Improving Spectral Efficiency of Wireless Networks through Democratic Spectrum Sharing

Abstract: Wireless devices need spectrum to communicate. With the increase in the number of devices competing for the same spectrum, it has become nearly impossible to support the throughput requirements of all the devices through current spectrum sharing methods. In this work, we look at the problem of spectrum resource contention fundamentally, taking inspiration from the principles of globalization. We develop a distributed algorithm whereby the wireless nodes democratically share the spectrum resources and improve their spectral efficiency and throughput without additional power or spectrum resources. We validate the performance of our proposed democratic spectrum sharing (DSS) algorithm over real-world Wi-Fi networks and on synthetically generated networks with varying design parameters. Compared to the greedy approach, DSS achieves significant gains in throughput (~60%), area spectral efficiency ($\sim$50\%) and fairness in datarate distribution (~20%). Due to the distributed nature of the proposed algorithm, we can apply it to wireless networks of any size and density.