Advertisment

Will software-defined radio change the tech landscape?

Given the extensive deployment of software-defined radio technology across verticals, the concept is slowly but surely gaining ground

author-image
Sonal Desai
New Update
Software defined radio

Sonal Desai

Advertisment

The software-defined radio (SDR) technology is currently at the forefront of numerous advances within the wireless sector, enabling new applications considered unrealizable only a decade ago.

Given the extensive deployment of SDR across a growing number of applications such as defense, public safety, connected vehicles, education, and scientific research and development activities, it is important that the community at large understands the features, advantages, and limitations of this technology.

James Lyke, Senior Member, IEEE, and an Associate Fellow of the American Institute of Aeronautics and Astronautics (AIAA) who also serves on the AIAA Computer Systems Technical Committee, demystifies the concept with CIOL in an email interview.

Advertisment

What is a software-defined radio? How is it beneficial vis-à-vis hardware-based radio devices?

A software-defined radio (SDR) at the most basic level is circuitry designed to receive or transmit radio-frequency (RF) signals where circuitry normally built with classical fixed components (resistors, capacitors, and active circuitry including amplifiers, filters) can be tuned by software definition.

As in any reconfigurable system, we can imagine that one level that we have embedded a number of software knobs that can be turned under program control. In the most radical cases, much of the entire functional block diagram can be rendered as software description.

In the original vision of software radio, we would digitize all RF energy it intercepts the antenna and is manipulated by using digital signal processors (DSP) or field programmable gate arrays (FPGAs). Reprogramming the DSPs or FPGAs would in this case amount to redefining the entire radio.

Advertisment

As such, most real-world software radios perform some of the initial amplification and filtering operations in dedicated hardware, resulting in a signal more easily digitized and therefore managed through software manipulation. In comparison, previous radios implemented their functions with largely fixed patterns of circuitry, it is possible to render these patterns in software and change them without physically changing internal hardware.

How has the concept evolved in the last two years?

Most of the evolution has been in a growing awareness that it is possible to create circuitry even in inexpensive consumer devices based on SDR principles, and these concepts have begun to infuse discussions about future cellular telephony standards for example (5G).

Now the idea of software radio has become more tenable in many more application fields than previously possible before.

Advertisment

Networking and connectivity have reached a new dimension all together with Bluetooth, Wi-Fi and Wireless taking over the wired and analog systems.

The advent of software defined networking (SDN) makes it possible to restructure conductivity in application-specific ways on-demand.

We see the power of this technique is being driven by cloud computing, and been accelerated by the movement of the industry away from proprietary infrastructure in running routing equipment toward open systems.

When functions and applications can be rendered in such a way to eliminate the physical constraints of tethers or cables, it opens up many interesting use cases. Some of these are fairly mundane such as being able to relocate equipment without laying down physical conduits. Others are far more profound such as the ability to provision more bandwidth dynamically in response to immediate needs and to alter protocols to improve efficiency and security (for example).

Advertisment

Also with IoT, wearables and smart cities gaining ground, SDR will be critical. Have you mapped the various opportunities for SDR?

Given the phenomenal projections for the sheer quantity of Internet devices (estimates in some cases above 200 billion individual things, far more than the number of people on the planet), it is clear we’re going to have problems in terms of energy management and bandwidth congestion in wireless devices.

Here, we are concerned about the tension between flexibility (provided by reconfigurable systems, such as software radio) and energy efficiency (the optimality of customized circuitry versus the overhead of software-definable versions). Which of these will win is unclear, and IoT may drive hybrid realizations that seem at one level oxymoronic, such as application-specific software radios, designed to respond to the cost and energy pressures of IoT, while still providing adequate agility and exploiting negotiating spectrum (and to accept/evolve new protocols).

While the semiconductor industry is very excited over the sheer volume of silicon required to fulfill this vision of IoT, it is important to evolve the protocols (and many different concepts are being proposed) responsibly.

Advertisment

What are these opportunities? Can you give me specific examples with the market size and use case examples?

We believe that there are tremendous opportunities in the markets affecting the design of systems to be reconfigurable (e.g., the creation of platforms ranging from computers and FPGAs to software radios and future software defined x) and exploiting these reconfigurable systems and end-user applications.

Smart phones with software radios will be able to replace multiple transceivers with a single configurable RF chain, and then multiple copies of these reconfigurable radios can be co-integrated in the same system. Such a shift will allow any of these software radios to serve a multiplicity of purposes, and the multiple copies will allow multiple simultaneous instantiations as well as providing additional resilience (in the event of failure of one copy). When the entire radio infrastructure of a cell phone can be replaced with the software-defined version, it will no longer be necessary to work within the confines of legacy protocols.

We can envision the advent of apps that, for example, create application-specific waveforms and protocols for purposes both important and mundane. Some of these may include new interpretations of location-based services, new types of mesh-based communications, possibly to include new forms of Internet, in which groups of devices self-organize and self-scale without the need of traditional cellular base stations using protocols that achieve internet-like connectivity but in ways more optimized for dynamic, roving networks.

Advertisment

Internet protocols were originally created for static, scale-free network models. It has been argued that re-interpreting these protocols for dynamically-evolving random networks might lead to better formulations.

A reconfigurable infrastructure, in which even the primitive physical layer protocols can be software-defined, might lead to better results, such as ad hoc, peer-to-peer communications without centralized hubs.

What are the tactical and strategic decisions a CXO has to keep in mind while writing SDR strategy?

One of the more important considerations is spectrum management. Software radios in many ways transcend the environment of century-old thinking in the policies of allocating spectrum.

The popularity of wireless creates many challenges in the responsible use of spectrum, no matter what the policy constraints may be. There will be pressures to innovate increased bandwidth efficiency in protocols and more aggressive spectrum sensing and dynamic spectrum allocation technologies which typify the research in cognitive radio.

But reconfigurability can also be a challenge…

Reconfigurability can be a two-edged sword in another important respect. It is possible to create new classes of software-defined services over the air, to allow waveforms to redefine the radio itself.

The idea of placeholder services can be established which may permit the aggressive creation of platforms for which the details protocols have yet been worked out under the hope that they can be defined even after systems are deployed in the field.

The drawback will be if the adequate adaptiveness (expressive capacity) is inadequate to pull off the originally imagined services. Worse than that may be the consequence of having the inability to adequately reconfigure future flexible systems in a way to avoid undesired collisions in spectrum and interference.

How can CIOs and CFOs associate TCO and RoI with SDR?

By establishing flexibility (through reconfigurability the ability to dynamically defined functions in software) and opening the programming model and hardware APIs, we open also the possibility where the most innovation can win.

For those who believe the landscape of new start-ups will be in software-based applications and services, the advent of reconfigurability and software defined x, will add fuel to this fire. Innovators can simply regard reconfigurable radios and other types of reconfigurable subsystems as extensions of the computer and FPGA.

More details can be found on Proceedings of the IEEE

must-read iot-hub