By Chet Babla
User demand for more functionality and multimedia capability from portable consumer electronics (CE) technology is perpetually at full throttle. Such functionality convergence is driven by consumer need for convenient “always-on” wireless connectivity, and the desire to consume multimedia content of their choosing, and at a time convenient to them.
With today’s multiplicity of connectivity standards and methods to consume audio and video content, portable CE manufacturers are facing significant challenges when trying to deliver featureset convergence, convenience of use, and consumer affordability. Multiple wireless technologies, a small device form factor, long battery life, and an intuitive user interface must all be delivered at a price-point acceptable to the mass market, especially with the current economic environment.
The current architecture to support wireless and feature-set convergence in portable devices involves integrating multiple integrated circuits (ICs). Each new wireless application to be supported (e.g., cellular, Bluetooth, Wi-Fi, NFC, WiMax, broadcast TV and radio, etc.) typically requires the addition of another IC, or, at a minimum, an extra IP block on an existing chip.
The former solution will necessitate more board space, the latter option an IC re-design and new manufacturing cycle. Both of these options will certainly add cost to the solution and incur a time-tomarket latency.
And it gets even worse for the CE manufacturer: Though markets are global, implementations of wireless standards may not be—in fact, they often exhibit significant regional variations. As a result, manufacturers cannot easily deploy a global platform solution to leverage economies of scale and, hence, deliver the lowest-cost solution.
Combine the evolving demands of wireless technologies and consumer functionality with the growing consumer popularity of the ultraportable “netbook” PC and mediarich feature-phone, and you will see manufacturers continually scrambling to devise new approaches to deliver convergence.
An alternative method to addressing wireless convergence and consumer demand for media-rich functionality in CE devices is now gaining early commercial traction. Termed the “software modem,” this approach has the potential to simultaneously address the formfactor and cost constraints of delivering functional convergence.
The software modem concept is simple: a reconfigurable RF front-end is coupled with a digital baseband implemented entirely in software (rather than in the traditional form of a digital IC). If the front-end is designed with sufficient reconfigurability, then the inherent flexibility of software can deliver a true multi-standard wireless modem, giving consumers the converged wireless and feature-set choice they demand, but at a potentially paradigm-shifting price point.
The concept of replacing hardware functionality with software does have prior art: dial-up computer modems and MPEG video decoders are two common commercial examples. However, it’s the recent availability of powerful and cheap host processing power that enables the software modem concept to be seriously considered for mainstream CE applications.
Note also that today’s CE devices now often feature a secondary processing resource (e.g., a graphics processing unit or multimedia processor) in addition to the central host processor to address highdefinition video playback and 3D gaming. The availability of such secondary processors can also be leveraged by software modems to relax host processor loading.
Mirics recently demonstrated the commercial viability of the multimode software modem concept, and other companies are now working on commercialising software modems for CE applications. In its first step to deliver wireless convergence, Mirics implemented broadcast TV and radio reception on portable PCs, leveraging a software demodulator to support reception of multiple broadcast standards (see the figure).
Of course, broadcast is just the tip of the iceberg of how software modems can be leveraged. The flexibility of the concept is that other wireless technologies such as Wi-Fi, Bluetooth, NFC, and WiMAX can all conceivably be implemented as a single reconfigurable RF front-end, with softwaredefined basebands.
Clear thought must be given to analysing various wireless use cases. However, even if more than one RF path must be implemented on the analog front-end to support parallel radio operation (for example, broadcast and Wi-Fi), a software approach is still able to deliver cost and size savings by obviating the need for multiple digital basebands.
Cellular radios can also be integrated with the wireless technologies listed above. Nonetheless, it’s more likely that the various 2G through 4G cellular technologies will become a software modem subset in their own right. That’s mainly due to the complex resource-intensive application software and type approval requirements for cellular receivers.
As consumer demand for communications convergence and multimedia content consumption on ever-shrinking CE devices continues unabated, a new technology approach will be needed to meet functionality and cost goals. The software modem becomes a very real solution.