIG. 2 is a diagram showing an example of a tree table according to an embodiment of the present invention. [Figure 6] FIG. 3 is a diagram illustrating an example of a routing table according to an embodiment of the present invention. [Figure 7] FIG. 4 is a diagram illustrating an example of a sensor table according to the embodiment of the present invention. [Figure 8] FIG. 3 is a diagram showing an example of a communication schedule according to an embodiment of the present invention. [Figure 9] 4 is a flowchart showing a scan process according to the embodiment of the present invention. [Figure 10] FIG. 2 is a diagram illustrating an overview of a learning unit according to an embodiment of the present invention. [Figure 11] FIG. 10 is a diagram showing an example of an intensity data list according to the embodiment of the present invention. [Figure 12] FIG. 10 is a diagram showing an example of a success rate data list according to the embodiment of the present invention. [Figure 13] FIG. 4 is a diagram showing an example of a scan data list according to the embodiment of the present invention. [Figure 14] FIG. 10 is a diagram showing an example of a multipath data list according to an embodiment of the present invention. [Figure 15]10 is a flowchart illustrating an update process according to an embodiment of the present invention. [Figure 16] FIG. 10 is a diagram showing an example of an updated tree table according to an embodiment of the present invention. [Figure 17] FIG. 10 is a diagram showing an example of an updated routing table according to an embodiment of the present invention. DETAILED DESCRIPTION OF THE INVENTION

[0017] A wireless communication system (a connection destination learning device, a connection destination learning program, and a connection destination learning method) according to an embodiment of the present invention will be described with reference to the drawings. Note that the configuration of the present invention is not limited to the embodiment. Also, the order of various processes constituting the flow described below is random as long as no inconsistencies occur in the process content.

[0018] FIG. 1 is a schematic diagram of a wireless communication system 100 according to an embodiment of the present invention.

[0019] The wireless communication system 100 includes a data station 10, a plurality of repeaters 20, and a plurality of sensors 30. The data station 10, the repeaters 20, and the sensors 30 are communication terminals that communicate with each other wirelessly and autonomously construct a network. That is, connection destinations for transmitting and receiving signals between the plurality of communication terminals are defined. In the wireless communication system 100, a multi-hop wireless network is formed. The data station 10 functions as a parent device. The repeaters 20 and the sensors 30 function as child devices. Basically, the data station 10 communicates with the repeaters 20, and the sensors 30 communicate with the repeaters 20. The number of sensors 30 is greater than the number of repeaters 20.

[0020] The data station 10 and the repeater 20 form a tree-type network topology with the data station 10 at the apex (top). In this embodiment, the data station 10 side of the network is referred to as the upstream side or upper side, and the terminal side of the tree is referred to as the downstream side or lower side. When the data station 10, the repeater 20, and the sensor 30 are not distinguished, they may be simply referred to as communication terminals. When distinguishing between the repeaters 20, they are distinguished by adding an alphabetical character after the symbol "20." Similarly, when distinguishing between the sensors 30, they are distinguished by adding an alphabetical character after the symbol "30."

[0021] In the wireless communication system 100, the sensor 30 detects a predetermined physical quantity of an object, and the detected value, i.e., the detected data, is collected in the data station 10 via the repeater 20. For example, the wireless communication system 100 is installed in a factory having a steam system. The steam system has a plurality of steam traps T (only one is shown in FIG. 1 ). The object is the steam trap T. The sensor 30 detects the vibration frequency and temperature of the steam trap T.

[0022] In addition, the wireless communication system 100 utilizes AI (Artificial Intelligence) machine learning (e.g., reinforcement learning) to learn how to improve the success rate of communication between the parent device and the child device, thereby optimizing the connection destinations (communication paths) between multiple communication terminals for sending and receiving signals.

[0023] <Data Station Configuration> 2 is a block diagram of the data station 10. The data station 10 collects and manages the detection values ​​of the sensors 30 of the wireless communication system 100. The data station 10 is also connected to a higher-level server 90 (FIG. 1) or the like via an external network or the like. The data station 10 transfers the detection values ​​of the sensors 30 to the server 90 as necessary. Furthermore, the data station 100 functions as a connection destination learning device and establishes a communication path for the wireless communication system 100.

[0024] The data station 10 includes a CPU 11, a memory 12, a storage unit 13, a wireless communication circuit 14, a clock circuit 15, a host interface unit 16, and a power supply circuit 17. In addition, the data station 10 may include a GPU.

[0025] Various programs and various information are stored in the storage unit 13. The CPU 11 performs various processes by reading and executing the various programs from the storage unit 13. For example, the storage unit 13 stores a program for forming a communication path of the network, a program for collecting detection values ​​of the sensors 30, a tree table that defines the tree structure of the network, a sensor table that defines which repeater 20 each sensor 30 is connected to, a routing table that defines the route to the final destination, a program for creating the routing table from the tree table, schedule information that defines a schedule for communication with the repeater 20, collected detection values, etc.

[0026] The storage unit 13 also stores a program for calculating a communication success rate. The communication success rate is the rate at which replies are received in response to reply requests sent from the master device (data station 10) to each slave device (repeater 20 and sensor 30) in wireless communication. The data station 10 (CPU 11) calculates the communication success rate as {(number of replies received) / (number of replies requested)}. The data station 10 (CPU 11) calculates the communication success rate for each slave device. For example, the data station 10 (CPU 11) stores the communication success rate for a predetermined number of replies requested from the present time (for example, the past five requests). Note that the "number of replies received" refers to the number of times the data station 10 receives a response. Therefore, if the sensor 30 returns a detection value but the intermediate relay device 20 fails to reply, this is not counted as the "number of replies received." The communication success rate is the rate at which communication is successful with a slave device that has requested a reply. Therefore, for example, when a data station 10 requests a reply from a certain sensor 30, even if a repeater 20 relays the signal along the way, the communication success rate between the data station 10 and the certain sensor 30 is calculated. Data such as the "number of replies received" used for the calculation, and the calculated communication success rate are also stored in the memory unit 13.

