Energy Consumption Anatomy of 802.11 Devices and its Implication on - PowerPoint PPT Presentation

Energy Consumption Anatomy of 802.11 Devices and its Implication on Modeling and Design A. Banchs 1,2 , G. Bianchi 3 1: Universidad Carlos III de Madrid 2: Institute IMDEA Networks 3: CNIT / Università Tor Vergata A. García-Saavedra 1 , P. Serrano 1 ,

Energy Consumption Anatomy of 802.11 Devices and its Implication on Modeling and Design What we wanted we need to understand the power consumption • To design an energy efficient comm. protocols • Previous experimental work ‒ Per-packet analysis of the wireless interface Wireless ¡interface ¡ (WiFi ¡NIC) ¡ Linksys ¡WRT54GL ¡WiFi ¡router ¡HW ¡ Rantala ¡et ¡al. ¡“Modeling ¡energy ¡efficiency ¡in ¡wireless ¡internet ¡communica7on” ¡, ¡ACM ¡Mobiheld, ¡2009 ¡

Energy Consumption Anatomy of 802.11 Devices and its Implication on Modeling and Design What we wanted we need to understand the power consumption • To design an energy efficient comm. protocols • Previous experimental work ‒ Per-packet analysis of the wireless interface ‒ Per-state measurements of the device A.Rice, ¡S. ¡Hay ¡“Measuring ¡mobile ¡phone ¡energy ¡consump7on ¡for ¡802.11 ¡wireless ¡networking”, ¡PMC. ¡2010 ¡

Energy Consumption Anatomy of 802.11 Devices and its Implication on Modeling and Design What we found Card App O.S. Device Non-card operations Card operations • Non-card can dominate the consumption • Questions previous schemes ‒ E.g. relaying in multihop • Enables new designs ‒ E.g. packet batching

Energy Consumption Anatomy of 802.11 Devices and its Implication on Modeling and Design Linksys 500 MHz AMD LX800 256 MB SDRAM Ubuntu 10.04 Kernel 2.6.29 WRT54GL Energy Consumption Anatomy Broadcom BM4320 (11b/g) 200 MHz BCM5352 16 MB RAM OpenWrt Backfire Kernel 2.6.32 BCM4319 (11b/g) Broadcom Alix 2d2 net4826-48 Device WiFi chipset CPU Memory Software / OpenBSD 5.1 Soekris Atheros AR5414 (11a/b/g) 233 MHz AMD SC1100 128 MB SDRAM Gentoo 10.0 Kernel 2.6.24 • Hardware used Device running PCE PA-6000 Protek 3033B controlled experiments

Energy Consumption Anatomy of 802.11 Devices and its Implication on Modeling and Design Methodology Baseline power consumption • With one device ‒ Results are not very precise (e.g. ~6%) Power ‒ We added more devices (~2%)

Energy Consumption Anatomy of 802.11 Devices and its Implication on Modeling and Design Power consumption: Transmission 6,25 6,25 6,25 UDP, no ACKs, no retx. 5,75 5,75 5,75 24Mbps, ¡400fps, ¡15dBm ¡ Power (Watts) Power (Watts) Power (Watts) 5,25 5,25 5,25 6Mbps, ¡400fps, ¡15dBm ¡ 4,75 4,75 4,75 4,25 4,25 4,25 6Mbps, ¡400fps, ¡5dBm ¡ 3,75 3,75 3,75 3,25 3,25 3,25 0 10 20 30 40 50 60 70 80 90 100 0 0 10 10 20 20 30 30 40 40 50 50 60 60 70 70 80 80 90 90 100 100 airtime (%) airtime (%) airtime (%) • Varying frame length -> Airtime = T plcp +(H+L)/R

Energy Consumption Anatomy of 802.11 Devices and its Implication on Modeling and Design P(fps) + P tx (MCS,power) x Airtime Power consumption: Transmission 24Mbps, ¡1200fps, ¡15dBm ¡ 6,25 6,25 6,25 6,25 6,25 6,25 UDP, no ACKs, no retx. 5,75 5,75 5,75 5,75 5,75 5,75 24Mbps, ¡400fps, ¡15dBm ¡ Power (Watts) Power (Watts) Power (Watts) Power (Watts) Power (Watts) Power (Watts) 5,25 5,25 5,25 5,25 5,25 5,25 6Mbps, ¡400fps, ¡15dBm ¡ 4,75 4,75 4,75 4,75 4,75 4,75 4,25 4,25 4,25 4,25 4,25 4,25 +1.2 W 6Mbps, ¡400fps, ¡5dBm ¡ 3,75 3,75 3,75 3,75 3,75 3,75 +0.4 W 3,25 3,25 3,25 3,25 3,25 3,25 0 0 0 0 0 0 10 10 10 10 10 10 20 20 20 20 20 20 30 30 30 30 30 30 40 40 40 40 40 40 50 50 50 50 50 50 60 60 60 60 60 60 70 70 70 70 70 70 80 80 80 80 80 80 90 90 90 90 90 90 100 100 100 100 100 100 airtime (%) airtime (%) airtime (%) airtime (%) airtime (%) airtime (%) • P = P base +

