Connectivity optimization in interference-limited random DSA networks

Abstract

Abstract Network connectivity in dynamic spectrum access (DSA) networks has been well studied–most of which are under the unit disk model. However, the disk model does not capture the primary-secondary and secondary-secondary interference; hence signal to interference and noise ratio (SINR) based models are more appropriate. Moreover, in the SINR regime, there is no unique way to characterize connectivity and hence its maximization becomes even more challenging. In this paper, we develop the long eluding network connectivity objective function which we use to build three connectivity optimization techniques each of which targets a particular network setup. The proposed techniques are: i) Fittest deployment density, ii) Fittest receive-only ratio, and iii) Fittest TDMA slotting. To develop the aforementioned objective function, we start by addressing the lack of any relation between deployment density and network connectivity in interference-limited DSA networks. Next, percolation theory in conjunction with the Boolean model are utilized to develop such a relationship between the density of the percolation visible nodes and the network connectivity in interference limited environments. Finally, we use that relation to build the objective function for connectivity maximization along three optimization techniques. Theoretical findings are validated by simulating networks under various scenarios. Results provide a blueprint to establish and maximize connectivity using physical layer parameters (density, coverage radius, etc.) which can be used in conjunction with higher layer techniques. Also, tackling the connectivity problem at the physical layer relieves the other higher layers like the MAC layer from excess signaling and complex protocol designs.