As discussed in the last chapter, in a Gaussian or flat fading channel, multicode channels are orthogonal. In a realistic wireless channel, however, the orthogonality no longer maintains. Thus, MCI is caused. In this chapter, a novel detection method, called hybrid MCI cancellation and MMSE detection, is proposed and compared with pure MMSE detection. The system performance is analytically studied with imperfect channel estimation to show how it is affected by parameters such as the window size in the channel estimation, Doppler shift, the number of stages of the hybrid detection, the power ratio of pilot to data channels, spreading factor, and so on.

Introduction

Although MMSE detection can suppress MCI as discussed in the last chapter, so far MMSE detection has not been investigated for the OFCDM system with interference cancellation. Figure 10.1 illustrates the situation when MMSE detection is employed with MCI cancellation. It can be seen that the input to the MCI canceller is a combination of the useful signal, MCI and background noise. After MCI cancellation, the useful signal and background noise remain unchanged, but MCI is reduced. The combined signal is input to the MMSE detector with estimated signal plus interference power. Then the weighted output is generated after MMSE detection. Since the weight of MMSE is related to the input signal plus interference power, it should be updated stage by stage because the input power of MCI changes due to MCI cancellation.