Lowenstein's View: The state of mobile broadband


Mark Lowenstein

I spent the better part of last week in Chicago at 4G World, which provided a good opportunity to assess the "state of the network." To me, there are three meta-questions about the evolution of mobile broadband:

  1. Will wireless networks be able to meet the demand for data, given the pace of device innovation and the growing requirement for near constant connectivity?

  2. Are there any "disruptive innovations" on the horizon that change the way networks are used or deployed, dramatically improve network performance, or alter the economics of data delivery?

  3. What is the operator mindset these days, not only from the standpoint of network investment but in leveraging their network resources and offering new features and services?

Before proceeding, a quick aside: I'd like to ask for a six-month moratorium on starting a presentation at a conference with: "Data demand is growing by eight gazillion yotta-terra-super bytes per day" or "There will be ten million quintillion new connected devices before the end of this presentation, all wanting to stream Netflix over mobile 24 hours a day." Thank you.

There is no question that spending on wireless networks will contribute its fair share to global GDP for the foreseeable future. However, the march toward true mobile broadband is going to be a decade-long phenomenon. It will be highly dependent on country/operator/spectrum/population density/investment density/data demand. Along the way, there will continue to be "flavors" of 4G, where we will see, for example, HSPA networks that outperform LTE networks, based on a host of factors. Numerous entities have tried to define what a "4G" experience is, and operator marketing organizations have been rather liberal in their interpretation of 4G. My own benchmark for what should be called a "mobile broadband experience" is 5 MB down, 2 MB up, below 100 milliseconds of latency, 75 percent of the time.

Keeping up with data demand will be an ongoing challenge. What is changing is that once a critical mass of macro base stations are enabled and fiber/Ethernet backhaul deployed to the majority of sites, network deployments get a lot more tactical. It's the wireless equivalent of door-to-door combat. Having spent some private time at 4G World with a couple of senior operator network executives, two things impressed me: the varied and vast arsenal of tools they are using to help meet demand, improve coverage and expand capacity; and how they have moved from reactive to proactive mode on a number of fronts. A great example is the Small Cell Forum. Whereas a few years ago, operators would sit, Solomon-esque, while the vendor ecosystem organized itself, now they are driving many of the Forum's initiatives.

There was in-depth discussion of three areas critical to the evolution of mobile broadband: spectrum, small cells, and backhaul. My quick take on each:

  • Small Cells. The needle on small cells has moved from whether, at 4G World 2011, to when, at 4G World 2012. Femtocells for homes and small businesses represent the vast majority of small cell deployments today. As for microcell and picocell deployments, and the broader "HetNet" vision, the technology has come a long way, the economics have improved significantly, and major operators are taking a very serious look. One validation point is AT&T's trials with the self optimizing network (SON), which helps shift traffic to neighboring towers and different size cells, as needed. The next year will see important small deployments and trials. Key issues around interference, backhaul, and zoning need to be tackled. Make no mistake--HetNet deployments are complex.

  • Spectrum. "Critical Part of 21st Century Infrastructure" is now part of any spectrum stump speech. There is a TON of activity going on regarding spectrum these days. OK, hold your breath: NPRM on incentive auctions, AT&T WCS, Dish decision, hearing on using 3550-3650 MHz band for small cell, PCAST report on spectrum sharing, TV White Space, federal/DOD spectrum reallocation. Now exhale. The bottom line: there is some relief on the way, but outside of private market solutions (trading, M&A, etc.), no new spectrum will be commercially available before 2016. There is growing consensus that what's needed is a combination of more spectrum (as in, more roads), and more effective use of spectrum that's already out there (spectrum-sharing, channel aggregation). It's all politically complex and expensive. Commissioner Pai, live via video from Washington, D.C., was typical FCC: earnest but non-committal.

  • Backhaul. This is probably the un-sexiest big issue in the wireless industry. Tremendous investment has been made over the past several years in upgrading backhaul in the macro network. One of the hottest innovation sectors in mobile is the number of companies focused on providing high capacity backhaul on a more favorable economic basis, everywhere from ex-urban and rural areas to urban markets and their combination of cell sites and Wi-Fi hotspots. Microwave, historically a small part of the backhaul market in North America, is making a comeback.

Rather than a particular "outside the box" product innovation, I was most impressed by the relentless effort to improve efficiencies in network components. There is also notable progress in other key areas:

  • Size and economics of small cells, and even the ability to run an eNodeB (LTE base station) off a PC, as demonstrated by Amari.

  • Better integration of commercial-grade Wi-Fi into an overall network plan, as envisioned in the Hotspot 2.0 initiative.

  • Microwave backhaul, "backhaul as a service,"  and Alcatel-Lucent's demonstration of bundling backhaul with a small cell.

  • Video optimization for mobile.

  • MIMO antennas.

  • Carrier aggregation, which allows operators to combine channels across their holdings in different parts of the radio spectrum.

Two larger companies impressed me with the steady progress they are making in mobile: Broadcom, which is becoming to wireless network infrastructure what Qualcomm is to devices; and Cisco, which is playing an ever-broadening (and profitable!) role across the portfolio of local and wide area wireless networks.

From the standpoint of service innovation and the "smart vs. dumb pipe" discussion, I'd say it was a mixed bag. Some wireless operators have gone from talk to action and are opening some network APIs to the developer community. We are likely to see this exposed in services and applications related to location, analytics, and payments, over the next year. Less visibly, this initiative should help address network-related security issues.

On voice over LTE (VoLTE), outside MetroPCS' announcement of commercial availability, the operators were notably cautious. It looks to me that some are taking a wait-and-see approach, while others have pushed large-scale launches into 2014.

There was also meaningful discussion around Quality of Service (QOS), which can mean a lot of things but to me represents opportunities to offer some level of higher grade or priority service; a way to partner with OTT providers; and a possible approach to video. Samsung alluded to this capability in its discussion of the "smart network." Some of this clearly gets us into network neutrality territory--so we're going to have find some loopholes.

It's easy to complain about imperfect cellular coverage and overburdened networks. But a week spent in Chicago-land, in wireless infrastructure-land, reveals tremendous innovation and huge spend focused on improving wireless network speed, capacity and efficiency.  

Mark Lowenstein, a leading industry analyst, consultant, and commentator, is Managing Director of Mobile Ecosystem. Click here to subscribe to his free Lens on Wireless monthly newsletter, or follow him on Twitter at @marklowenstein.