4th International ICST Conference on Wireless Internet

Research Article

Design and Deployment Considerations for High Performance MIMO Testbeds

Download302 downloads
  • @INPROCEEDINGS{10.4108/ICST.WICON2008.4927,
        author={Konstantinos Pelechrinis and Ioannis Broustis and Theodoros Salonidis and Srikanth V. Krishnamurthy and Prasant Mohapatra},
        title={Design and Deployment Considerations for High Performance MIMO Testbeds},
        proceedings={4th International ICST Conference on Wireless Internet},
        publisher={ICST},
        proceedings_a={WICON},
        year={2010},
        month={5},
        keywords={IEEE 802.11n MIMO Communications Testbed Deploy- ment Topology.},
        doi={10.4108/ICST.WICON2008.4927}
    }
    
  • Konstantinos Pelechrinis
    Ioannis Broustis
    Theodoros Salonidis
    Srikanth V. Krishnamurthy
    Prasant Mohapatra
    Year: 2010
    Design and Deployment Considerations for High Performance MIMO Testbeds
    WICON
    ICST
    DOI: 10.4108/ICST.WICON2008.4927
Konstantinos Pelechrinis1,*, Ioannis Broustis2,*, Theodoros Salonidis3,*, Srikanth V. Krishnamurthy4,*, Prasant Mohapatra5,*
  • 1: Dept. of CS&E, UC Riverside
  • 2: Dept. of ECE, University of Thessaly
  • 3: Thomson Research, Paris, France
  • 4: CS&E, UC Riverside
  • 5: Dept. of CS, UC Davis
*Contact email: kpele@cs.ucr.edu, broustis@uth.gr, theodoros.salonidis@thomson.net, krish@cs.ucr.edu, prasant@cs.ucdavis.edu

Abstract

MIMO (Multiple Input Multiple Output) enabled systems are characterized by higher reliability and transmission rates, as compared to conventional SISO (Single Input Single Out- put) systems. However, unless administered properly, the MIMO technology may not facilitate very high throughputs on point-to-point wireless links. Therefore, it becomes im- perative for the network architect to design such networks in ways that fully exploit the inherent properties of MIMO. In this paper, we first conduct an extensive experimental study, using a powerful hardware platform, in order to understand the behavior of MIMO links in different topological scenar- ios. Our experiments involve scenarios with MIMO links in isolation, as well as in competition with other MIMO and SISO links. Second, we perform measurements with different commercial platforms towards assessing the ability of each platform to efficiently support the MIMO technology. Based on our experimental observations we deduce that the CPU processing speed of the underlying hardware platform is an important factor that can hide the performance benefits of a MIMO enabled tranceiver. We comment on the applica- bility of the different hardware choices that we test; further- more, we suggest the most appropriate choice for building a MIMO testbed, taking into account the cost, the extend- ability and the re-usability of the selected platform. Finally, having adopted this choice in our testbed design, we provide a description of our testbed architecture.