Figure 1

Coherence of measurement matrices as a fraction of nt. The networks size N is shown in Table 1, P is the number of experiments, and we select the measurement data from the time \(t=350\). (a) When the input vector \(u=0\), the coherences of six measurement matrices for these six networks, i.e. Celegansneural, Dolphins, Football, Jazz, ZK, Polbooks vary with the increment of nt (now \(nt=P/N\), \(0.2N\le P\le N\)). (b) When the input vector u is the standard Gaussian noise, the coherences of six measurement matrices for these six real networks change with the increment of nt (now \(nt=(P-N)/M\), \(1.2N\le P\le 2N\), and we fix \(M=N\)). (c) When the input vector \(u=0\), the coherence values of measurement matrices for these four networks (i.e. Erdos-Renyi random network (ER), Newman-Watts small-world network (NW), Barabási-Albert scale-free network (BA) and Watts-Strogatz small-world network (WS)) created through the linear network system change with the increment of nt (now \(nt=P/N\), \(0.2N\le P\le N\)). (d) When the input vector u is the standard Gaussian noise, the coherence values of these four measurement matrices for these four model networks change with the increment of nt (now \(nt=(P-N)/M\), \(1.2N\le P\le 2N\), and we fix \(M=N\)).