Friday, October 27, 2006

Cognitive Radio as Metaphor

Cognitive radio is the integration of software radio and machine intelligence. The term, coined by Joe Mitola, describes a radio communications paradigm where a network or individual nodes changes communications parameters to adapt to a changing environment that includes both external factors like spectrum usage, and internal factors like user behavior. It blends the promise of artificial intelligence with the excitement of software-defined radio. Science often advances through the exploration of inspired metaphors like this one.

Here are some examples of the language of Cognitive Radio, taken from a review article by Mitola and Maguire [1]:
Part of the processor might be put to sleep. The radio and the network could agree to put data bits in the superfluous embedded training sequence. A radio that knows its own internal structure to this degree does not have to wait for a consortium. Cognitive radio, then, matches its internal models to external observations to understand what it means to commute to and from work. A cognitive radio can infer the radio-related implications of a request. The radio also warns the network.

Computing research is often informed by anthropomorphism, which is obvious here. The anthropomorphism fits easily with the dominant mental model of wireless communications, described in Mental models for wireless spectrum:
  • Spectrum as Land
  • Signals as Moving Objects, or as Sound
  • Radios as Sentient Agents
Here are some (lightly edited) examples from Mitola and Maguire, and Haykin [2]:

  • The electromagnetic radio spectrum is a natural resource (Spectrum-as-Resource)
  • The underutilization of the electromagnetic spectrum leads us to think in terms of spectrum holes (Spectrum-as-Resource, Spectrum-as-Space)
  • White spaces, which are free of RF interferers except for ambient noise (Spectrum-as-Space, Signals-as-Sound)
  • The stimuli generated by radio emitters (Radios-as-Agents)
  • In some bands, cognitive radios will simply compete with one another. (Radios-as-Agents)
  • The game board is the radio spectrum with a variety of RF bands, air interfaces, smart antenna patterns, temporal patterns, and spatial location of infrastructure and mobiles. (Spectrum as Space)
It’s a small step from Radios as Sentient Agents to Radios as People. The “cognitive radio” moniker works in part because the groundwork for the concept has been laid by the model of radios as sentient agents.

Cognitive radio technology presumes that networks as well as individual radios have intellective behavior, as in “The network then knows that this user ...” The image of a sentient network does not fit into the existing model as readily as that of a sentient radio. One can thus expect that lay people will not take up the concept of cognitive network as readily. To the extent that the default spectrum/signals/radios model determines the scope of wireless regulation, networks (as distinct from the radios that comprise them) will fall beyond the ambit of regulators.

Cognitive radio thinking has some affinities with another mental model, “Wireless Communications as Internet.” This model is more recent and less well developed than the dominant spectrum/signals/radios model; it has been advanced by Open Spectrum advocates like Benkler, Werbach, and Weinberger [3]. Its salient elements are open standards, information transport, and decentralization of control. Since cognitive radio work focuses on system design, and the figure/ground relationship of devices in spectrum is less important than system architecture than in earlier, more passive technologies, the fit with the Wireless-as-Internet model is not surprising.

Even though Cognitive Radio is used by opponents of the traditional approach to argue for a complete rethinking of spectrum regulation, it is premised on the very structures they want to remove. The radio forms a hinge between the old and the new mindsets. The dominant perspective focuses on spectrum, then signals, and finally radios. The “Internet perspective” starts with radios as smart edge devices, and considers how they enable efficient information transport. Spectrum still plays a role as a means of communication, but it is no longer primary.


[1] Mitola, Joseph III, and Gerald Maguire (1999), “Cognitive Radio: Making Software Radios More Personal,” IEEE Personal Communications, August 1999, pp. 13-18

[2] Haykin, Simon (2005), “Cognitive Radio: Brain-Empowered Wireless Communications”, IEEE Journal on Selected Areas in Communications, Vol. 23, No. 2, February 2005, pp. 201-220

[3] See e.g. Yochai Benkler, “Some Economics of Wireless Communications,” (2002); Kevin Werbach, “Supercommons,” (2004); David Weinberger, “Why Open Spectrum Matters,” (2003)

No comments: