Connectivity of Clustered and Multi-type User CR Network: A Percolation Based Approach

For the clustered, large scale ad hoc cognitive radio network with multi-type users, we address the percolation-based connectivity problem, in which the existence of a communication link between two secondary users depends on not only the distance between them, the transmitting and receiving activities of nearby primary users, but also the neighboring user’s type. From the mean-field approximation perspective, we firstly give the sufficient condition for the single type (here “type” means transmission radius) clustered secondary users on how the marginal nodes in the clusters are correlated to provide the critical percolation radius. Then the connectivity of the secondary users with inhomogeneous node distribution is studied, where two types of sub-critical secondary users are migrated into a super-critical cognitive user network through a multi-type branching process in random environment, which essentially is a percolation parameter optimization problem. Various simulations are performed to show the percolation is effective both in theory and practice in guidance of the deployment of the wireless network.