MIMO OFDM Interference Alignment Testbed

MIMO OFDM Interference Alignment Testbed

Interference Alignment (IA) is a revolutionary wireless transmission strategy that uses cooperation among transmitters to share the wireless channel more efficiently. The ability of IA  to work in large-scale, real-world scenarios depends heavily on whether associated practical challenges can be overcome, such as how to provide the required transmit power and number of additional signal dimensions, how to perform channel estimation, channel feedback and distributed coordination, and  how to synchronize to establish a coherent interference channel. Assuming these practical challenges are overcome, the data rate of real-world IA systems is also affected by the computing capability of (embedded) system platforms on top of which an IA system is realized. To study and optimize the practically achievable performance of IA in real-world settings, we have built a flexible and extensible MIMO IA testbed.


Detailed Testbed Description

To study practical issues in the implementation of IA, we are prototyping a MIMO OFDM system with three pairs of users, each having two transmit antennas and two receive antennas. The prototype is implemented on a PC-based Software Defined Radio (SDR) platform. The baseband digital signal processing is performed on the PCs using LabVIEW, a widely-used GUI-based software toolkit for digital system design from National Instruments (NI). The ADC/DAC and RF components are on low-cost NI USRP-2921s, which are flexible software radios developed for SDR system design. For each transmit or receive antenna, a USRP-2921 is used. Transmission occurs in the unlicensed 2.4GHz frequency band.



Recent Results

With our testbed setup, we are able to demonstrate real-time distributed three user, 2×2 MIMO OFDM with IA. All nodes in our setup are physically distributed, and, therefore, there is no wired cooperation between any two nodes in the testbed. We had taken several steps to build the testbed. We started from 6×6 MIMO OFDM testbed, and evolved it to three user, 2×2 MIMO OFDM testbed, but with wired synchronization and CSI feedback as presented in our previous work [3]. Then, it has been modified to work in a distributed manner as our recent results in [4].

To cope with the distributed operation, we have investigated two items: time and frequency synchronization, and sharing CSI feedback information among the nodes. For the synchronization of nodes, we have adopted a master-slave synchronization protocol, in which one of the nodes works as a master and enables the other nodes’ synchronization to the master by broadcasting a known training signal. For the wireless CSI feedback, we have implemented analog feedback, which is known to be a suitable feedback strategy in high SNR regions. The analysis of analog feedback with IA was studied in our previous work [1]. Finally, to find IA precoding matrices at transmitters and combining matrices at receivers, we have exploited the general and low-complexity alternating minimization scheme which we formulated in the previous work [2].

Using our testbed, we can study how the challenges in the real-world limit the performance of IA and deviate from ideal theoretical performance. In the process, we explore and develop novel methods to overcome practical limitations of IA. Furthermore, we use the testbed to analyze performance requirements of digital signal processing chains for IA, with the goal of exploring the design space to find efficient embedded system implementations that maximize the performance of IA under tight computing and power constraints. Recent results along these research topics are summarized in [4]. Future and ongoing work is concerned with IA issues in network layer and the optimal embedded system design.


[1] O. El Ayach and R. W. Heath, Jr., “Interference Alignment with Analog Channel State Feedback”, IEEE Trans. on Wireless, vol. 11, no. 2, pp. 626-636, February 2012.

[2]  S. W. Peters, and R. W. Heath, Jr., “Interference alignment via alternating minimization,” Proc. of the IEEE Int. Conf. on Acoustics, Speech, and Signal Proc.,  April 2009.

[3]  J. W. Massey, J. Starr, S. Lee, D. Lee, A. Gerstlauer, and R. W. Heath Jr., “Implementation of a real-time wireless interference alignment network,” Proc. of the Asilomar Conference on Signals, Systems and Computers, Nov 2012.

[4] S. Lee, A. Gerstlauer, and R. W. Heath, Jr., “Distributed Real-time Implementation of
Interference Alignment with Analog Feedback”
, To appear, IEEE Trans. on Vehicular Technology.


We have been fortunate to have several outstanding sponsors of our work including the DARPA IT-MANET program, Grant W911NF-07-1-0028,and the Army Research Labs under grant Grant W911NF1010420. Any opinions, findings, and conclusions or recommendations expressed in this material are those of the author(s) and do not necessarily reflect the views of the aforementioned sponsors.

Comments are closed.