Industrial networking embraces wireless connectivity

Historically industrial network designers have shunned wireless systems in favor of wired connectivity. The reasons for this are that safety, security and reliability are paramount in a factory environment. However, attitudes are now shifting due to advances in wireless technology.

By Paul Dillien April 28, 2015

Wired systems have a proven pedigree in traditional factories, where production is based on long product runs with infrequent changes. The cabling or fiber supports secure, real-time, reliable links with high data rates with a well-understood latency. The local area networks (LANs) can be organized as Ethernet star, tree, daisy chain, segmented branches, mesh or ring topology. Each has benefits and drawbacks. For example, a star configuration allocates equipment with its own dedicated link, which is expensive in terms of cabling, but gives excellent availability. A ring topology features a shared resource, but has resilience because it can route around the ring in the event of a broken cable.

Wireless transmissions rely on the electromagnetic propagation characteristics within the factory complex. This environment can be very harsh, as the machinery will generate radio frequency noise and the transmissions will be reflected off the metal equipment and walls to distort the signal. Nonetheless, it offers many benefits for industrial networking if these issues can be overcome.

The main advantage is that the equipment is not "tethered" in place by the cabling and can be moved and repositioned within the machine floor. Today, many production lines are required to be highly flexible and to rapidly switch to different tasks. This may include reconfiguring the line to remain efficient, and reprogramming the machine for the new task. 

Technology development, low cost have driven demand

So what has changed to make wireless a viable technology? There are two main reasons. Firstly, the technology used in modern systems such as cellular connectivity has made rapid advances in both efficiency and resilience. The second factor is the launch of cost-effective, low-power, highly integrated field-programmable radio-frequency (FPRF) devices.

The combined development of many companies involved in cellular activities has spurred significant advances to improve wireless technology. This has led to enhancements in how effectively the radio spectrum is used. This efficiency is measured in bits per second per Hertz (bit/s/Hz) of bandwidth, and reflects how much information can be transferred in a given frequency band. The early 2G struggled to achieve 0.5 bit/s/Hz, while the 3G system, rolled out in the past decade supported less than 3 bit/s/Hz. In contrast, the Long Term Evolution (LTE) specification, which is sometimes called 4G, pushes this maximum to 5 bit/s/Hz of bandwidth, giving the headline speeds of 100 Mbps in a 20 MHz spectrum. This dramatic increase of an order of magnitude is testament to the skills and ingenuity of the developers.

Further development work has also resulted in advances in modulation techniques. The most significant new scheme is called Multiple-Input Multiple-Output (MIMO). This is a complex configuration that uses two or more antennae, which are separated by a physical distance. A configuration using two antennae at the transmitter and two at the receiver is termed a 2 x 2 MIMO system. This can be extended to 4 x 4 or 8 x 8 MIMO by adding more antennae and transceivers. 

MIMO changes the game

MIMO techniques can be adopted to improve the spectral efficiency and to achieve a diversity gain that improves the link reliability. Transmitting different data on each antenna and decoding and combining these at the receiver achieve spectral efficiency. For industrial applications, the diversity gain that improves the link reliability is the dominant reason to consider MIMO. The transmitters each propagate the same data patterns, which add up constructively at the receiver. MIMO can provide better performance due to fading and interference or under multi-path transmission conditions, where the signals reflect off machinery or buildings resulting in a distorted signal at the receiver. Moreover, where there is a line of sight between the two radios, a directional beam can be formed. This is done by adjusting the phase of the signals sent to each antenna at the transmitter so the signals are additive. The multiple antennae can provide a composite signal with a beam forming capability that can "illuminate" a user or machine with a stronger signal.

The MIMO transmissions use orthogonal frequency-division multiple (OFDM), which is a technique that uses multiple carriers, where each is modulated with a low-data rate signal. The carrier frequencies are organized so that they do not interfere with each other, because of the orthogonality designed into the system. However, the individual signals from each carrier are aggregated at the receiver and the combination of the combined signal provides a higher data rate. Added to that, the baseband signals are typically pre-processed with forward error correction codes such as Reed-Solomon where the algorithm can correct transmission errors.

The latest field programmable radio frequency (FPRF) device, called LMS7002M, is "MIMO-ready" as it has two identical wireless transceivers. These are programmable over the frequency range 0.1 to 3800 MHz, so encompass all the ISM bands defined by the ITU-R of 6.765 MHz to 2.45 GHz. Frequencies above this, such as the 5.8 GHz band can be accommodated, too, with the addition of a simple external circuit.

An important attribute of a MIMO system is that the channels should be closely matched in terms of phase relationship. Data streams of in-phase and quadrature (I&Q) components are used to modulate the carrier. Achieving matched phase in the two channels can be very tricky with some wireless architectures, requiring extensive adjustment and complex interactions with the baseband logic to get the system aligned for a single frequency. In contrast, the LMS7002M includes an on-chip microcontroller running industry standard 8051 code that simplifies the calibration of the chip. It calibrates dc offset, the TX/RX LPF bandwidth tuning, transmit local oscillator leakage feed-through, IQ gain and phase mismatch in both transmit and receive chains, as well as handling on-chip resistor and capacitor calibration. To supplement the calibration facility, internal RF loop-back paths can be enabled to test out the operation of the device as part of the BIST on the device.

