Viral processes by random walks on random regular graphs (Q2341622)

This paper studies the SIR model with infections carried by \(k\) particles making independent random walks on a random regular graph on \(n\) vertices with \(k\leq n^{\varepsilon}\). Let \(S(t)\), \(I(t)\), \(R(t)\) be the set of susceptible, infected, recovered particles at step \(t\), respectively. The following rules are observed. (i) if \(x\), \(y\) are on the same vertex \(v\), there is an \(xy\) interaction with probability \(\rho\); if they are on different vertices, they cannot interact; each particle pair (non)interaction is independent of every other particle pair (non)interaction. (ii) If \(x\in S(t-1)\), then \(x\in S(t)\), unless there was an \(xy\) interaction with at least one particle \(y\in I(t-1)\). (iii) If \(x\) was infected at time \(t(x)\), then \(x\in I(t)\) for \(t=t(x)+1,\cdots, t(x)+\xi\), and \(x\in R(t)\) for \(t>t(x)+\xi\). It was shown that in the subcritical regime \(O(\ln k)\) particles are infected; in the supercritical regime, for a constant \(0<\beta<1\), \(\beta k\) get infected with probability \(\beta\), and \(O(\ln k)\) get infected with probability \((1-\beta)\). A completion time of the process for the SI case was also derived.