Networked Architecture and Characteristics of Intelligent Instruments

Intelligent instruments are the product of the combination of computer technology and testing technology. The instruments are equipped with intelligent software with strong processing capabilities. Instruments are no longer simple hardware entities, but a combination of hardware and software. In recent years, intelligent instruments have begun to develop from relatively mature data processing to knowledge processing, making their functions develop to a higher level.

1 Development of intelligent instruments

Since the 1990s, the intelligence of instruments and meters has been prominently reflected in the following aspects:

(2) Miniaturization. The comprehensive application of microelectronics technology, micromechanical technology, information technology, etc. has made the instrument a small-sized, fully functional intelligent instrument that can complete functions such as signal acquisition, processing, control signal output, amplification, and interface with other instruments. It plays a unique role in automation technology, aerospace, military, biotechnology, and medical fields.

(3) Multifunctionality. Multifunctionality is a characteristic of intelligent instruments. For example, a function generator with functions such as a pulse generator, frequency synthesizer, and arbitrary waveform generator not only has higher performance (such as accuracy) than dedicated pulse generators and frequency synthesizers, but also provides better solutions for various test functions.

(4) Intelligence. Modern detection and control systems tend to be more or less intelligent. The further development of intelligent instruments will contain a certain degree of artificial intelligence, so that they can complete detection or control functions autonomously without human intervention.

(5) Instrument virtualization. In a virtual reality system, data analysis and display are completed using PC software. As long as certain data acquisition hardware is provided, it can be combined with a PC to form a measuring instrument. This PC-based measuring instrument is called a virtual instrument VI (Virtual Instrument). In a virtual instrument, the same hardware system can be used to obtain a measuring instrument with completely different functions by applying different software programming. "Software is the instrument." The software system at the core of the virtual instrument is universal, popular, visible, scalable and upgradeable, representing a new direction for the development of today's instruments.

(6) Networking of instrumentation systems. Generally, intelligent instruments and meters have two-way communication functions, but this two-way communication function is still far from true network communication. With the rapid development of network technology, Internet technology has enabled instruments and meters to be networked on the basis of being intelligent, so that on-site measurement and control parameters can be accessed on the network nearby and have the necessary information processing functions.

2 Functional requirements and technical support of networked instruments

2.1 Support remote measurement and control requirements

Networked instruments, such as fieldbus intelligent instruments, are suitable for use in remote measurement and control. They are the result of the deep integration of instrument measurement and control technology, modern computer technology, network communication technology and microelectronics technology. Networked equipment can automatically measure, control, store and display measurement results and control status of related physical quantities according to set procedures like ordinary instruments; at the same time, it has important network application characteristics. Authorized instrument users can remotely operate the instrument functions, obtain measurement results, monitor the instrument in real time, set parameters and diagnose faults, and control it to dynamically publish information on the Internet through the Internet. Like computers, they have become independent nodes in the network. They can easily connect directly to the nearest network communication cable, and they are "plug and play" to directly send field test data to the Internet; users can browse this information (including processed data, panel images of instruments, etc.) in real time through browsers or standard applications.

2.2 Characteristics of networked instruments

基于Internet的测控系统中前端模块不仅完成信号的采集和控制，还兼顾实施对信号的分析与传输，因为它以一个功能强大的微处理器和一个嵌入式操作系统为支撑。在这个平台上，使用者可以很方便地实现各种测量功能模块的添加、删除以及不同网络传输方式的选择。其次，基于Internet的测控系统最为显著的特点，是信号传输的方式发生了改变。基于Internet的测控系统对测量、控制信号等的传输，是建立在公共的Internet上的。有了前端嵌入式模块，系统的测量数据安全有效的传输便成为可能。再有，基于Internet的测控系统对测得结果的表达和输出也有了较大改进，一方面，不管身在何处，使用者都可通过客户机方便地浏览到各种实时数据，了解设备现在的工作情况；另一方面，在客户端的控制中心，所拥有的智能化软件和数据库系统都可被调用来对测得结果分析，以及为使用者下达控制指令或作决策提供帮助。[page]

