Mimo Communication Precoding And Detection Technology

Modern wireless communication requires higher data rate transmission, MIMO(Multi-Input Multi-Output) technology has attracted lots of attentions these years as an effective measure to enhance radio channel capacity and reliability.MIMO Precoding is a kind of technic to optimize communication system's performance making use of channel status. Being a part of IEEE 802.16e standard, precoding has turned into a hotspot of mobile communication. Starting from mathematic modeling of wireless MIMO channel, this paper compares different precoding patterns in view of theirs effects to channel capacity and system performance.Signal detection is one of the most important factors defining MIMO system performance. While single antenna system transmitting a single symbol at a time slot, it is a vector composed of several symbols in MIMO system, which increases detection complexity greatly. Optimizing system performance with acceptable operation complexity is greatly concerned. This paper analyzes traditional signal detection criteria and introduces sphere decoding algorithm, which can greatly reduce complexity without performance loss.Channel code technology is applied widely in information technology domain as an important artifice which can boost up reliability of information transmission. Based on channel code, Turbo iterative detection algorithm is testified to perform excellently. Modified from sphere decoding, list sphere decoding (LSD) algorithm can find more than one most possible symbol vectors around the search center of sphere decoding. Then probability information of coded bits can be calculated form these vectors and provide to Turbo equalizer.In practical system especially wideband wireless communication scene, radio channel should be modeled as frequency selective fading pattern. This paper proposes MMSE equalization algorithm in frequency selective fading MIMO channel, which can turn a multi-path MIMO channel into flat fading one. This algorithm can work very well in Turbo iterative detection architecture.