Method of creating wireless ad hoc network in communication-denied environments using robotic systems and the network created by this method

Abstract

A method of creating wireless ad hoc network in communication-denied environments using robotic systems. The method comprises step of using a robotic system configured to carry a plurality of communication nodes to deploy one node at the starting point suitable for entering the area where the network shall be established. Further, the method comprises navigating the mobile robotic vehicle to following point to deploy static node or dynamic node, while upon navigating to this following point, checking that the distance from this following point to any of the preceding node(s) deployed does not extend the range in which the signal from any of the preceding node(s) is secured to establish the wireless connection through any of those preceding node(s). The method is carried out by repeating the second step until reaching the final destination point of the network and deploying the node to establish the wireless communication in the selected area. The invention also relates to a wireless ad hoc network created by this method. The network utilizes the flood routing approach and constrained randomized broadcasts to propagate messages through the ad hoc network. Such network is particularly suitable for controlling mobile robotic vehicles, in particular in multi-robot teaming in an autonomous exploration in communication-denied environments.

Description

Title: Method of creating wireless ad hoc network in communication-denied environments using robotic systems and the network created by this methodTECHNICAL FIELD

[0001] The present invention relates to the field of wireless communication systems, specifically to a Method of creating wireless ad hoc network in communication-denied environments using robotic systems and the network created by this method and the deployable wireless ad hoc network system with dynamic topology designed for multirobot systems in communication-denied environments without physical communication or wireless infrastructure, such as non-trivial underground and complex subterranean environment.BACKGROUND ART

[0002] Communication is crucial in multi-robot systems, enabling efficient coordination and cooperation among team members. However, environments without existing communication infrastructure, such as underground caves, mines, or tunnels, necessitate developing rapidly deployable wireless network solutions to maintain situational awareness among multi-robot teams representing communication networks with highly dynamic topology.

[0003] The following patent documents represent the known solutions related to this invention. Document EP1480387A1 describes an ad hoc communication network protocol using an ad hoc network coordinator, locally stored complete communication routes, and identification of other devices for message forwarding. The solution relies on the network topology with a network controller that controls a web of network devices, thus requiring a dedicated node coordinating the routing.

[0004] The routing is based on incremental source routing, where the address field within the message might increase and thus limit a payload part of the communication message in packet-driven communication networks with minimal data payloads, such as IEEE 802.15.4 2006.[0005} An explicit network coordinator might be replaced by the so-called flooding routing in an ad hoc multicast network described in the document EP2068500A2. However, the role of the communication node is determined dynamically, and each node needs to include a local topology processor, network topology processor, and local multicast processor. The network topology processor maintains network topology databases or routing tables for establishing dedicated backbone routing links. The multicast processor is designed for the role of a supernode to maintain the backbone multicast registration table.

[0006] Resource demands on the nodes can be decreased by flood-relaying, which requires low computational power and memory requirements. Hybrid flood-relying and routing are presented in document EP3267729A1 with three types of network nodes. Namely, the edge nodes, the relay nodes, and the high-end nodes. The edge nodes serve as the source and destination nodes of the network packets. The relay nodes are capable of flood-routing the packet, while the high-end nodes convert the propagation-unspecified packet to a routing-propagation packet. As such, the solution requires performing route discovery and creating and maintaining the routing table, which brings communication overhead and suffers from poor performance with large topology changes in the ad hoc networks.

[0007] The proposed deployable wireless ad hoc network system offers a simple, scalable, and cost-effective solution for multi-robot team coordination in challenging environments. Its versatile nature makes it suitable for various applications, including underground exploration, disaster management, vehicular networks, and Internet of Things (loT) distributed sensory networks. The solutions to ad hoc wireless networks need to deal with a priory unknown topology that can be either dynamically discovered, which brings communication overhead, or precise synchronization of the modules can be used, which either requires communication overhead or dedicated hardware adjustments. In our approach, we utilize a simple flood routing approach to propagate the systolic messages through the network.[0D08]The proposed deployable wireless ad hoc network system eliminates complex hardware components and software communication overhead required by the art.Hence it represents an affordable alternative to complex systems while being scalable and sufficiently reliable.

