Dynamic resource allocation in radio-over-fiber enabled dense cellular networks

Abstract

Network densification has emerged as a powerful paradigm to boost spectral efficiency and accommodate the continual rise in demand for wireless capacity. The corresponding reduction in cell sizes also results however in greater spatial and temporal uncertainty and variation in traffic patterns and more extreme and unpredictable interference conditions. These features create unprecedented challenges for efficient allocation of spectral resources compared to conventional cellular networks. As a further challenge, the allocation of spectral resources needs to be jointly optimized with the assignment of wavelengths in the optical backhaul of Radio-over-Fiber (RoF) networks, which are increasingly used in dense deployments and indoor environments. Motivated by these issues, we develop online algorithms for joint radio frequency and optical wavelength assignment in RoF networks. The proposed algorithms rely on load measurements at the various access points, and involve configurable thresholds for triggering (re)assignment of spectral resources. We provide a detailed specification of a system implementation, and conduct extensive simulation experiments to examine the behaviour in various scenarios and assess the impact of key parameters. The results in particular demonstrate that the proposed algorithms are capable of maintaining adequate load levels for spatially heterogeneous and time-varying traffic conditions, while providing favourable throughput performance.