JP7601735B2 - Mesh network system and communication control method - Google Patents

Description

The present invention relates to a wired/wireless mesh network system that configures a mesh network using both wireless and wired connections.

The progress of robotics technology in recent years has been remarkable, and robots are increasingly being used to solve various social issues. Many of these robots are unmanned mobile objects, such as unmanned aerial vehicles and autonomous vehicles. Unmanned mobile objects need to be equipped with a communication system to transmit control command data for remote operation and autonomous control, and video data captured by cameras mounted on the unmanned mobile object. In such cases, if the unmanned mobile object is not a machine that moves along a predetermined route such as a rail, wireless communication suitable for movement is often used.

For example, Patent Document 1 discloses an invention in which a mobile base station and a terminal station are equipped with a long-distance communication function for preparing short-distance communication and a short-distance communication function for data transmission, and the timing of short-distance communication is scheduled by communication using the long-distance communication function. Patent Document 2 discloses an invention in which, in a relay system using an unmanned aerial vehicle, the relay position of the unmanned aerial vehicle is searched for based on the relay communication quality, the planned relay time, and the power supply state of the unmanned aerial vehicle (available power supply amount).

International Publication No. 2017/018021 JP 2019-169848 A

One type of communication network that can be used to communicate with unmanned vehicles is a mesh network. A mesh network is made up of multiple nodes, and by connecting adjacent nodes that can communicate with each other, the entire network forms a mesh-like structure. In particular, when the connections between nodes are made wirelessly, it is called a wireless mesh network, and data is transferred wirelessly across multiple nodes in a bucket brigade style.

Mesh networks usually have multiple routes, and if communication becomes unavailable on one route, they can switch to an alternative route, making them more resistant to failures than other network topologies. However, the existence of multiple routes also means that there are loop-shaped routes in the network. When broadcasting IP packets in such a network, the spanning tree protocol is generally used to prevent the IP packets from repeating broadcasts even after they have traveled around the loop-shaped route, causing a broadcast storm.

The spanning tree protocol uses a spanning tree algorithm to automatically block communication between specific nodes, making it possible to prevent loops. In addition, when a failure occurs, it is possible to continue communication via an alternative route by releasing the blocking state.

However, in this type of mesh network, the route is determined not by centralized control by specific high-priority nodes, but by all nodes on an equal footing, making it less efficient than a centralized control method. In particular, in cases where high throughput and low latency are required, there is a risk that a highly fault-tolerant mesh network cannot be adopted.

A use case that requires high throughput and low latency in addition to fault tolerance is a wireless communication system that transmits video related to the remote control of an unmanned aerial vehicle (UAV). Figure 1 shows a configuration example of a conventional example 1 of such a wireless communication system. This system transfers video captured of the surroundings of a UAV when the UAV is flying beyond visual line of sight from the UAV to a server, allowing the pilot to check the video in real time.

In the wireless mesh network of conventional example 1 shown in FIG. 1, a first fixed node 111, a second fixed node 112, a third fixed node 113, and an n-th fixed node 114 are placed on the ground along the flight path of a mobile node 101 including a UAV body. A server 121 is also connected to the first fixed node 111 by wire. In the following explanation, the number of hops is defined as the number of nodes that communicated minus 1, such that one hop is defined as wireless communication between two nodes, regardless of whether the nodes are mobile or fixed, and two hops is defined as wireless communication between three consecutive nodes.

When the mobile node 101 is flying above the first fixed node 111, one-hop communication takes place between the mobile node 101 and the first fixed node 111, and data is transferred to the server 121 via the shortest route. Next, when the mobile node 101 moves above the second fixed node 112 and is no longer able to communicate directly with the first fixed node 111, two-hop communication takes place from the mobile node 101 to the first fixed node 111 via the second fixed node 112. Similarly, when the mobile node 101 moves above the third fixed node 113, the communication becomes three-hop communication, and when the mobile node 101 moves above the nth fixed node 114, the communication becomes n-hop communication. As the number of hops increases, the throughput decreases and the delay time also increases.