[0009] Within the multi-robot team, the proposed Hood routing approach together with the used broadcasting allows effective sharing of the state information about the individual mobile robotic vehicles and possibly their commanding, which improves the overall performance of the robotic team in the execution of the designated mission.SUMMARY OF THE INVENTION

[0010] The herein-described deployable wireless ad hoc network system addresses the need for wireless communication in otherwise communication-denied environments. It includes static nodes and dynamic nodes, a routing protocol, and a transmission scheduling policy that accommodates dynamic network topology changes caused by robot movement and environmental characteristics. The system resolves wireless access arbitration conflicts to ensure effective media access.

[0011] The described system is suitable for deployments with demands on continuously transmitting relatively short messages within the network to support information sharing between team members, such as mobile robotic vehicles. Thus, the delivery of every message is not critical, as in the target deployments, messages are repeatably replaced by new ones with more recent information to be shared.

[0012] The network comprises of communication nodes with similar functional capabilities of wireless receiving, transmitting, and carrier sensing, and sufficient computational power to handle the herein described transmission scheduling policy. In the disclosed embodiment, all the nodes are homogeneous in functionality and hardware and software implementation.

[0013] All communication nodes within the network act as relays. Further, each communication node can be a source of new messages in the network, in practice, and as shown by the disclosed exemplary embodiments, the new messages areemanated to the network by mobile robotic vehicles to maintain their situational awareness and receive commands.

[0014] During the building of the ad hoc network in the environment, the mobile robotic vehicles are expected to drop the nodes in the environment to extend the communication range. These dropped nodes are referred to as static nodes while the nodes attached to the robotic vehicles are referred to as dynamic nodes. It is expected that the static nodes act as static relays within the dynamically created network. The exact position of the static nodes can be accidentally changed due to environment dynamics or intentionally changed by the mobile robotic vehicles throughout the mission.

[0015] Different deployment strategies of the static nodes in the environment are part of the presented invention,

[0016] Messages might get lost mainly due to the signal propagation characteristics of the deployment environment and a hidden terminal problem when two nodes not within the communication range start to broadcast simultaneously toward a node within the communication range of both nodes. However, the described network creation and wireless media access mechanism allows sufficient information sharing within the network for efficient coordination in scenarios, such as multi-robot teaming in an autonomous exploration of otherwise communication-denied, a priory unknown environments.

[0017] In the disclosed embodiment, the system employs a packet-based broadcast- enabled wireless communication approach suitable for harsh environments. A transmission scheduling policy is implemented to mediate packet loss through constraint-randomized arbitration of wireless transmission access. The architecture of the communication system comprises communication modules equipped with wireless transceivers and computational power. Data synchronization is achieved through broadcast and flood-routing-based protocol, ensuring robustness and supporting information sharing in highly dynamic network topology. The transmission scheduling policy utilizes asynchronous transmission with constraint randomized delays,eliminating the need for additional communication overhead, time synchronization among modules, or building and maintaining information about the network topology.

[0018] According to this invention, the method and the deployable wireless ad hoc network system presented herein provide a rational approach to wireless communication in oommunication-denied environments. Its unique architecture, transmission scheduling policy, and adaptability to dynamic network topology changes make it an innovative solution for multi-robot teaming and situational awareness building in scenarios with an absence of physical communication or wireless infrastructure. The main assumption behind the invention is that all the messages do not have to be delivered, which means that a certain drop-out of messages is expected and practically inevitable within the proposed invention.

[0019] While the invention has been described in connection with specific embodiments, variations, and modifications are apparent to those skilled in the art. Therefore, the scope of the invention should not be limited to the described embodiments but rather defined by the appended claims and their equivalents.

