In this paper, we analyze the outage probability and the long-term average throughput (LAT) of a decode-and-forward (DF) cooperative scheme with hybrid automatic repeat request (HARQ) over Nakagami-m fading channels. To derive the outage probability, the distribution of the weighted sum of two independent gamma random variables is needed. We approximately obtain the distribution by using gamma approximation and derive a closed-form expression of the outage probability of the DF cooperative scheme with high SNR assumption. By exploiting the outage probability, we obtain the LAT of the DF cooperative scheme. Numerical results show that the derived LAT of the DF cooperative scheme is able to predict the average throughput and the DF cooperative scheme achieves higher LAT than the direct transmission scheme because of higher diversity gains.

In cognitive radio networks, multiple secondary systems can access a licensed spectrum when none of the secondary transmitters cause harmful interference to the Primary Users (PUs). For spectrum sharing with coordination among co-located secondary systems, we propose both centralized and distributed beamforming algorithms. The proposed algorithms minimize the total transmit power of secondary systems, while maintaining the interference to PUs below a certain threshold and satisfying the Quality-of-Service (QoS) constraint for each secondary system. The centralized algorithm achieves the optimal transmit power by exploiting the virtual uplink-downlink duality using the knowledge of the channel state information for all the secondary links. However, the assumption of global channel knowledge at each secondary system may not be allowed in practical applications for multi-cell coordination. To address this problem, we design a distributed transceiver beamformer that satisfies the interference constraint to protect PUs. On the basis of this distributed beamformer, we also propose power allocation algorithms that guarantee the QoS for secondary systems. Distributed beamforming and power allocations operate iteratively to minimize the total transmit power. Simulation results show that the distributed algorithms achieve a near-optimal transmit power while satisfying both the QoS and interference constraints.

In this paper, we study a cooperation-based dynamic spectrum leasing mechanism via multi-winner auction of multiple bands. Based on a second-price auction mechanism, the primary users independently conduct auctions to determine winners who are then granted access to leased bands and prices for those bands. Before auctions, each secondary user jointly chooses bands which they want to lease and generates bids for those bands with a limited transmit power budget. To this end, we determine the feasibility conditions for band selection and for power and time allocation. Further, we propose a low complexity bidding algorithm that iteratively selects a band and calculates the bid for that band. For the paying price among winners after auctions, a multi-objective optimization problem is considered. We derive Pareto optimal solutions and propose paying schemes that prioritize the objectives with pre-defined weighting rules. In addition, the payment and secondary utility are investigated with regard to the number of winners. Simulation results show that secondary users achieve significantly increased utility as more winners are chosen, while primary users are guaranteed a non-negative benefit from spectrum leasing. The power consumed by the secondary users for cooperative transmission is compared for the proposed paying schemes.

In cognitive radio networks, imperfect spectrum sensing causes harmful interferences or low utilization of unused bands, depending on an access policy of the secondary system. Since these losses have a trade-off relationship, it is hard to decide proper sensing and access policies for the licensed bands. To deal with this problem, we propose a joint algorithm for sensing adaptation and opportunistic resource allocation. The joint optimization problem minimizes the total expected cost of the losses and utilities which are likely generated for the secondary system. To protect the primary system and ensure the secondary system's utility, we employ an average interference threshold constraint and a QoS constraint for the secondary system. Two types of average secondary utilities are considered for the QoS constraint; one is averaged over only transmitted frames and the other is averaged over total frames. Based on energy detection, we derive feasible ranges of sensing threshold to satisfy the constraints. Simulation results show that the proposed scheme dynamically adapts sensing threshold depending on the network environment and minimizes the total cost of the secondary system. In addition, we investigate the performance of the proposed scheme with the two types of the secondary utility.

This paper proposes a spectral efficient multiuser scheme for single-carrier transmission. In the proposed scheme, a combination (hybrid) of localized and distributed subcarrier mapping types is employed, where the signals of users with different subcarrier mapping types are superimposed on a common set of subcarriers. We first present mathematical formulations of multiuser detection and derive the equalization coefficients for the hybrid type. In fading environments, the proposed scheme allows users to employ any subcarrier mapping types (localized, distributed, hybrid types), which can support low and high speed users. To estimate the performance of the proposed scheme, we develop a density evolution technique and show that the predicted performances are consistent with simulated performances. Further, based on the performance estimation of the proposed scheme, we discuss channel-dependent resource allocation algorithms: opportunistic scheduling and adaptive modulation and coding (AMC) schemes. These resource allocation algorithms increase the spectral efficiency of the proposed scheme with guaranteed quality-of-service (QoS). Simulation results show that the proposed multiuser scheme achieves both multiuser and frequency diversity gains in fading environments.

In this paper, we propose power budget allocation algorithms for bidding of auctions in cooperation-based spectrum leasing. The cooperation-based spectrum leasing allows secondary users to dynamically access licensed bands while providing the benefit of cooperative diversity to primary users as a reward. For efficient allocation of the licensed bands to the secondary users, a second-price based multiple auction mechanism is considered. In this mechanism, the secondary users select bands which they want to use and calculate bids for those bands, under a limited transmit power. For the optimal solution of the bidding problem, however, computational complexity increases exponentially with the number of bands. To overcome this problem, we find feasibility conditions for band selection and propose a suboptimal algorithm. The suboptimal algorithm makes bids consisting of the maximum achievable rates for cooperative transmissions. Simulation results show that the primary and secondary users achieve significantly increased achievable rates through spectrum leasing with the proposed algorithms. It is also shown that the performance of the suboptimal algorithm approaches that of the optimal solution as the invested power budget increases.

Spectrum sensing is a key enabling technology of cognitive radio. Reliable detection increases access opportunity to temporarily unused bands and prevents harmful interference to the licensed users. Due to the receiver noise, signal attenuation, and multi-path fading effect, however, it is usually not possible to determine the existence of primary signal with absolute certainty. By extracting a global decision from shared local sensing results, cooperative sensing achieves high reliability over fading channels. In this paper, we assume that the traffic statistic of primary system is logged into the radio environment map (REM) and can be accessed by the secondary systems. The threshold of each energy detector is dynamically adapted according to the utility values and a priori information from REM. Then, decision results and corresponding operating points are collected by a fusion center, which makes a global decision with high confidence.

Spectrum sensing is a key enabling technology of cognitive radio. Reliable detection increases access opportunity to temporarily unused bands and prevents harmful interference to the licensed users. Due to the receiver noise, signal attenuation, and multi-path fading effect, however, it is usually not possible to determine the existence of primary signal with absolute certainty. Without the information of primary user activity, Neyman-Pearson criterion has been commonly used to minimize the missed detection probability for a given false alarm rate. In this paper, we assume that the traffic statistic of primary system is logged into the radio environment map (REM) and can be accessed by the secondary system. Considering sensing errors, Bayes criterion is adopted for total utility maximization of primary and secondary systems. The threshold of energy detector is adapted according to the utility values and a priori information from REM, i.e., both false alarm and detection probabilities are dynamically adjusted.

In this letter, we propose a new multiuser detection scheme for a space-frequency block code (SFBC) single-carrier modulation with frequency-domain equalization (SC-FDE). The iterative multiuser detection method used for the single transmit antenna system cannot be directly applied to SFBC SC-FDE, because there exists the off-diagonal components of channel matrices corresponding to multiuser interferences after maximal ratio combining. The output equations of SFBC combiner are reformulated, and the equalizer coefficients are derived for the SFBC SC-FDE system. Simulation results show that the proposed SFBC IMD-FDE efficiently provides the diversity gain, in the presence of multiuser interferences.