Clearing the Ethernet mist

Market analyst firm IMS Research forecasts that global shipments of industrial Ethernet nodes will grow at almost 22% over the next five years to overtake traditional fieldbuses as the most popular means of networking in manufacturing and process applications. The move to industrial Ethernet is happening across all sectors of industry, where it offers a departure from the standardisation on the various open and proprietary fieldbus protocols of the 1990s, IMS noted.

A common perception of industrial Ethernet is that it is simply a more ruggedised version of standard Ethernet, providing an alternative to a fieldbus. This is wrong on several counts, not least in that it assumes Ethernet can operate as a fieldbus.

To understand Ethernet, the best place to start is with the OSI (Open System Interconnection) seven-layer model. Layer 1, the Physical Layer, defines the electrical and physical specifications for devices, in particular the relationship between a device and the physical medium.

Layer 2, the Data Link Layer, provides the functional and procedural means to transfer data between network entities and to detect and possibly correct errors that may occur in the Physical Layer. It is here that Ethernet is defined as a network protocol under the IEEE 802.3 standard.

Over the years, Ethernet has become synonymous with the TCP/IP suite, but one does not necessarily imply the other. IP is defined under the Network Layer (Layer 3) of the OSI model. This layer provides the functional and procedural means of transferring variable length data sequences from a source to a destination via one or more networks. The Transport Layer (Layer 4) provides transparent transfer of data between end-users, and defines the likes of TCP and UDP.

The Session Layer (Layer 5) controls the connections between computers, whilst the Presentation Layer (Layer 6) transforms the data to provide a standard interface for the Application Layer (Layer 7) at the top of model. It is here that you find typical applications such as FTP, HTTP, RTP, SMTP, SNMP and others. In short, when it comes to operating as a communications architecture in industrial networks, Ethernet is capable of very little without the layers that sit above it.

Industrial Ethernet is the use of Ethernet as the data link layer protocol with one of the fieldbus protocols as the application layer, in much the same way that the various fieldbus options are actually the use of RS232/485 as the data link layer, with the fieldbus protocol itself at the application layer. Industrial Ethernet provides a vast increase in speed, up from the sub-10kbps typical with RS232 to the gigabit and beyond potential. Overall performance is also increased, and costs are reduced through the ability to use standard access points, routers, switches, hubs and cables that are far cheaper than equivalent serial port devices.

For end-users, however, the biggest advantage aside from cost is the significantly greater interoperability between devices. Industrial Ethernet offers the potential for a seamless flow of data from field devices all the way up to higher level business systems, and back.

Meanwhile, the evolution of Ethernet technology from a 10Mbps bus/tree topology to a gigabit, switch-based topology has paved the way for its use to support time-critical applications in industrial networks.

This switch-based topology makes the implementation of an industrial Ethernet network very different from implementing a device-level network.

The infrastructure of Layer 2 and Layer 3 switches is the core of the industrial Ethernet network, providing the determinism - the ability to ensure that a packet of information is sent and received in a specific period of time - and throughput required for control applications. When Ethernet was first discussed for the likes of machine control tasks on the plant floor, its inherent lack of determinism was a major issue.

Rather than being simply a junction box that links the disparate nodes in a network, the Ethernet switch provides a myriad of functions that define the capabilities of the network and increase the flexibility of the overall system.

The switch's ability to eliminate collisions is the most important mechanism to provide real-time capability for Ethernet-based control systems. Switches can be added to split the data load between segments, resulting in higher performance. In addition, managed switches can prioritise traffic, allowing the preferential handling of real-time traffic over supervisory traffic.

The Ethernet switch also makes it possible to build redundancy into the industrial Ethernet network — something that was very difficult and/or expensive to do with standard fieldbus networks. At field device level, dual connectivity is the standard approach for providing redundancy, but the lack of PLCs and field devices with dual connectivity built-in for sensor and controller applications has made this almost impossible to achieve cost effectively.

A key issue to be addressed when implementing an industrial Ethernet network is security. To date, network security has focused on making Layer 2 secure, but the very fact that industry is embracing Ethernet as its network of choice, and thereby potentially opening up the plant floor to attack from the outside world, means that network security requirements are evolving, driving a need to look beyond Layer 2.

Under the seven-layer model, all it takes is for one layer to fall to an attack before the whole communications system is compromised - potentially without the other layers even being aware that there is a problem. Security is only as strong as the weakest link, and Layer 2 can be a very weak link indeed.

Today's switches and routers, however, create enhanced integrated solutions for security for Ethernet networks, meeting the emerging demand for Layer 3 (IP) and Layer 4 (TCP or UDP) security mechanisms, and enabling a seamless security system to be established.

To a great extent industrial Ethernet frees up the user from the danger of being locked into a single solution, and enables the system integrator to focus on installing the network without having to worry about the protocol that will eventually be used over it.

With industrial Ethernet, the investment made in the network infrastructure is a flexible decision with room for evolution. An industrial Ethernet infrastructure will accommodate multiple industrial Ethernet protocols in use at the same time, with the switch network routing the data packets appropriately.

Developments in industrial Ethernet will make the case for investment even more impressive. New Power over Ethernet (PoE) switches can supply power to connected devices, so eliminating the need for costly cabling back to a power source. Then there is the emergence of wireless Ethernet for industrial applications, using RF transmitters and receivers to send and receive data. And on the horizon is Power over Wireless Ethernet.

A question of protocol
"Industrial Ethernet appears to be following the same path trodden by fieldbus protocols, with a growing number of variations on the theme," said John Morse, a senior market analyst at IMS. His research suggests that Profinet will ultimately be dominant in the EMEA region and Ethernet/IP will enjoy the lion's share in the Americas — reflecting the relative strength of Siemens and Rockwell

Automation in EMEA and the Americas, respectively. However, noted Morse, the technology mix will be more evenly spread in Asia.

While there are around 50 protocols, Stefan Knauf, of Mitsubishi Electric, said the market profile is directly related to the PLCs market: "Profinet, EthernetIP and [the recently launched] CC LInk IE will be the most popular networks worldwide. The other networks are very specialist so there are always users that want to use Ethercat, for example, because it is special for motion applications."

According to Dave Cook of GarretCom Europe, a lot of these other protocols are running over the top of Ethernet: "The term 'Ethernet' is used all the time to indicate TCIP and Ethernet," he said. "We have got to be very careful that people are not confusing these upper layer protocols with the lower layer protocol of Ethernet. It is much more flexible than it first looks. Different protocols offer a slightly different flavour, but most are going to run over standard Ethernet."