Topology inference for CSMA/CA wireless network based on inter frame space

Abstract

In this paper, we propose a novel approach called Wntiifs to infer the topology of CSMA/CA wireless network without the need to cooperate in the network or decode the network's messages. The approach divides RF frames observed into session atoms according to an inter frame space threshold, and ensures that a session atom consists only of frames transmitted by two nodes. Then, it extracts the link in each atom in different observation time window. Finally, it merges the links in different time window to reconstruct the complete network topology. Experimental results demonstrate that, compared with the existing wireless network topology inference algorithms, Wntiifs is of higher topology inference accuracy.