[0027] Furthermore, the memory unit 13 stores a connection destination learning program. The CPU 11 executes the connection learning program to function as a learning unit DL (learning model). The learning unit DL outputs output information relating to the connection destinations between multiple communication terminals that transmit and receive signals. The CPU 11 (data station 10) establishes (changes) communication paths of the wireless communication system 100 based on this output information. Details will be described later.

[0028] The wireless communication circuit 14 performs wireless communication with other communication terminals such as the repeater 20. The wireless communication circuit 14 operates under the control of the CPU 11, converts various signals into wireless signals by processing such as encoding and modulation, and transmits them via an antenna. The wireless communication circuit 14 also converts signals received via the antenna into appropriate signals by processing such as demodulation and decoding. Furthermore, the wireless communication circuit 14 measures the received signal strength (RSSI: Received Signal Strength Indicator) based on the received signal. The received signal strength can be, for example, the average value of the received signal. The received signal strength is stored in the memory unit 13 in association with the identification information of the terminal that transmitted the received signal.

[0029] The timing circuit 15 generates a predetermined clock and keeps track of the reference time of the data station 10. The upper interface unit 16 performs interface processing with the server 90. The power supply circuit 17 is connected to an external power supply (not shown) and supplies power to each element of the data station 10.

[0030] <Repeater configuration> 3 is a block diagram of the repeater 20. The repeater 20 transmits the detection value of the sensor 30 to the data station 10 in response to a command from the data station 10. The repeater 20 has a CPU 21, a memory 22, a storage unit 23, a wireless communication circuit 24, a timer circuit 25, a power supply circuit 26, and a battery 27.

[0031] Various programs and various information are stored in the storage unit 23. The CPU 21 performs various processes by reading and executing the various programs from the storage unit 23. For example, the storage unit 23 stores a program for forming a communication path of the network, a program for relaying detection values ​​of the sensors 30, a tree table, a routing table, a program for creating a routing table from the tree table, sensor connection information for identifying the connected sensors 30, and detection values ​​acquired from the sensors 30.

[0032] The storage unit 23 also stores a program for estimating the remaining capacity of the battery 27. The CPU 21 executes the program to integrate the power consumption according to various processes and estimate the remaining capacity of the battery 27.

[0033] The wireless communication circuit 24 performs wireless communication with other communication terminals. The wireless communication circuit 24 operates under the control of the CPU 21, converts various signals into wireless signals by processing such as encoding and modulation, and transmits them via an antenna. The wireless communication circuit 24 also converts signals received via the antenna into appropriate signals by processing such as demodulation and decoding. Furthermore, the wireless communication circuit 24 measures the received signal strength based on the received signal. The received signal strength can be, for example, the average value of the received signal. The received signal strength is stored in the memory unit 23 in association with the identification information of the terminal that transmitted the received signal.

[0034] The timing circuit 25 generates a predetermined clock and measures the reference time of the repeater 20. A battery 27 is connected to the power supply circuit 26. The power supply circuit 26 supplies power to each element of the repeater 20.

[0035] The repeater 20 is configured to be switchable between an active state in which various processes, such as transmitting and receiving signals to and from other communication terminals, can be performed, and a sleep state in which processes, such as transmitting and receiving signals, cannot be performed but power consumption is reduced compared to the active state. When the repeater 20 transitions from the active state to the sleep state, the CPU 21 sets the time to transition to the active state in the clock circuit 25 and transitions to the inactive state. In the sleep state, the clock circuit 25 continues timing. When the set time arrives, the clock circuit 25 notifies the CPU 21 of the arrival of the time, and upon receiving this notification, the CPU 21 transitions from the inactive state to the active state. In addition, the CPU 21 in the active state permits power supply from the power supply circuit 26 to the memory 22, the storage unit 23, and the wireless communication circuit 24. In this way, the repeater 20 transitions from the sleep state to the active state.

[0036] <Sensor configuration> 4 is a block diagram of the sensor 30. The sensor 30 detects the vibration frequency and temperature of the steam trap T and transmits the detected values ​​to the corresponding relay device 20. The sensor 30 has a sensor unit 40 that detects a predetermined physical quantity of the object, and a processing unit 50 that transmits the detected value of the sensor unit 40 to another communication terminal.

[0037] The sensor unit 40 includes a vibration sensor and a temperature sensor, and detects the vibration frequency and temperature of the steam trap T. The sensor unit 40 is installed so as to be in contact with the casing of the steam trap T (for example, the inlet portion into which steam and condensate flow in), and detects the vibration frequency and temperature of the contact portion. The sensor unit 40 outputs an electrical signal corresponding to the detected vibration frequency and temperature to the processing unit 50.

[0038] The processing unit 50 includes a CPU 51, a memory 52, a storage unit 53, a wireless communication circuit 54, a timer circuit 55, a sensor interface unit 56, a power supply circuit 57, and a battery 58. There are.

[0039] Various programs and information are stored in the storage unit 53. The CPU 51 performs various processes by reading and executing the various programs from the storage unit 53. For example, the storage unit 53 stores a program for forming a network communication path, a program for acquiring vibration frequency and temperature from the sensor unit 40 and transmitting the detected values ​​to the repeater 20, repeater connection information for identifying the connected repeater 20, the detected values, etc.

[0040] The storage unit 53 also stores a program for estimating the remaining capacity of the battery 58. The CPU 51 executes the program to integrate the power consumption according to various processes and estimate the remaining capacity of the battery 58.

[0041] The wireless communication circuit 54 performs wireless communication with other communication terminals. The wireless communication circuit 54 operates under the control of the CPU 51, converts various signals into wireless signals by processing such as encoding and modulation, and transmits them via an antenna. The wireless communication circuit 54 also converts signals received via the antenna into appropriate signals by processing such as demodulation and decoding. Furthermore, the wireless communication circuit 54 measures the received signal strength based on the received signal. The received signal strength can be, for example, the average value of the received signal. The received signal strength is stored in the memory unit 53 in association with the identification information of the terminal that transmitted the received signal.

