Building cognitive radios in MATLAB Simulink

Cognitive Radio (CR) is a future radio technology that is aware of its environment, internal state and can change its operating behavior (transmitter parameters) accordingly. It is intended to coexist with primary users (PUs) for using the underutilized spectrum without any harmful interference. Its key domains are sensing, cognition and adaptation. This paper presents a detailed survey of coexistence techniques used in IEEE 802 wireless network standards family to reduce the interference between different types of networks. It also presents a prototype system for designing and testing cognitive radios built on top of software defined radio in a MATLAB/-Simulink and interfaced with a universal software radio peripheral-2 (USRP2) main-board and RFX2400 daughter board from Ettus Research LLC.

The philosophy behind the prototype is to sense, predict and adjust the operating parameters to achieve the desired objectives. The PUs detection (sensing) is performed using the statistical (non-parametric) spectrum estimation techniques, which are more useful in a noisy environment. Regression-based statistical time-series modeling & analysis are carried for PU’s channel behavior prediction for spectrum decisions. Software-defined radio is used for reconfigurability to adjust the operating parameters of the cognitive radios.