Real-Time Implementation of 4x4 MIMO-OFDM System for 3GPP-LTE Based on a Programmable Processor
MIMO-OFDM is currently being considered as a promising technology for future wideband wireless systems. Meanwhile, the detection of MIMO-OFDM system forms one of the most intensive tasks in baseband signal processing. There have been a great number of alg
- PDF / 2,641,006 Bytes
- 11 Pages / 439.363 x 666.131 pts Page_size
- 89 Downloads / 179 Views
Abstract. MIMO-OFDM is currently being considered as a promising technology for future wideband wireless systems. Meanwhile, the detection of MIMO-OFDM system forms one of the most intensive tasks in baseband signal processing. There have been a great number of algorithms proposed for MIMO-OFDM detection. This paper investigates the real-time implementation of MIMO OFDM systems using programmable multi-SIMD (single instruction multiple data) processor targeting for software defined radio. We analyze the computational characteristic and cost of major algorithms of LPF, symbol synchronization, OFDM modulation, channel estimation and MIMO detection of a 4x4 MIMO-OFDM system based on 3GPP-LTE R-12 standard. The evaluation and implementation results show that our SDR architecture supports 300Mbps LTE system with 20MHz bandwidth. Finally, we discuss the architecture perspective of our SDR processor for future wireless communication standards with higher throughput. Keywords: MIMO-OFDM, Multi-SIMD, SDR, 3GPP-LTE.
1
Introduction
Multiple Input Multiple Output (MIMO) and Orthogonal Frequency Division Multiplexing (OFDM) have been adopted to increasing the spectrum efficiency by emerging wireless broadband standards, such as the 3rd generation partnership project long term evolution (3GPP-LTE) and WiMax [1]. However, the new techniques bringing the enhancement of data rate in wireless communication always have high computational complexity, a large number of algorithms about OFDM modulation, channel estimation (CE) and MIMO equalization have been proposed [2][3]. Unfortunately, some of them are too complicated to implement in the state-of-the-art digital signal processors (DSP) in real-time though they can provide excellent performance. Consequently, trade-off must be made between performance and complexity for practical and efficient implementation.
Corresponding author.
W. Xu et al. (Eds.): NCCET 2013, CCIS 396, pp. 33–43, 2013. c Springer-Verlag Berlin Heidelberg 2013
34
T. Chen, H. Liu, and J. Wan
Many works have been carried out to implement the MIMO-OFDM physical layer by ASIC and VLSI [4][5]. However, they are difficult to be upgraded for the evolving communication standards. Recently, a lot of programmable processor is proposed, e.g., MSC8156 [6], PC205 [7] and TMS320TCI6618 [8]. But they still rely on some accelerators to offload parts of the wireless standard processing to achieve high throughputs. Data movement between accelerators and main memory may push these platforms toward inflexible solutions. This paper investigates the real-time and hardware-efficient implementation of low pass filter (LPF), symbol synchronization, OFDM modulation, channel estimation and MIMO detection for MIMO-OFDM systems using fully programmable hardware aimed for software defined radio (SDR). A real-time 4x4 MIMO-OFDM system of 3GPP-LTE standard is implemented on a programmable multi-SIMD SDR processor. The remainder of this paper is structured in the following way. In section II, the MIMO-OFDM system is outlined. Then, some algorithms
Data Loading...
