Adaption Overhead in Time-Varying Cognitive Radio Channels

The ability to adapt to a changing RF environment is a major feature of cognitive radios. As the interference and noise characteristics change, the radio should be able to select a waveform optimal for a particular scenario. However, more frequent adaptation requires more frequent distribution of radio parameters, increasing the overhead and decreasing available proportion of capacity for transmitting data. This paper presents a basic model for analyzing modulation adaption overhead in a point-to-point cognitive radio link, and simulates a few specific cases of interest. We examine how quickly a radio can adapt without that overhead actually causing worse overall performance. Our major results are as follows: to decrease overhead cognitive engines should be placed at receivers, not transmitters; modulation adaptation in Rayleigh fading environments is only beneficial under high-SNR conditions; the higher the average SNR, the more frequently we can adapt to fading channels and see rate gains; and adaptation to interference channels offers a significant performance gain.