Asynchronous Gossip in Multiplex Networks

In this work, we take a first step to investigate information spreading in multiplex networks, adopting the gossip (randomwalk) based model. Two key features of multiplex networks allow potentially much faster information spreading: availability of multiple channels and communication actions for each user, and more choices on neighbor contacting. As a first work in this area, we explored the impact of layer number, layer similarity, and average node degree on the efficiency of information spreading, and theoretically proved our results. Another observation is that multiplex network structure can improve network connectivity. Based on the analytical results, we also propose a new recommendation algorithm to build up new social layer that have low similarity with previous layers from the social network operator perspective. From the user’s aspect, we suggest a new modified contacting strategy applied in gossip algorithm in order to help the user spread their information more efficiently in a multiplex social network environment. Network simulations are performed to justify our theoretical results and the advantage of the new proposed algorithms.