By now, I’m sure you've read that mobile data demand is growing quite rapidly. It's no surprise, really: We're increasingly reliant on our mobile devices—be they laptops, netbooks, tablets, or smartphones. And the things we do with these devices increasingly require network access, whether it's access to the corporate VPN to collect email and trade presentation slides, access to the airline web site to change our flight reservation, or access to social networks and news sources to keep up with what's going on in the world.
If you're a highly mobile person and rely heavily on your Internet connection, searching for accessible, affordable WiFi connections everywhere you go just doesn't cut it. The cellular network is much simpler. I have a Qualcomm WWAN modem built into my laptop, and have used it for years on Verizon's network. Virtually anywhere I go within the U.S., I can get a data connection within 60 seconds. (Outside of the U.S., it’s a non-starter, unfortunately. Then I get to appreciate what life is like without WWAN service.)
Lately, though, I’m finding that while I can still get a connection virtually anywhere I go within the U.S., it's often a slow connection. Last week, for example, I was on the 18th floor of a hotel in downtown Austin, Texas. I could see practically to Dallas. I suspect that there were multiple Verizon towers within my line of sight. But my WWAN connection was painfully slow. My guess is that this was a symptom of too much demand relative to the available data bandwidth.
Beefing up the established cellular infrastructure with 4G technology, more-directional antennas, and the like can certainly help with this dilemma. But another obvious piece of the solution is small base stations. And, indeed, these so-called "picocells" and "femtocells" are now starting to proliferate after many years of discussion.
Why is it that things often move so slowly in the cellular world? One can argue that with the rollout of a new generation of cellular technology (e.g., 2.5G to 3G, 3G to 4G), there are many complex obstacles to implementation, including securing spectrum, installing new base stations, etc. But small cells don't suffer from these obstacles. They’re small, after all, so they can often just be bolted onto a wall. And they use the same spectrum as their larger "macrocell" base station cousins. And yet, while the industry has been talking energetically about small cells for a decade, it’s only now that we're starting to hear about large-scale deployments. Well, better late than never.
Small, lower-capacity base stations can provide coverage with a home, office building, or airport terminal. Small cells are attractive to wireless network operators because they enable operators to expand their coverage and their data bandwidth at much lower cost compared to deploying traditional base stations. And small cells are attractive to wireless network users (consumers and businesses) because they provide a simple, inexpensive way to improve coverage and data bandwidth.
A year ago I wrote about how cellular base stations have become a volume driver for digital signal processors, as wireless network operators build networks in new regions and beef up their existing networks. At the time, I was talking about traditional macrocell base stations. Today, small cells are emerging as another huge chip opportunity. Of course, the chips inside these small cells are not the hulking, multi-core, multi-hundred-dollar chips found in macrocell equipment; an entire femtocell unit typically sells for under $200 at retail, and they're often subsidized by network operators.
Chips for femtocells therefore must be highly integrated and cost-effective. But femtocells have the potential to ship in volumes much larger than those of macrocells. Market research firm Informa Telecoms & Media reports that multiple network operators have already deployed hundreds of thousands of femtocells each, with some operators expected to reach a million cumulative units deployed within the next two years.
So, yet again we see a wireless application becoming a high-volume consumer of digital signal processing chips. But, also yet again, getting to the point of widespread deployment has taken a lot longer than many people predicted.
Add new comment