Top 3 Challenges in Assuring the Internet of Things

December 13, 2016

Crunching big data, tracking yottabytes in the wild

Shhh… the toaster is listening. You’re likely already using an IoT device. More and more, you’re offered options and upgrades for “smart” cars, homes, and devices. The obvious question is, when do the machines take over now that they’re talking to each other?

Skynet is probably not something we need to worry about yet, although don’t let us stop you from stocking your bunker. The real stuff that should be keeping you up at night has to do with reliably available critical services. What do you do when your smart city is down? How do you call IT when a hospital crashes? Where’s your kid going to go if the house won’t let her in?

As a service provider, you’re wondering how these smart, chatty machines let you know when something is wrong. Because no matter how innovative they are, they might not know how to say “I’m broken,” especially if they’re, you know, broken.

Ultimately, it comes down to keeping your IoT infrastructure operational and on its best behavior 24/7. Of course, consistently perfect functionality doesn’t happen by itself. RADCOM’s solutions detect anomalies as they occur, and help you keep even the most complex systems working at peak performance from endpoints A to Z, and at all points along the line. We alert you in real-time, so that you can jump right on any issues as they’re happening, before they become a problem. Because RADCOM’s MaveriQ family of applications is completely scalable, we can cover every fine-grained detail of even the biggest virtual or physical grid. That means nobody’s cat or dog buddy is crying by the IoT-controlled pet feeder: you can trust your IoT devices to work reliably with MaveriQ on guard duty.

What is IoT?

IoT, M2M or MTC, CIoT or NB-IoT, IIoT, HIoT, and IoE: we sometimes use these terms for remote device access interchangeably. But they aren’t identical in meaning. They are cogs in a partially unified mechanism: they can work together, and often overlap.

Some will tell you IoT, the Internet of Things, cannot yet exist in the real world. With the advent of 5G, they say, true IoT will emerge. Perhaps, with 5G, IoT may advance to a much higher standard, but it is effectively live in various current instances.

The earliest IoT device we know of was a very popular Coke machine modified by students at Carnegie Mellon University sometime in the 1970s. Programming students relied on it heavily for sweet, sweet caffeination. When a department expansion moved offices further away, they found it very frustrating to go all the way down to the third floor to discover an empty machine, or worse, to spend carefully hoarded student coins only to get a warm Coke. So, students installed micro-switches to detect the number of bottles present. They also wrote a server program to keep track of each bottle in the machine, including the length of time it had spent cooling off. The display showed each slot with an EMPTY or COLD status, to let you know which button to push. The last step was modifying the university’s server to run the Coke status program whenever someone typed the command anywhere in the university, or in fact, on the newly developing internet.

IoT devices use an open connection. Insight accumulated by IoT is sent to a server, the web, or the cloud, rather than to another device in a linked array. IoT communication is sent via both wireless and wired systems. Because it is highly automated, service quality assurance is critical.

Many endpoint devices in IoT technologies are not actually computing machines: they really are passive “things” that are set up to provide information to connected sensors or portals. A fun instance of a “thing” is an RFID-tagged wristband, used by a luxury hotel to share streaming video of a user’s vacation experience to Facebook or Instagram.

IoT devices integrate gathered intelligence with Big Data, analytics, social media and other applications. They perform a more open-ended function, as opposed to M2M’s closed loop.

M2M (machine to machine) communication, also called MTC (machine type communication) appeared around 1973, and has been applied to great effect since. MTC fully automates the exchange and processing of communication between machines, with minimal or zero human contact. With the constant addition of more smart machines and services, analysts predict 50 billion (yes, billion) connected devices by 2020. This Hokusai-magnitude wave of MTC is known as MMTC (massive machine type communication), and it will rely on evolving network structures and the advent of 5G connectivity.

M2M or MTC is designed to work with no need for human intervention in its communication or actuation processes. A device or sensor in the field receives information and relays it back to a “manager” machine over a closed network. The data is received and processed automatically. The managing machine can simply collect information, or it can trigger a command, relayed back to the “worker” machine. The in-field machine executes the task, and reports back as necessary. M2M refers to machines talking only to other machines in its system. They may connect to each other via internet, but often use proprietary connective systems or existing telecom networks.

M2M or MTC assurance is critical: systems and their deployed units must be tracked closely and continually. Simple peripherals cannot inspect themselves. If they’re broken, they’re simply broken, and no one will know unless there is also a monitoring handler as well as the vital automatic report function. As a case in point, a minor element in an M2M system regulating environmental controls for a chicken farm malfunctioned, cooling the temperature to freezing. Because the farm’s monitoring and reporting did not occur in real-time, no action was taken until after the weekend. By the time repair personnel showed up on Monday, all the chickens had died.

It’s also important to keep track of machine anomalies as an excellent proactive way to avoid issues. If a connected machine acts erratically or in a way that does not match the typical performance of other similar machines, we can tag this as a “bad” or “rogue” machine that should immediately be scrutinized and assigned for repair or replacement.

CIoT (cellular IoT) and rival NB-IoT (narrow-band IoT) use existing telecom technologies such as 2G, 3G, and 4G to add perpetual connectivity to mobile devices. The standards and specifications are hotly contested by various telecom and equipment providers, and are still being hammered out: it remains to be seen which one, or which combination of the two, will be chosen by 3GPP, the mobile broadband decision-making body. In basic terms, it’s just IoT for your cell phone or tablet.

IIoT, the Industrial Internet of Things, is an aspect of IoT that concentrates on improving efficiency, safety, and productivity in business operations. It can incorporate M2M, as well. Some folks use the term IIoT as distinct from another term: HIoT (Human Internet of Things). HIoT refers to items that are interactive or reactive and which enable lifestyle improvement, or enhance human needs and home or health elements. An illustration of HIoT could be smart toys, smart appliances, or home health monitoring. IIoT might include robotics, fleet tracking, or security systems.

