Steady-State Rate-Optimal Power Adaptation in Energy Harvesting Opportunistic Cognitive Radios with Spectrum Sensing and Channel Estimation Errors

We consider an opportunistic cognitive radio network, consisting of Nu secondary users (SUs) and an access point (AP), that can access a spectrum band licensed to a primary user. Each SU is capable of harvesting energy, and is equipped with a finite size battery, for energy storage. The SUs operate under a time-slotted scheme, where each time slot consists of three non-overlapping phases: spectrum sensing phase, channel probing phase, and data transmission phase. The AP feeds back its estimates of fading coefficients of SUs-AP link to SUs. To strike a balance between the energy harvesting and the energy consumption, we propose a parameterized power control strategy that allows each SU to adapt its power, according to the feedback information and its stored energy. Modeling the randomly arriving energy packets during a time slot as a Poisson process, we establish a lower bound on the achievable sum-rate of SUs-AP links, in the presence of both spectrum sensing and channel estimation errors. We optimize the parameters of the proposed power control strategy, such that the derived sum-rate lower bound is maximized, subject to an interference constraint. Via simulations, we corroborate our analysis and explore spectrum sensing-channel probing-data transmission trade-offs.