Despite the ubiquity of Ethernet networks, communication technologies based on DSL do not lose their relevance to this day. Until now, DSL can be found in last mile networks for connecting subscriber equipment to ISP networks, and recently the technology is increasingly used in building local networks, for example, in industrial applications, where DSL acts as an addition to Ethernet or to field networks based on RS-232/422/485. Such industrial solutions are actively used in developed European and Asian countries.
DSL is a family of standards that were originally conceived for the transmission of digital data over telephone lines. Historically, this was the first broadband Internet access technology, replacing DIAL UP and ISDN. The wide variety of currently existing DSL standards is due to the fact that many companies, since the 80s, have tried to develop and market their own technology.
All these developments can be divided into two broad categories - asymmetric (ADSL) and symmetrical (SDSL) technologies. Asymmetric refers to those in which the speed of the incoming connection differs from the speed of the outgoing traffic. By symmetrical it is understood that the speeds for receiving and transmitting are equal.
The most famous and widespread asymmetric standards are, in fact, ADSL (in the latest edition - ADSL2 +) and VDSL (VDSL2), symmetrical - HDSL (obsolete profile) and SHDSL. All of them differ from each other in that they operate at different frequencies, use different methods of coding and modulation on a physical communication line. Error correction methods are also different, which provides a different level of noise immunity. As a result, each technology has its limits in the speed and distance of data transfer, including depending on the type and quality of the conductor.
Limits of various DSL standards
In any DSL technology, the data rate drops as the length of the wire increases. At extreme distances, it is possible to get a speed of several hundred kilobits, but when transferring data over 200-300 m, the maximum possible speed is available.
Among all technologies, SHDSL has a serious advantage that makes it possible to use it in industrial applications - high noise immunity and the ability to use any type of conductor for data transmission. In asymmetric standards, this is not the case, and the quality of communication is highly dependent on the quality of the line used for data transmission. In particular, it is recommended to use twisted telephone cable. In this case, a more reliable solution is to use an optical cable instead of ADSL and VDSL.
Any pair of conductors insulated from each other - copper, aluminum, steel, etc. - is suitable for SHDSL. The transmission medium can be old electrical wiring, old telephone lines, barbed wire fences, etc.
Dependence of the SHDSL data transfer rate on the distance and type of conductor
From the plot of data transfer rate versus distance and type of conductor given for SHDSL, it can be seen that conductors with a large cross section allow information to be transmitted over a long distance. Thanks to the technology, it is possible to organize communication over a distance of up to 20 km at the maximum possible speed of 15.3 Mb / s for a 2-wire cable or 30 Mb for a 4-wire cable. In real applications, the baud rate can be set manually, which is necessary in conditions of strong electromagnetic interference or poor line quality. In this case, to increase the transmission distance, it is necessary to reduce the speed of SHDSL devices. Free software tools such as .
Due to what SHDSL has a high noise immunity?
The principle of operation of the SHDSL transceiver can be represented as a block diagram, in which a specific and independent (invariant) part from the point of view of the application is distinguished. The independent part consists of PMD (Physical Medium Dependent) and PMS-TC (Physical Medium-Specific TC Layer) functional blocks, while the specific part includes TPS-TC (Transmission Protocol-Specific TC Layer) and user data interfaces.
The physical link between transceivers (STUs) can exist in the form of single-pair or multiple single-pair cables. In the case of multiple cable pairs, the STU contains multiple independent PMDs associated with a single PMS-TC.
SHDSL Transceiver (STU) Functional Model
The TPS-TC module depends on the application in which the device is used (Ethernet, RS-232/422/485, etc.). Its task is to convert user data to SHDSL format, perform multiplexing / demultiplexing and time adjustment of several user data channels.
The PMS-TC layer performs SHDSL framing and synchronization, as well as scrambling and descrambling.
The PMD module performs the functions of encoding/decoding information, modulation/demodulation, echo cancellation, matching parameters on the communication line and establishing a connection between transceivers. It is at the PMD level that the main operations that ensure the high noise immunity of SHDSL are performed, including TCPAM coding (Trellis coding with analog pulse modulation), a joint coding and modulation mechanism that improves the spectral efficiency of the signal compared to a separate method. The principle of operation of the PMD module can also be represented in the form of a functional diagram.
Block diagram of the PMD module
TC-PAM is based on the use of a convolutional encoder that generates a redundant bit sequence on the side of the SHDSL transmitter. At each cycle of work, each bit entering the encoder input is assigned a double bit (dibit) at the output. Thus, at the cost of a relatively small redundancy, the noise immunity of the transmission is increased. The use of Trellis modulation makes it possible to reduce the used data transmission bandwidth and simplify the hardware while maintaining the same signal-to-noise ratio.
