Reliable spectrum sensing for resource allocation of cognitive radio based WiMAX femtocells

Femtocells are perceived as having a strong potential to provide high data-rate services with increased coverage at low cost. In this paper, we consider a WiMAX network with both macro base stations (BS) and low-power, low-cost Femto Access Points (APs). We assume that Femto-APs share the same bandwidth as macro BS and Femto-APs are connected to the macro BS by wired backhaul such as cable or DSL. One of the important technical challenges in this framework is to mitigate the interference from macro users by avoiding their spectrum. Robust and dynamic spectrum sensing for channel occupancy by macro users is therefore necessary to schedule resources for femto users. An efficient cyclostationary-based spectrum sensing method is presented to facilitate cognitive radio (CR) aided resource allocation in femtocell networks for maximizing system throughput while minimizing inter-tier interference. Contrary to the classical energy detection (ED) approach, a cyclostationary-based spectrum sensing is more effective in dealing with imperfect knowledge of in-band noise variance in practical radio signals. Also, this method is more efficient than ED in low signal-to-noise ratio which is a very important requirement in WiMAX based macrocell-femtocell overlay network.