A PCI card is fine for developers but it's not the most convenient form factor, and, at $750, Noctar isn't cheap. But this is by no means where the Per Vices vision ends, with Wollesen explaining that their "ultimate goal is to build a device and position it as a Wi-Fi router and femtocell, with the capability of having applications developed for it which take advantage of the SDR capabilities."
Having outlined their route into a much larger market, Wollesen goes on to say: "Software-defined radio is really an infrastructure technology, so we're trying to incorporate it into a way that is easy for consumers to use and appreciate, and that allows software developers to take advantage of this platform as a means of accessing wireless information and manipulating it."
The implications of such a device could be huge and, for the end user, it may mean that hardware upgrades become few and far between, with the addition of support for a new wireless system being simply a matter of installing additional software. This would open up all manner of possibilities for software developers, as they become able to directly manipulate large swathes of radio spectrum, instead of being marshalled through multiple restrictive, and much narrower bandwidth, single-purpose systems.
Per Vices is aiming high and Wollesen describes their ambition using a Cisco analogy: "Unbeknownst to the average consumer most of the internet traffic is handled by Cisco routers. Cisco routers don't care about whether you're sending e-mail or looking at YouTube videos, their job is to handle IP traffic in all its forms. By extension that's the role we're trying to set our company as: to handle wireless traffic in all its forms, without necessarily being focused on the application – quite the opposite."
Before this SDR-enabled wireless panacea can become a reality there are additional hardware challenges that must be overcome, such as how to process the enormous volume of data involved within practical cost and power budgets. But, as with the early days of computing, motivation must first be provided by software which demonstrates the opportunity at hand, and this isn't quite there yet.
The multi-system stack
GNU Radio has provided the fundamental building blocks for the open source SDR ecosystem and an opportunity now exists to use these to develop a multi-system stack. When combined with hardware such as the Per Vices Noctar, this would be capable of replacing many discrete wireless systems.
This would require substantial engineering effort, but projects such as OpenBTS – which is used to create an SDR GSM base station – have demonstrated what small, talented development teams and open source are capable of. Open source SDR implementations in varying degrees of completeness are also available for GPS, DECT, Bluetooth, 802.11 and LTE (4G), among others.
Given the sheer complexity of many systems and the expansive nature of the landscape, this is likely to be too much for a single project to address. But this is where open source presents an opportunity to minimise risk and to succeed where costly initiatives such as the Joint Tactical Radio System (JTRS) – a US military multi-system SDR program reported to have cost $6 billion – have failed.
It now remains for two or more popular wireless systems to be integrated via a common SDR platform in order to underline the opportunity. In practice, it would make sense for these to be selected based on their use of shared or adjacent spectrum, so as to limit the bandwidth and therefore the volume of raw data that would need to be processed in real time.
A wideband multi-system SDR would not only enable device consolidation and bring upgrade convenience, but would also provide an immensely powerful platform for wireless innovation. It would make it possible to rapidly deploy new systems, and open the door to "cognitive radio" that is spectrally aware and able to far better coordinate access to an increasingly scarce resource.
The component architecture of radio systems is well suited to distributed development, with open source providing a means of addressing a major opportunity that comes with certain challenges. The risks associated with a proprietary multi-system SDR are extremely high, and the move to software-driven innovation and market forces means that commoditisation is inevitable.
End user wireless devices must be certified and this would provide commercial opportunities along with those typically associated with open source, such as customisation, integration and support. Key patent encumbered standards may need to be licensed for use in products, and performance demands particular to wideband SDR would create new opportunities for hardware manufacturers.
The SDR-powered Wi-Fi router and femtocell may be a little way off yet, but when it arrives it will be so much more than this, and the impact on how wireless systems are developed will be profound.