Figure 2 shows an example of the configuration of a wired and wireless mesh network (hereinafter referred to as a "wired/wireless mesh network") as conventional example 2. In the system of Figure 2, similar to the system of Figure 1, a first fixed node 211, a second fixed node 212, a third fixed node 213, and an nth fixed node 214 are placed on the ground along the flight path of a mobile node 201 including the UAV body, and a server 221 is further connected to all the fixed nodes by wires.

Therefore, at first glance, it appears that one-hop communication between the mobile node 201 and the server 221 can be achieved regardless of which fixed node the mobile node 201 is wirelessly connected to. However, because a loop is formed between the wired network and the wireless network, communication over a wired line is permitted only between one of the fixed nodes and the server 221, and wired lines to all remaining fixed nodes are blocked.

In other words, only the connection between the first fixed node 211 and the server 221 is not blocked. Therefore, when the mobile node 201 moves from above the first fixed node 211 to above the second fixed node 212 and is no longer able to communicate directly with the first fixed node 211, two-hop communication is performed from the mobile node 201 to the first fixed node 211 via the second fixed node 212, and the node is connected to the server 221.

In addition, since loop detection is generally performed by broadcasting loop detection packets, if the network is large or the communication load is heavy, it may take a long time to detect the loop. Furthermore, since the loop detection function itself increases the network load, it is best to build a network that does not form loops as much as possible.

Figure 3 shows another example of the configuration of a wired/wireless mesh network as conventional example 3. In the system of Figure 3, similar to the system of Figure 2, a first fixed node 311, a second fixed node 312, a third fixed node 313, and an nth fixed node 314 are placed on the ground along the flight path of a mobile node 301 including the UAV body, and a server 321 is further connected to all the fixed nodes by wires. However, wireless communication between the fixed nodes is prohibited.

Specifically, the MAC address of the second fixed node 312, the MAC address of the third fixed node 313, and the MAC address of the nth fixed node 314 are specified for the first fixed node 311, and connections to these fixed nodes are prohibited by MAC address filtering. This allows the first fixed node 311 to wirelessly connect only to the mobile node 301. Similar settings are made for the other fixed nodes.

The system in FIG. 3 also differs from Conventional Example 2 in that the maximum number of hops is set to 1. This can be achieved by setting the TTL (Time To Live) in the IP header of the packet to be sent to 1. When a packet is transferred from wireless node to wireless node, 1 is subtracted from the TTL, immediately making it 0. If the packet has not reached the destination node at this point, it is discarded, resulting in a wireless network with a maximum number of hops of 1. In other words, since the path shown by the dashed line in FIG. 3 can only be passed once, no loop is formed even if the wireless network and wired network are connected by multiple nodes, and there is no blocking between any fixed node and server 321.

In this way, if the wireless network exceeds a certain number of hops, packets are forcibly discarded to prevent the formation of a loop, thereby preventing blocking on the wired network side. Furthermore, by setting a route so that the wired network can always be reached within a certain number of hops from the wireless network, communication is always carried out via the shortest route regardless of which fixed node the mobile node 301 is connected to. This makes it possible to build a network that provides high throughput and low latency.

However, such mesh networks have the problem that, due to the constraints on the number of hops on the wireless network side, the smaller the allowable number of hops, the more difficult it is to expand the communication area to locations farther away from the wired network. Also, the larger the allowable number of hops, the more difficult it is to place fixed nodes while considering the number of hops so as not to form a loop. Also, for example, if the mobile node 301 can only communicate with the second fixed node 312, if a failure occurs between the second fixed node 312 and the server 321, the mobile node 301 will be unable to access the server 321, and so on, thus sacrificing the high fault tolerance that is inherent to mesh networks.

The present invention was made in consideration of the above-mentioned conventional circumstances, and aims to provide control in a wired/wireless mesh network having mobile nodes and fixed nodes so that loops are not formed between the wireless network and the wired network.

