Throughput enhancement for secondary users in cognitive network Imen Sahnoun Higher School of Communication of Tunis, Sup’Com Ariana - Tunisia sahnounimen@supcom.rnu.tn Ines Kammoun National School of Ingineering of Sfax, ENIS Sfax - Tunisia ines.kammoun@ieee.org Mohamed Siala Higher School of Communication of Tunis, Sup’Com Ariana - Tunisia mohamed.siala@supcom.rnu.tn Abstract—Cognitive radio is an exciting emerging technology that has the potential of dealing with the stringent requirement and scarcity of the radio spectrum. Such revolutionary and transforming technology represents a paradigm shift in the design of wireless systems as it will offer the ability of radio sensing, self adaptation and dynamic spectrum sharing. In this paper, we are interested in the case where secondary users are allowed to communicate concurrently with primary users provided that they do not create harmful interference to the licensed users. More precisely, the secondary transmission will depend on a variable cost that reliably quantifies the interference power generated by the secondary user at the primary base stations. Here, we aim to improve the unlicensed system performance in terms of throughput. For this aim, we propose to use an adaptive modulation as well as an energy optimization at the unlicensed user in order to maximize its data rate. We show that the system throughput depends on several parameters such as the maximum allowed cost and the number of base stations in the primary network, the secondary user location and its transmission energy. Index Terms—cognitive network, adaptive modulation, inter- ference cost constraint, throughput, transmission energy optimi- sation I. I NTRODUCTION The explosive growth in wireless technologies over the past several years illustrates the huge and growing demand of the spectrum resources. On the other hand, conventional fixed spectrum allocation cannot successfully cope with the scarcity of radio frequency spectrum. A recent survey of spectrum utilization made by the Federal Communications Commission (FCC) has indicated that the actual licensed spectrum is largely underutilized in vast temporal and geographic dimensions [2]. This was the reason for allowing non-legitimate users to use licensed bands assuming that it would not disturb and impact the primary transmissions. Such paradigm has been originated by Mitola [3] to promote the efficient use of the limited spectral resource [4] by seeking and opportunistically utilizing radio resources in time, frequency and space domains on a real time basis. To do so, some important abilities should be provided by the cognitive users which include spectrum sensing, dynamic frequency selection and transmit power control [1]. Among these various cognitive radio imple- mentation techniques, spectrum-sharing is especially appealing for practical deployment since it does not involve complex spectrum sensing mechanisms [5] [6] [7] that still a time consuming and power intensive process. More specifically, we propose in this paper a spectrum-sharing cognitive radio where the coexistence of simultaneous primary and secondary communications in the same frequency channel may occur as long as the aggregated interference generated by the secondary users is below some acceptable threshold [8]. Thus, we will define a maximum cost value belong it the secondary user is not authorized to transmit. On the other hand, under such constraint, throughput maximization stills a key challenge in such cognitive radio where the available resource and the transmission authorization may change randomly from time to time. For this reason, we propose in this paper to use an adaptive modulation technique in order to compensate the throughput loss and improve secondary user transmission quality [9]. Furthermore, to ensure data rate maximization a transmission energy optimization is proposed at the secondary user with respect to the tolerable interference power at the primary receiver. The rest of this paper is organized as follows: In section II, we present the principle of our adopted cognitive approach and introduce the system model. In section III, we derive analytical formulation for our constraint imposed on secondary users. In section IV, we provide some simulation results and analyze the system performance in terms of throughput. Finally, we end the paper with a brief conclusion and some perspectives for our work. II. SYSTEM MODEL As shown in Figure 1, we consider a cognitive radio network consisting of a secondary terminal denoted by SU, a secondary base station denoted by BSs and n primary base stations denoted by BSp  where  = {1,.., }. All links between terminals are assumed to be independent. Each link consists of Rayleigh, slow fading channel, so that we can consider its coefficients as constant during the transmission of at least one frame. We denote respectively the propagation channel attenuations between SU-BSs and SU- BSp  by ℎ  and ℎ  . Let   and   are the additive white Gaussian noise terms corrupting respectively the two links SU- BSs and SU-BSp  with a common variance  0 . Furthermore, to investigate the effect of the path loss into the channel model, the coefficients   and    corrupting respectively SU-BSs and SU-BSp  channels, follow a propagation model of 1/  , where  denotes the path loss factor. We assume that the