Skip navigation
All Places > Keysight Blogs > Insights Unlocked > Blog > 2016 > December
2016

Future projections for the Internet of things (IoT) are staggering: billions of devices around the world, with 500 residing in a typical home. That’s why the wireless designers I know are preparing their devices for the Interference of Things—and they’re doing it early in the development process. This is especially important for IoT widgets intended for mission-critical applications.

 

A recent customer visit drove this point home. They had released a new healthcare-related IoT device that uses Wi-Fi to send information to a centralized database. Even though testing in the lab went well, the device was consistently losing connections due to interference from myriad onsite devices—literally 913 devices on Wi-Fi alone in one hospital. When we met, our customer had an urgent need to resolve this problem and recover from the figurative black eye it had received from hospital administrators.

 

Troubleshooting with 20:20 hindsight

After the fact, avoidable problems in receivers, transmitters and the RF environment may suddenly seem obvious. In a receiver, blocking is related to its ability to tolerate strong or local signals on adjacent channels that use the same standard or protocol. When an adjacent signal is too strong, the receiver becomes so desensitized that it can’t recognize an on-channel signal.

 

An ill-behaved transmitter has a poor adjacent-channel power ratio (ACPR) when too much of its signal spills into neighboring channels. That’s one reason why we see ACPR in virtually every wireless standard. In a noisy wireless environment like a hospital, better ACPR helps improve overall system behavior.

 

In today’s airwaves, one issue keeps cropping up: the overabundance of standards- and non-standards-based devices clogging up the industrial, scientific and medical (ISM) band at 2.4 GHz. Devices that use different standards simply see each other as interference and thus don’t play well together.

 

Using foresight to design for interference

Long before a wireless device enters field trials, developers can explore potential issues using the latest design, simulation and test tools. These enable such extensive wringing-out—ranging from obvious to insidious—that a manufacturer can send its prototype into the field with much greater confidence that it will perform as intended.

 

Of course, there’s an important reality check: Where does your IoT device reside on the continuum between mission-critical and throwaway? The economics change significantly if your device trends towards either end of that spectrum.

 

The IoT device in my story would have fared much better in the Interference of Things if it had been debugged early on using a more thorough approach to design, simulation, emulation, test and analysis. An integrated IoT solution is a crucial step toward overcoming likely IoT challenges—and setting up your IoT end users for success.

 

What approaches have you used to set yourself up for success in IoT and other wireless products?

Nearly every manufacturing executive I talk to is looking for ways to improve product quality and reliability, and they often think that drastic measures are required. They assume they’ll have to reinvent their engineering design processes, implement expensive upgrades on the manufacturing floor, or restructure their supply chain. They’re often surprised to learn that the key to improving product quality is already in place in their organization, and it can be summed up in one word:  Accountability.   

                     

For most large manufacturers, the responsibility for designing and building a product often rests with one part of the organization while the responsibility for providing warranty support lands elsewhere. Each area has its own budget and reporting structure, so costs incurred in warranty support are rarely traced back to the design/build process where defects often originate. By creating a culture of shared accountability between upstream and downstream teams, you can dramatically improve product quality using the people and processes you have in place today.

 

The concept is simple, intuitive, and nearly free to implement. But putting it in place takes three key steps.

 

 

1. Identify exactly where your costs are.

When a product fails under warranty, you might repair it, replace it, or even provide a discount on a future purchase to secure your customer’s business. Who covers those costs? Many manufacturers have corporate accounts that are funded annually to cover warranty service—that was Keysight’s approach up until a few years ago. Other manufacturers track warranty expenses at the business-unit level or the product level. The problem is, those costs are often hidden from view. In fact, many of the executives I work with have only a vague idea of who actually pays for warranty service. The first thing I encourage them to do is follow the money: Trace warranty expenses to the payer so you know exactly who’s footing the bill and what it’s really costing the company. The next thing I have them do is look upstream and identify the source of the problems that are costing them money. Typically, when a product is returned under warranty, the issue can be traced to a design flaw, a manufacturing process, or a perceived flaw—meaning the product is functioning exactly as intended but does not meet the expectations of the buyer. Each scenario is expensive for manufacturers. By looking upstream and downstream, you get a clear understanding of where problems come from and who pays, so you know where to focus your efforts in the next two steps.

 

2. Empower your teams to fix the problem. 

Product teams are under dire pressure to meet schedule and cost targets. They’re probably also thinking about sales targets, competitive threats, supply chain logistics, environmental issues, you name it. It’s no surprise that product reliability sometimes falls off their radar, especially if a product can pass inspection and ship on time. But here’s the thing. When product teams are held financially responsible for warranty repairs that occur a few years down the road, product quality becomes a priority. Give them the power to solve the problem. Let them decide how to allocate budgets and resources, and make the process changes and business tradeoffs that need to be made. Ultimately, their goal is not to ship a product on time or on budget but to make the company more profitable and successful. In that sense, empowering product teams is a lot like parenting teenagers: instead of telling them exactly what to do, tell them what the goal is, make sure they understand the consequences, then let them figure it out.    