[0042] The timing circuit 55 generates a predetermined clock and keeps track of the reference time for the sensor 30. The sensor interface unit 56 performs interface processing with the sensor unit 40. A battery 58 is connected to the power supply circuit 57. The power supply circuit 57 supplies power to each element of the sensor 30.

[0043] Like the repeater 20, the sensor 30 is configured to be switchable between an active state in which it can perform various processes such as sending and receiving signals with other communication terminals, and a sleep state in which it cannot perform processes such as sending and receiving signals but consumes less power than the active state.

[0044] <Connections between communication terminals> In the wireless communication system 100, the connection destination of each communication terminal is determined, and signals are propagated based on the connection relationship. The wireless communication system 100 holds a tree table, a routing table, a sensor table, sensor connection information, and repeater connection information as the connection relationships between the data station 10, the repeater 20, and the sensor 30.

[0045] The tree table is a table that defines the tree structure of the wireless communication system 100 and defines the upper communication terminal of each repeater 20. One tree table is created in the wireless communication system 100, and the data station 10 and all repeaters 20 hold a common tree table. Fig. 5 shows a tree table corresponding to the wireless communication system 100 of Fig. 1. The upper column of the tree table lists the target repeater 20, and the lower column specifies the upper communication terminal (data station 10 or repeater 20) to which each repeater 20 is connected.

[0046] The routing table defines the correspondence between all reachable final destination communication terminals from a given communication terminal and the next (one hop away) communication terminal on the communication path from the given communication terminal to the final destination, i.e., the communication terminal one hop below the given communication terminal on the communication path to the final destination. The routing table is created based on a tree table. The data station 10 and each relay device 20 each have their own unique routing table.

[0047] 6 is a routing table of relay 20a in wireless communication system 100 of FIG. 1. The upper column of the routing table lists the target final destination, and the lower column specifies relay 20, which is one hop away when relay 20a is the starting point. Since relays 20b, 20c, 20d, 20e, 20f, 20g, 20h, and 20i can be reached from relay 20a, relays 20 that are one hop below relay 20a on the communication path to those relays 20 are specified. Since relays 20j and 20k cannot be reached from relay 20a, the routing table of relay 20a does not specify relays 20, which are one hop away when relays 20j and 20k are the final destination.

[0048] The sensor table defines the connection relationship between the sensors 30 and the repeaters 20 (i.e., which repeater 20 each sensor 30 is connected to). One sensor table is created in the wireless communication system 100 and is held by the data station 10. Fig. 7 shows the sensor table corresponding to the wireless communication system 100 of Fig. 1. The upper column of the sensor table lists the target sensors 30, and the lower column defines the repeater 20 to which each sensor 30 is connected.

[0049] The sensor connection information is information held by each relay device 20, and is information that identifies the sensor 30 connected to each relay device 20 (for example, the communication address of the sensor 30).

[0050] The relay connection information is information held by each of the sensors 30, and is information that identifies the relay 20 to which each sensor 30 is connected (for example, the communication address of the relay 20).

[0051] In the wireless communication system 100, signals are propagated using these connection relationships. First, a case where a signal is transmitted in the downlink direction from the data station 10 will be described. For example, when the data station 10 transmits a signal to the sensor 30g, the data station 10 determines the relay 20h to which the sensor 30g is connected based on the sensor table. Then, the data station 10 determines, based on its own routing table, that when the final destination is the relay 20h, the relay 20 one hop away is the relay 20a. The data station 10 sets the relay 20h as the final destination and transmits a signal with the relay 20a set as the one hop away destination.

[0052] Thereafter, each relay 20 that receives this signal changes the relay 20 one hop away based on its own routing table and propagates the signal to relay 20h. Specifically, relay 20a sets relay 20d as the one hop away destination and forwards the signal. Relay 20d that receives the signal sets relay 20g as the one hop away destination and forwards the signal. Relay 20g that receives the signal sets relay 20h as the one hop away destination and forwards the signal. When relay 20h, the final destination, receives the signal, it sets both the final destination and the one hop away destination to sensor 30g based on its own sensor connection information and transmits the signal. In this way, the signal is finally received by sensor 30g.

[0053] Next, a case where a signal is transmitted in the uplink direction to the data station 10 will be described. For example, when the sensor 30g transmits a signal to the data station 10, the sensor 30g transmits a signal in which both the final destination and the one-hop destination are set to the relay 20h based on its own relay connection information. The signal is received by the relay 20h. The relay 20h determines, based on the tree table, that the relay 20 one hop above is the relay 20g. The relay 20h transmits a signal in which the data station 10 is set as the final destination and the relay 20g is set as the one-hop destination.

[0054] Thereafter, each relay 20 that receives this signal changes the relay 20 one hop away based on the tree table and propagates the signal to the data station 10. Specifically, relay 20g sets relay 20d as the destination one hop away and forwards the signal. Relay 20d sets relay 20a as the destination one hop away and forwards the signal. Relay 20a sets data station 10 as the destination one hop away and forwards the signal. In this way, the signal is finally received by data station 10.

[0055] In this way, the data station 10, the repeater 20, and the sensor 30 transmit signals based on the connection relationships of the communication terminals (tree table, routing table, tree table, sensor connection information, and repeater connection information).

[0056] <Communication Schedule> The wireless communication system 100 configured as described above performs a collection process as a normal operation, collecting the detection values ​​of the sensors 30 into the data station 10. The data station 10 communicates with each repeater 20 according to the communication schedule shown in Fig. 8, and collects the detection values ​​of the sensors 30 corresponding to, i.e., connected to, each repeater 20.

[0057] The communication schedule in Fig. 8 shows one cycle of the collection process, and the communication schedule in Fig. 8 is executed repeatedly. The communication schedule is divided into multiple time slots. Each repeater 20 is assigned a specific time slot. Each repeater 20 communicates with the data station 10 in the corresponding time slot, and transmits the detection value from the sensor 30 connected to that repeater 20 to the data station 10 (hereinafter, this process is also referred to as "reply process").