The principle of operation of the Trellis encoder (TC-PAM 16)
The double bit is generated by a logical modulo 2 addition (exclusive or) operation based on the input bit x1(tn) and the bits x1(tn-1), x1(tn-2), etc. (there can be up to 20 in total), which were received at the input of the encoder before and remained stored in memory registers. At the next cycle of encoder operation tn+1, bits will be shifted in memory cells to perform a logical operation: bit x1(tn) will move into memory, shifting the entire sequence of bits stored there.
Convolutional Encoder Algorithm
Truth tables for modulo 2 addition
For clarity, it is convenient to use the state diagram of the convolutional encoder, by which you can see what state the encoder is in at times tn, tn+1, etc. depending on the input. The encoder state in this case means a pair of input bit values x1(tn) and a bit in the first memory cell x1(tn-1). To build a diagram, you can use a graph whose vertices contain possible states of the encoder, and transitions from one state to another are indicated by the corresponding input bits x1(tn) and output dibits $inline$y ₀y ₁(t ₀)$inline$.
State Diagram and Transition Graph of Transmitter Convolutional Encoder
In the transmitter, based on the received four bits (two encoder output bits and two data bits), a symbol is formed, each of which has its own amplitude of the modulating signal of the analog-pulse modulator.
Status of the 16-bit AIM depending on the value of the four-bit character
On the side of the signal receiver, the reverse process occurs - demodulation and extraction from the redundant code (double bits y0y1(tn)) of the required sequence of encoder input bits x1(tn). This operation is performed by the Viterbi decoder.
The decoder algorithm is based on calculating the error metric for all possible encoder states. The error metric is the difference between the received bits and the expected bits for each possible path. If there are no receive errors, then the true path error metric will be 0 because there is no bit mismatch. For false paths, the metric will be different from zero, constantly increasing, and after some time the decoder will stop calculating the erroneous path, leaving only the true one.
Encoder state diagram computed by the receiver's Viterbi decoder
But how does this algorithm provide noise immunity? Assuming that the receiver received the data in error, the decoder will continue to calculate two paths with an error metric of 1. A path with a metric of 0 will no longer exist. But the conclusion about which path is true, the algorithm will make later on the basis of the next received double bits.
When the second error occurs, there will be several paths with a metric of 2, but the correct path will be revealed later based on the maximum likelihood method (that is, the minimum metric).
Encoder state diagram computed by the Viterbi decoder when receiving erroneous data
In the case described above, for example, the algorithm of the 16-bit system (TC-PAM16) was considered, which provides the transmission of three bits of useful information in one symbol and an additional bit for error protection. In TC-PAM16, data rates from 192 to 3840 kbps are achievable. With an increase in bit depth to 128 (modern systems work with TC-PAM128), six bits of useful information are transmitted in each character, and the maximum achievable speed is from 5696 kbps to 15,3 Mbps.
The use of analog pulse modulation (PAM) links SHDSL to a number of popular Ethernet standards such as 1000BASE-T gigabit (PAM-5), 10GBASE-T 10 gigabit (PAM-16) or 2020BASE-T10L industrial single-pair Ethernet for 1 (PAM-3).
SHDSL over Ethernet
There are managed and unmanaged SHDSL modems, but this classification has little in common with the usual division into managed and unmanaged devices, which exists, for example, for Ethernet switches. The difference lies in the configuration and monitoring tools. Managed modems are configured via a web interface and can be diagnosed via SNMP, while unmanaged modems can be diagnosed using additional software via the console port (for Phoenix Contact, this is a free PSI-CONF program and a mini-USB interface). Unlike switches, unmanaged modems can operate in a network with a ring topology.
Otherwise, managed and unmanaged modems are absolutely identical, including functionality and the ability to work on the Plug&Play principle, that is, without any prior configuration.
Additionally, the modems can be assigned the functions of surge protection with the possibility of its diagnostics. SHDSL networks can form very long segments, and conductors can run in places where surge voltages (induced potential differences caused by lightning or short circuits in nearby cable lines) can occur. The induced voltage can cause discharge currents of kiloamperes to flow. Therefore, to protect equipment from such phenomena, SPDs are built into modems in the form of a removable board, which, if necessary, can be replaced. It is to the terminal block of this board that the SHDSL line is connected.
Topologies
Using SHDSL on Ethernet, it is possible to build networks with any topology: point-to-point, line, star and ring. At the same time, depending on the type of modem, both 2-wire and 4-wire communication lines can be used for connection.
Ethernet network topologies based on SHDSL
It is also possible to build distributed systems with a combined topology. Each segment of the SHDSL network can have up to 50 modems and, taking into account the physical capabilities of the technology (the distance between modems is 20 km), the segment length can reach 1000 km.
If a managed modem is installed in the head of each such segment, then the integrity of the segment can be diagnosed using SNMP. In addition, managed and unmanaged modems support VLAN technology, that is, they allow you to divide the network into logical subnets. The devices are also capable of working with data transfer protocols used in modern automation systems (Profinet, Ethernet/IP, Modbus TCP, etc.).
