Boston, Mass.--Leading off the keynote speakers at the Embedded Systems Conference, cell phone inventor Dr. Martin Cooper (now Chairman of ArrayComm LLC, San Jose, CA) talked about the five problems of the wireless industry, along with his suggested solutions.
Dr. Cooper began by noting that the cell phone and wireless connectivity have not only made us mobile, they have changed our way of thinking: with wired communications, you called a phone, hoping to get the right person, but with wireless, you call the person directly. He sees these issues:
Despite what you assume, most cell calls are made indoors, yet the basestations are outdoors. This wastes power and impedes bandwidth reuse. His solution: increased use of femtocells and microcells installed in buildings. The backhaul and cellular management infrastructure will see some additional burden, but the benefits more than make up for it.
The Internet proves that an open network is good, but the mindset of cell phone industry's service providers is a holdover from the days of AT&T and wired phones, providing a "walled garden". This impedes innovation and makes the customer (the user) adapt to the product, instead of vice versa. He said we need open methods where different devices can be connected, not beholden to the service provider. As a positive example, he cited the ultra-simple Jitterbug cell phone he carries: it's a basic cell phone, no menus, no camera, simple keys for each of its few functions (a "power" key turns it on and off, not the "hang up" key). The recently announced Android platform may be a step in this direction.
Supporting wide bandwidth and high data rates costs real power (and money), yet providers splash the power around in all directions. Smart antennas, which dynamically focus their pattern to match user needs, offer a better alternative (ArrayComm's business).
Service providers have conditioned customers that they can get something (the handset) for nothing, and then try to persuade them to throw away that something for another something that is even better. A better solution is to allow users to buy just the phone the want, one that fits their needs.
The trend towards a combination-function, universal handset that does voice, video, music, data, and more burden both the handset designer and the end-user, and the product doesn't do any of these really well. Again, allowing the use to select a unit that more closely fits their needs and priorities is a better approach. User manuals which run 100+ pages of fine print, even for basic products, are indicative of the problem.
Dr. Cooper derided the fact that service providers have used their spectrum monopolies to control users and increase apparent capacity, while not actually delivering promised bandwidth or quality of service, although he conceded the later is improving.
His closing admonition was to remind designers that "good technology is transparent; the best technology is invisible." He cited the oft-derided automotive industry as an example of the latter: you can get into a car anywhere in the world and drive it, and the complexity of the processors managing the automatic transmission--itself a complex system of mechanical, hydraulic, and electronic functions, closely linked--is reduced to a simple shift control for the user: forward or reverse.
David Patterson, known for his pioneering research that led to RAID, clusters and more, is part of a team at UC Berkeley that recently made its RISC-V processor architecture an open source hardware offering. We talk with Patterson and one of his colleagues behind the effort about the opportunities they see, what new kinds of designs they hope to enable and what it means for today’s commercial processor giants such as Intel, ARM and Imagination Technologies.