In order to achieve the above object, a mesh network system, which is one aspect of the present invention, is configured as follows. That is, the mesh network system according to the present invention is a mesh network system having a mobile node and multiple fixed nodes as wireless nodes forming a wireless network, where at least two of the multiple fixed nodes are also connected to a wired network, and the fixed nodes connected to both the wireless network and the wired network are characterized in that they discard the received packet when the sender receives a packet from the wireless network side from the wired network side, or when the sender receives a packet from the wired network side from the wireless network side.

Here, in the mesh network system according to the present invention, a fixed node connected to both a wireless network and a wired network can be configured to perform address resolution between itself and a network device connected to the wired network, and to notify the mobile node of the address resolution result.

In addition, in the mesh network system according to the present invention, when a mobile node receives address resolution results from multiple fixed nodes, the mobile node can be configured to designate the fixed node with the lowest communication cost as the final node in the wireless network.

A communication control method according to another aspect of the present invention is configured as follows. That is, in a communication control method in a mesh network system having a mobile node and multiple fixed nodes as wireless nodes forming a wireless network, two or more of the multiple fixed nodes are also connected to a wired network, and a fixed node connected to both the wireless network and the wired network discards a received packet when the sender receives a packet from the wireless network side from the wired network side, or when the sender receives a packet from the wired network side from the wireless network side.

According to the present invention, in a wired/wireless mesh network having mobile nodes and fixed nodes, it is possible to control the network so that loops are not formed between the wireless network and the wired network.

FIG. 1 is a diagram illustrating a configuration example of a wireless mesh network according to a first conventional example. FIG. 11 is a diagram showing a configuration example of a wired/wireless mesh network according to a second conventional example. FIG. 11 is a diagram showing a configuration example of a wired/wireless mesh network according to a third conventional example. 1 is a diagram showing an example of the configuration of a wired/wireless mesh network according to an embodiment of the present invention; 5 is a diagram showing an example of a route in the wired/wireless mesh network of FIG. 4.

An embodiment of the present invention will now be described with reference to the drawings.
4 shows an example of the configuration of a wired/wireless mesh network according to an embodiment of the present invention. The wired/wireless mesh network of this example includes a mobile node 401, a first fixed node 411, a second fixed node 412, a third fixed node 413, an nth fixed node 414, and a server 421. The server 421 is connected by wires to a plurality of fixed nodes other than the second fixed node 412, i.e., the first fixed node 411, the third fixed node 413, and the nth fixed node 414. In this example, a UAV is assumed as the mobile node 401, but any mobile radio, such as a radio mounted on a train or automobile, or a radio portable by a person, can be applied.

Here, the routes between the mobile node 401 and all the fixed nodes 411 to 414 are determined by the spanning tree protocol. In addition, the first fixed node 411, the third fixed node 413, and the nth fixed node 414, which are connected to the wired network, are equipped with a MAC address filtering function to prevent the formation of loops between the wireless network and the wired network.

Specifically, when a packet having a source address belonging to a wireless network is received from the wired network, the packet is discarded. Similarly, when a packet having a source address belonging to a wired network is received from the wireless network, the packet is discarded. This makes it possible to avoid broadcast storms without blocking the route on the wired network side. Note that fixed nodes connected to a wired network have different MAC addresses, i.e., source addresses, on the wired and wireless sides.

In addition, the first fixed node 411, the third fixed node 413, and the nth fixed node 414, which are connected to the wired network, periodically communicate directly with the server 421, and also have the function of monitoring the address resolution status by, for example, sending a ping at regular time intervals, and reporting the results to the mobile node 401.

Next, the operation of the wired/wireless mesh network of this example will be explained using Figure 5. Figure 5 has the same basic configuration as Figure 4, and shows how the fixed nodes that can wirelessly connect to the mobile node are limited depending on the location of the mobile node. Also, in Figure 5, the metric value between nodes is represented by a numerical value in parentheses. For example, the metric value between the first fixed node 411 and the second fixed node 412 is 4, as shown by (4) in the figure.