[0058] Basically, each repeater 20 is in an active state during a specific time slot (hereinafter also referred to as a "specific slot") assigned to it, and is in a sleep state outside the specific slot. However, a repeater 20 existing on a communication path between another repeater 20 and the data station 10 must perform relay processing when a lower-level repeater 20 communicates with the data station 10, and therefore also becomes active in a time slot (hereinafter also referred to as a "relay slot") assigned to the lower-level repeater 20 to perform relay processing. In addition, a sensor 30 transmits a detection value to the repeater 20 in the specific slot of the repeater 20 to which it is connected, and therefore is in an active state during the specific slot of the repeater 20. When there is no need to transmit a detection value to the repeater 20, the sensor 30 is basically in a sleep state.

[0059] In the communication schedule of Figure 8, time slots are defined in a matrix. Basically, time slots are assigned according to the hierarchy of the tree-structured communication path. More specifically, a hierarchy of the tree structure is assigned to each column. For example, data station 10 is assigned to column L0, the first hierarchy (i.e., 1 hop) is assigned to column L1, and the second hierarchy (i.e., 2 hops) is assigned to column L2. The same applies to the third hierarchy and beyond.

[0060] Typically, each repeater 20 is assigned one of the time slots. The repeaters 20a and 20j in the first layer are assigned the time slots in column L1. The repeaters 20b, 20c, 20d, and 20k in the second layer are assigned the time slots in column L2. The repeaters 20e, 20f, and 20g in the third layer are assigned the time slots in column L3. The repeaters 20h and 20i in the fourth layer are assigned the time slots in column L4.

[0061] On the other hand, since the data station 10 has more processing content than the repeater 20, multiple time slots (all of the time slots in row L0 in FIG. 8) are assigned to the data station 10 instead of one time slot. Note that the number of time slots in a row is different from the number of repeaters 20 in each layer (usually, the number of time slots in a row is greater), so some time slots in a row do not have a repeater assigned to them.

[0062] As described above, in a specific slot of a certain repeater 20, the sensor 30 connected to that repeater 20 is also active, so that each sensor 30 is essentially assigned a specific time slot. However, since multiple sensors 30 can be connected to a repeater 20, in such a case, multiple sensors 30 are assigned to the specific slot of that repeater 20. For example, in the example of Figure 8, repeater 20e is assigned to the time slot in column L3, row N1. Since two sensors 30d and 30e are connected to repeater 20e (see Figure 1), those two sensors 30 are essentially assigned to the time slot in column L3, row N1.

[0063] In a communication schedule, time slots are processed in the column direction. For example, in a certain column (e.g., column L1), time slots are processed in ascending order of row number (i.e., from row N1 to Nm), and when the time slot in the last row number (row Nm) in that column is processed, processing continues in the same order from the time slot in the first row number (row N1) in the next column (e.g., column L2).

[0064] <Establishing connection relationships and allocating time slots> When establishing a communication path of the network, the data station 10 determines the connection relationship of the communication terminals and allocates time slots to complete the communication schedule. Specifically, during the initial setting for establishing the communication path of the network, the data station 10 determines the connection relationship of the communication terminals and allocates time slots to complete the communication schedule.

[0065] For example, the data station 10 determines which communication terminals are connected to each other, i.e., the connection relationships between the data station 10, the repeaters 20, and the sensors 30, and creates a tree table and a sensor table. Furthermore, the data station 10 creates a routing table based on the tree table. Once the connection relationships are determined, the data station 10 assigns time slots to each repeater 20 and completes the communication schedule. The data station 10 stores the tree table, sensor table, routing table, and communication schedule (i.e., the assignment of time slots to repeaters 20) in the memory unit 13.

[0066] When the data station 10 has completed the allocation of time slots, it notifies each repeater 20 of the tree table and the slot number of the specific slot. At this time, a repeater 20 that needs to perform relay processing for a lower-level repeater 20 is notified of the slot number of its own specific slot as well as the specific slot of the lower-level repeater 20, i.e., the slot number of the relay slot. Furthermore, the data station 10 notifies each repeater 20 of sensor connection information and each sensor 30 of relay connection information based on the sensor table.

[0067] Each repeater 20 creates a routing table based on the tree table. Furthermore, the repeater 20 notifies the connected sensor 30 of the slot number of the specific slot of the repeater 20. The repeater 20 stores the slot numbers of the specific slot and the repeat slot, the tree table, the routing table, and the sensor connection information in the memory unit 23.

[0068] The sensor 30 stores the slot number of the specific slot of the connected repeater 20 and repeater connection information in the storage unit 53 .

[0069] After the initial setting, the communication path is updated based on the output information (tree table and sensor table) output from the learning unit DL. Details will be described later.

[0070] System Operation Next, various processes performed by the wireless communication system 100 will be described.

[0071] -Collection and processing- In the collection process, the data station 10 proceeds with the process according to the communication schedule. Specifically, the data station 10 performs the process required for the data station 10 in the time slot assigned to it. Then, the data station 10 communicates with the repeaters 20 assigned to the time slots in order of the time slots. At this time, the signal sent from the data station 10 to each repeater 20 includes at least a request signal requesting the return of the detection value of the sensor 30.

[0072] Meanwhile, the repeater 20 becomes active at the timing of a specific slot according to the communication schedule and waits for a request signal from the data station 10. The repeater 20 also acquires a detection value from the sensor 30 connected to the repeater 20 according to the specific slot. When the repeater 20 receives a request signal, it returns the detection value from the sensor 30 to the data station 10 as a response to the request signal. The repeater 20 also becomes active during the repeat slot and performs relay processing between the data station 10 and the lower repeater 20.

[0073] The sensor 30 becomes active in accordance with the specific slot of the repeater 20 to which it is connected and transmits its detection value to the repeater 20. When multiple sensors 30 are connected to one repeater 20, all of the multiple sensors 30 are active at least at the start of the specific slot of the repeater 20. The multiple sensors 30 receive request signals from the repeater 20 in turn and transmit their detection values ​​to the repeater 20. The multiple sensors 30 enter sleep mode in the order in which they have completed transmitting their detection values ​​to the repeater 20.