2.3 How to access the Internet or Ethernet

The design method of networked instruments is to embed embedded systems into instruments and make them the core of measurement and control. Generally, there are three ways to connect embedded instruments to the Internet or Ethernet to become network instruments:

(1) Embedded instruments are composed of 32-bit high-end MCUs. Because there are sufficient resources for expansion and utilization, the entire TCP/IP protocol family can be built into the system, so it can become a network instrument directly connected to the Internet, but the development is difficult;

(2) For embedded instruments composed of low-end 8-bit computers, a dedicated network (such as RS-232, RS-485, Profibus, etc.) is used to connect several embedded instruments to a PC. The PC is used as a gateway and the PC converts the information on the network into TCP/IP protocol packets and sends them to the Internet to achieve information sharing. However, a dedicated PC must be equipped for protocol conversion.

(3) An embedded network instrument composed of an 8-bit single-chip microcomputer that is directly connected to the Internet. The advantage of this solution is that it can utilize the previous measurement equipment based on 8-bit single-chip microcomputers and directly drive the network interface chip through an external network chip. However, it occupies more resources (ROM, RAM, CPU) and requires the single-chip microcomputer to have a fast enough running speed.

2.4 Network interface chip support

The network interface chip uses RTL8019AS from RELTEK, which is an ideal chip for Ethernet communication due to its excellent performance and low price.

(1) Main performance

It complies with EthernetⅡ and IEEE802.3 standards. It is a full-duplex communication interface, and the transmission and reception can reach a rate of 10Mbps at the same time. It has a built-in 16K SRAM for transmission and reception buffering, reducing the speed requirement for the main processor. It supports 8/16-bit data bus, 8 interrupt request lines and 16 I/O base address selections. It can complete the formation, encoding and decoding of physical frames, the formation and verification of CRC, the transmission and reception of data, etc., and can send and receive data simultaneously on the twisted pair through a switch.

(2) Internal structure

The RTL8019AS can be divided into remote DMA interface, local DMA interface, MAC (media access control) logic, data encoding and decoding logic and other ports. The remote DMA interface refers to the bus for the microcontroller to read and write the internal RAM of the RTL8019AS, that is, the interface part of the ISA bus. The microcontroller only needs to operate the remote DMA to send and receive data. The local DMA interface is the connection channel between the RTL8019AS and the network cable, completing the data exchange between the controller and the network cable.

(3) Internal RAM address space allocation

There are two RAM areas inside RTL8019AS. One is 16K bytes, with addresses from 0x4000 to 0x7fff; the other is 32 bytes, with addresses from 0x0000 to 0x001f. RAM is stored in pages, with 256 bytes per page. Generally, the first 12 pages (i.e. 0x4000 to 0x4bff) of RAM are used as the send buffer; the last 52 pages (i.e. 0x4c00 to 0x7fff) are used as the receive buffer. Page 0 has an address of 0x0000 to 0x001f, which is used to store Ethernet physical addresses.

(4) I/O address allocation

RTL8019AS has 32-bit input and output addresses, and the address offset is 00H~1FH. Among them, 00H~0FH has a total of 16 addresses, which are register addresses. The register is divided into 4 pages: PAGE0, PAGE1, PAGE2, PAGE3. The page to be accessed is determined by the PS1 and PS0 bits in the CR (Command Register) of RTL8019AS. The remote DMA address includes 10H~17H, which can be used as a remote DMA port. Just use one of them. The reset port includes 18H~1FH, a total of 8 addresses, with the same function, used to reset RTL8019AS. [page]

3 Architecture and Implementation of Networked Instruments

3.1 Abstract Model

Networked instruments are an organic combination of electrical and electronic, computer hardware and software, as well as network, communication and other technologies. They have a relatively complex structure, and often use architecture to represent their overall framework and system characteristics. The architecture of networked instruments includes basic network system hardware, application software and various protocols. Figure 1 is a simple model of the networked instrument architecture. This model divides the networked instrument into several logical layers, which can more essentially reflect the principle characteristics of information collection, storage, transmission and analysis and processing of networked instruments.