In a mesh network, the communication cost is generally calculated for each route, and the route with the lowest communication cost is selected. For example, in IEEE 802.11s, which is the standard for wireless LAN mesh networks, the following metric value Ca is calculated as the communication cost.
Ca=[0+Bt/r]/(1-ef)
Here, 0 is the channel access overhead, Bt is 8192 (the bit size of the test frame), r is the data rate (Mbit/sec), and ef is the frame error rate. The smaller the metric value, the better the wireless quality of the route.

This metric value can be considered as airtime, that is, the time required to wirelessly transmit a given amount of data. Furthermore, for a route that involves multiple hops, the sum of the metric values between each node can be considered as the metric value of the entire route. For example, when mobile node 401 communicates with first fixed node 411, there is a one-hop route and a two-hop route that passes through second fixed node 412, with the metric value of the one-hop route being 7 and the metric value of the two-hop route being 5 + 4 = 9. In this case, the one-hop route with the smaller metric value will be selected. Note that in FIG. 5, the metric value is set to a value one digit smaller to simplify the explanation.

Furthermore, FIG. 5 also clearly shows that the route starting from the mobile node 401 in the wired/wireless mesh network is constructed using a spanning tree protocol, and that some of the routes are blocked. Specifically, the route between the first fixed node 411 and the second fixed node 412 is blocked, and the route of the tree topology is determined so that a loop is not formed when the mobile node 401 communicates with other fixed nodes in the wired/wireless mesh network.

As shown in FIG. 5, the mobile node 401 is flying midway between the first fixed node 411 and the second fixed node 412, and the mobile node 401 is in a state in which it can directly communicate (one hop) only with the first fixed node 411 and the second fixed node 412. In addition, the first fixed node 411 and the second fixed node 412 are blocked by the spanning tree protocol, and the second fixed node 412, the third fixed node 413, and the nth fixed node 414 are connected by a single route.

In this state, when a broadcast packet is sent from the mobile node 401, the broadcast packet reaches the first fixed node 411 and the second fixed node 412. The broadcast packet that reaches the first fixed node 411 then reaches the server 421, the third fixed node 413, and the nth fixed node 414 via the wired network. However, the third fixed node 413 and the nth fixed node 414 discard packets received from the wired network with a source address of the mobile node 401 by MAC address filtering, preventing a broadcast storm.

On the other hand, the broadcast packet that arrives at the second fixed node 412 from the mobile node 401 reaches the third fixed node 413. The broadcast packet that arrives at the third fixed node 413 then reaches the server 421, the first fixed node 411, and the nth fixed node 414 via the wired network. However, the first fixed node 411 and the nth fixed node 414 discard packets received from the wired network with a source address of the mobile node 401 by MAC address filtering, preventing a broadcast storm.

The broadcast packet that reaches the third fixed node 413 passes through the nth fixed node 414, and then reaches the server 421, the first fixed node 411, and the third fixed node 413 via the wired network. However, the first fixed node 411 and the third fixed node 413 discard packets received from the wired network with a source address of the mobile node 401 by MAC address filtering, preventing a broadcast storm.

The same applies when a broadcast packet is sent from server 421. For example, assume that a broadcast packet that arrives at third fixed node 413 from server 421 reaches nth fixed node 414 via a route on the wireless network side. In this case, nth fixed node 414 discards the packet received from the wireless network side with a source address of server 421 by MAC address filtering, thereby preventing a broadcast storm.

In this way, the wired/wireless mesh network system of this example is configured to perform separate MAC address filtering on the wireless network side and the wired network side. This makes it possible to build a wired/wireless mesh network in which wireless and wired networks are integrated without causing blocking between multiple fixed nodes and the server 421, even if the wireless network and the wired network are connected by multiple nodes to form a loop.

