Industrial networks begin to move toward standardization

In general, the capabilities of an industrial network are limited, the topology is closed, and the protocols and data it carries are dedicated. However, these solutions are being replaced by more widely used standards (such as Ethernet) that run TCP/IP and even Web protocols instead of more real-time or more specialized interfaces.

The problem with moving to a standardized network is that these networks are not designed for a strict real-time industrial environment. But in any case, the market has such a demand, so the designer's responsibility is to overcome these obstacles to ensure the best performance.

Traditional serial networks and control area networks (CANs) have been widely used to transfer data from a device or sensor to a recorder or reading device. Although they are still popular today, due to several factors, these networks are being replaced by newer Ethernet or TCP/IP and Web-based networks. One of the main factors is that people have realized the cost advantages of Ethernet, and the TCP/IP protocol stack is not only easy to configure but also can be widely used for various transmission purposes.

Figure 1: Low-cost, industrial-use embedded Ethernet card.

In some cases, end-users get compelling information: they may pass real-time processing information to management. However, the reality is that engineers have the technical means to deliver any size of data to any location, but this ability is crippled by certain facts related to the side effects of configuration.

The Ethernet interface is mainly used to support general-purpose networking applications, rather than being designed for the extreme physical conditions of industrial or manufacturing environments. In addition, the TCP/IP stack is designed for reliable delivery of static content and does not guarantee the speed or order of packet delivery.

Despite the high flexibility of hardware interfaces and software protocols, designers need to understand their limitations in an industrial environment and know how to work around them to provide a reliable and lasting solution for process monitoring and control. .

As the data acquisition equipment, programmable logic controllers, sensors, meters, and other systems used by the factory are increasingly interdependent, end-users are looking for more extensive network and network applications that can be used in highly integrated environments. Although it seems ridiculous that office workers want to monitor and optimize the production process on PCs during breaks, there are indeed many applications that can use a single network standard.

However, designing such a system and building a device that can take advantage of it poses significant engineering challenges. The most stringent constraints on the network may come from extreme environmental conditions: manufacturing facilities and industrial sites are often exposed to high temperatures, high humidity, and high electrical interference. Although the shielding measures taken by traditional Ethernet may be able to avoid most of the electrical interference, hardware reliability and redundant backups are also important.

An attractive solution is integrated device networking, which implements Ethernet networking in a certain part of the chip to obtain a reliable and high-performance solution.

Designers can integrate at several levels. For example, designers can integrate at the chip level through custom ASICs or commercial integrated processors (such as NetSilicon's NET+ARM, etc.). The NET+ARM processor uses an ARM7 or ARM9 processor core and integrates 10/100Base-T Ethernet and numerous peripheral connections such as USB and PCI. NetBurner also offers a similar configuration product with the Motorola ColdFire 5270 microprocessor, flash memory, SDRAM, and 10/100 Ethernet.

Designers can also integrate at the board level by integrating standard processors, Ethernet, and other peripheral devices on the board. The board can generally add more processing power and external interfaces, such as digital I/O, analog-to-digital and digital-to-analog conversion, and serial ports supporting RS-232 and RS-485/422. For example, AMD offers Net186 embedded Ethernet reference designs, including Am186ES microcontrollers, flash and SDRAM, serial ports, and network controllers.

At the software level, Ethernet with TCP/IP as a protocol stack has some less desirable features. In particular, these protocols can reduce but not completely avoid conflicts, and there is no guarantee that packets will be delivered sequentially or within a specified time.

There are two ways to alleviate these protocol issues. The first is to optimize the protocol or the hardware in the Ethernet stack to compensate. Another approach is to use the basic hardware and network layers of Ethernet, but use a protocol designed specifically for real-time industrial systems instead of TCP/IP.

Many vendors, including NetBurner and NetSilicon, have chosen to optimize the protocol and interface hardware. In general, users have the right to choose the protocol used to transmit data, so that they can use protocols that have higher performance or best meet the needs of the application.

