A new probabilistic approach for adaptive frequency hopping

This paper proposes Utility Based Adaptive Frequency Hopping (UBAFH), an adaptive hopping technique implementing a new paradigm for frequency hopping systems. Traditional adaptive algorithms aim at identifying bad channels that are consequently removed from the hopset. UBAFH instead utilizes all the available frequencies but assigns different usage probabilities to different channels according to the experienced channel conditions. Opportune upper and lower limits are used in order to bound these usage probabilities and achieve a desired level of frequency diversity. We simulate the behavior of UBAFH over frequency selective fading channels and compare the achieved packet error rate and throughput with the ones of IEEE 802.15.1 and with the adaptive frequency hopping implementation proposed by IEEE 802.15.2 showing that UBAFH outperforms both approaches.