[0020] The method according to this invention comprises following basic steps:A) using a robotic system configured to carry a plurality of communication nodes to deploy one node at the starting point suitable for entering the area where the network shall be established. The wireless network comprises at least three nodes, wherein each node is adapted to receive and retransmit received message to establish wireless communication. Further, at least one of these nodes is actively emanating messages into the network according to defined transmission schedule (such node is further defined as dynamic node);B) navigating the mobile robotic vehicle to following point to deploy node, while upon navigating to this following point, checking that the distance from this following point to any of the preceding node(s) deployed does not extend the range in which the signal from any of the preceding node(s) is secured to establish the wireless connection through any of those preceding node(s);C) repeating step B) until reaching the final destination point of the network and deploying the node to establish the wireless communication in the selected area.

[0021] Preferably, in step B), checking the range in which the signal from any of the preceding nodes is secured is performed by sending control messages through the network and / or by computing the estimated safe communication range between the node to be deployed and any of the preceding deployed nodes.

[0022] In another preferred variant, in step B), navigating the robotic system to the point where communication interference with at least a plurality of previously deployed nodes is avoided and deploying the node in such point.

[0023] It is also advantageous that the nodes are power activated during their deployment using the etectro-magnetioal principle.

[0024] Furthermore, the steps of the method can preferably be carried out using an unmanned aerial vehicle UAV and / or unmanned ground vehicle UGV and / or unmanned underwater vehicle UUV.BRIEF DESCRIPTION OF THE DRAWINGS

[0025] Figure 1 shows a schematic illustration of an exemplary ad hoc network with communication ranges and connections achievable via broadcast outlined. Nodes depicted as circles denoted NID3, NID4, NID5, and NID6 represent static nodes deployed in the environment, typically by the mobile robotic vehicles. Dynamic nodes depicted as triangles are denoted NID1 and NID2. The gray disks denote communication range coverage in ideal conditions, and the black straight-line segments connecting the nodes denote existing communication links between the nodes. Note that due to different platform specific reasons and environmental factors, the communication range of individual nodes may not be similar as depicted in the figure.

[0026] Figure 2 shows schematic illustration of the network topology for the first embodiment example. Static nodes NID11, NID12, and NID13 were deployed of the mobile robotic vehicle. Dynamic nodes NID1, NID2, NID3, and NID4 were mounted on the mobile robotic vehicles operating in the environment. The gray area shows theunderground tunnel outline and the black segments connecting the nodes denote existing communication links between the nodes. Note that the topology has been created dynamically, and the shown topology is the final topology of the network at the end of one hour mission.DETAILED DESCRIPTION OF THE INVENTION

[0027] The network consists of individual nodes. Each node has a wireless receiving, transmitting, and Carrier Sensing Multiple Access (CSMA) capabilities. Each node has a computational and memory capacity to receive and temporarily store several messages broadcasted via the network. Each node has a power source. Each node has a unique identification number within the network, denoted the node ID (NID).

[0028] The network nodes are either dynamic or static. The dynamic nodes are the mobile robotic vehicles, which are expected to be dynamic with respect to the environment and other network nodes, and thus experience often network topology changes with respect to them. The static nodes are expected to be statically deployed in the environment. The static nodes can be deployed manually or automatically by mobile robotic vehicles. The exact position of the static nodes in the environment can be accidentally changed due to environment dynamics or intentionally changed by a mobile robotic vehicle.

[0029] Each node serves as a relay in the network. The communication within the network is message-based, where the message size is according to the used transport signal and the underlying wireless technology. Broadcasting of a single message is atomic.

[0030] Every communication in the network is broadcast to all nodes within the wireless communication range.

[0031] Nodes may emanate new messages into the network. Only a single new message within the defined transmission period may be emanated. The messages propagate through the network by flood-routing protocol. Each network node broadcasts the message If it has not done it before. Hence, each node maintains a listof messages (limited in size) that have been transmitted. Within each message, two data fields are reserved for the identification of the message. Namely, the NID and Message ID (MID).

