Survivable millimeter-wave mesh networks

  • Authors:
  • Abdul Jabbar;Justin P. Rohrer;Victor S. Frost;James P. G. Sterbenz

  • Affiliations:
  • Information and Telecommunication Technology Center, Department of Electrical Engineering and Computer Science, The University of Kansas, Lawrence, KS, USA;Information and Telecommunication Technology Center, Department of Electrical Engineering and Computer Science, The University of Kansas, Lawrence, KS, USA;Information and Telecommunication Technology Center, Department of Electrical Engineering and Computer Science, The University of Kansas, Lawrence, KS, USA;Information and Telecommunication Technology Center, Department of Electrical Engineering and Computer Science, The University of Kansas, Lawrence, KS, USA

  • Venue:
  • Computer Communications
  • Year:
  • 2011

Quantified Score

Hi-index 0.24

Visualization

Abstract

Millimeter-wave mesh networks have the potential to provide cost-effective high-bandwidth solutions to many current bandwidth-constrained networks including cellular backhaul. However, the availability of such networks is severely limited due to their susceptibility to weather, such as precipitation and humidity. In this paper, we present a rigorous approach to survivable millimeter-wave mesh networks based on experimentation, modeling, and simulation. Individual link performance is characterised using frame error-rate measurements from millimeter-wave transmissions on test links over a period of one year. A geometric model based on radar-reflectivity data is used to characterise rain storms and determine their impact on spatially correlated links of a mesh network. To mitigate the impact of link impairments on network services, we present two cross-layered routing protocols to route around the failures: P-WARP (predictive weather-assisted routing protocol) and XL-OSPF (cross-layered open shortest-path first). We conduct a performance analysis of the proposed mesh network under the presence of actual weather events as recorded by the US National Weather Service. Results show that the proposed approach yields the highest dependability when compared against existing routing methods.