Optimizing for sparse training of Cognitive Radio networks

Abstract

In order to find a configuration suitable to fulfill its needs, Cognitive Radios search the parameter configuration search space using one or more particular algorithms or heuristics. While each individual configuration tested uses a similar cost for evaluation (for example in airtime, computational cost for evaluation or power), many configurations will not yield any value to the radio and their exploration turns out to be a waste of resources. This paper introduces the application of fractional factorial designs to Cognitive Radios (CR), a technique to drastically prune the parameter search space while still yielding good results, thus enabling CRs to find the best possible configuration fast while using less resources for the search. We show that by using this technique CRs can evaluate a fraction of configurations while still correctly estimating the factors influencing its performance.

Local control of cognitive radio networks

“…Although physical proximity would certainly achieve this objective, a cognitive radio network has a multitude of other parameters it can adapt to shield itself from outside influences and strengthen its links to neighboring peers, for example, by modifying its transmission power, frequency setting, modulation scheme, or encoding parameters. The cognitive radio can then statistically analyze how each parameter setting and the resulting parameter interactions would affect the target variable it intends to improve, which could be efficiently learned through the use of fractional factorial designs [12]. If it has the objective to improve network communication by reducing the impact of outside interference, a cognitive radio could, for example, increase its transmission power.…”