[0032] The NID identifies the source of the message, which is typically a dynamic node. The MID encodes the gradually incremented message counter from the source. Whether the message has been retransmited or not is determined based on the combination of the NID and MID. Every broadcast of the message is scheduled with a constrained randomized delay within the defined transmission period. The CSMA is employed to sense if the wireless medium is available for transmission at the scheduled time. The message is broadcasted at the earliest after a randomized delay passes and the wireless medium is available.

[0033] The minimum bitwise representation of the NID equals the bit length of the binary representation of the number of nodes in the network. Further, the maximum lifespan of a single message in the flood network equals the number of all the nodes in the network multiplied by the predefined transmission period, as each node may retranslate the message with the predefined transmission period only once. Hence, the minimum bitwise representation of the MID equals the bit length of the binary representation of double the number of all the nodes in the network to distinguish between new and repeating messages.

[0034] The predefined transmission period is common for the whole network and is set by a user before the field deployment. The predefined transmission period is calculated with respect to the expected drop-out rate of the network, given the number of all the nodes in the network as follows. Each broadcast of a message takes constant broadcast time. Hence, the transmission period can be approximately divided by broadcast time into individual broadcast slots. Although there is no prior time synchronization in the network, the CSMA capability of the nodes ensures that two nodes do not start broadcasting simultaneously unless they are outside of their mutual communication range.

[0035] The number of broadcasts of a single message emanated into the fully connected flood routing-based network equals number of all the nodes within thenetwork. Hence, the overall number of broadcasts within each transmission period throughout the whole network during its continuous operation equals the number of all the nodes within the network times the number of expected emanated messages within a single transmission period.

[0036] Hence, the probability of two nodes starting to broadcast simultaneously in the network, given the number of broadcast slots, can be mathematically approximated using the generalized birthday problem. It is used to determine the expected drop-out rate and the probability of the message delivery, and thus the quality of the service of the network. Note that the mathematical model applies to the network topology of a fully connected graph; in less connected network topologies, the performance is inherently better.

[0037] Mobile units, such as autonomous mobile robotic vehicles, can carry a set of nodes in addition to their own node. During the mission, the robots deploy the nodes according to the build situational awareness and deployment strategy. Since the carried nodes might interfere with each other, the nodes are power activated during their drop-off using the electro-magnetical principle. Such an automated activation improves the performance at the beginning of the mission but does not affect the performance in the long run.

[0038] Example embodiments of the system include the deployment of the system in the communication-restricted subterranean environments of caves, mines, and tunnels. In the first exemplary embodiment, the system was used in a coal mine-like subterranean environment. The deployed system consisted of twenty nodes. The hardware of the network nodes was based on an off-the-shelf low-cost communication transceiver modules of 915 Mega Hertz industrial, scientific, and medical communication frequency, with packet-oriented messages. For message processing, an off-the-shelf low-cost RISC architecture-based microcontroller with 48 Mega Hertz CPU frequency and 6 kilo Bytes of RAM has been used. The node Is powered by a single-cell battery of 18650 form factor and 10.08 Watt hour battery capacity that provides continuous operation of the module for 11 hours. The module uses a quarter- wavelength-long whip antenna. The whole module is enclosed in a 3D printed, sealed enclosure of 30 x 30 x 100 milli meter size that is IP64 protected while being RFtransparent. An estimated price of a single node was at the time of development about 12 USD.

[0039] The used message size, equaling the packet size of the utilized wireless technology, was 64 bytes, with estimated 57 transmission slots per second. Hence, the predefined transmission period was set to 1.0 second.

