SPECTRUM EFFICIENT POWER ALLOCATION FOR OFDM COGNITIVE RADIO

Abstract

Recently, due to an explosive growth in the demand for new mobile cellular services and other wireless applications, the radio spectrum has become one of the most scarce and valuable resources for wireless communications. The scarcity of available spectrum and the under utilization of the statically allocated spectrum bands have led to much interest in the problem of opportunistic spectrum access over the past decade. Cognitive radio (CR) has emerged as a promising technology to achieve this opportunistic spectrum access. CR users have the capability to sense the radio environment and opportunistically access the spectrum that is originally licensed to primary users (PUs), provided the interference caused to licensed PUs is kept below their specified thresholds. Thus, CR arises as a new communication paradigm that has a great potential to improve the utilization efficiency of the radio spectrum. Orthogonal frequency division multiplexing (OFDM) has been recognized as an appropriate modulation scheme for CR systems because it meets most of the physical layer requirements of a CR system. However, despite its numerous advantages, conventional OFDM suffers from high out-of-band radiated interference (OBRI) due to the sidelobes of sinc-shaped spectrum on each subcarrier of the transmitted OFDM signal. Consequently, in an overlay scenario, where the OFDM-based CR users and the PUs coexist side-by-side in adjacent frequency bands, the OBRI generated by the transmissions from OFDM-based CR system may significantly affect the performance of the PU system. Therefore, it becomes necessary to keep this OBRI below some tolerable interference thresholds of the PUs. Fortunately, there are techniques available in literature to shape the spectrum of the CR transmitted OFDM signal. These spectral shaping techniques, when employed at the OFDM CR transmitter, may help in reducing the OBRI inflicted on the PUs operating in adjacent bands. Power allocation is an important issue for OFDM-based CR systems which helps in improving the spectral efficiency and the transmission rate of CR users. In literature, considerable works have been done on power allocation for maximizing the transmission capacity of OFDM-based CR systems. Most of the existing (conventional) power allocation strategies focus on enhancing the spectrum efficiency of the OFDM-based CR system while keeping the OBRI experienced by PUs below predefined interference thresholds (so that the performance of the PUs operating in adjacent frequency bands is not affected). It is noted that the existing power allocation schemes do not consider any means for reducing OBRI. Consequently, by using these conventional power allocation schemes, an OFDM-based CR system may not be able to best harness its available power resources since a higher OBRI may force the CR system to operate in an interference limited scenario. This suggests that a better utilization of the limited power budget and the available i radio spectrum may be achieved if the power allocation is augmented with some spectral shaping technique at the OFDM-based CR transmitter. Motivated by the fact that dynamic power allocation together with spectral shaping for OFDM-based CR systems may lead to even more efficient utilization of the available radio spectrum, in this thesis, we propose spectrum efficient power allocation schemes augmented with appropriate spectral shaping techniques for an OFDM-based CR system coexisting side-by-side with multiple PU bands in a spectrum overlay fashion. The proposed schemes have been designed under different conditions and criteria, but all with the same objective of maximizing the downlink transmission rate of an OFDM-based CR system subject to the constraint(s) on the interference introduced to the PU system. In the first part of this thesis, we propose optimal and suboptimal power allocation schemes augmented with active interference cancellation (AIC)-based spectral shaping for an OFDM-based CR system such that the downlink capacity of a CR user is maximized while the total interference introduced to the PU system is maintained below the specified threshold. We consider that CR transmitter (base station) is equipped with enough power budget, and therefore, we investigate the performance of the proposed schemes in an interference limited scenario. Presented simulation results demonstrate the effectiveness of the proposed schemes over the conventional power allocation schemes available in the literature for the similar scenario. Further, the proposed schemes provide a good trade-off between the computational complexity and the achievable CR user throughput. In the second part of this thesis, we consider a scenario where instantaneous values of the channel fading gains between CR transmitter and active PU receivers (that are operating in adjacent frequency bands) are not perfectly known at the CR transmitter, and there is only a limited amount of power budget available at the CR base station transmitter. For such an OFDM-based CR system scenario, we propose spectrum efficient AIC-based optimal and suboptimal power allocation schemes that maximize the downlink transmission capacity of a CR user subject to total power constraint at the CR base station transmitter as well as the individual statistical interference constraints of each of the PUs. In addition, an optimal power allocation scheme that is augmented with raised cosine (RC) windowing-based spectral shaping is also studied. The performance of the proposed schemes is investigated in both the scenarios, viz, the power limited scenario as well as the interference limited scenario. The presented simulation results show that both the AIC-based as well as the RC windowing-based power allocation schemes lead to significantly higher transmission rates for the CR user compared to the conventional (without any spectral shaping) optimal power allocation scheme. The computational complexity analysis of the proposed schemes is also presented. ii Spectral shaping in OFDM-based CR systems can also be achieved using precoding techniques which have attracted a great deal of attention recently. In particular, orthogonal precoding in OFDM-based CR systems can significantly reduce the OBRI inflicted on the adjacent PU band(s). In the last part of this thesis, we consider a multiuser scenario and study the problem of power allocation for multiuser precoded OFDM-based CR systems. We propose an optimal power allocation scheme that is augmented with orthogonal precoding-based spectral shaping. Here, the aim is to maximize the downlink sum data rate of all users in the OFDM-based CR system while the interference introduced to PUs (operating in adjacent frequency bands) is maintained below their prescribed thresholds and the power transmitted at the CR base station transmitter is kept within its available power budget. The performance of the proposed scheme is investigated in terms of the achievable sum data rate of CR users as well as on the basis of the computational complexity involved. Presented simulation results demonstrate that the proposed precoded OFDM optimal power allocation scheme can deliver much higher sum data rate of CR users in comparison to the conventional (uncoded OFDM) optimal power allocation scheme and the precoded OFDM uniform power allocation scheme.