[0074] During the collection process, the data station 10, the repeater 20, and the sensor 30 each measure the received signal strength based on the received signal.

[0075] For example, when the sensor 30 receives a request signal from the repeater 20 to which it is connected, it measures the strength of the received signal and returns the measurement result together with the detection value to the repeater 20. Furthermore, when each repeater 20 performs the process of returning the detection value in its own specific slot during the collection process, it measures the strength of the received signal when transmitting and receiving signals with the upper communication terminal (data station 10 or repeater 20) and the lower communication terminal (sensor 30), and returns the measurement result together with the detection value and received signal strength of the sensor 30 to the data station 10.

[0076] The received signal strength may be associated with the identification information of the measuring terminal and the identification information of the transmitting terminal of the received signal.

[0077] During the reply process, the data station 10 also measures the strength of the received signal from the downstream repeater 20. The data station 10 stores the received signal strength measured by itself and the received signal strength measured by other communication terminals along with the detected value, together with the time of acquisition (date and time) in the memory unit 13 (see the strength data list in FIG. 11). The data station 10 updates the data in the memory unit 13 every time it acquires a new measurement result.

[0078] Furthermore, in the collection process, the repeater 20 transmits a synchronization signal during a specific slot. The repeater 20 broadcasts a synchronization signal during the specific slot and synchronizes with other repeaters 20 or sensors 30 that receive the synchronization signal. The data station 10 also broadcasts a synchronization signal during an assigned time slot and synchronizes with other repeaters 20 or sensors 30 that receive the synchronization signal.

[0079] Furthermore, during the collection process, the data station 10 calculates the communication success rate for each sensor 30. In the collection process, the communication success rate for one sensor 30 is calculated as {(number of replies received) / (number of replies requested)}. The "number of replies requested" is the number of times the data station 10 transmitted a request signal to a certain sensor 30. The "number of replies received" is the number of times the data station 10 received a detection value from the certain sensor 30. In one cycle of the collection process, the communication success rate is calculated for each of all sensors 30. As described above, the calculation is performed for the number of times a reply was requested up to the past five times from the current time. Each communication success rate is stored in the memory unit 13 in association with the identification information of the sensor 30 (see, for example, the success rate data list in FIG. 12). Each time the communication success rate is calculated, the data in the memory unit 13 is updated.

[0080] In this way, in the basic processing of the collection process, the data station 10 collects the detection values ​​of the sensors 30 corresponding to each specific slot in accordance with the communication schedule, thereby collecting the detection values ​​of all the sensors 30. Note that the collection process differs depending on the number of sensors 30, but is executed, for example, once every 30 minutes.

[0081] -Scan processing- In the wireless communication system 100, in a connection destination update process described later, connection destinations (communication paths) for transmitting and receiving signals between multiple communication terminals are updated. At this time, information such as the received signal strength and communication success rate of each slave terminal is used. Therefore, the wireless communication system 100 executes a scan process to scan the received signal strength, etc. In the collection process described above, the received signal strength, etc. is acquired for communication terminals with a connection relationship, but in the scan process, the received signal strength, etc. is also acquired for communication terminals with no connection relationship.

[0082] The scanning process is performed, for example, once every 12 hours (twice a day). FIG. 9 shows a flowchart of the scanning process. The data station 10 performs the scanning process when the time based on the period arrives. First, the data station 10 transmits a scanning signal to all repeaters 20 and sensors 30 (step S10). The data station 10 transmits the scanning signal in parallel with the collection process. Specifically, the data station 10 transmits a scanning signal together with a detection value request signal to the corresponding repeater 20 in each time slot. That is, the scanning process is performed together with the collection process.

[0083] Upon receiving the scan signal, the repeater 20 and the sensor 30 perform a scan operation. Specifically, in the collection process, the repeater 20 becomes active not only in the specific slot but also in other time slots, and waits for signals from other communication terminals.

[0084] During the collection process, the repeater 20 performing the scanning operation waits for a synchronization signal from another communication terminal (the data station 10 and the repeater 20) in a time slot other than its own specific slot. If the repeater 20 receives a synchronization signal, it measures the strength of the received signal.

[0085] In the collection process, the timing at which each repeater 20 transmits a synchronization signal is determined within each specific slot. That is, the repeater 20 transmits a synchronization signal at a predetermined timing within the specific slot (for example, at the beginning or end of the specific slot). In addition, in each specific slot, a period during which the corresponding repeater 20 does not transmit any signal is also determined. Therefore, in time slots other than its own specific slot, the repeater 20 performing the scanning operation is in an active state only during a portion of the period that includes the timing at which the synchronization signal is transmitted and during a portion of the period during which no signal is transmitted, and measures the communication signal strength of the synchronization signal, and is in a sleep state during the rest of the period. This allows for reduced power consumption during the scanning operation.

[0086] The repeater 20 performs this process (scanning operation) for one cycle of the communication schedule to acquire the received signal strength of the synchronization signal in all time slots except for the specific slot. The repeater 20 stores data in the memory unit 23 as scan data, which associates the identification information of the transmitting terminal that transmitted the synchronization signal with the received signal strength of this synchronization signal.

[0087] The sensor 30 also performs a scanning operation similar to that of the repeater 20. In time slots other than the specific slot of the repeater 20 to which it is connected, the sensor 30 is in an active state only for a portion of the period including the timing when a synchronization signal is transmitted and for a portion of the period when no signal is transmitted. If the sensor 30 receives a synchronization signal, it measures the strength of the received signal. Outside of the above-mentioned period in the time slot, the sensor 30 is in a sleep state. The sensor 30 performs this process (scanning operation) for one cycle of the communication schedule.

[0088] The sensor 30 also measures the received signal strength of the synchronization signal in a specific slot of the connected repeater 20. That is, the sensor 30 acquires the received signal strength of the synchronization signal in all time slots. The sensor 30 stores data in the memory unit 53 as scan data, which associates the identification information of the transmitting terminal that transmitted the synchronization signal with the received signal strength of this synchronization signal.