Industrial network resiliency

An industrial network should be self-healing, have a high MTBF and high redundancy. The wireless system offers the ability to be configured in a star or mesh topography, which allows it to reconfigure in the event that a node fails or is taken out of service and also be easily scalable to add new nodes. The network must also provide real-time access and control, and the wireless transceivers provide a high data rate with an RF modulation bandwidth up to 120 MHz, which is equivalent to 60 MHz baseband IQ bandwidth.

The programmable nature of the FPRF can be utilized in an industrial environment in several ways. Firstly, the devices can be tuned to local needs, such as 902 – 928 MHz for the Americas or 433 – 434 MHz for the EMEA region. However, the flexibility can be exploited way beyond a simple factory setting option. Being re-tunable on the fly, over a wide spectrum, allows the wireless subsystem to intelligently change channels to mitigate interference. A baseband chip controls the FPRF, and where the communication link is unsatisfactory, it can agree with the far end to move to an alternative frequency. In fact, the FPRF is capable of supporting a fast frequency hopping capability which can be used for noisy environments. This is analogous to schemes used in military radios to circumvent intentional attempts at jamming, except in an industrial environment, the noise is from the factory machinery. An additional benefit of hopping is that it greatly enhances the system security by frustrating attempts to snoop on the data.

It is also possible to design a wireless hub that can communicate with more than just the factory machines. For example, it can be re-tuned to the Private Mobile Radio (PMR) or General Mobile Radio Service (GMRS) frequencies that may be used by mobile workers such as forklift drivers. Similarly, the FPRF covers all the cellular bands, so it is equally adept at reconfiguring for asset tracking or to contact truck drivers on the road. The cellular option also allows direct access to factory workers with BYOD handsets. This level of flexibility allows the factory management simple "one touch" communication across the enterprise.

The transmitted data packets can be a technology such as WirelessHART, ISA100.11a or 6loWPAN, as the wireless transmission is agnostic to the payload. With the enhanced reliability that a MIMO system provides, technologies such as IEEE 1588v2 can also be included to provide accurate time stamps for the control instructions. The data can also be encrypted and, if required, authenticated on a packet basis.

Baseband chipsets hold the key to power management

MIMO needs complex processing at both the transmitter and receiver, but this is within the capabilities of modern baseband solutions. A processor solution is possible, but to create the modulation patterns to support high data rates would require a very fast processor that would be expensive and consume a lot of power. Modern FPGAs such as an Altera Cyclone V SE are ideal for wireless nodes or Arria V SX for high performance hubs. Both provide a very flexible programmable solution, as each family supports a rich mix of logic, memory, DSP functions, as well as embedded ARM processors. The 32-bit ARM cores provide an integrated facility that can interface with the logic fabric or run external code.

The function of the baseband chip is to encode or decode the I&Q data streams and to program the FPRF via the SPI interface. The baseband device also controls the modulation scheme used and can be, for example, quadrature phase-shift keying (QPSK), quadrature amplitude modulation (QAM), wideband code division multiple access (W-CDMA) in addition to OFDM. For example, the logic fabric, memory and DSP elements can be configured to perform discrete fourier transforms in the form of FFT or DFT and forward error correction (FEC) and generate the modulation patterns for MIMO techniques. These can be configured as required by the ARM cores on the chip, which can instantiate the required combination of functions. The logic can also be used to encrypt the data, for example, using Advanced Encryption Standard (AES) for data confidentiality, or the more complex AES-GCM that also ensures that the data has not been maliciously altered or corrupted in transmission.

The FPGAs are configured on power-up. The configuration file is typically held in an external non-volatile memory (NMV), such as Flash, which programs both the logic fabric and the ARM cores. The ARM executable will typically also be stored in the NMV. The NVM can also hold a copy of the SPI configuration data for the FPRF. This data provides a simple way for the processor to store and access the address and data for programming the frequency, bandwidth, gain and the DSP elements on the FPRF. This allows the customer the ability to update the NVM with new firmware files in the field to add new capabilities or support new standards.

The latest highly integrated wireless FPRF devices and FPGAs provide a cost-effective and robust solution. The capabilities that they offer can be exploited in a modern production facility to provide flexibility industrial communications.

– Paul Dillien, chartered engineer and member of The Institution of Engineering and Technology (MIET), is principal consultant at High Tech Marketing. Edited by Eric R. Eissler, editor-in-chief, Oil & Gas Engineering, eeissler@cfemedia.com.

ONLINE extra

– See related articless on wireless and industrial networks below.

Read more about industrial networks on the Control Engineering industrial wireless tutorials blog.