The fastest-growing sector of electronics is desperate for communication standards.
Ed.: This is the fourth of an occasional series by the authors of the 2017 iNEMI Roadmap. This information is excerpted from the Roadmap, which is available from iNEMI (http://community.inemi.org/content.asp?contentid=51).
The Internet of Things (IoT) is an infrastructure of networked objects (cyber-physical devices, information resources, and people) that interact with the physical world through sensors and actuators.
It is an intelligent, interconnected system made of networks of communicating actuators, sensors, smart objects and services that can be quickly provisioned into systems to achieve goals for the system users. These systems may in turn be part of another IoT system.
The number of networked IoT devices is growing and predicted by Gartner to reach at least 20 billion devices globally by 2020. The IoT revolution will bring about a global transformation (Industry 4.0) using technologies such as machine learning/vision, autonomy, and next-generation communication 5G access to provide a multi-trillion dollar economic impact.
IoT can be categorized at the highest level as consumer-facing, which includes networked electronics such as cellphones, wearables, and appliances; and industrial-facing, which in the broadest sense includes heavy equipment, the manufacture of consumer IoT devices, and networked infrastructure such as smart cities, smart highways and self-driving cars. IoT systems are characterized by choreographed relationships amongst and between machines and people, with varying degrees of human-independent decision making.
The evolution of IoT is complicated by market pressures; manufacturers strive to be first to market with new IoT-enabled functionality, leading to an unmanaged and unstandardized plethora of consumer products and protocols which can go obsolete or unsupported before products hit their usual end-of-life. Industrial IoT is likely to enjoy more reliable support, but is hindered by the lack of a coherent IT infrastructure which ties IoT devices back through the manufacturing ecosystem that created them.
Standards. Over 20 different organizations developing standards for the Internet of Things are aimed at creating home network protocols, reducing power consumption, creating more robust security, condensing and delivering data ready for analytics, industrial environment use cases, routing and networking, etc. Not only are these standards unevenly adopted and not designed to work together, many of these standards have a domain-specific focus (i.e., some focus on the consumer domain, while others focus on the manufacturing domain). They are focused on sharing data and components within a domain but not between domains. For instance, few if any of the consumer IoT standards are designed to share data backward with the high-tech manufacturing systems that create the devices and code that makes IoT work, and will be called upon to support those devices when they fail.
Interoperability is another key aspect for these devices that should be considered here. Ideally, the same standards needed to manufacture IoT devices should be used for those same devices to communicate and operate in the field, to make support as seamless as possible.
IoT provides a new opportunity for a more direct relationship between manufacturers and the end-user, but not without manufacturers having visibility into the richness of the data created by the web of IoT devices in the field. Improved data feedback could allow the industry to make important improvements, like manufacturing-in better security.
Timing. Timing is particularly important for networked sensing, computing and actuating, which characterizes IoT. The data from sensors need timestamps accurate enough to make aggregated data useful. Computing needs to finish fast enough to enable control actions, or actuation, in time to meet requirements.
All of this must happen on networks and in hardware that, until recently, did not incorporate timing capabilities. Generally speaking, networked devices that require precise timing currently overcome this fundamental conflict of modern system design by using dedicated hardware and customized software for timing-critical systems. Things that require precision timing are processed as much as possible with systems that do little or no data processing. However, in many cases such devices must also include significant data processing.
In many IoT systems, precision timing needs may not seem critical, with timing requirements being in the range of seconds to minutes, rather than milliseconds and below. For example, traffic congestion data are not needed within milliseconds of an observation; rather, one or two minutes are reasonable. If the traffic observation data are 20 minutes old, then it is unlikely to be useful. In these situations, it is still important to understand the timing capabilities of a component or system, and be able to relate them to system requirements. This is not currently being done for the vast majority of IoT systems; instead, it is just assumed the data are good enough.
Large-scale environments, characterized as Systems of Systems, present another set of challenging timing issues. For example, “smart highways” will involve many different systems, some in the vehicle, some in the infrastructure, some in a traffic management center, etc. Each will have its own timing requirements that must be met while sharing network bandwidth, and in some cases, computation bandwidth on servers. Today the technology for managing the timing in such systems and understanding the timing characteristics of the composed system are works in progress.
Timing presents its own security issues as well. Requirements for secure and resilient time exist at all layers of a network, from the physical to the application layer. While time is physical, its abstraction into networks and complex information systems transform its security into both cyber and physical concerns. Timing may be vulnerable to unintentional threats, such as interference, space weather impacts, network anomalies, or intentional threats, such as jamming and spoofing (counterfeiting via RF signal injection or a cyber-attack).
Civil Global Navigation Satellite Systems (GNSS) signals, such as GPS, are the primary worldwide timing distribution mechanism, and are inherently vulnerable to jamming and spoofing. Jamming refers to the denial of the signals-in-space by illegally broadcasting energy in the radio navigation spectrum. Low-power (<1W) jammers are widely available to consumers and are marketed and used as “personal privacy devices.” High-power jammers are generally used to intentionally deny GNSS receivers over a wide area. Though the effects of denial can be damaging, robust timing receivers should enter predefined holdover, mitigation, or failure modes when detected.
GNSS spoofing is the RF injection of counterfeit or recorded GNSS signals into a receiver. Spoofing attacks may be data- (e.g., replace the navigation data on the GNSS signal) or timing-oriented (e.g., induce a delay). Jamming may be intentional or incidental. Generally spoofing is intentional, though it may be possible for incidental spoofing to occur (e.g., through legal GNSS repeaters).
Satish Parupalli is senior engineering manager at Intel (intel.com) and Chair of the IoT Product Emulator Group (PEG) of the 2017 iNEMI Roadmap. Barbara Goldstein is associate director of the Physical Measurements Laboratory, National Institute of Standards and Technology (NIST) and co-chair of the IoT Product Emulator Group (PEG) of the 2017 iNEMI Roadmap.
