Cross-Layer Estimation and Control for Cognitive Radio: Exploiting Sparse Network Dynamics

Abstract

In this paper, a cross-layer framework to jointly optimize spectrum sensing and access in agile wireless networks is presented. A network of secondary users (SUs) accesses portions of the spectrum left unused by a network of licensed primary users (PUs). A central controller (CC) schedules the traffic of the SUs, based on distributed compressed measurements collected by the SUs. Sensing and access are jointly controlled to maximize the SU throughput, with constraints on PU throughput degradation and SU cost. The sparsity in the spectrum dynamics is exploited: leveraging a prior spectrum occupancy estimate, the CC needs to estimate only a residual uncertainty vector via sparse recovery techniques. The high complexity entailed by the POMDP formulation is reduced by a low-dimensional belief representation via minimization of the Kullback-Leibler divergence. It is proved that the optimization of spectrum sensing and access can be decoupled via dynamic programming. A partially myopic access strategy is proposed, proving that it allocates SU traffic to likely idle spectral bands. Simulation results show that this framework balances optimally the resources between spectrum sensing and data transmission. More in general, this framework defines sensing-scheduling schemes most informative for network control, yielding energy efficient resource utilization.