2018 will be a particularly testing time for IoT bandwidth planners
via Flickr © Lic. Marcos Gasparut (CC BY-SA 2.0)
- Even the best planning unlikely to survive prolonged contact with real-world data
- "Per node" traffic will be the ultimate testing ground
- Difference between theory and practice
- OEM telematics will dominate per node data traffic
The Chief of the Prussian General Staff, the catchily named Field Marshal Helmuth Karl Bernhard Graf von Moltke (the elder), was the lead architect of the seven years (1864 - 1871) of war that eventually resulted in the unification of Germany. He was so archetypically a stiff and aristocratic Prussian officer that it was commonly believed that rather than keep his field marshal's baton in his knapsack as per military tradition, he preferred to secrete his down the back of his tight uniform trousers, the better to keep him standing rigidly to attention at all times. Who knows what happened when he sat on a horse?
The elder von Moltke famously observed that "no battle plan survives the first contact with the enemy", which, in a roundabout way, leads us to the subject of the Internet of Things (IoT) and the planning of message bandwidth and a new report from ABI Research that concludes that even the best and most comprehensive such planning will probably not survive long after first contact with real-world data.
The ABI research says that, despite the best intentions and efforts, many first generation IoT solutions will be tried and found wanting when exposed to the harsh realities of dealing with massive amounts of real-world data traffic on a "per node" basis. The report adds that this is because network planners and system integrators are trying to square the circle of continuing to develop IoT projects both as intellectual scientific and technological concepts whilst simultaneously attempting to determine and provide the means to meet a plethora of real-world data traffic use cases.
The ABI report examines and analyses the growth factors driving IoT data traffic per node, including bandwidth requirements, in 23 key market segments in an effort to provide insights into likely new business opportunities and services which, in turn, will drive the future requirements of IoT data traffic across multiple markets and applications.
Commenting on the release of the new report, Kevin McDermott, Principal Analyst at ABI Research said, IoT system design needs to consider multiple factors around the critical communication links including messaging size, frequency, and data types. However, by anticipating the detailed use case scenarios including data bandwidth growth, the uncertainty between design expectations and real-world experiences can be minimised".
It is self-evident that while requirements and use-case considerations are application dependent, they all seek to interpret the meaning of raw sensor data and adduce value-added information from them. Even as real-world IoT systems and networks are deployed and come on-stream, standards and protocols are still being developed and in some cases a marked and deleterious time lag is evident. After all, and as the ABI report points out, "message structure and information flow depend on the use-case around the data traffic of the target application".
The new paper reveals that the fastest growing segment for data traffic per node is OEM telematics. This sector is forecast to increase at a compound annual growth rate (CAGR) of 540 per cent between now and 2023. This market, ABI says, "will drive the highest IoT solution revenues per connected vehicle - partly driven by the additional security and integrity measures needed to protect the vast amounts of data exchanged".
As Kevin McDermott says in ABI Research’s latest "IoT Data Traffic: Application and Market Analysis report, “Urgent messages, alerts and alarm indicators may require both priority attention and assured bandwidth allocation, but planning also needs to consider peak demand and exception utilisation. System's Integrators, IoT system designers, and network planners can utilise the data traffic per node approach to anticipate the use case and application factors that are expected to drive growth over the next five years”.
The upside: much IoT traffic can survive high latency conditions
In essence it's the usual and ultimate answer to any question concerning digital communications: bandwidth. It is the lack of bandwidth that always forces compromise and IoT devices attempting to transmit over a bandwidth links that are inherently incapable of carrying the amount of data required to be sent will in some way or another, have to filter out less important data and compress what is left.
Most current forecasts would have us believe that there will be 25 billion IoT devices connected by 2020 although a particularly bullish one from Cisco's Internet Business Solutions Group reckons that figure will be 50 billion. And it doesn't really matter which one you believe because it is evident that the production, transmission, archiving and manipulation of data is going to grow by unprecedented amounts over the next few years and as a report from Computer Sciences Corporation points out, history shows that the industry always underestimates its predictions.
Given that, one of the greatest problems facing the IoT industry is that every solution might be different one from the other and thus require different approaches to solve specific network latency and bandwidth problems. Latency always affects performance when data is moved over the Internet or through the cloud but there is an upside for parts of the IoT sector in that many IoT data payloads are small and not necessarily time-critical and so can and even thrive under high latency conditions. Time will tell and this year will be the toughest testing time.