ABSTRACT
Contents 4.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 112 4.2 UMTS Downlink and HSDPA . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 114
4.2.1 Multiple Access in UMTS FDD Downlink. . . . . . . . . . . . . . . . . . . 114 4.2.2 UMTS Services . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 116 4.2.3 HSDPA Features. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 116 4.2.4 Downlink Transmission Model . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 120
4.3 Downlink Channel and HSDPA Signal Models . . . . . . . . . . . . . . . . . . . . . 122 4.3.1 Channel Impairments and Mitigation . . . . . . . . . . . . . . . . . . . . . . . 123 4.3.2 HSDPA Signal Model. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 126
4.4 Suppression of Intra-cell Interference in HSDPA. . . . . . . . . . . . . . . . . . . 128 4.4.1 RAKE Receiver and LMMSE Chip Equalizer. . . . . . . . . . . . . . . . . 128 4.4.2 HSDPA Performance Analysis of RAKE Receiver
and Chip Equalizers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 134 4.4.2.1 Hypothetical Receiver Models . . . . . . . . . . . . . . . . . . . . . 135 4.4.2.2 Parameter Modeling . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 136 4.4.2.3 Simulations Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 138
4.5 Advanced Receivers for Interference Cancellation . . . . . . . . . . . . . . . . . 142 4.5.1 Symbol-Rate Signal Model . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 142 4.5.2 Optimal Receiver . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 144 4.5.3 Decorrelating Receiver . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 145
4.5.3.1 Projection Interpretation of Decorrelating Receiver . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 146
4.5.4 LMMSE Receiver . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 149 4.5.5 Linear Parallel Interference Cancellation
(LPIC) Receiver . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 150 4.5.6 Iterative Receivers Based on Chip Equalizers . . . . . . . . . . . . . . 151
4.5.6.1 Polynomial Expansion Receiver . . . . . . . . . . . . . . . . . . . 153 4.5.6.2 Intercell Interference Cancellation Expansion. . . . . 161 4.5.6.3 Simulations and Conclusions . . . . . . . . . . . . . . . . . . . . . . 163
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 169
4.1 Introduction Over-the-air communication is interference limited. Deployment of wireless networks therefore needs resource planning. Neighboring base stations in a cellular network could, for instance, transmit in different frequency bands. For such deployment, hexagonal cell geometry is assumed in cellular systems (see Figure 4.1), and resource planning is done by allocation of disjoint chunks of spectrum to neighboring base stations. The number of neighboring cells in which a certain frequency can be used only once is known in the cellular literature as the frequency reuse factor. A frequency reuse factor of 7 results when isotropic antennas are deployed at cell-site (base station) and of 3 when using antenna sectorization of three sectors per hexagonal cell. Such frequency planning obviously aims interference avoidance, but it makes network deployment cumbersome and is wasteful of spectral resources. A direct sequence (DS)-CDMA (Code Division Multiple Access) system, in principle, boasts a frequency reuse factor of 1. In fact, this factor was purportedly one of the main advantages of the first CDMA-based cellular network, IS-95 [41]. Users can coexist in the same frequency spectrum. A resource is a spreading code and in the specific case of downlink communications, codes belong to a binary orthogonal set, the Walsh-Hadamard set. A base station has the entire set of codes at its disposal.
