Research On Downlink Receiver Technology In GPP-based 5G Massive Distributed MIMO Test Bed

With the fast development of mobile communication technology and widely deployment of 4G system,people start to research next generation of mobile communication(5G).Massive MIMO,as one of the most important technologies in 5G,can improve the capacity of system and reduce energy consumption at the same time.Massive MIMO is the most popular topic of wireless communication in recent years.In massive MIMO system,how to get precise signals under interference and noise by users is the most important problem of massive MIMO receiver.This thesis is concentrated on research of the downlink receiver in distributed massive MIMO.First,the thesis introduces the research background,the development history of massive MIMO tech-nology and the topic related current research status.Secondly,the thesis system model of massive distributed MIMO is given,we introduce synchronization and transmission of downlink and uplink.According to the system model,we also give the system parameters and frame structure.In this chapter,the system model refers to some reference signals.We introduce the usage and design ideas of these reference signals.Adaptive transmission scheme is an important party of the system,and we discuss the adaptive transmission scheme.In distributed massive MIMO system,the most common detection includes linear detection and non-linear detection.The complexity of linear detection is very low and the performance is poor.Comparing to linear detection,non-linear detection has a better performance.However,the complexity of non-linear detection is very high.To decrease error propagation probability,we research the iterative soft decision interference cancellation algorithm.In downlink massive MIMO system,the interference of other users plus noise can be seen as colored noise.We also analysis the SISO detection under colored noise in this chapter.According to the simulation results,SISO detection under colored noise can effectively improve detection performance.In this thesis,the base stations use wideband channel state information and multi-user pre-coding.User side will be interfered by residual multi-user interference.The interference plus noise received by users is no longer the gauss white noise.In this chapter,we research the downlink transmission scheme in 5G platform.Then,we use soft decision interference cancellation cooperated with turbo iteration decoding.Simulations show that turbo iteration receiver can significantly enhance system performance.Finally,according to the characteristic of general purpose processor,we make some optimizations in our 5G massive distributed MIMO.Our platform use multi-core processors,so we can use multi-threading technique to accelerate uplink and downlink transmission and receiver speed.We also use Intel MKL and some other technologies to improve the performance.Intel MKL can bring great benefit in matrix operation.At last,we use intel performance analysis tools to analysis the downlink receiver.The test results show that multi-threading parallel technique and compiler optimization can significantly increase the speed of test program.