Bench Talk

Figure 1: International Microwave Symposium attendees try out virtual reality.  Image: WTWH Media, LLC.

 “We are at the beginning of the golden age of wireless,” one test and measurement executive was overheard to say at last month’s International Microwave Symposium (IMS) 2016 in San Francisco. “It’s probably the most exciting time I’ve experienced in the 30-year period of my career.”

And so seemed the feeling amongst the 600+ exhibitors and over 9,000 attendees at one of the industry’s premier events for technologists involved in all aspects of microwave theory and practice. Between the sessions, papers, workshops, and exhibit halls you got the sense that suddenly the super-hyped world of IoT, M2M, and V2V have finally intersected, and perhaps breathed new life into, the world of microwave, Millimeter wave (MMW), and 5G. At a time where the number of people in the world (6 billion) is exceeded by the number of cellphones…now what? IoT is what. M2M, V2V, IoT. Talk around Narrowband IoT (NB-IoT) or what some refer to as 4.5G, is seen as an evolutionary step towards 5G driven largely by need to make those billions of connections possible.

Following are some of the interesting observations as gathered from two key sessions and booth visits.

Implementing 5G creates new business models in unexpected places

Over 500 terabytes of new data touches Facebook databases every day. That Boeing 737 some attendees took on their flight from the East Coast to San Francisco? It generated potentially 240 terabytes of flight data. More frequency, more bandwidth, more channels. 5G might be the future but is it everything to everyone? This was a question up for debate from both exhibit floor and podium. 5 GHz signals hold the promise of pretty serious data rates up to 10GB/sec on a wireless device. That’s an order of magnitude over “4.5G,” which runs at 1G – 1.5 Gb/sec. And what do you want that for? The consensus seemed to center around “densification applications” – a sports stadium, for instance, where you have 100K people tapping into their networks.

Figure 2: “The 5G IoT Conundrum” panel. From left to right: Dr. Thomas Cameron, Analog Devices; Eric Starkloff, National Instruments’ Satish Dhanasekaran, Keysight Technologies; Tony Ofperman, Rohde & Schwartz; Robert Donahue, Anokwave; and moderator Eric Higham, Strategy Analytics. Image: WTWH Media, LLC.

As for user devices, 5G design is apparently expected to take us to 2040 or 2050. In a panel session entitled “The 5G IoT Conundrum”, National Instruments’ Eric Starkloff posed the question “What will a 5G user device even look like 20 years from now? When 4G was being conceived, I doubt anyone was thinking of the applications we are using now.”

Beyond the majority of low-battery applications that populate IoT today, there are plenty of big data applications that may speed us to 5G by sheer necessity. Tesla and soon, Toyota, are using cameras and radar data on vehicles to create better maps. It’s like a “sensor or even a data center on wheels” as one speaker put it. This means an order of magnitude of Gbps per car, scaling to millions of cars. What do Tesla and Toyota get out of it? Highly accurate maps and a business model that didn’t previously exist.

Coexistence and the network

But how does the network handle a projected 30 – 50 billion devices? While there may not yet be 50 billion connected devices, they are growing. The schedule for 5G broad commercialization deployment is somewhere around 2020. That provocative question elicited some equally provocative answers. But panelists generally agreed that 5G will handle that many, because there won’t be just one network handling conflicting requirements or a single technology dominating the network. It will be a mix of technologies that brings its own challenges. “An interesting thing that will continue to happen is around coexistence,” said Starkloff. “5G will have multiple radio access technologies and it will also have to coexist with other standards, including standards that are in unlicensed bands.”

Are you there, operator?

Panel members asked: “Is your next data plan a data plan of connected devices? Telco operators are reinventing themselves beyond being merely cell phone service providers. They are asking the questions ‘Where do we want to go? What applications do we want to develop?’”

Keysight Technology’s Shanasekaran said, “5G’s adoption has more to do with the operator business models than it has to do with the broadband use case or latency. In many ways, it is an outside view of technology adoption versus where we’ve been in the last decade.” This would be a heartily welcomed revenue stream for the operators, perhaps making that alone a major driver of 5G. However, from the test and measurement perspective, the cost of testing these systems “… hasn’t scaled down at the same rate as Moore’s law from a cost point of view,” said Starkloff.  “Test costs of devices are going to have to approach the cost of testing, say, audio. If I have a cell phone five years from now that has a MMW radio, I expect Moore’s Law to give me 10x better bandwidth and all these new features without paying more. But we are going to have to scale test costs down at MMW in order for that to be commercially viable.” The MMW market, including various segments, is projected to reach US $7.14 Billion by 2022. 

Keep up spec

The general view is that the electronics industry is pushing traditional standards bodies much harder than in past decades, driven by intense consumer demand. So 802.11ax and 5G can’t come fast enough. Manufacturers are already designing devices ahead of spec.  (Recall that 802.11ax may realistically deliver speeds above 2Gbps with a possibility of 10Gbps in perfect conditions.) “Something very new and different for us as engineers,” said Anokiwave CEO Bob Donahue, “is looking at the architectural solutions behind that and how do you get there first. It’s like running in an earthquake – how do you create those solutions when standards are changing and all the frequency bands are in chaos and they’re not set? But that’s also where the opportunity is.”

Millimeter Wave – the undeveloped band

Conversations around MMW wave applications leaned toward the replacement of “last-mile cable” – a well-recognized headache for cable companies attempting to deliver high-speed services to customers. In this scenario, the cable company would transmit a MMW wave beam at a house instead of digging up the front lawn. But the lack of connections with the integrative, active array antennas used in MMW frequencies poses a significant testing challenge. Panelists agreed that measuring over-the-air puts totally new requirements on production environments.

New products featured at IMS

Infineon demonstrated their radar-based sensing chips including the BGT24ATR12, what they refer to as their “self-contained radar on a chip” designed for short-range radar ADAS applications. Infineon also featured a demonstration of the new 60 GHz gesture sensing technology – Soli radar technology – a collaborative effort with Google Advanced Technology and Projects (ATAP) with multiple HMI applications.

Figure 3: Demonstration of Infineon’s 60 GHz gesture sensing technology in collaboration with Google ATAP. Image: WTWH Media, LLC.

Texas Instruments was an anticipated exhibitor, featuring a new RF-sampling analog-to-digital converter (ADC). The Texas Instruments ADC32RF45 is a 14-bit dual-channel ADC that enables direct RF signal conversion up to 4 GHz, giving engineers access to the highest dynamic range and input bandwidth.

Figure 4: TI’s ADC32RF45 supports out up to 4 GHz of bandwidth for direct RF sampling. Image: WTWH Media, LLC.

Integrated Device Technology’s (IDT) growing RF business supports the demand for products such as cellular base stations, distributed antenna systems, broadband repeaters, and microwave backhaul network equipment. IDT’s Dirk Petrini stated, “We are moving from GaAs to silicon as silicon is more reliable in the 60GHz space. It’s a different realm we are playing in where new markets are opening up in IoT, cable modems, and aerospace.” IDT showed their latest low insertion loss, absorptive RF switch, the F2923 SP2T RF switch that covers a broad frequency range from 300kHz to 8000MHz. In addition to providing low insertion loss, industry leading isolation at 2GHz and excellent linearity, the F2923 also includes a patent pending constant impedance (K_Z) feature, “providing constant impedance on all ports during transitions without compromising isolation, linearity, or insertion loss.”

Figure 5: IDT’s F2923 is designed for wireless and RF applications that cover a broad frequency range from 300kHz to 800MHz. Image: IDT.