Those using the second method can use the LonWorks protocol, a hierarchical, packet-based peer-to-peer communications protocol designed to meet the requirements of a control system rather than a data processing system. Similar to related Ethernet and Internet protocols, this protocol is also an open standard that complies with the Open Systems Interconnection Reference Model of the Internet Standards Organization. The LonWorks protocol effectively eliminates packet collisions, which is a major cause of system performance degradation. By using LonWorks to IP routers, Ethernet can connect to the Internet or the wide-area backbone.

Although this alternative is different from the Ethernet-only solution, it allows connection to an Ethernet-based network for data transmission beyond the plant level. This approach can achieve the best of both worlds: better real-time performance in the industrial application of the network; allowing easier access to data in the general application of the network for monitoring and evaluation.

safe question

In fact, all end users know that viruses, worms, or hackers can compromise the security of basic services or disrupt their systems. However, as Ethernet and TCP/IP-based solutions become more and more popular, industrial networks are also Begin to be exposed to these dangers.

Figure 2: Customizable optimized Ethernet protocol stack.

Viruses, worms, or other automatic intrusion techniques can impede or invade systems that are running monitoring applications. Although these malicious codes do not necessarily affect the meter or the device itself, they can damage the servers and desktop systems used to monitor and record data.

Another risk is the automatic denial of service (DoS) attack, which does not affect the network itself, but it does not allow the Internet to access the network. If the software solution contains remote monitoring capabilities, this attack will cut off the connection on the communication link. DoS attacks are sometimes deliberate and their purpose is to retaliate or extort.

The last risk is man-made invasion. This kind of intrusion is deliberate malicious behavior that can cause business interruptions or theft of commercial secrets. Smaller equipment networks based on industry standards can provide some degree of protection through obfuscation, and from an intruder's perspective, it is difficult to get anything. But for a widely distributed Ethernet network that covers workshops and corporate offices, the gains of intruders will be greater, and accordingly, the technical challenges will be even greater.

Another potential threat to software is that people are increasingly using Linux and its derivatives as embedded real-time operating systems. Although there is nothing inherently insecure in Linux, the increasing use of such systems in computer systems and industrial equipment makes it a target for attackers who understand and exploit their vulnerabilities. Other real-time operating systems may be more secure, but they are less commonly used for general purposes.

How to design industrial control systems and take these and other new security issues into account? Most designers implement this mainly through software, although hardware-based authentication and access are also feasible. The software standard for authentication is Kerberos, which can run on non-secure networks. When a user needs access to a network device, he first requests a "passport" from the Kerberos server, and then the server creates a packet containing the requested content, the current time, and the length of time the "pass" will remain valid.

The server adds random keys and identification information outside the encrypted data packet, encrypts the data packet again using the requester's key, and sends it back to the requester. The customer then decrypts the “pass” and sends it back to the server. After receiving the “pass”, the server decrypts the “pass” with its own key. If decryption is successful, the server checks the timestamp to confirm if the "pass" is valid and authenticates its identity.

Kerberos for embedded networks may only require 25 kB of code and data, which is relatively low overhead for 32-bit processors. Using Kerberos for authentication helps ensure that non-authorized software or users cannot access the protocol stream.

Design considerations

Standard network stacks such as Ethernet and TCP/IP may not necessarily reduce the cost of industrial networks, but their use in such applications will continue to increase. Ethernet solutions require hardware and supporting materials that provide good performance and reliability in extreme environmental conditions, as well as software solutions that ensure real-time data delivery.

However, if the embedded Ethernet solution provides the right combination of cost, standards, and functionality, it is possible to achieve unprecedented connectivity between the shop floor and the general network. Currently, the most common application of this configuration is remote monitoring, usually through a web browser. Matching the real-time data in the manufacturing process with other business intelligence can greatly increase productivity and efficiency. Making the device connect properly is the most important step.


500-2000mah Lithium Battery

Brand MIGHTY MAX BATTERY
The battery is composed of nickel -cadmium batteries
Recommended purpose of the product IROBOT
Voltage 14.4 volts
Rechargeable

500-2000Mah Lithium Battery,2000Mah Rechargeable Battery,5000 Mah Power Bank,5000Mah 26650 Battery,Cylindrical Lithium Battery

Langrui Energy (Shenzhen) Co.,Ltd , https://www.langruienergy.com