A Dynamic Interference-Avoidance Algorithm for Frequency Hopping Systems

In this paper we investigate an algorithm for the Adaptive Frequency Hopping mechanism that is used by frequency dynamic systems to mitigate interference from othersystems. With this algorithm we introduce several improvements in relation to the existing algorithms that are based on the approach of using Packet Error Rate as the means for channel classification. One is the use of a single criterion for channel classification regardless of the dynamics of interfering systems, which adds more flexibility and reduces the risk of erroneous channel classification. The second is the introduction of the concept of channel probing which ensures that channels that areexcluded from the hopset are not used until they are clear from interference. The third improvement is the parameterization of the algorithm, which enables the control of the trade off between the main achievements of the algorithm: throughput and quickness of adaptation to changing interference. We show these achievements of the proposed algorithm through simulation