Transmission and Detection Techniques for Large MIMO Systems

Wireless data traffic has been growing at an exponential speed in the past decade, largely due to the proliferation of smart phones and tablet computers. Reducing cell size is a conventional way to support this growth, but deploying new base station (BS) sites can be very costly in many situations. Multiple-input multiple-output (MIMO) technology potentially provides an alternative solution. A MIMO system involves multiple antennas mounted at both transmitter and receiver, and has long been identified as a promising technology for future cellular systems.A large MIMO system refers to one involving many (tens or even hundreds of) antennas, typically at the BS side. The large MIMO concept has recently attracted considerable research interest. In principle, a large MIMO system provides a large degree of freedom and can potentially achieve very high capacity.However, there are many challenges that must be addressed in order to realize large MIMO systems. Channel modeling, antenna design, and asymptotic capacity analysis are active research topics, but there are also many open issues at the system design level. In particular, channel state information (CSI) is a crucial issue for large MIMO systems. It has been revealed recently that the so-called pilot contamination effect imposes a limit on CSI accuracy in large MIMO systems. This has serious consequences for system design, since many existing transmission and detection techniques rely on accurate CSI and so may not work well in the presence of CSI errors. There are transmission schemes, such as space-time coding, that do not require CSI at the transmitter (though they still need reliable CSI at the receiver for detection). However, such schemes cannot make full use of the available degrees of freedom in a large MIMO system.In this project, we will study system design techniques for large MIMO systems. We will develop new strategies for channel estimation, up-link and down-link transmission, error control coding and linear pre-coding. In recent years, we have worked on many aspects related to MIMO systems and we will exploit this accumulated experience to make real progress on the challenging issues in large MIMO systems. Our main aim is to develop low-cost but efficient solutions for large MIMO that will be useful in practice.