The first is the hardware layer, which mainly refers to the remote sensor signal acquisition unit, including the microprocessor system, signal acquisition system, hardware protocol conversion and data stream transmission control system. The realization of the hardware layer functions benefits from the technological progress of embedded systems and the development of large-scale integrated circuit technology in recent years. Hardware protocol conversion and data stream transmission control rely on FPGA/CPLD.

Another logical layer is the embedded operating system kernel. The main function of this layer is to provide a platform for controlling signal acquisition and data stream transmission. The main resources of the front-end module unit of this platform are processors, memory, signal acquisition units and information; the main function is to reasonably allocate and control the processor, control the signal acquisition unit to make it work normally, and ensure the effective transmission of data streams. This logical layer is mainly composed of link layer, network layer, transport layer and interface. Depending on the application, the specific implementation of this layer may be different and can be simplified to a certain extent.

3.2 Peripheral Hardware Design

There are two hardware designs for Internet or Ethernet communications.

(1) Use a dedicated CPU as the controller and use C language programming to implement TCP/IP communication. The advantage is that the dedicated CPU has strong processing power and is easy to implement other functions of the test instrument. The disadvantage is that the cost is slightly higher and the hardware is slightly more complicated.

(2) Use 51 series single-chip microcomputer as the CPU of the controller, do not use embedded operating system, directly use C51 programming to implement data link layer protocol and TCP/IP protocol. The advantage is that the hardware is relatively simple and the price is low. The disadvantage is that the software workload is large and the difficulty is also high. The basic structure of the network instrument composed of single-chip microcomputer as the core and RTL8019 Ethernet interface chip as the network instrument interface is shown in Figure 2.

3.3 Protocol and Design

The system performs initialization operations, mainly configuring the network interface chip. After configuration, the system is in a waiting state until the client sends data. Data reception is achieved through the network interface chip, which can perform packet filtering on the physical frames on the network. When an Ethernet station's information frame is sent to a shared signal channel or medium, all Ethernet interfaces connected to the channel read the frame and check the first 48-bit address field of the frame, which contains the destination address. Each interface compares the destination address of the frame with its own 48-bit address. If the address is the same as the destination address of the frame, the Ethernet station will continue to read the entire frame and send it to the upper-layer network software running on the computer. The upper-layer network software reads the type field of the frame to determine whether the information frame is an ARP packet or an IP packet, and then hands it over to different protocol stacks for processing. When other network interfaces find that the destination address is different from their addresses, they will stop reading the frame.

When sending data, the data to be sent is encapsulated in a frame format and sent to the send buffer in the RTL8019AS through the remote DMA channel, and then a transmission command is issued to complete the frame transmission. It is necessary to set the Ethernet destination address, Ethernet source address, and protocol type, and then set the data segment according to the set protocol type. After that, the remote DMA is started, the data is written to the RAM of the RTL8019AS, and then the local DMA is started to send the data to the Internet. The RTL8019AS cannot store the data packet into the FIFO through the DMA channel at one time, so before forming a new data packet, it must wait for the previous data packet to be sent. In order to improve the transmission efficiency, the design divides the 12-page send buffer into two 6-page send buffers, one for data packet transmission and the other for data packets at the construction end, which are used alternately.

4 Conclusion

With the advancement and continuous expansion of computer technology and network technology, the concept of instrumentation in the 21st century will be an open system concept. Based on PCs and workstations, forming a practical measurement and control system through the establishment of a network, improving production efficiency and sharing information resources has become the development direction of modern instruments and meters. The concept of networked instruments is a breakthrough in the concept of traditional measuring instruments. In a sense, computers and modern instruments and meters are mutually inclusive, and computer networks are also universal instrument networks. If more different types of intelligent devices in the measurement and control system become network nodes like computers and workstations and are connected to the network, they will make full use of the relatively mature Internet network equipment, which will not only realize more resource sharing and reduce the cost of building the system, but also improve the function of the measurement and control system and expand its scope of application. The concept of "network is instrument" accurately outlines the development trend of networked instruments.