The MIMO Test Bed
Multiple-input multiple-output (MIMO)
systems are a research breakthrough
that has revolutionized the communication
industry. By making use of multiple
transmit and receive antennas, MIMO
systems not only provide robustness against multipath
fading impairments but also promise significant
increase in the throughput rates over frequency
selective channels. Orthogonal Frequency Division Multiplexing (OFDM)
on the other hand, is a modulation scheme that
converts a wideband channel into
parallel narrowband channels allowing
relatively simple channel estimation and frequency domain
equalization. By combining MIMO technology with OFDM,
it is possible to build high data
rate communication systems with reasonable
receiver complexity. Because of these
advantages MIMO-OFDM combination has emerged as a dominant physical
layer technology. MIMO-OFDM based physical layers are
being considered for various wireless
systems like IEEE 802.11n (WLANs), IEEE
802.16e-2005 (WiMAX) and 3rd Generation Partnership Project’s (3GGP)
Long Term Evolution (LTE).
Real-time MIMO-OFDM testbed is a research initiative of the Iqra
University sponsored by ICT R&D Fund. The aim is to develop a
testbed that can help in the design, analysis and rapid prototyping of
future MIMO-OFDM systems. With all their advantages MIMO-OFDM systems,
when compared with its single antenna counterparts, are not only more
sensitive to channel estimation and synchronization errors but also
require a lot more computational power and implementation
effort. MIMO systems demand perfect
estimate of the channel matrix at
the receiver to successfully decode the
received signal. Channel estimation is
a challenging task because in MIMO
systems different signals are transmitted from transmit antennas and
received signal at each antenna is the superposition of all the
transmitted signals. Sensitivity to synchronization errors comes from
the fact that a MIMO-OFDM system requires simultaneous
transmission and reception of parallel
spatial signal streams from its
antennas. A slight misalignment caused by
a timing synchronization error can
lead to serious degradation in the
performance. Optimum Maximum Likelihood (ML)
MIMO decoder can become almost
impossible to implement when constellation size and
number of antennas approaches a certain threshold. There has been a
significant research to resolve these issues over the last few
years and various solutions have been proposed but practical MIMO-OFDM
system design still pose a challenge for communication system designers.
Real time MIMO-OFDM testbed development work at IQRA University is focused on achieving the following two key objectives.
- 1. To resolve MIMO-OFDM system design related issues by designing simple, robust and computationally efficient algorithms for synchronization, channel estimation and MIMO detection etc. The main emphasis will be on reducing the computational complexity of MIMO detection algorithms so that more number of antennas, at both Tx and Rx ends, can be supported with dense constellations like 64 QAM.
- 2. To develop a cost-effective MIMO-OFDM testbed platform that can facilitate researchers/developers in the design, analysis and rapid prototyping of MIMO communication systems like WLANs, WiMAX etc in real-time.
The overall development of MIMO-OFDM testbed will be completed in two phases. In the first phase a low-scale 2x2 MIMO-OFDM prototype for the indoor environment will be developed. Prototype hardware will consist of a PC and a transceiver board carrying an analogue front end (AFE), high speed analogue and digital converters (ADCs) and digital to analogue converters (DACs), and a digital down-converter (DDC). A USB interface will be used to communicate between transceiver card and the host PC where all the signal processing will be performed. The main focus of this low-scale prototype phase is to optimize the design and development of algorithms by minimizing the implementation effort. The outcome of this phase will be a real-time but low-scale 2x2 MIMO-OFDM prototype for the indoor environment that will be able to process a bandwidth of approximately 1 MHz. As a reference design a scale down version of IEEE 802.11n WLAN standard will be implemented on the prototype hardware. The design will be tested and verified in simulation using the channel models recommended by the IEEE 802.11 task group.
In the 2nd phase a full-scale real-time 4x4 MIMO-OFDM testbed that could be able to process a bandwidth up to 20 MHZ will be developed. The hardware will consist of an AFE, high speed ADC and DCAs, DDCs and a baseband processing board that will carry a combination of high speed DSPs and FPGAs. Testbed hardware will be configured and controlled from a host PC where a part of signal processing will also be performed. A high speed interface like Gigabit interface or PCI Express will be used for data transfer between platform hardware and host PC. Full-scale testbed shall enable real-time development of MIMO-OFDM communication system like IEEE 802.11n (WLANs), IEEE 802.16e-2005 (WiMAX) and 3GGP’s Long Term Evolution (LTE).
A tentative specifications/capabilities list for the both tested platforms is provided in the table 1 below. These specifications are subject to change as new challenges can arise as along with the time.
Specification |
Low-scale Prototype |
Full-scale Testbed |
Antennas (Tx × Rx) |
2x2 |
4x4 |
Bandwidth |
Upto 1 MHz |
Upto 20 MHz |
Processing Platform |
GPP |
FPGA+DSP+GPP |
RF Frequency Band |
Dual ISM Band (2.4 and 5.8 GHz) |
Dual ISM Band (2.4 and 5.8 GHz) |
Data Interfaces |
USB |
Gigabit Ethernet/PCI Express |
Air Interface |
Dipole Antennas |
Dipole Antennas |
Operating Environment |
Indoor |
Outdoor (mobile channel) |
Applications |
WLAN, WiMAX, LTE etc (Scale down versions) |
WLAN, WiMAX, LTE, Cognitive Radio, Co-operative MIMO etc. |