A 2×2–16×16 Reconfigurable GGMD Processor for MIMO Communication Systems

The generalized geometric mean decomposition (GGMD) is a recently proposed matrix decomposition which can be viewed as a computationally efficient counterpart of the conventional geometric mean decomposition (GMD). As the GMD is the core algorithm in many high performance precoders and equalizers such as the Tomlinson-Harashima precoder and decision feedback equalizer, GGMD facilitates a more computationally efficient implementation while exhibiting identical performance. This work presents the first GGMD processor in the open literature, supporting various matrix sizes by leveraging the reconfigurable processing element (PE) using coordinate rotation digital computers (CORDICs). The implemented GGMD processor supports matrix sizes of 2ⁿ which ranges from 2×2 to 16×16, and the throughput performance is maximized through a PE array architecture. The chip integrates 326.9K gates in an area of 1.65 mm² in a 90-nm CMOS technology with the maximum throughput achieving 450K matrices/sec for a 16×16 matrix at 125 MHz. It dissipates 20.7-28.5 mW at 125 MHz from a 1V supply. Compared to previous GMD designs, this work supports a larger MIMO system with lower hardware complexity and power consumption.

I. Introduction

Multiple-input-multiple-output (MIMO) communications have now become a standard technology for realizing high data rate in modern wireless communication systems [1]. To achieve the huge capacity available from the MIMO channel, matrix decompositions such as singular-value-decomposition (SVD), QR decomposition, and etc. [2] are often used in MIMO transceiver designs [3]. Geometric mean decomposition (GMD) [4], [5] can be viewed as an extension of the QR decomposition which not only triangularizes the MIMO channel but also equalizes the effective subchannel gains. The optimal bit-allocation strategy in a GMD-assisted transceiver therefore boils down to the simple equal bit-allocation, which is more advantageous from the implementation perspective. More recently, the Generalized Geometric Mean Decomposition (GGMD) are proposed in [6], [7]. Compared to the GMD which triangularizes the channel matrix into one upper triangular matrix, the GGMD transforms the channel matrix into multiple upper triangular matrices with smaller dimensions.