[0089] The scan data stored in the storage unit is updated every time new data is acquired.

[0090] Next, the data station 10 transmits a reply signal to the repeaters 20 and the sensors 30 requesting the return of the scan data (step S11). After one cycle of the communication schedule has elapsed since the data station 10 transmitted the scan signal in the process of step S10, the data station 10 transmits the reply signal to all the repeaters 20 and the sensors 30 in sequence according to the time slot. That is, the data station 10 transmits the reply signal when executing the collection process next to the collection process of one cycle in which the scan signal was transmitted.

[0091] Furthermore, if the data station 10 has transmitted a scan signal only to a specific repeater 20 or sensor 30 in the processing of step S20, it only needs to transmit a reply signal to only the repeater 20 or sensor 30 that transmitted the scan signal.

[0092] The repeater 20 and the sensor 30 that received the reply signal return, for example, the top five scan data with the highest received signal strengths, excluding the communication terminals with which a connection has been established, to the data station 10. In other words, scan data from other communication terminals with which communication is possible but not established as a connection is returned. Note that the repeater 20 and the sensor 30 may return all scan data to the data station 10.

[0093] The data station 10 stores the scan data returned from the repeaters 20 and the sensors 30 in the storage unit 13 in association with the time of acquisition (see, for example, the scan data list in FIG. 13 ) (step S12). Furthermore, in the processing of step S12, the data station 10 calculates the communication success rate with each of the repeaters 20 and the sensors 30 that transmitted the reply signals, based on the reply results. Here, the “number of times a reply was requested” refers to the number of times the data station 10 transmitted a reply signal requesting a reply of scan data to a certain repeater 20 or sensor 30. The “number of times a reply was received” refers to the number of times the data station 10 received scan data from the certain repeater 20 or sensor 30. Then, the data station 10 updates the corresponding communication success rate (success rate data list) in the storage unit 13. Therefore, while the collection processing described above updates only the communication success rate for each sensor 30, the scanning processing updates the communication success rate for each of all the slave devices.

[0094] -Connection destination update process- As described above, in the wireless communication system 100, connection relationships between communication terminals are determined, and signals are propagated based on these connection relationships. However, the communication environment may change, and previously good communication between communication terminals may become poor. Therefore, the data station 10 uses machine learning (e.g., reinforcement learning) to update the connection destinations (communication paths) between multiple communication terminals for transmitting and receiving signals so as to improve the communication success rate.

[0095] In this embodiment, the hierarchy of the relay device 20 in the tree-type network topology is not changed, and the slot allocation of each relay device 20 in the communication schedule is not changed.

[0096] First, the learning unit DL of the data station 10 will be described in detail. FIG. 10 is a diagram showing an overview of the learning unit DL. When input information relating to the communication status, including the received signal strength and the communication success rate, is input, the learning unit DL outputs output information relating to the connection destination. In this embodiment, a tree table and a sensor table are output as output information. The learning unit DL is generated by reinforcement learning using a neural network.

[0097] 10, the learning unit DL is composed of an input layer DL1, an intermediate layer DL2, an output layer DL3, etc. The input layer DL1 receives input information related to the current communication state of the wireless communication system 100. Specifically, received signal strength data of a currently connected terminal, communication success rate data, received signal strength data of other nodes with which communication is possible, multipath data, etc. are input as input information related to the current communication state.

[0098] The received signal strength data of a connected terminal is data on the received signal strength during communication between communication terminals that have established a connection relationship. For example, the strength data list shown in FIG. 11 is used as the strength data. The strength data list includes the received signal strength values ​​measured by each communication terminal when a detection value request is made in the collection process described above. Therefore, the strength data list does not include the received signal strength from other communication terminals with which a communication terminal can communicate but with which no connection relationship has been established.

[0099] The strength data list is composed of a measuring terminal ID, a transmitting terminal ID, an acquisition time, a received signal strength, etc. The measuring terminal ID indicates the identification information of the communication terminal that measured the received signal strength. The transmitting terminal ID indicates the identification information of the communication terminal that transmitted a signal to the corresponding measuring terminal. When a repeater 20 corresponds to a measuring terminal, there are multiple transmitting terminal IDs (transmitting terminals). This is because the repeater 20 communicates with a higher-level repeater 20 or data station 10 and a lower-level repeater 20 or sensor 30 during the collection process. The acquisition time is the time when the data station 10 received (acquired) the received signal strength data. The received signal strength indicates the received signal strength (average value) (dBm) of the corresponding transmitting terminal (transmitting terminal ID).

[0100] For example, in FIG. 11, the sensor 30a with the measuring terminal ID: 30a stores the received signal strength of the signal from the sensor 20a with the transmitting terminal ID: 20a.

[0101] The communication success rate data is data that sets the communication success rate between the data station 10 (parent device) and each of the repeater 20 and the sensor 30. For example, the success rate data list in FIG. 12 is used as the communication success rate data. The success rate data list is composed of a target terminal ID, a communication success rate, etc. The target terminal ID indicates the identification information of the communication terminal to which the data station 10 has requested a reply. The communication success rate indicates the communication success rate between the data station 10 and the target terminal (target terminal ID).

[0102] The received signal strength data of other nodes with which communication is possible is the scan data described above. That is, it is the top five scan data with the highest received signal strength, excluding communication terminals with which a connection relationship has been established. For example, the scan data list shown in FIG. 13 is used as the received signal strength data of other nodes with which communication is possible. In addition to the received signal strength, the scan data list is also referenced as information for each communication terminal to identify other communication terminals with which communication is possible.

[0103] The scan data list has the same format as the strength data list described above, and is composed of a measuring terminal ID, a transmitting terminal ID, an acquisition time, a communication signal strength, etc. The measuring terminal ID indicates the identification information of the communication terminal that measured the received signal strength. The transmitting terminal ID indicates the identification information of another communication terminal that is not connected to the corresponding measuring terminal but can communicate with this measuring terminal and that transmitted a signal to this measuring terminal. Therefore, some measuring terminals may have multiple transmitting terminal IDs or no transmitting terminal ID. The acquisition time is the time when the data station 10 received (acquired) the received signal strength data. The received signal strength indicates the received signal strength (average value) (dbm) of the corresponding transmitting terminal (transmitting terminal ID).

