We focus on the direct information exchange problem in the presence of an
adversary. In the direct information problem, a group of wireless nodes use a
lossless wireless channel to exchange a set of packets, such that each node
has
a subset of packets available to it as a side information and needs to obtain
all other packets in the set. We consider a setting in which some of the nodes
are controlled by the adversary and propose an algorithm that allows each node
to obtain correct packets in this setting. We analyze the computational
complexity of this problem and propose an approximation algorithm that
minimizes
the required number of transmissions.