 

3. Extend P&L accountability across the product lifecycle.

In many OEM environments, R&D designs a product, manufacturing builds it, the support team fixes it, and everyone’s responsible for their own budget. If you’re serious about improving quality, that model needs to change. Have all teams operate under a single P&L structure that spans the entire product lifecycle, from design and build through end of warranty. With a unified accounting structure, everyone shares the cost of repairing a faulty product, and everyone gains when quality is improved.

 

This isn’t theoretical. At Keysight we implemented a companywide accountability program in the mid-2000s. Most divisions were able to make the transition in less than a year, producing a 50 percent reduction in failure rates across the company. In fact, warranty repairs declined so dramatically that we were able to extend our standard warranty coverage from one year to three years, creating a powerful competitive advantage at zero cost to our company.

 

It’s never easy to change corporate culture. It takes vision and fortitude in the C-suite and buy-in down the line. But changing your company’s culture of accountability is one of the best competitive moves you can make, and it’s available to any manufacturer.

 

DuaneLowenstein is a Test Strategy Analysis Manager for Keysight Technologies.  Read his bio.

Only a few years ago, six billion hertz was plenty to manage Facebook, WeChat, and YouTube. But mobile wireless owns a fraction of that six billion so we are driving to frequencies far beyond what many of us consider our radio comfort zone.

I have seen multiple radio engineering labs coming to grips with these new frequencies. As 5G mmWave goes from obscure to elite to mainstream, the number of engineers doing component, subsystem, and radio design in these rarified wavelengths will skyrocket.

 

Many will have little experience with wavelengths no wider than their thumbs, and with bandwidths that sound like carrier frequencies. How will you set up your lab to ensure success? I offer the following questions and suggestions to those braving this territory.

 

1.  Which frequency bands are you targeting?

While 3GPP’s new radio (NR) development is aimed at carriers up to 100 GHz, I do not see a 5G wireless future in which this entire range will be used for access. So you not only have to anticipate which bands the policy groups will stipulate, you must speculate on which will be used for your target application spaces.

 

I also have doubts about 5G mobile multi-user access above 45 GHz. 802.11ad/ay will occupy the current 60 GHz band (and possibly the FCC’s extension of this band to 71 GHz). Point-to-point for backhaul, distributed antenna systems, and fronthaul will be implemented above 45 GHz. There is also early work in high-speed train communications up to 100 GHz (for on-board Wi-Fi “backhaul”).

 

Do you need to cover this entire range? Only part of it? Consider carefully because the tools and accessories become more expensive as you get closer to triple-digit gigahertz.

 

2.  What bandwidth do you need to support?

While there is talk about information bandwidths of 2 GHz, consider the following for frequencies below 45 GHz:

  •  Licensed bands will be divided between at least two licensees.
  • The widest in the FCC’s recent announcement is 425 MHz (28 GHz band)
  • The new air interface access designs are aimed at aggregating carriers modulated to no more than 200 MHz.

 

Notwithstanding your potential need to manage aggregated carriers and perhaps do work above 45 GHz, consider how wide—and thus how complex and expensive—you will want your lab to go.

 

3.  How will you connect to your device-under-test?

I have yet to see a serious design of a commercial mmWave transceiver system that includes a connector between the antenna and the amplifier. Thus, the labs I have seen all include anechoic chambers equipped with directional antennas with varying styles of positioners, and (often) open on one side. Smaller wavelengths and antenna apertures, highly directional propagation, and the lower likelihood of interfering signals allow for a different approach. But the requirement to make calibrated measurements in free space without violating regulations means a mix of enclosure, positioner, antenna, measurement equipment, and the necessary software.

 

4.  What software tools will your team need?

Your software arsenal ought to include six items: EDA; system design and simulation; EM simulation, measurement and analysis; device and test-equipment control; data manipulation and management; and mathematics tools. While the associated learning curve for your engineers is substantial, the productivity gains of working in the virtual world, particularly in the uncharted seas of small waves, quickly repay this investment. Then, the software-enabled power to generate test stimuli for radio components and systems, and analyze measurement and sample results, will give your designers new insights in time for your target release date.

 

5.  How do you future-proof your investments?

The world of commercial wireless is not for the faint of heart, and the foray into millimeter wave technology is an expensive step deeper into fraught territory. Short wavelengths mean exotic materials; tighter mechanical tolerances; and bandwidth, sampling rates, and digital speeds requiring significant CAPEX for a productive lab. CAPEX also implies these purchases must hold their value throughout and even after your depreciation period.

 

Successful organizations will future-proof these investments. Things to look for include capabilities to serve needs that arise during your depreciation period; modular software and hardware with an upgrade path; and proven vendors with technology-upgrade programs and expert support staff.

 

Wrapping up

Lastly, stay close to what is going on in the market. Your view of the considerations listed above will clarify as new policy emerges, 3GPP standardizes, and innovators make new and more capable technology available for your own building blocks.

 

Get more mmWave resources!