Optimization of cognitive radio systems with energy harvesting
Date of Issue2017-05-18
School of Electrical and Electronic Engineering
Cognitive radio (CR) enables dynamic spectrum access and offers a promising solution to the impending spectrum scarcity problem. It senses the licensed spectrum and identifies spectrum opportunities which can be utilized by the secondary users (SUs) without disrupting the primary user (PU) transmission. Hence, it improves the spectrum efficiency by opportunistic spectrum access. Power is another critical resource. Recently, a lot of work has been focused on utilizing energy-harvesting techniques to achieve self-sustaining wireless networks. Adding energy-harvesting capability will further enhance the ability of CR systems. However, the random and sporadic nature of the energy arrival poses challenges in the realization of such self-reliant CR devices. We identify and address some new interesting research problems that arise in different types of energy-harvesting CR systems. Our research is focused on finding the optimal spectrum sensing and access policies to maximize the opportunistic spectrum utilization while considering the energy uncertainty at the CR users and preventing intolerable interference to the PUs. We first consider an opportunistic energy-harvesting CR system. Low-power CRs are equipped with rechargeable batteries to harvest radio-frequency (RF) energy from the PU transmission. A stochastic geometry based multichannel CR system is considered where each PU channel has a different PU density offering varied opportunities for harvesting energy or transmission for the SU. At any time, SUs can access a PU channel to exploit either the spatiotemporal spectrum opportunities for transmission or the PU busy periods for RF energy harvesting. A decentralized channel-access strategy is proposed for the multichannel energy-harvesting CR system. An optimal channel-selection probability vector is obtained to achieve the maximum SU throughput by finding sufficient spectrum-access and energy-harvesting opportunities while accounting for SU contention on the same channel. Numerical results show that the proposed channel-selection policy enables an energy-harvesting SU to achieve better expected throughput by utilizing greater spectrum-opportunities through an enhanced probability of having a charged battery as compared to the case when only the spectrum opportunities are optimized. Furthermore, we investigate the opportunistic RF energy-harvesting CR system in the context of an unslotted PU system. We optimize the sensing intervals to balance between energy harvesting and spectrum access, considering the PU traffic statistics that affects both the energy arrival and consumption in addition to the opportunistic spectrum access. We propose that in such a scenario, different sensing intervals should be used if the CR user decides to transmit or harvest energy. The harvesting and transmission durations can be used to optimize the tradeoff between maximizing the utilization of PU spectrum opportunities and harvesting sufficient energy to transmit. Optimal transmission and harvesting durations have been found to maximize the CR throughput while protecting the PU. Furthermore, we enhance the system capability by considering that the SU can choose to purchase energy from the PU during the PU idle periods in addition to the energy harvested opportunistically. It can further exploit the disparity in the PU busy and idle periods as per its energy needs. The average energy cost spent in a time slot is minimized to achieve a given SU transmission rate while adhering to the PU collision constraint. Next, we study the advantages of employing cooperation in an energy-harvesting CR system. We consider a CR system where each SU harvests energy to recharge its battery. Cooperative spectrum sensing is employed to enhance the sensing performance. We propose that in this energy-constrained scenario, the SUs can be coordinated to achieve higher opportunistic spectrum utilization and improved PU detection. For a heterogeneous CR system with different sensing environment and energy-arrival process at each SU, the optimal probability of sensing is obtained for each SU to maximize the SU network throughput while satisfying the PU collision and energy-causality constraints. Moreover, under a homogeneous scenario, it is shown that with a minimum number of SUs cooperating, the CR throughput performance is no longer limited by the energy-causality and PU collision constraints. Furthermore, we propose a dynamic sensing-access policy for the cooperative CR system. We employ centralized cooperative spectrum sensing, and the optimal sensors are chosen from among the SUs with varied received PU signal power levels and energy-arrival rates. We consider a discrete Markov chain based PU traffic and the proposed policy exploits the correlation in PU occupancy states to make decisions in each time slot based on the SU energy levels and the believed PU channel state. The finite-horizon partially observable Markov decision process (POMDP) framework is employed to optimally schedule the SUs for sensing and access to maximize throughput with the available energy while satisfying the PU detection constraint. The sensing-access policy decides whether to sense, the set of sensing SUs, the sensing detection threshold, and the SU that accesses the spectrum opportunities. The optimal policy achieves an optimum tradeoff between SU transmission, preventing energy shortage, and obtaining sensing information for the future gains.