summary
The reliable operation of an elevator IoT system requires a mature and stable network architecture and a reliable data transmission mechanism. This paper focuses on the sensing layer of the elevator IoT system, introducing the structure of the sensing layer network and the roles of each component; it also describes the data transmission mechanism of the sensing layer, including data transmission methods and data packet routing principles.
Abstracts:
Thereliableoperation of theinternetofelevatorneedsmatureandstablenetworkstructureandreliabledatatransmitmechanism.
Keywords:
Elevator IoT WEN sensor layer routing data transmission
Keywords:
InternetofElevatorsWirelessElevatorNetworkSensorLayerRoutingDataTransmit
Introduction
Existing elevator IoT systems typically employ a single-point network architecture for the sensing layer, with each elevator acting as a separate network. This typically uses GPRS or 3G networks, resulting in higher costs and placing greater demands on IoT servers. Since a connection needs to be created for each elevator, this presents a significant challenge to server hardware and deployment architecture, making it unsuitable for widespread application.
The elevator IoT sensing layer network structure proposed in this paper adopts a distributed design method and uses WEN elevator-specific wireless technology to build wireless local area networks on a regional basis, and centralizes data to the data aggregation unit. This network structure is convenient for expansion and greatly improves cost-effectiveness.
This paper also introduces the data transmission mechanism of the sensor layer of the elevator Internet of Things. Based on actual needs, a variety of data transmission methods are designed, and the principle of minimum routing loss that conforms to the actual application of wireless networks is adopted as the basis for optimal routing link selection, which ensures the stability of the sensor layer WEN network operation and the reliability of data transmission.
II. Elevator Internet of Things Network Structure
2.1 Network Structure
The design principle of the elevator IoT sensor layer LAN architecture is "distributed control, centralized management." The network includes both wired and wireless communication methods. The design prioritizes high cost-effectiveness and high reliability, with the concept of building a LAN at the sensor layer and aggregating data into a data aggregation unit. WEN technology is used to construct the sensor layer wireless network.
The elevator main controller is the object of data acquisition. The data acquisition unit includes a wired communication unit and a radio frequency transmission unit. The data acquisition unit communicates with the elevator main controller using wired communication methods such as RS232, RS485 or CANBUS. The acquired data is transmitted to the WEN wireless network through the radio frequency transmission unit and reaches the data aggregation unit through one or more data relay units, as shown in Figure 1.
The WEN wireless network is the core of the sensor layer LAN. While the WEN architecture doesn't fully conform to the OSI 7-layer network model, it shares some common components, such as the physical layer, MAC layer, and network layer. The WEN network model unifies layers 4-7 into an application support layer, as shown in Figure 2. The physical layer is responsible for converting data packets into radio frequency bits, or vice versa. The MAC layer primarily provides network identification and network probing via beacons; it also provides point-to-point step acknowledgments. The network layer is mainly responsible for building the mesh network, including broadcasting data packets within the WEN network, determining routing links for unicast data, and ensuring the reliability of point-to-point data communication. The application layer is responsible for determining the sending and receiving of data packets based on the addresses or group numbers specified in the program.
Figure 1. Elevator IoT Sensing Layer Network Structure
Figure 2 WEN network model
2.2 Network Roles
In the elevator IoT system, the WEN wireless LAN network roles are mainly divided into four categories based on their functions: network creation unit, data aggregation unit, data relay unit, and data acquisition unit. The network creation unit is responsible for creating and activating the WEN network. Only the network creation unit can activate the WEN network. After successful network creation, the network creation unit can leave the network, as its participation is not required for the operation of the WEN network except for activation. However, the network creation unit determines whether nodes can join the network, as it is the only one with a trust center. To address this issue in the elevator IoT system, a network retention function is designed. After successful network activation, simply activating the network retention function on the data aggregation unit allows the network creation unit to completely leave the network.
As the second-in-command of the WEN network, the data aggregation unit is responsible for network maintenance, network management, and data aggregation, serving as the connection between the Internet and the local area network. Any data from the IoT server must first pass through the data aggregation unit, which retrieves the target address and then delivers it to the target node via the WEN network. Any data from the data collectors is ultimately collected by the data aggregation unit. After analysis, if it is network data, it is stored in the aggregation unit; if it is elevator status data, it is directly transmitted to the Internet.
The data relay unit has a relatively simple function: it is responsible for data routing and forwarding, delivering data packets from the parent node to the data aggregation unit through the optimal routing link.
The data acquisition unit is responsible for collecting data from the elevator's main controller. This data can be acquired via protocol or dry contacts. The acquired data is then transmitted to the data aggregation unit via the radio frequency transmission unit. Normally, the data acquisition unit is passive, only returning data when the IoT platform requests it. However, in the event of an elevator malfunction or entrapment, the data acquisition unit will proactively send an anomaly request data packet to the data aggregation unit, informing the IoT platform of the elevator malfunction.
III. Data Transmission Mechanism
The core focus of wireless networks is how to reliably transmit data between nodes. WEN networks can automatically plan the optimal route to ensure data exchange between nodes. WEN networks employ a standard data transmission mechanism, including three aspects:
(1) Data request, i.e., sending data
(2) Data confirmation, that is, the requested data receives a response.
