Your SlideShare is downloading. ×
On Effectively Exploiting Multiple Wireless Interfaces in ...
Upcoming SlideShare
Loading in...5
×

Thanks for flagging this SlideShare!

Oops! An error has occurred.

×
Saving this for later? Get the SlideShare app to save on your phone or tablet. Read anywhere, anytime – even offline.
Text the download link to your phone
Standard text messaging rates apply

On Effectively Exploiting Multiple Wireless Interfaces in ...

1,190
views

Published on


0 Comments
0 Likes
Statistics
Notes
  • Be the first to comment

  • Be the first to like this

No Downloads
Views
Total Views
1,190
On Slideshare
0
From Embeds
0
Number of Embeds
0
Actions
Shares
0
Downloads
23
Comments
0
Likes
0
Embeds 0
No embeds

Report content
Flagged as inappropriate Flag as inappropriate
Flag as inappropriate

Select your reason for flagging this presentation as inappropriate.

Cancel
No notes for slide

Transcript

  • 1. On Effectively Exploiting Multiple Wireless Interfaces in Mobile Hosts Cheng-Lin Tsao and Raghupathy Sivakumar Georgia Institute of Technology ACM CoNEXT ‘09 Dec 4, 2009 Rome, Italy
  • 2. Introduction • Multi-homed wireless devices – Laptops equipped with Wi-Fi, WWAN, Bluetooth, WiMAX – Smartphones equipped with Wi-Fi and 2.5G/3G • BlackBerry, iPhone, Google Android – Only one interface used at a time • Priority-based switching • What is the best approach to leverage the multiple interfaces available at a mobile device in terms of user performance? 2
  • 3. Experimental Testbed • Real-life heterogeneous wireless testbed – Laptop Ethernet AT&T DSL • Atheros 802.11a/b/g Internet Desktop server PCMCIA Network Nightmare WAN emulator • Verizon USB727 Ethernet EVDO Georgia Tech Verizon T-Mobile – Google Android G1 Wi-Fi network 3G network 3G network • Embedded 802.11g 802.11b/g EVDO 802.11g HSDPA • T-Mobile HSDPA – Desktop server Laptop Smartphone (Google Android) • WAN emulator on Wi- Fi to mimic Internet 3
  • 4. Motivation • Simple aggregation – Ex. pTCP [ICNP`02] B1 – Sum of available BW B1+B2 – Marginal benefits B2 • Achieving performance 16 Wi-Fi 14 better than the sum of TCP Throughput (Mbps) 12 3G the parts? 10 8 simple aggr 6 • Leveraging heterogeneity 4 2 – Wi-Fi: high bandwidth 0 laptop (802.11b) laptop (802.11g) google android – 3G: high availability, Wireless interfaces allocated resources 4
  • 5. Super-Aggregation • Concept: transferring data over multiple cooperating interfaces w/ high-layer knowledge • Design – Three generic principles – TCP-specific throughput enhancement in Wi-Fi + 3G – Extension to other protocols/wireless interfaces • Realization – Client software changes that work w/ legacy servers – Layer-3.5 TCP acceleration in Wi-Fi + 3G networks – Prototyped and evaluated on laptop and smartphone 5
  • 6. Principle 1: Selective Offloading • Concept – Selectively offload certain portions of the transferred data to the low-bandwidth interface • Relation to TCP: self-contention Impact from self-contention in Wi-Fi – Between uplink ACK and 25 downlink DATA 20 TCP Throughput (Mbps) UDP – PHY/MAC overheads 15 biUDP – Degrading throughput by 10 30%~70% 5 – Verified with bidir. UDP 0 laptop (802.11b) laptop (802.11g) google android • 1464B data and 32B ack Wireless interfaces 6
  • 7. Solution 1: Offloading-ACK • Idea – Diverting uplink ACKs from Wi-Fi to 3G • Challenges – Insufficient bandwidth on 3G – Long RTT on 3G degrades overall TCP throughput • Solution – Fractional offloading – Opportunistic operation 7
  • 8. Offloading-ACK Details • Fractional offloading – Offloading sustainable fractions on 3G – Discarding ACKs carrying no additional information • Opportunistic operation – Offloading when RTT inflation has little impacts, like cwnd more than a threshold – Heuristic: ssthresh value of TCP • Default value: 20 mss 8
