Application of IP Multicast Technology in Distributed Control System (DCS)

With the development of economic globalization, the decentralized control system (DCS) has gradually formed the integration of information technology (IT), the integration of multiple buses, the use of functional safety technology to integrate control, instrumentation, and electronic control. At the same time, DCS has higher requirements for dealing with complex large-scale system architecture and large-capacity data exchange.

2.1 Advantages of Multicast Transmission

There are three types of TCP/IP transmission in IP networks: unicast, broadcast, and multicast.

The traditional unicast and broadcast communication methods cannot achieve the problems of single-point transmission and multi-point reception with minimal network overhead. The emergence of IP multicast technology solves this problem in time.

Compared with unicast, the advantage of multicast is that since the transmitted information starts to be copied and distributed at the network node as far as possible from the information source, the increase of users will not result in an increase in the load of information sources and network resources. A significant increase in consumption.

Compared with broadcasting, multicast has the advantage that because the transmitted information is only sent to the receivers who need the information, it does not cause a waste of network resources and can improve the security of information transmission; in addition, broadcasting only Can be carried out in the same network segment, while multicast can realize transmission across network segments.

2.2 Multicast Communication

In multicast, users are divided into different user groups according to different applications. Group members are required to register with a multicast server (usually a router). A user host sends a request packet to indicate the group to be added. Each multicast group has a unique Class D address. Its address ranges from 224.0.0.0 to 239.255.255.255. IP can provide up to 228 simultaneous multicast group addresses. Therefore, the actual number of groups is limited by the size of the selected routing table rather than the addressing.

3.1 Overall System Design

In the distributed control system (DCS) application process, there are a large number of DPU uplink data sent to multiple human-machine interface (HMI) systems. It can just take advantage of the advantages of IP multicast and implement a mixed transmission method that combines unicast and multicast. By designing a reasonable data exchange solution, it not only ensures that data security reaches the HMI workstation, but also reduces the DPU and network device CPUs. The network load improves system efficiency and achieves good performance and economy. Figure 1 shows the system architecture.

Figure 1: System Structure

The system mainly consists of four parts: I/O acquisition card, distributed processing unit (DPU), system IP network, and human-machine interface (HMI).

As shown in Figure 1, the distributed processing unit (DPU) completes the bidirectional data exchange with the I/O collection card through the bus protocol; when the system is designed, as long as the specific data acquisition requirements of the system are reasonably divided into different multicast groups, each will be A DPU (a multicast source) joins different multicast groups as required. It can join multiple multicast groups and publish real-time data in the form of multicast on the system IP network. Each Human Machine Interface (HMI) workstation (group) The recipient of the broadcast data) joins different multicast groups according to the data acquisition requirements, and may also join multiple multicast groups to complete the reception of the multicast data. The system IP network is composed of routers, hubs, switches, and firewalls that are equipped with multicast functions to jointly complete the forwarding of multicast data. Multicast mainly completes the DPU to HMI uplink data, and other data such as control commands are supplemented by unicast.

3.2 IP Multicast Winsock Programming Method

In practical applications, the method of implementing IP multicast programming differs with the DPU embedded support system and the HMI operating system platform; programmers usually use the network data communication programming interface provided in the operating system to implement the IP on the network. Multicast communication features. The following describes a specific method of IP multicast programming implementation using the Windows Socket (Winsock) in the Microsoft Windows environment.

• Call the socket() function to create a datagram socket, which is the initial socket of the multicast group, and later data is sent on this socket.

● Call the setsockopt( ) function to set the local interface (IP_MULTICAST_IF) and the multicast address (for example, 239.100.1.1) and the port number (for example, 20120) to be sent.

● Use the sendto() function to send data. The destination address is the multicast address set in the second step.

● Close the socket and release the Winsock resource.

3.2.2 IP Multicast Data Reception Process (HMI)

● Invokes the WSAStartup() function, which initializes the Winsock service.

l Call the socket() function to create a datagram socket, which is the initial socket of the multicast group, and the data is received on this socket later.

● Call the bind() function to bind the local address (such as INADDR_ANY) and the port number (same as the sender).

● Call the setsockopt( ) function to set the IP_ADD_MEMBERSHIP option and join the multicast group.

● Call the recvfrom() function to receive data.

IP multicast packets typically use the User Datagram Protocol (UDP), which is a "best-effort" protocol. Therefore, IP multicast applications must encounter packet loss and out-of-order problems.

Secure multicast means that only registered senders can send data to the group; only registered receivers can receive multicast data. However, IP multicast is difficult to guarantee this.

The Internet is a heterogeneous network. Therefore, when designing an IP multicast network, it is necessary to fully consider the heterogeneity of the network.

IP multicast technology can effectively solve the problem of single-point transmission to multi-point, multi-point transmission to multiple points, and realize efficient point-to-multipoint data transmission in IP networks, which can effectively save network bandwidth and reduce network load. In the actual application process, as long as the problem of multicast is faced, and the heterogeneity of the network is fully considered, a safe and reliable system can be constructed by designing a reasonable network architecture and data exchange protocol. Keyuan's NT6000 distributed control system has been successfully developed to fully prove this point; in the application and promotion process of the NT6000 distributed control system, the application of the IP multicast communication technology in the NT6000 distributed control system has also been fully affirmed by customers.

references:

[1] TCP/IP Network Interconnection Technology (1) Principle, Protocol and Architecture (3rd Edition) Tsinghua University Press, 2001.

[2] IP Multicast and Security (1st Edition) National Defense Industry Press, 2006

[3] Application Status of DCS in China's Large-scale Thermal Power Plants CONTROL ENGINEERING China, 2009

[4] IP Multicast Communication and Its Implementation in VxWorks Operating System Computer and Digital Engineering, 2003