Reservation of communication channels using SHDSL
SHDSL is used to create redundant communication channels in an Ethernet network, most often optical.
SHDSL and serial interface
SHDSL modems with a serial interface allow you to overcome the limitations in distance, topology and quality of the conductor that exist for traditional wired systems based on asynchronous transceivers (UART): RS-232 - 15 m, RS-422 and RS-485 - 1200 m.
There are modems with serial interfaces (RS-232/422/485) for both universal and specialized applications (eg Profibus). All such devices are classified as "unmanaged", therefore they are configured and diagnosed using special software.
Topologies
In networks with a serial interface using SHDSL it is possible to build networks with a point-to-point, line and star topology. As part of a linear topology, it is possible to combine up to 255 nodes into one network (for Profibus - 30).
In systems built using only devices on the RS-485 interface, there are no restrictions on the used data transfer protocol, but line and star topologies are not typical for RS-232 and RS-422, so the operation of end devices in the SHDSL network with similar topologies is only possible in half-duplex mode. At the same time, in systems with RS-232 and RS-422, addressing of devices must be provided at the protocol level, which is not typical for interfaces most often used in point-to-point networks.
When connecting devices with different types of interfaces via SHDSL, it is necessary to take into account the fact that there is no single mechanism for establishing a connection (handshake) between devices. However, it is still possible to organize an exchange in this case - for this, the following conditions must be met:
- communication coordination and data transfer control should be performed at the level of a single information data transfer protocol;
- all end devices must operate in half-duplex mode, which must also be supported by the information protocol.
The Modbus RTU protocol, which is the most common for asynchronous interfaces, allows you to avoid all the described limitations and build a single system with different types of interfaces.
Serial network topologies based on SHDSL
When using two-wire RS-485 on equipment more complex structures can be built by linking modems through a single bus on a DIN rail. A power supply can be installed on the same bus (in this case, all devices are powered through the bus) and optical converters of the PSI-MOS series to create a combined network. An important condition for the operation of such a system is the same speed of all transceivers.
Additional features of SHDSL in RS-485 network
Application examples
SHDSL technology is actively used in urban utilities in Germany. More than 50 city utility companies are using old copper wires to link facilities across the city. First of all, control and accounting systems in water, gas and energy supply are built on SHDSL. Among these cities are Ulm, Magdeburg, Ingolstadt, Bielefeld, Frankfurt an der Oder and many others.
The largest system based on SHDSL was created in the city of Lübeck. The system has a combined structure based on optical Ethernet and SHDSL, integrates 120 remote objects and uses more than 50 modems . The entire network is diagnosed via SNMP. The longest segment from Kalkhorst to Lübeck Airport is 39 km long. The reason why the client company chose SHDSL was that it was not economically viable to implement the project entirely on optics, given the presence of old copper cables.
Data transmission via slip ring
An interesting example is the transfer of data between moving objects, such as is done in wind turbines or in large industrial twisting machines. A similar system is used for information exchange between controllers located on the rotor and stator of the installations. In this case, a sliding contact via a slip ring is used for data transmission. Examples like this show that it is not necessary to have a static contact for data transmission over SHDSL.
Comparison with other technologies
SHDSL vs GSM
If we compare SHDSL with data transmission systems based on GSM (3G / 4G), then in favor of DSL is the absence of operating costs associated with a regular fee to the operator for access to the mobile network. With SHDSL, we do not depend on the coverage, quality and reliability of mobile communication in an industrial facility, including immunity to electromagnetic interference. In SHDSL, there is no need to configure the equipment, which speeds up the commissioning of the facility. Wireless networks are characterized by large delays in data transmission and difficulty in transmitting data using multicast traffic (Profinet, Ethernet IP).
Information security speaks in favor of SHDSL due to the absence of the need to transfer data over the Internet and the need to configure VPN connections for this.
SHDSL vs WiFi
Much of what has been said for GSM can also be applied to industrial Wi-Fi. Wi-Fi is opposed by low noise immunity, limited data transmission distance, dependence on the topology of the area, delays in data transmission. The main drawback is the information security of Wi-Fi networks, because anyone has access to the data transmission medium. With Wi-Fi it is already possible to transmit Profinet or Ethernet IP data, which would be difficult for GSM.
SHDSL vs optical
Optics overwhelmingly has a big advantage over SHDSL, but in a number of applications SHDSL allows you to save time and money on laying and welding optical cables, reducing the time to commission the facility. SHDSL does not require special connectors, because the communication cable is simply connected to the modem terminal. Due to the mechanical properties of optical cables, their use is limited in applications related to the transfer of information between moving objects, where copper conductors are more common.
Source: habr.com