When the mobile node 401 first communicates with the server 421 and broadcasts an ARP request packet from the mobile node 401 to the server 421, the ARP request packet reaches the server 421 via multiple possible communication routes. The server 421 then unicasts an ARP response packet to the mobile node 401, thereby performing address resolution and enabling communication between the mobile node 401 and the server 421. However, because the address resolution information is overwritten by the last ARP response packet received out of the multiple ARP response packets, the shortest route is not necessarily selected. Also, if communication between the mobile node 401 and the server 421 becomes impossible, a processing delay occurs due to address resolution when the ARP request packet is sent again.

Therefore, in order to select the best route while minimizing processing delays, it is desirable to perform address resolution using a method that does not broadcast an ARP request packet from the mobile node 401. Therefore, in this example of a wired/wireless mesh network, address resolution between the mobile node 401 and the server 421 is performed using the method described below.

First, the first fixed node 411, the third fixed node 413, and the nth fixed node 414, which are connected to the wired network, periodically send pings to the server 421. If address resolution has not been achieved, these fixed nodes broadcast an ARP request packet from the wired network side, that is, with the MAC address of the wired side set as the source address.

When the first fixed node 411 broadcasts an ARP request packet to the wired network, the ARP request packet reaches the server 421, the third fixed node 413, and the nth fixed node 414. The server 421 then returns an ARP response packet to the first fixed node 411, and address resolution is performed. On the other hand, for example, assume that the ARP request packet that arrived at the third fixed node 413 reaches the nth fixed node 414 via the wireless network. In this case, the nth fixed node 414 discards packets received from the wireless network whose source address belongs to the wired network by MAC address filtering, thereby preventing a broadcast storm.

In this way, even if a fixed node connected to the wired network broadcasts, MAC address filtering prevents a loop from being formed. This makes it possible to maintain an unblocked state between the fixed nodes 411-414 and the server 421. Furthermore, except in cases where a failure occurs in the wired network that makes it impossible to communicate with the server 421, thereafter, arrival confirmation is performed using only periodic pings without broadcasting an ARP request packet.

If no particular fault occurs in the wired network, the server 421 returns a ping response to each of the first fixed node 411, the third fixed node 413, and the nth fixed node 414. Upon receiving the ping response, the first fixed node 411, the third fixed node 413, and the nth fixed node 414 send a control packet with the MAC address of the server 421 added to the mobile node 401 to notify it that it can communicate with the server 421. Conversely, if a fault occurs in the wired network and no ping response is obtained from the server 421, no control packet is sent to the mobile node 401.

When the mobile node 401 receives the control packet, it obtains the MAC address of the server 421 added to the control packet to perform address resolution, and recognizes the fixed node that can communicate with the server 421 from the source address of the control packet.

Meanwhile, the mobile node 401 is aware of the metric values when communicating with each node during route search within the wired/wireless mesh network. For example, when calculated in accordance with the metric values for each route shown in FIG. 5, the metric value with the first fixed node 411 is 7, the metric value with the second fixed node 412 is 5, the metric value with the third fixed node 413 is 5+3=8, and the metric value with the nth fixed node 414 is 5+3+6=14. In other words, it can be determined that the fixed node with the best wireless quality among the fixed nodes connected to the server 421 by wire is the first fixed node 411.

Therefore, when communicating with the server 421, the mobile node 401 can select the best route by making the first fixed node 411 the final node in the wired/wireless mesh network. Also, if that route becomes unavailable, the mobile node 401 can instantly switch to the best route by selecting the third fixed node 413, which has the second lowest metric value, as the final node in the wired/wireless mesh network.

In this example, the second fixed node 412 is the only fixed node not connected to the wired network, but generality is not lost as long as the condition that multiple fixed nodes are connected to the wired network is met. In this example, a ping is sent from the fixed node to the server, but this is not limited to ping. The server may also be configured to individually check whether communication with multiple fixed nodes connected to the wired network is possible. In this example, a MAC address filtering function provided in the fixed node connected to the wired network is used to avoid loops, but a network switch with a function to avoid loops may be added.