IoE, the Internet of Everything, is a term that considers how connectivity brings together human interaction, processes, data, devices, and things, as part of a more comprehensive flow. After all, actual people do have input and participation at various points in the global data lifecycle.

So, how do we keep the IoT and all its digital citizens up and running?

Challenge 1: More Devices, More Data, More Problems

As mobile and connected device use explodes, data transmission is expanding exponentially. Most devices are now operating on older networks, and 5G is on its way. With the sheer amount of information roaring along the internet expressway, someone’s got to sort through a whole lot of data packets to ensure quality of service.

As data gets bigger, information storage gets smaller, cheaper, and more efficient. Have you seen the tiny 5D quartz coin that can store 360 TB for 14 billion years? Service providers are muscling up their capabilities, too, and are already processing more data streams from more users than ever before.

With IoT running on mobile networks, we have to consider possible data stream bottlenecks. Picture rush hour with self-driving cars. Now picture re-routing them all at once to avoid a jackknifed tractor trailer up ahead. You really don’t want them all stalling or crashing due to QoS monitoring that simply can’t keep up with the barrage of bytes.

We can talk about the multiple smaller, closer nodes of fog computing at the edge of the cloud. Sure, that’s helpful when it comes to gathering real-time data from, say, self-driving cars. But while that’s a great development, what’s sorting through the output and making sure each component stays on track?

When 5G officially hits, whether that’s in 2020 or earlier, it’s going to bring a domino effect of updates and upgrades. It’s also going to herald a tremendous amount of opportunities for new telecom revenues in a saturated market. Paired off with the potential wave of 5G revenue are the possible signaling storms that could cause network outages: the key to balancing the two, and ensuring your company comes out on top, is service assurance.

RADCOM’s MaveriQ is ready for the challenge. Its service assurance and customer experience solutions run off a scalable, open-architecture platform, and its steady ultra-high performance ensures that your devices perform well and consistently, too. MaveriQ is a pro at holding chaos at bay: we track the sneakiest little glitches and warn you before they have a chance to wreak havoc.

Challenge 2: That Grating Latency

IoT is enabling some ambitious projects, including smart cities. Smart cities are built on the concept of an organically threaded tapestry of linked intelligent devices connected to each other and to the cloud.

Smart city services are managed centrally and intelligently. Smart parking offers real-time information on available spaces throughout the city, as well as alerting electric cars to charging stations. Street lighting monitoring and management helps deal with equipment breaks or weak points effectively, avoid energy waste, and reduce costs. But your smart city can’t afford wait time as information leaps from point to point. You need all services functioning and active at once, and you need to know right now if all the lights in the city go out. Latency is an IoT killer, and keeping it down takes know-how. Aside from the hardware, you’d better make sure your smart city has a data officer on the beat, checking that there’s not even a second’s slowdown in active problem triage.

RADCOM’s MaveriQ suite of customer experience management components keeps your smart projects on track. You might not consider a delay of a fraction of a second “grinding to a halt”, but we will notice the hiccup and alert you to what’s going on, and where the hold-up is happening.

Challenge 3: Secure the Perimeter

Wait, where is the perimeter? It’s kind of endless, isn’t it? What exactly are the boundary lines you might define or lock down, when it comes to IoT?

Your first gateway is obviously your IoT device itself. But, because your widget is now actively connected to the cloud, borders are blurred at best. It’s a long and sometimes intricate series of interconnected gates rather than one doorway. Can you put a digital big-armed bouncer on each one? Should you? Will that even shut out determined intruders?

Everything is, ultimately, hackable. The goal is to make it as tough to hack as possible, because there are a lot of real-life consequences riding on the security of your connected devices.

On October 21, 2016, denial-of-service attacks took down major websites in the United States, mostly on the East Coast, including Twitter and Spotify. Through the day, other sites such as Reddit, Airbnb, Tumbler, and Amazon were apparently knocked out by the same DDoS attack using malware known as Mirai, all through domain registration service provider Dyn. The entry point? Hacked IoT devices. Ouch.

Some basic (and yeah, obvious) security precautions would include keeping the software for your IoT device or system updated to prevent penetration at possible weak points, as well as changing your default usernames and passwords. As with any tech, you will also want to have some sort of antivirus. But, in addition, there should be some higher-level monitoring of performance and integrity for all the devices in an array. You can’t afford that kind of breach in your network. You have a responsibility to your subscribers, and a business reputation to uphold.

Checking your IoT network’s exposed flank should also mean keeping an eye out for those rogue machines and machine anomalies we mentioned earlier. That’s something MaveriQ can help with.

Answering the Challenges

Are you ready for IoT and all its challenges and rewards? We are.

The building burst of new technology can feel like a threat, or you can see it as the fantastic catalyst for new revenue that it is. Keep churn down, and your revenue up, with next-level IoT security and functionality patrols from RADCOM’s MaveriQ application set. MaveriQ’s consistent, consolidated, context-savvy perspective on every type of network topology means we’re equipped to support your dynamic, growing network.

MaveriQ monitors and correlates IoT control plane messages as well as user plane data traffic in real-time, allowing your engineers to quickly find the root cause of any issues as they arise. Aside from the usual subscriber frustration with any delay in service correction, IoT devices that control critical functions such as health or traffic cannot and should not be experiencing downtime. Nobody can afford that kind of chaos, least of all your subscribers who depend on these services.

And, for MTC supervision, consider using MaveriQ Analytics. Our analytics layer automatically detects anomalies in M2M communications patterns, and sets off alarms if a failure or service degradation is detected. We’d like to help you keep your chickens, metaphorical and otherwise, alive and clucking.

Share this article
Skip to content