System-level Simulation And Research On Key Technology In LTE-Advanced

LTE offers high space multiplexing gain and diversity gain with the adoption ofMIMO-OFDM as physical layer technology. As a future evolution of LTE, LTE-Advancedintroduce more new technologies such as multi-point cooperating, MIMO enhancement,and carrier aggregation. Roughly speaking, introduction of such technologies willcertainly bring a rise in the two key performance metrics of wireless communication:frequency efficiency and power efficiency. But technically, to understand whether therewill be any improvement and how much improvement when such new technologies areadopted, quantitive analysis is needed. There are generally two ways of quantitive analysis,one way is to perform theoretic analysis and pursue a close-form solution, and suchmethod is usually applied when evaluating a certain algorithm or a local performanceimprovement. As for evaluation of system-level performance, with multiple protocollayers involved, and multiple modules cooperating and constraining with each otherwithin a layer, a theoretic model is usually hard to build. Especially for LTE-A, due to thecomplexity of wireless channel model, diversity of MIMO open-loop and close-looptransmission modes, synchronization of HARQ retransmission scheduling, theoreticanalysis is hard to obtain an accurate close-form solution when facing such a system-levelsimulation. Therefore, system-level simulation has played an important role whenanalyzing complex systems such as LTE, and it has been broadly adopted in the researchand enterprise community.In this thesis, we start with an introduction of LTE-A system architecture. We useC/C++to build the simulation platform, which has the mechanism of message-driven. Wesimulated many kinds of events in real system, and we perform these tasks in real time. Bydoing so we avoid the shortage of MATLAB, which is a descriptive language and veryslow when dealing with system-level simulation, and we avoid the low efficiency ofpolling. We abstract the physical layer in the simulation, including modeling of MIMOwireless channel, abstracting of MIMO transmission modes, and modeling of LTE-Aphysical channel and HARQ gain. We also modeled and implemented wireless data-linklayer, including PDCP, RLC, and MAC. As to HARQ retransmission of cell-edge cooperative user, we compare cooperative retransmission scheme and non-cooperativeretransmission scheme, and get an insightful conclusion, which also lays foundation offuture research of LTE-A.By building this LTE-A system-level platform, the proposals submitted by our groupconcerning HARQ retransmission and codebook designing are supported with strongsystem-level evaluation, and the proposals are accepted by the LTE-A standardizationsubgroup, which gives us stronger motivation for future research. This simulation platformnot only gives us a thorough understanding of LTE-A performance analysis, but also offersa test platform when introducing new module-level technologies. The platform makesperformance evaluation evolves from module-level to system level, which has muchsignificance.