Distributed versus cluster-based cooperative linear networks: A range extension study in Suzuki fading environments

This paper studies two topologies for cooperative multi-hop linear networks; a distributed equi-distant node topology and a co-located group of nodes topology such that both topologies operate under composite shadowing-fading environment. The transmission from one hop to another in both topologies is modeled as a Markov process where the underlying channel model is drawn from a Suzuki distribution. The distribution of the sum of multiple Suzuki random variables (RVs) is obtained by a moment generating function (MGF)-based approximation technique. It is shown that the coverage of both topologies is different contingent upon the severity of shadowing and the transmit power of the radios. The optimal level of cooperation between nodes is shown to have a dependency on the path loss exponent such that a large number of cooperators is optimal for a small path loss exponent and vice versa for obtaining the maximum range of the network. Monte Carlo simulations are performed to validate the analytical model.