Energy Consumption Anatomy of 802.11 Devices and its Implication on Modeling and Design Per-Packet “energy toll” Soekris: 0.93 mJ/frame (Linux), 1.27 mJ/frame (OpenBSD) Linksys: 0.46 mJ/frame Alix: 0.11 mJ/frame slope ≈ 1 W / 1000 fps = 1mJ/frame

Energy Consumption Anatomy of 802.11 Devices and its Implication on Modeling and Design Energy Consumption Anatomy Transmission “Cross-Factor” 6,25 6,25 Power (Watts) Power (Watts) 5,75 5,75 • (a) App.: disc. before the OS 5,25 5,25 • (b) TCP/IP: disc. before driver 4,75 4,75 4,25 4,25 • (c) Driver: disc. after driver 3,75 3,75 • Total 3,25 3,25 0 0 20 20 40 40 60 60 80 80 100 100 airtime (%) airtime (%)

Energy Consumption Anatomy of 802.11 Devices and its Implication on Modeling and Design Anatomy Results 2.6 2.4 2.2 6 Mbps 1400 B/Pkt 400 fps 48 Mbps ≈ ≈ 1400 B/Pkt 48 Mbps 1200 fps 750 B/Pkt 1200 fps 48 Mbps 48 Mbps 100 B/Pkt 100 B/Pkt 300 fps 1200 fps

Energy Consumption Anatomy of 802.11 Devices and its Implication on Modeling and Design TCP/IP 22% 21% 33% 24% NIC Driver App The Cross Factor • Energy toll to handle a frame ‒ Independent of frame size ‒ Total power > base power + card power • Energy split: • Very far from negligible (vs. Tx Power) ‒ Previous slide: 37% ~ 97% energy/frame

Energy Consumption Anatomy of 802.11 Devices and its Implication on Modeling and Design Retransmissions (and control frames) • E.g. 2 retries, but only 1 cross factor

Energy Consumption Anatomy of 802.11 Devices and its Implication on Modeling and Design Retransmissions ReTx Tx X-factor

Energy Consumption Anatomy of 802.11 Devices and its Implication on Modeling and Design Model for the power consumption • Similar results for reception. • Model: Baseline ¡ TX ¡air7me ¡ “Classical” New RX ¡air7me ¡ Packet ¡processing ¡ • Parametrization for the Soekris, Linksys, Alix

Energy Consumption Anatomy of 802.11 Devices and its Implication on Modeling and Design Validation of the model AP Sta Sta Sta • General scenarios

Energy Consumption Anatomy of 802.11 Devices and its Implication on Modeling and Design Ok, but... classical model detailed anatomy Cross factor: energy / frame • Does it matter? 37% ~ 97% ∆ • What are the implications? • 1. Revisit old proposals based on the • 2. Design of new schemes building on the

Energy Consumption Anatomy of 802.11 Devices and its Implication on Modeling and Design 250 Old: Packet relays 2 hops 1 hop Power (Watts) 1500 1250 10,7 750 500 1000 0 11,1 11,6 10,9 11 10,8 11,2 11,3 11,4 11,5 6Mbps ¡ STA ¡1 ¡ AP ¡ STA ¡2 ¡ 48Mbps ¡ 48Mbps ¡ Classical ¡ Packet Size (B)

Energy Consumption Anatomy of 802.11 Devices and its Implication on Modeling and Design 11 iface Fwd. Old: Packet relays Measurement 2 hops 1 hop iface 12,6 11,4 11,8 12,2 6Mbps ¡ STA ¡1 ¡ AP ¡ STA ¡2 ¡ 48Mbps ¡ 48Mbps ¡ New ¡ Power (Watts) 0 250 500 750 1000 1250 1500 Packet Size (B)

Energy Consumption Anatomy of 802.11 Devices and its Implication on Modeling and Design the protocol stack New: Packet batching • Group n packets before they transverse ‒ Fixed energy cost per bundle ‒ Same information over the medium

Energy Consumption Anatomy of 802.11 Devices and its Implication on Modeling and Design New: Packet batching savings (~80%) according to the classical model • Substantial • No savings

Other implications Energy Consumption Anatomy of 802.11 Devices and its Implication on Modeling and Design App. iface TCP Driver App. • Data compression in multihop ‒ Old model: savings ‒ New model: not • Directed Multicast ‒ Where to generate frames • Use of raw sockets ‒ E.g., skipping TCP/IP: 0.2 mJ/frame

Energy Consumption Anatomy of 802.11 Devices and its Implication on Modeling and Design Summary consumption of a wireless device • Per-packet analysis of the energy ‒ Parametrized for various devices • Characterization of the cross factor • Two-fold impact ‒ Revisit previous schemes ‒ Enable new designs

Energy Consumption Anatomy of 802.11 Devices and its Implication on Modeling and Design Thanks!