(3) Data indication, i.e., receiving data
Data requests are initiated by the originating node, as shown in Figure 3. Data acknowledgment is a direct result of the data request; the originating node receives an acknowledgment packet each time it sends a data request to indicate whether the request was successful or failed; however, the arrival time of the data acknowledgment is not definite. Data requests can originate from multiple sources, including unicast with a response mechanism, unicast without a response mechanism, broadcast, and multicast.
Figure 3 Data transmission mechanism
3.1 Data transmission method
The main data transmission methods in WEN networks include unicast, multicast, and broadcast.
Unicast is primarily used for point-to-point communication. If a point-to-point communication is responded to, unicast will attempt to determine the final route three times. During this process, if the route is disrupted, such as by electromagnetic interference, obstacles, or severe weather, a new route detection mechanism will be activated. If the point-to-point nodes are neighboring nodes, meaning the communication distance is within radio frequency range, the route link can be determined after the first communication.
Broadcasting is a point-to-all communication method in a network, provided the receiving node is within the radio frequency radius specified by the sending node. There are three types of broadcasts, distinguished by their broadcast addresses: 0xffff for broadcasting to all nodes; 0xfffd for broadcasting to non-dormant nodes; and 0xfffc for broadcasting to routing nodes. Broadcast communication uses a packet relay method, where all nodes within the specified broadcast radius perform packet reception, address determination, and data delivery. This means a single packet may arrive at the same node multiple times. If the broadcast frequency is high, such as 10 times per second, it will inevitably cause network congestion, leading to packet loss, packet errors, and other anomalies.
Multicast is a special case of broadcast. It is a broadcast to a specific group of nodes. Any other node that receives multicast data will discard the data packet. If a node wants to send multicast data to a specific group, the data packet will reach the node identified by the specified group.
3.2 Packet Routing
The WEN network adopts a network structure. When data packets are transmitted between points, route detection is first required to find a route link. The source node can reach the designated target node through multiple steps. If the route link is damaged during the route detection process, such as being blocked by obstacles or power failure, the route link will be automatically re-planned and route detection will continue until a new route link is found, as shown in Figure 4.
Figure 4. Schematic diagram of data packet rerouting in a mesh network
In a mesh network, all nodes are equal, and any node in the network can discover routing links. During this process, nodes become potential routing links; this process is called route detection. Nodes maintain a route detection table while searching for the most efficient routing link. To find a routing link to a target node, the source node first broadcasts data packets, as it doesn't yet know the target node's location and can only search through broadcasting. On the routing link, route loss indicates the energy of each step. The source node broadcasts data in all directions, accumulating the route loss with each step. This determines the route loss of each link, and finding the one with the minimum loss locks in the optimal routing link. As shown in Figure 5, source node 3 (with a blue background) sends data packets to target node 9 (with a red background). Source node 3 performs route detection, broadcasting data in all directions. Assuming equal route losses between nodes (e.g., a route loss of 1 in a favorable environment), after route detection, the source node calculates the path with the minimum route loss, finding 3→5→9 to be the most efficient path. It is worth noting that the principle of minimizing routing loss is not the same as the principle of minimizing routing steps, because communication quality is not necessarily good when the number of steps is small. As shown in Figure 5, assuming that the routing loss between nodes 3 and 5 is 3 and the loss of all others is 1, then according to the principle of minimizing routing steps, the routing link between source node 3 and destination node 9 is 3→5→9, and the optimal routing link calculated according to the principle of minimizing routing loss is 3→4→6→9.
Figure 5 shows that the mesh network determines routing links based on the principle of minimum routing loss.
If the previously working optimal route link encounters a problem, such as link damage, the network will send a failed data packet delivery message to the source node, as shown in Figure 5. Previously, the source node detected the optimal route link as 3→4→6→9. If at some point the communication link between node 6 and node 9 is damaged, causing the data packet delivery to fail, node 6 will return a routing error data packet to source node 3. Source node 3 will then re-submit a routing request to find a new optimal route link.
It's important to note that in the elevator IoT WEN network, the optimal route is not proactively replanned for each data packet. As long as a previously found optimal link exists in the network, the next data packet will use this link for delivery. Because each route detection takes 1-2 seconds and requires broadcasting to detect the link with the least routing loss, this would significantly increase the load on the WEN network. The usual practice is to save the optimal path found in the first route detection to the next cycle and even the entire network cycle, unless there is link failure, communication stoppage, or node offline.
IV. Conclusion
Elevator IoT sensor layer network structures vary, but they all revolve around a single-point architecture. Single-point network structures have many drawbacks. From a practical perspective, their high cost deters customers. Single-point structures result in a limited data transmission mechanism, and implementing broadcast communication requires significant bandwidth investment at the application layer, without the ability to perform secondary control over broadcasts. This paper proposes a distributed network structure for the elevator IoT sensor layer that overcomes the shortcomings of single-point structures. Furthermore, the distributed structure allows for the use of WEN wireless technology, multiple data transmission methods, and optimal routing principles, enabling the construction of a cost-effective elevator IoT system with significant potential for widespread adoption.
References
[1] Wang Xue. Wireless Sensor Network Measurement System [M], Beijing: Machinery Industry Press, 2007: 212
[2]CarmanDW,KruusPS,MattBJ.Constraintsandapproachesfordistributedsensornetworksecurity[R].NAILabs,2005.