[0104] Multipath data is data relating to multipaths between communication terminals at two points. For example, the multipath data list shown in FIG. 14 is used as the multipath data. In wireless communication, multipath refers to phenomena such as reflection (reflected waves) that occur when a signal propagates through space via two or more propagation paths. Therefore, if the reflected waves are strong, they may collide with the original direct waves, causing data corruption or the like. As described above, the wireless communication system 100 is installed in a factory or the like that has a steam system, and therefore there is a risk of multipaths occurring due to reflected waves from factory equipment such as tanks and piping. Therefore, using multipath data as input information is also useful for improving the communication success rate.

[0105] The multipath data list consists of terminal ID-1, terminal ID-2, number of multipaths, multipath signal strength, etc. Terminal ID-1 and terminal ID-2 indicate the identification information of the communication terminals, and specify which two points of the communication terminals are involved. Terminal ID-1 indicates the identification information of the communication terminal that transmits the signal between the two points. Terminal ID-2 indicates the identification information of the communication terminal that receives the signal. It is sufficient to prepare at least for combinations for which a connection relationship may be established.

[0106] The number of multipaths indicates the number of propagation paths (paths) that occur between two points, terminal ID-1 and terminal ID-2. propagation Occurs excluding the path (direct wave) propagation Indicates the number of paths (reflected waves). Multipath signal strength is indicated by the number of multipaths. propagation Shows the received signal strength (average value) (dBm) for each path (reflected wave).

[0107] The above-mentioned number of multipaths and multipath signal strength may be generated using radio wave simulation software. The multipath data may be stored in advance in the storage unit 13 of the data station 10. Data such as the number of multipaths may be prepared for all combinations of repeaters 20 and sensors 30.

[0108] 10, the middle layer DL2 is subjected to reinforcement learning so that information about connection destinations to improve the communication success rate is output from the output layer DL3. Note that the middle layer DL2 also references the current tree table and sensor table.

[0109] The intermediate layer DL2 is configured to perform reinforcement learning using the success rate of communication between the parent device and each child device as a reward. That is, when the connection destination (communication path) is updated based on the tree table and sensor sable output from the output layer DL3, parameters (weights, etc.) are updated so as to improve the success rate of communication between the parent device and each child device. Examples of reinforcement learning include deep Q-learning and AC (Actor-Critic). Also, instead of reinforcement learning, unsupervised learning using algorithms such as GAN and clustering may be performed.

[0110] In this embodiment, the learning unit DL is initially in a state where reinforcement learning has not progressed, but reinforcement learning may be performed in advance, for example, by running a simulation of a model of a wireless communication system in advance.

[0111] When the input information is input to the input layer DL1, the output layer DL3 outputs a tree table and sensor tables as output information related to the connection destination. The data station 10 changes the connection relationships using the output tree table and sensor tables.

[0112] 15 shows a flowchart of the connection destination update process. The data station 10 executes the update process, for example, after the collection process is completed. For example, the update process is executed during a predetermined period between the completed collection process and the next collection process.

[0113] First, the data station 10 inputs the above-mentioned input information to the learning unit DL (input layer DL1) (step S20). After that, the data station 10 identifies the output information output from the learning unit DL (output layer DL3) (step S21).

[0114] Next, the data station 10 determines whether to update (change) the connection relationships based on the output information (step S22). Specifically, the data station 10 compares the current tree table and sensor table with the output information, and determines to change the connection relationships if there is a change in the connection destination.

[0115] If it is determined that the connection relationship is not to be updated (step S22: NO), the data station 10 proceeds to the process of step S24. On the other hand, if it is determined that the connection relationship is to be updated (step S22: YES), the data station 10 performs update execution processing (step S23). In the update execution processing, processing for setting a new connection destination is executed.

[0116] Specifically, the data station 10 changes (updates) the tree table and sensor table stored in the storage unit 13 to the tree table and sensor table output in the processing of step S21. In addition, the data station 10 creates a new routing table based on the tree table, and changes (updates) the routing table stored in the storage unit 13.

[0117] Furthermore, the data station 10 notifies the repeater 20 and the sensor 30 of the new connection destination. Specifically, the data station 10 notifies all repeaters 20 of the updated tree table. Furthermore, when the connection destination of a sensor 30 is updated, the data station 10 notifies the repeater 20 to which the new sensor 30 is connected of sensor connection information, i.e., information identifying the connected sensor 30, and notifies the sensor 30 whose connection destination has been changed of repeater connection information, i.e., information identifying the repeater 20 to be connected and the slot number of the repeater 20. Figure 16 shows the tree table after the update. In Figure 16, information updated in the tree table of Figure 5 is underlined. The connection destination of repeater 20c has been updated to repeater 20b.

[0118] Relay device 20, which has received the updated tree table, stores the tree table in storage unit 23 and updates the routing table based on the tree table. Figure 17 shows the updated routing table of relay device 20a. In Figure 17, information that has been updated with respect to the routing table in Figure 6 is underlined. When the final destination is relay device 20c or relay device 20f, the destination one hop away has been updated to relay device 20b.

[0119] Furthermore, the repeater 20 that has received the new sensor connection information updates the sensor connection information, and in subsequent processing (for example, the next collection processing), the repeater 20 executes processing for the newly connected sensor 30.

[0120] Upon receiving the new repeater connection information and the new slot number, the sensor 30 updates the repeater connection information and the slot number. The sensor 30 becomes active in accordance with the specific slot of the new repeater 20 and transmits the detected value to the repeater 20.

[0121] During the connection destination update process, all of the repeaters 20 and sensors 30 may be in an active state. The data station 10 may notify the repeaters 20 and sensors 30 of the new connection destinations during the period in which they are in this active state.

