Distributed Power Control for Cognitive Radio Networks, based on Incumbent Outage Information Olasunkanmi Durowoju, Kamran Arshad and Klaus Moessner Centre for Communication Systems Research University of Surrey, United Kingdom E-mail :{O.Durowoju, K.Arshad, K.Moessner}@surrey.ac.uk Abstract—The interference management problem in cognitive radio networks is in this paper, tackled from the transmitter power control perspective so that transmissions by cognitive radios does not violate the interference level thresholds at incumbent users. We modify distributed power control algorithms to suit the cognitive radio framework by exploiting spectrum use and radio environment knowledge for incumbent user location estimation in worst-case scenario. Most literature employs worst-case analysis to guarantee robustness thereby trading off optimality. We therefore, propose a stochastic approach which allows the cognitive radio network, access the extra capacity based on incumbent user outage information with guarantees on interference protection to the incumbent user at all times. This paper therefore shows that the proposed distributed power control strategy is robust with the benefit of increased spectral efficiency compared to its worst case counterpart. I. INTRODUCTION Much literature advocates cognitive radio technology has a viable solution to the impending spectrum underutilization problem [1] leading to significant spectrum gains. However, with such spectrum gains comes the risk of increased interference to the licensed user of the spectrum called the Incumbent user (IU). Therefore ways of appropriate interference mitigation/avoidance techniques to IUs becomes exigent. In this paper we focus on Transmitter Power Control (TPC) for cognitive radios (CRs) as a way of curtailing excessive interference to IUs due to spatial co-existence while maintaining a reasonable quality of service (QoS) within the cognitive radio network (CRN). TPC [2-4] is an age long technique proven viable for interference mitigation in conventional cellular networks. Imperative is therefore the need to modify some of these algorithms to suit the cognitive radio framework [5-7]. Conventional TPC algorithm was first adapted to the CR environment in [5], where a sensor in proxy of the IU reports interference violations to the CRN. Such feedback reporting channel techniques are prone to inherent failures and expose the IUs to transient interference degradations. This anomaly was corrected in [6], where an autonomous power control (PC) algorithm was formulated with the ability of maintaining the IU interference environment unperturbed at all times given stringent interference conditions. Spectrum sensing based prototypes for interference mitigation in TV white space has been investigated and found viable in a series of experiment conducted by FCC [8] without serious degradation in TV signal quality given cognitive radio access. This led to the investigation of power control algorithms (PCAs) based on spectrum sensing information [9]. In our previous work, we extended the approach presented in [6] for the case of CRN, where multiple CR terminals individually estimate their channel to the worst case IU and coupled with some readily available information to formulate respective TPC algorithm [7]. The algorithm in [7] were fully distributed PCAs with primary protection via spectrum sensing and has the ability of ensuring that the IU environment remains unperturbed at all times with the opportunity of increased number of CR users. The approaches in [5-7] however considered worst case placements of incumbent receivers. Worst case scenarios considers, the aggregate interference contributions by multiple CRs to the nearest co-channel incumbent receiver to it, therefore CRs are denied access to the extra capacity achievable when IU system is able to tolerate more interference or in partial outage due to its planning i.e. grade B or noise limited contours are usually planned to cover 50% of locations, 90% of time following the F(50,90) curve [10], therefore the remaining 50% of locations which may be in outage can therefore be accessed by CRs. Worst case analysis generally trades optimality for robustness, we therefore in this paper eschew such conservative approach and provide a more efficient solution with strict interference consideration to IUs. Worst case analysis requires that the incumbent outage must be greater than 90% before CRs can be allowed to access licensed spectrum. Such power control strategy only provides an upper bound on the total number of CR terminals that can be supported. In this paper, we present a stochastic power control procedure based on incumbent outage information, which allows CRs to access spectrum when the incumbent is not totally in outage. To allow for fully distributed framework, we enable CR terminals with the prerogative of estimating incumbent outage information using spectrum sensing results. We show that even when we eschew worst-case conditions, a degree of CR terminals can still be accommodated depending on incumbent outage probability (IOP) without degrading the interference limit at the nearby receivers. We further advance the model to incorporate shadow fading which was not modeled in [5-7] and assume that fast fading is compensated by using appropriate coding and interleaving process. The benefits of this approach are manifolds: (i) it is a purely distributed approach and therefore very suitable for the CRN (ii) it provides for explicit incumbent protection at all times (iii) it is “optimal” as well as robust since the TPC strategy is formulated based on IOP (iv) it provides for increased 978-1-61284-231-8/11/$26.00 ©2011 IEEE This full text paper was peer reviewed at the direction of IEEE Communications Society subject matter experts for publication in the IEEE ICC 2011 proceedings