ADMM-Based Infinity-Norm Detection for Massive MIMO: Algorithm and VLSI Architecture

Abstract

We propose a novel data detection algorithm and a corresponding very large scale integration (VLSI) design for massive multi-user (MU) multiple-input multiple-output (MIMO) wireless systems. Our algorithm uses alternating direction method of multipliers (ADMM)-based infinity norm constrained equalization and is called ADMIN. ADMIN is an iterative algorithm that outperforms linear detectors by a large margin when the ratio between the numbers of base-station (BS) and user antennas is small. In the first iteration, ADMIN computes the linear minimum mean-square error (MMSE) solution, which is sufficient when the ratio between the numbers of BS and user antennas is large. We develop time-shared and iterative VLSI architectures for LDL-decomposition based soft-output ADMIN supporting 16-and 32-user systems. We present applicationspecific integrated circuit (ASIC) designs for 16 to 64 antenna base stations in 28 nm complementary metaloxidesemiconductor (CMOS) that supports up to 64 quadrature amplitude modulation (QAM). The 16-user ADMIN ASIC achieves 303 Mbps while dissipating 85 mW. The 32-user ADMIN ASIC achieves 287 Mbps and 241 Mbps while dissipating 121 mW and 135 mW for 32 and 64 BS antennas, respectively. ADMIN has also been implemented on a Xilinx Virtex-7 field-programmable gate array (FPGA) and is compared with state-of-the-art massive MIMO data detectors.

Likelihood ascent search‐aided low complexity improved performance massive MIMO detection in perfect and imperfect channel state information

“…Also, Table 2 shows the required multiplications for different iterative MMSE methods. This section shall compare the proposed detectors with other existing detectors like EPA, 34 ADMIN, 33 and OCDBOX 35 . Figure 12 shows the BER performance of different detectors for 64 × 256 MIMO system with 16‐QAM, 64‐QAM, and 256‐QAM modulation.…”

“…Rayleigh flat fading channel model. The performance of the proposed algorithms are compared with existing state-of-arts detectors like CG, 16,17 GS, 14 Chebyshev, 19 ST-JA, 13 hybrid-RI, 21 exact MMSE, alternating direction method of multipliers-based infinity-norm-constrained equalization (ADMIN), 33 expectation propagation with approximation (EPA), 34 and optimized coordinate descent with box equalization (OCDBOX). 35…”

Efficient QAM Signal Detector for Massive MIMO Systems via PS-ADMM Approach

“…In this subsection, the BER performance of the proposed PS-ADMM detector was evaluated and compared with conventional and state-of-the-art MIMO detectors by numerical simulations, which are the classical MMSE detector, Neumann detector [14], GS detector [17], OCD-BOX detector [20], and two ADMM-based detectors ADMM and ADMIN in [23] and [24] respectively. The termination criterium is that iteration number K reaches 30 or the residual in ( 16) is less than 10 −5 .…”

“…The ADMM strategy was first introduced in MIMO detection by Takapoui et al in [23]. In [24], a detection algorithm based on ADMM with infinity norm or box-constrained equalization, named ADMIN, was proposed, which outperformed the state-of-the-art linear detectors by a large margin if the BS-to-user-antenna ratio is small. The ADMM detection algorithm was applied in various scenarios for MIMO systems [25], [26] and improved for massive MIMO systems [27]- [31].…”

Dynamic branch pruning aided low switching fixed complexity sphere decoding for small scale and massive MIMO detection

“…In massive MIMO application, linear detectors such as MMSE/ZF can achieve optimum performance when $$$ {N}_T\ll {N}_R $$$ 25‐27 . However, when the $$$ {N}_R/{N}_T $$$ ratio is close to one, the performance of MMSE/ZF is not close to optimal.…”