[0122] Next, the data station 10 determines whether the learning execution condition is met (step S24). An example of the learning execution condition is that five scan processes have been executed since the previous learning process was executed in step S25. Since the communication success rate is the value for the past five times, if the learning execution condition is met, the communication success rate for all the slave units is updated to the communication success rate after the previous learning process was executed. Note that the learning execution condition may also be that five collection processes have been executed since the previous learning process was executed. Alternatively, the learning execution condition may also be that one collection process has been executed.

[0123] If it is determined that the learning execution condition is not met (step S24: NO), the data station 10 ends the change process. On the other hand, if it is determined that the learning execution condition is met (step S24: YES), the data station 10 executes the learning process (step S25) and ends the change process. In the learning process, the latest communication success rate (success rate data list) is given to the learning unit DL as a reward, and the hidden layer DL2 undergoes reinforcement learning so as to improve the communication success rate.

[0124] As described above, learning is performed to improve the communication success rate, and information regarding each other's connection destinations between multiple communication terminals is automatically updated through learning, thereby reducing communication failures and enabling the optimization of communication routes between multiple communication terminals.

[0125] The input information in the above embodiment is not limited to the above embodiment. It is sufficient if it includes at least the received signal strength data (strength data list, scan data list) and the communication success rate of each node. Therefore, multipath data does not have to be included in the input information.

[0126] In the above embodiment, the connection destination update process is executed each time one collection process is completed, but this is not particularly limited. Learning and connection destination update may be executed at any timing. For example, they may be executed once a week.

[0127] In the above embodiment, the received signal strength is measured in the collection process to generate the strength data list, but this is not particularly limited. In the scan process, the received signal strength may be measured in all communication terminals in response to a synchronization signal, and the strength data list and scan data list may be generated. [Industrial Applicability]

[0128] The present invention is useful for optimizing communication routes between a plurality of communication terminals. [Explanation of symbols]

[0129] 10 Data Station (communication terminal, parent device, connection learning device) 20 Repeater (communication terminal, handset) 30 Sensors (communication terminals, slave units) 100 Wireless Communication System DL Learning Department

Claims

1. A wireless communication system including a plurality of communication terminals including a master device and a slave device, in which connection destinations for transmitting and receiving signals between the plurality of communication terminals are specified, the slave device measures the strength of a received signal from another communication terminal among the plurality of communication terminals; the parent device calculates a communication success rate with each of the child devices; a learning unit that learns to improve the communication success rate, and outputs output information about the connection destination when input information about a communication state including the received signal strength and the communication success rate is input; the parent device updates the connection destination based on the output information; The input information includes, for each combination of the plurality of communication terminals, a number of paths, which is the number of reflected waves of a signal in wireless communication between a terminal device that transmits a signal and another communication terminal that receives the signal, and a communication signal strength of each of the reflected waves. Wireless communication system.

2. the learning unit learns by reinforcement learning using the communication success rate as a reward.

10. The wireless communication system of claim 1.

3. the master device communicates with the slave devices assigned to each of the plurality of time slots according to a communication schedule including a plurality of time slots; each of the slave devices measures the received signal strength based on a signal transmitted by the master device or the other slave device in a time slot assigned to the master device or the other slave device among the plurality of time slots; 3. The wireless communication system according to claim 1 or 2.

4. each of the parent device and the child device is configured to transmit a synchronization signal in an assigned time slot among the plurality of time slots; each of the slave devices measures the received signal strength based on the synchronization signal transmitted by the master device or the other slave device in a time slot assigned to the master device or the other slave device among the plurality of time slots; 4. The wireless communication system according to claim 3.

5. the communication success rate is a rate at which replies are received in response to reply requests sent from the parent device to the child device in wireless communication; the parent device calculates the communication success rate for each of the child devices; 5. The wireless communication system according to claim 1.

6. A connection destination learning device for use in a wireless communication system in which a plurality of communication terminals including a master device and a slave device are provided and connection destinations for transmitting and receiving signals between the plurality of communication terminals are specified, an acquisition means for acquiring input information on a communication state including a received signal strength from another communication terminal measured by the slave device and a communication success rate between the master device and each of the slave devices; learning means for learning so as to improve the communication success rate; an output means for outputting output information relating to the connection destination when input information relating to the communication state is input; Equipped with The input information includes, for each combination of the plurality of communication terminals, a number of paths, which is the number of reflected waves of a signal in wireless communication between a terminal device that transmits a signal and another communication terminal that receives the signal, and a communication signal strength of each of the reflected waves. Connected learning device.

7. A connection destination learning device for application to a wireless communication system in which a plurality of communication terminals including a master and a slave are provided and connection destinations of the plurality of communication terminals for transmitting and receiving signals between the plurality of communication terminals are specified, an acquisition means for acquiring input information on a communication state including a received signal strength from another communication terminal measured by the slave device and a communication success rate between the master device and each of the slave devices; learning means for learning so as to improve the communication success rate; an output means for outputting output information relating to the connection destination when input information relating to the communication state is input; It functions as The input information includes, for each combination of the plurality of communication terminals, a number of paths, which is the number of reflected waves of a signal in wireless communication between a terminal device that transmits a signal and another communication terminal that receives the signal, and a communication signal strength of each of the reflected waves. Connected learning programs.

8. A connection destination learning method to be applied to a wireless communication system in which a plurality of communication terminals including a master device and a slave device are provided, and connection destinations for transmitting and receiving signals between the plurality of communication terminals are specified, an acquisition process for acquiring input information regarding a communication state including a received signal strength from another communication terminal measured by the slave device and a communication success rate between the master device and each of the slave devices; a learning process for learning so as to improve the communication success rate; an output process for outputting output information regarding the connection destination when input information regarding the communication status is input; Including, The input information includes, for each combination of the plurality of communication terminals, a number of paths, which is the number of reflected waves of a signal in wireless communication between a terminal device that transmits a signal and another communication terminal that receives the signal, and a communication signal strength of each of the reflected waves. How to learn where to connect.

Images