FPGA-based Router Virtualization: A Power Perspective

Both Internet and semiconductor technology have advanced dramatically over the past decade. These advancements have made great impact on the conventional Internet infrastructure where networking equipment is dedicated on a per network basis. Router virtualization allows a single hardware router to serve packets from multiple networks while ensuring the same throughput and Quality of Service (QoS) guaranteed originally. In this paper, we study the effect of router virtualization, from a power consumption perspective, on the widely used Field Programmable Gate Array (FPGA) platform. An analytical model is proposed to estimate Layer-3 power consumption under different virtual router configurations. The analytical model is verified using post place-and-route results obtained using state-of-the-art FPGA and the models stand accurate with only a ±3% maximum error. Low power FPGA families are explored in this work to highlight the benefits of using such platforms in networking environments. Our experimental results show that by virtualizing, power savings proportional to the number of virtual networks can be achieved compared with non-virtualized routers.

[1]  Viktor K. Prasanna,et al.  Towards On-the-Fly Incremental Updates for Virtualized Routers on FPGA , 2011, 2011 21st International Conference on Field Programmable Logic and Applications.

[2]  Walid Dabbous,et al.  Survey and taxonomy of IP address lookup algorithms , 2001, IEEE Netw..

[3]  Viktor K. Prasanna,et al.  Memory-efficient and scalable virtual routers using FPGA , 2011, FPGA '11.

[4]  Viktor K. Prasanna,et al.  Multiroot: Towards Memory-Efficient Router Virtualization , 2011, 2011 IEEE International Conference on Communications (ICC).

[5]  Javier Jiménez,et al.  Network virtualization: a view from the bottom , 2009, VISA '09.

[6]  Viktor K. Prasanna,et al.  Multi-Way Pipelining for Power-Efficient IP Lookup , 2008, IEEE GLOBECOM 2008 - 2008 IEEE Global Telecommunications Conference.

[7]  Bin Liu,et al.  An ultra high throughput and power efficient TCAM-based IP lookup engine , 2004, IEEE INFOCOM 2004.

[8]  Fang Hao,et al.  Building Scalable Virtual Routers with Trie Braiding , 2010, 2010 Proceedings IEEE INFOCOM.

[9]  Viktor K. Prasanna,et al.  Towards Green Routers: Depth-Bounded Multi-Pipeline Architecture for Power-Efficient IP Lookup , 2008, 2008 IEEE International Performance, Computing and Communications Conference.

[10]  Lixin Gao,et al.  Scalable network virtualization using FPGAs , 2010, FPGA '10.

[11]  Jing Fu,et al.  Efficient IP-address lookup with a shared forwarding table for multiple virtual routers , 2008, CoNEXT '08.

[12]  Stefanos Kaxiras,et al.  IPStash: a Power-Efficient Memory Architecture for IP-lookup , 2003, MICRO.