ABSTRACT
CONTENTS 10.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 240 10.2 System Model . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 241 10.3 Maximization of the Ergodic Capacity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 242
10.3.1 Average Interference Power Constraint . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 242 10.3.1.1 Full CSI [P(g1,g0,h1)] . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 243 10.3.1.2 Partial CSI: Reduced Only CSI h1[P(g1,g0)] . . . . . . . . . . . . . . . . . 244 10.3.1.3 Partial CSI: Reduced Only CSI g0[P(g1,h1)] . . . . . . . . . . . . . . . . . 245 10.3.1.4 Partial CSI: Reduced Only CSI g1[P(g0,h1)] . . . . . . . . . . . . . . . . . 246 10.3.1.5 Without CSI [Constant P] . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 247
10.3.2 Peak Interference Power Constraint . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 247 10.3.2.1 Full CSI [P(g0)] . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 248 10.3.2.2 Without CSI [Constant P] . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 248
10.3.3 Numerical Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 248 10.3.3.1 Average Interference Power Constraint . . . . . . . . . . . . . . . . . . . . . . 248 10.3.3.2 Peak Interference Power Constraint . . . . . . . . . . . . . . . . . . . . . . . . . 250
10.4 Minimization of the Bit Error Rate (BER) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 250 10.4.1 Minimum BER Under Average Interference Power Constraint . . . . . . . . . . . . . 251
10.4.1.1 Minimization of Exp(-SINR) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 251 10.4.1.2 Special Case 1: The Effect of Having Extra CSI g0 at ST . . . . . 252 10.4.1.3 Special Case 2: The Effect of Having Extra CSI g0 and h1 at
ST . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 253 10.4.1.4 Minimization of aQ(√bSINR) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 254
10.4.2 Minimum BER Under Peak Interference Power Constraint . . . . . . . . . . . . . . . . . 256 10.4.2.1 Minimization of Exp(-SINR) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 257 10.4.2.2 Minimization of aQ(√bSINR) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 257
10.4.3 Numerical Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 257 10.4.3.1 Average Interference Power Constraint . . . . . . . . . . . . . . . . . . . . . . 258 10.4.3.2 Peak Interference Power Constraint . . . . . . . . . . . . . . . . . . . . . . . . . 260
10.5 Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 262
Abstract
This chapter describes a cross-layer design for reliable transmission in cognitive radio networks under spectrum sharing approach. In a cognitive radio scenario that consists of the primary users and the secondary users, secondary users opportunistically are allowed to access the existing spectrum without adverse effect on primary users. A cognitive radio network is allowed to detect its communication environment, replace the parameters of its communication scheme to raise the qualityof-service (QoS) for secondary users, and decrease the interference to primary users. Alternatively, the other approach involved in designing a cognitive radio network is to allow simultaneous transmission of primary users and secondary users, which is also termed as spectrum sharing. In this technique, a secondary transmitter can transmit while its maximum interference to the primary receiver is smaller than the predefined threshold. However, a secondary user must control its transmit power to get a reasonable transmission rate. In cognitive radio networks, the main issue is how to guarantee QoS in different applications. High capacity and minimum bit error rate (BER), varying as a function of the channel quality, are two of the major QoS requirements and are interferencelimited in mobile communication systems. Therefore, in this chapter, we maximize the ergodic capacity and minimize the average BER under different constraints at the primary users. The effect of reducing channel state information (CSI) at the secondary transmitter is discussed for both optimization problems. Finally, simulation results are presented to support analytical results.
