While the age of mobility brought its own set of networking challenges, it also brought a degree of homogeneity. However, this degree of consistency won’t last much longer; with the rise of the internet of things, network planning is going to get a whole lot more complicated.
Here comes IoT
One of the biggest components to network planning is looking at the “how much” and “when” as it relates to bandwidth. Traditionally, most devices — everything from servers and PCs to laptops and smartphones — had significant storage, compute and networking power. Lower-powered devices were linked to these devices, much in the same way smartwatches were initially tied to smartphones. This means that there were a finite number of devices with semi-predictable bandwidth requirements. So, that made network planning less daunting.
Today, with the rise of IoT, anything can be connected directly to the internet. So, what exactly does this mean for networking planning and optimization? Less predictability and more uncertainty.
More connections and fluctuations means more complexity
The number of connected devices and machines being added to the network is one of the most widely talked about impacts of IoT. However, beyond the sheer number, these devices have variable computing capabilities and variable bandwidth demands, and as a result, require more agile bandwidth adaptation. There has also been a shift in how technology creates, stores and sends data.
Let’s take a look at a few examples from the transportation industry. Bombardier introduced a new aircraft that has an engine with a staggering 5,000 sensors — that is an almost 2,000% increase compared to the less than 250 sensors most engines have today. This means significantly more data needs to be processed, and while some of the data is sent to the ground during flight, most of the data is stored until it connects to a Wi-Fi connection. This is because as of right now, sending data over a cellular network is extremely cost prohibitive. Similarly, Tesla’s cars collect data and video on the road with most of it being stored and relayed once connected to a home Wi-Fi connection.
Not only is there more data, but there is also less consistency in when and where data is produced. This can cause wild fluctuations in bandwidth requirements in any given area of the network, as well as the network core itself. These considerations require network operators to utilize powerful new software that enable network automation and orchestration. Network slicing is also a vital part of the equation as it allows for the traffic to be properly isolated and prioritized, as well as expand and contract based on specific needs.
Getting everyone to talk the same language
Another consideration is size. As IoT sensors and other devices shrink, they are constrained in their compute, storage and network capabilities either due to size, weight or cost (or a combination of all three).
The industry’s response to these constraints has been deployments of low-powered wireless technologies (LPWAN), including: ultra-wideband, near field communication, DASH7, Zigbee, Sigfox, Weightless, Nwave, Ingenu and long range (LoRa). While many of these new technologies overcome challenges around size, many do not rely on TCP/IP or Ethernet, but rather on derivatives or alternates.
As such, in the new Wild West of IoT, the gateway has emerged as the wrangler to pull these disparate technologies together. As defined by IoT Agenda, “a gateway is a network node that connects two networks using different protocols together. While a bridge is used to join two similar types of networks, a gateway is used to join two dissimilar networks.” Problem solved? Not quite. The range of many of these technologies requires that the network get much closer to the user. In fact, gateways may even need to be deployed closer to the premises than existing service provider pedestals. This means companies can not only rely on cell service towers, but must find ways to get devices closer to each other, as well as the end user.
One unexpected but common structure companies are turning to? Streetlights. Many cities across the U.S. are in the process of converting their existing streetlights to LED lights. One reason is for cost-savings. The other? They provide an opportunity to mount sensors and even deploy networking equipment directly inside the streetlights. Cities or public/private partnerships that deploy LED lighting may one day provide network providers the opportunity to build networks on and in these streetlights. If this is successful, streetlights have the potential to become the new cell tower.
However, this too comes with its own challenges. Compared to cell towers, the use of sensors means more installations and more technology to service, triggering a slower rollout and higher maintenance costs for providers.
What does this mean?
According to IDC, approximately 80 billion devices will be connected to the internet by 2025. That comes out to 152,000 new smart devices being connected to a network every minute — a whole lot of bandwidth to manage. With the rise of IoT, it is clear that network providers will need to gain a greater level of visibility into their networks, as well as the needs of their customers, in order to plan for networking demand today, and prepare for tomorrow.
All IoT Agenda network contributors are responsible for the content and accuracy of their posts. Opinions are of the writers and do not necessarily convey the thoughts of IoT Agenda.