This work presents the organization and operation of a hopping sensor wireless network. It defines two decentralized algorithms for the discovery of isolated nodes, aggressive and smart discovery, to propagate base station connectivity. Both algorithms rely on the usage of hopping sensors on the boundary of a cluster, without prior topology information. We studied the efficiency of the algorithms in terms of network energy consumption and connectivity propagation. A cluster to cluster (C-to-C) packet forwarding scheme, which relies upon boundary hopping sensor gateway nodes, is defined and simulated, and proves to have remarkable low energy consumption. In-field measurements were performed to obtain real data to serve as input parameters for simulations. Simulation results indicate significant savings in network wide energy consumption (57% for aggressive discovery and 79% for smart discovery).