An Improved Wireless Communication Fabric for Emerging Network-on-Chip Design

Michael Opoku Agyeman, Kenneth Tong, Terrence Mak

Research output: Contribution to JournalArticlepeer-review

Abstract

Existing wireless communication interface has free space signal radiation which drastically reduces the received signal strength and hence reduces the throughput efficiency of Hybrid Wired-Wireless Network-on-Chip (WiNoC). This paper addresses the issue of throughput degradation by replacing the wireless layer of WiNoCs with a novel Complementary Metal Oxide Semiconductor (CMOS) based wave guide communication fabric that is able compete with the reliability of traditional wired NoCs. A combination of a novel transducer and a commercially available thin metal conductor coated with a low cost Taconic Taclamplus dielectric material is presented to generate surface wave signals with high signal integrity. Our experimental results demonstrate that, the proposed communication fabric can achieve a 5dB operational bandwidth of about 60GHz around the center frequency (60GHz).Compared to existing WiNoCs, the proposed communication fabric a performance improvement of 13.8% and 10.7% in terms of throughput and average packet delay, respectively. Specifically, under realistic traffic patterns, the average packet latency can be reduced by 30% when the mm-Wave is replaced by the proposed communication fabric.
Original languageEnglish
Pages (from-to)415-420
Number of pages5
JournalProcedia Computer Science
DOIs
Publication statusPublished - 31 Jul 2015
EventInternational Conference on Future Networks and Communications(FNC 2015) / The International Conference on Mobile Systems
and Pervasive Computing (MobiSPC 2015)
-
Duration: 10 Apr 2015 → …

Keywords

  • Wireless Network-on-Chip
  • Communication Fabric
  • Surface wave
  • Performance Evaluation
  • Throughput
  • Reliability

Fingerprint

Dive into the research topics of 'An Improved Wireless Communication Fabric for Emerging Network-on-Chip Design'. Together they form a unique fingerprint.

Cite this