Autonomous opportunistic spectrum access in distributed femtocell networks

Femtocells randomly deployed in a given macrocell coverage area share the wireless spectrum available to the macrocell. The unplanned and ad hoc nature of the femtocell deployment in the macrocell environment renders centralized frequency planning inapplicable. Furthermore, the femtocells are unable to apriori know neither the channel assignment of their neighbours nor the impact of their transmissions on nearby macrocell users due to the lack of explicit coordination. In this paper, we present the distributed Autonomous Opportunistic Channel Access (AOCA) framework that allows the femtocells to share the available spectrum with the overlaying macrocell without any kind of coordination neither between the macrocell and femtocells nor amongst the femtocells. Furthermore, it provides statistical guarantees on the performance of the macrocell users. We formulate the AOCA constrained spectrum access problem as non-linear program to find its optimized parameters' settings. Simulation results show that AOCA achieves multiple folds improvement in the average femtocell network rate due to its probabilistic and non-greedy access that enables multiple neighbouring femtocells to simultaneously use a given channel.