ABSTRACT
Switched diversity offers one of the lowest-complexity solutions to mitigate the effect of fading in wireless communication systems. In this chapter, starting from traditional dual-branch switch and stay combining (SSC) schemes, we develop and analyze several switching-based diversity combining schemes. More specifically, we first present a Markov chain-based analytical framework for the performance analysis of various switching strategies used in conjunction with dual-branch SSC systems. This analysis leads to a thorough comparison and trade-off study between different SSC strategies. Then we generalize switched diversity to a multibranch scenario. We show that while SSC does not benefit from additional diversity paths, the performance of switch and examine combining (SEC) improves with additional branches. Finally, noting the deficiency of pure switched diversity schemes in taking advantage of available diversity paths, we
propose generalized switch and examine combining (GSEC) as a good candidate combining scheme for diversity-rich environments. A description of the GSEC mode of operation shows that this scheme conserves a fixed complexity as the number of diversity paths increases, and offers a considerable complexity savings compared to traditional maximum ratio combining (MRC) and even other low-complexity schemes such as generalized selection combining (GSC). For all of these schemes, we present closed-form expressions for the statistics of the combiner output, including the moment-generating function, cumulative distribution function, and probability density function. The resulting expressions allow for accurate performance analysis of these diversity techniques over most fading models of interest. As an illustration of the mathematical formalism, selected numerical examples are provided together with some related discussions and interpretations. It is observed that compared with the more popular selection-based combining schemes, switching-based combining schemes lead to simpler receiver structure, less signal processing, and lower power consumption at the cost of certain performance loss.