As described above, the wired/wireless mesh network system of this example has a mobile node 401 and multiple fixed nodes 411, 412, 413, 414 as wireless nodes that form a wireless network, and at least two of the multiple fixed nodes 411, 413, 414 are also connected to a wired network, and the fixed nodes 411, 413, 414 connected to both the wireless network and the wired network are configured to discard received packets when the sender receives a packet from the wireless network side from the wired network side, or when the sender receives a packet from the wired network side from the wireless network side. In this way, by configuring the wireless network side and the wired network side to perform separate address filtering, it is possible to control so that a loop is not formed between the wireless network and the wired network.

In addition, in the wired/wireless mesh network system of this example, the fixed nodes 411, 413, and 414 connected to both the wireless network and the wired network are configured to perform address resolution with the server 421 connected to the wired network and notify the mobile node 401 of the address resolution result. Therefore, address resolution can be performed without broadcasting an ARP request packet from the mobile node 401, and communication can be established between the mobile node 401 and the server 421. In this example, the server 421 is given as an example of a network device according to the present invention, but the address resolution method according to the present invention can be applied to address resolution with any other network device connected to a wired network.

In addition, in the wired/wireless mesh network system of this example, when the mobile node 401 receives address resolution results from multiple fixed nodes 411, 413, and 414, it is configured to designate the fixed node with the lowest communication cost as the final node in the wireless network. Therefore, the mobile node 401 can select the best route when communicating with the server 421.

As described above, a mesh network can be configured using both wireless and wired networks, with the wireless network side constructing routes using a normal spanning tree protocol, and the wired network side performing different address filtering on the wireless and wired sides to prevent loops from forming (i.e., to prevent blocking from occurring), while determining the final node of the wireless mesh network according to the best route, making it possible to construct a network with high throughput, low latency, high-speed route switching, and high fault tolerance.

Although the embodiments of the present invention have been described above, these embodiments are merely illustrative and do not limit the technical scope of the present invention. The present invention can take various other embodiments, and various modifications such as omissions and substitutions can be made without departing from the gist of the present invention. These embodiments and their modifications are included in the scope and gist of the invention described in this specification, etc., and are included in the scope of the invention described in the claims and their equivalents.

The present invention can be provided not only as the devices described above or as systems composed of these devices, but also as methods executed by these devices, programs for implementing the functions of these devices using a processor, and storage media for storing such programs in a computer-readable format.

The present invention can be used in a wired/wireless mesh network system that configures a mesh network using both wireless and wired connections.

101, 201, 301, 401: mobile node, 111, 211, 311, 411: first fixed node, 112, 212, 312, 412: second fixed node, 113, 213, 313, 413: third fixed node, 114, 214, 314, 414: nth fixed node, 121, 221, 321, 421: server

Claims (3)

In a mesh network system having a mobile node and a plurality of fixed nodes as wireless nodes forming a wireless network,
Two or more of the fixed nodes are also connected to a wired network;
a fixed node connected to both the wireless network and the wired network discards a received packet when a source of the packet is received from the wired network side, or when a source of the packet is received from the wireless network side ,
Furthermore, a mesh network system is characterized in that a fixed node connected to both the wireless network and the wired network performs address resolution between itself and a network device connected to the wired network, and notifies the mobile node of the address resolution result . 2. The mesh network system according to claim 1 ,
A mesh network system characterized in that, when the mobile node receives address resolution results from a plurality of fixed nodes, the mobile node designates the fixed node with the lowest communication cost as the final node in the wireless network. A communication control method in a mesh network system having a mobile node and a plurality of fixed nodes as wireless nodes forming a wireless network, comprising:
Two or more of the fixed nodes are also connected to a wired network;
a fixed node connected to both the wireless network and the wired network discards a received packet when a source of the packet is received from the wired network side, or when a source of the packet is received from the wireless network side ,
Furthermore, a communication control method characterized in that a fixed node connected to both the wireless network and the wired network performs address resolution between itself and a network device connected to the wired network, and notifies the mobile node of the address resolution result .

Images