[0040] Three mobile robotic vehicles and a single base station operated by a human operator were used as the dynamic nodes, emanating new messages to the network periodically with the given transmission period. In the presented embodiment, the stability of the connection between the robotic team and the human operator was critical for the success of the mission. A single mobile robotic vehicle was equipped with an autonomous node-dropping device that dropped a node every 60 meters of the traveled distance. Altogether three nodes were dropped. The resulting nodes topology is depicted in Fig. 2. The mobile robotic vehicles were moving along the created topology throughout their mission. The observed average drop rate for the whole team of robots within a one-hour mission was about 30.6 percent, sufficient for team monitoring and coordination.

[0041] The second exemplary embodiment consisted of eight communication nodes of the same type as in the first exemplary embodiment. There were six static nodes and two dynamic nodes, emanating messages to the network. The transmission period was set to 0.1 second. The nodes were manually deployed in a cave environment, forming a chain topology. The relay nodes covered a 280 meters long part of the unstructured cave environment. The observed drop rate for the communication between the source nodes on the opposite sides of the relay chain was measured to be about 28.7 percent.INDUSTRIAL UTILIZATION

[0042] The disclosed invention is suitable for deployment in scenarios requiring rapid and effective reactions in communication-restricted environments. These include but are not limited to exploration, search, or inspection in underground caves, mines, or tunnels by robot or human-robot teams as presented in the herein-presentedembodiment of the invention. Further suitable deployments include rapidly creating communication infrastructure in areas hit by natural disasters. The creation of the wireless network using a robotic system can take advantage of the situation awareness built by the robots to predict signal propagation using data-driven methods of machine learning and artificial intelligence techniques to satisfy the optimality constraints of building communication chains with a communication connectivity graph of the static nodes with a maximum of the degree two; thus, having only up to two neighborhood nodes within the communication range for each static node.

Claims

AMENDED CLAIMS received by the International Bureau on 24.10.24CLAIMS amended under ART 19 PCT, 24.10.20241. A wireless ad hoc network for communication-denied environments, wherein the network is created using robotic systems and using a method comprising steps:A) using a robotic system configured to carry a plurality of communication nodes comprising at least three nodes adapted to receive and retransmit received message to establish wireless communication, such nodes further defined as static nodes, of which at least one node is actively emanating messages into the network according to defined transmission schedule, such node further defined as dynamic node, to deploy one node at the starting point suitable for entering the area where the network shall be established;B) navigating the mobile robotic vehicle to following point to deploy static node or dynamic node, while upon navigating to this following point, checking that the distance, in the sense of communication accessibility using locally, online built situation awareness and data-driven communication signal propagation model, from this following point to any of the preceding node(s) deployed does not extend the range in which the signal from any of the preceding node(s) is secured to establish the wireless connection through any of those preceding node(s);C) repeating step B) until reaching the final destination point of the network and deploying the node to establish the wireless communication in the selected area, characterized in that the wireless ad hoc network comprises nodes that are either dynamic or static, further referred to as nodes, such that• nodes consisting of a transmitter, transmitting messages over a wireless medium;• a receiver, receiving messages over said wireless medium;• a local message processor, coupled with said receiver, maintaining the message queue of received and retransmitted messages;• a local stochastic schedule processor coupled with said transmitter and receiver, said local stochastic schedule processor maintaining and retransmitting received messages, capable of retransmitting received messages according to defined stochastic schedule;AMENDED SHEET (ARTICLE 19)• nodes that emanate messages into the network according to the defined schedule;• with minimum of three nodes, out of which at least one node emanates the messages into the network.

2. The wireless ad hoc network according to claim 1 , wherein the communication between the nodes employs use of message-based broadcast-enabled wireless signal where the packet size is selected according to the used transport signal and underlying technology and wherein each node maintains a list of messages that have been transmitted for a given time period which are limited in size and wherein the messages contain identifying data fields reserved for the identification of the message.

3. Use of the wireless ad hoc network according to any of claims 1 or 2 for controlling mobile robotic vehicles, in particular in multi-robot teaming in an autonomous exploration.AMENDED SHEET (ARTICLE 19)

Images