  • 9. Principle 2: Proxying • Concept – Use the low-bandwidth interface for critical control when the high-bandwidth one is temporarily down • Relation to TCP: blackouts Impact from blackout 120 – Blackout: fading or handoff 25 throughput Cwnd and ssthresh (mss) cwnd 100 20 • Vehicle net: up to 75 sec1 Throughput (Mbps) ssthresh 80 15 – Impacts 60 10 • RTO timeout blackout 40 5 20 • Unnecessary idle 0 0 • Slow slow-start 0 2 4 6 8 10 Time (sec) 1V. Bychkovsky, B. Hull, A. Miu, H. Balakrishnan, and S. Madden, “A measurement study of vehicular internet access using in situ wi-fi networks,” Mobicom `06. 9
  • 10. Solution 2: Proxying-blackout-freeze • Idea – Use the 3G link to notify the TCP sender about blackouts on Wi-Fi • Challenges – Real-time blackout detection with low overhead – Freezing a TCP connection during blackout • Solution – Hybrid blackout detection – Freezing TCP with flow control 10
  • 11. Proxying-blackout-freeze Details • Freezing TCP with flow control – Sending a zero-window advertisement on 3G to make TCP enter persist mode – Resuming the TCP connection w/ the original flow window via 3G • Hybrid blackout detection – Passive monitoring of received packets/beacons – Active probing when no activity for a certain period • ICMP probing if no packet for more than 200 ms 11
  • 12. Principle 3: Mirroring • Concept – Intelligently mirror the certain portion of the transferred data on the low-bandwidth interface • Relation to TCP: random losses Impact from random wireless losses – Caused by interference, 10 fading, or long distance 8 Throughput (Mbps) – Interpreted by TCP as 6 congestion 4 – 0.1% packet loss rate 2 reduces TCP throughput 0 0.0% 0.1% 0.3% 1.0% 3.0% by 49% Packet loss rate 12
  • 13. Solution 3: Mirroring-loss-fetching • Idea – Hide random losses in the original connection and fetch the lost packets in the mirror connection • Challenges – Decoupling TCP congestion control and reliability – Mirroring the TCP connection on 3G – Efficiently fetching lost packets on 3G • Solution – Loss distinction – Connection mirroring – Fast fetching 13
  • 14. Mirroring-loss-fetching Details • Loss distinction – Receiving corrupted frames indicates random losses • TCP connection mirroring – Replay messages exchanged in the original connection – Offset TCP sequence numbers – Verify identical data received in the mirror connection • Selected and fast fetching – Proactively acknowledge unneeded packets – Place a guard time before fetching the desired packets • 256 ms in prototype 14
  • 15. Super-Aggregation Architecture • Software Architecture TCP – Client-only changes Loss Hider – Layer-3.5 middleware Blackout ACK Marker – Transparency to TCP & Fast Fetcher Handler link layers Mirroring Manager • Integrated operations Offloader – Offloading-ACK – Proxying-blackout- Blackout Detector Intf . Characterizer freeze Wi-Fi Interface 3G Interface Data of upstream traffic Data of downstream traffic – Mirroring-loss-fetching ACK of upstream traffic ACK of downstream traffic 15
  • 16. Performance Evaluation • Performance metric – Overall throughput of bulk data transfer 20 – Improvement over simple default TCP simple aggregation Throughput (Mbps) 15 aggregation super-aggregation • Improvement on Android 10 – Offloading: 26% 5 – Proxying: 35% 0 Offloading-ACK Proxying- Mirroring-loss- • Blackouts (2 sec every 20 blackout-freeze fetching sec) – Mirroring: 52% • Random packet loss 0.3% 16
  • 17. Offloading-ACK Analysis • Avoiding TCP self- contention in Wi-Fi super-aggregation – Packets captured with tcpdump – Self-contention observed data ack in default TCP default TCP – Offloading-ACK utilizes the Wi-Fi downlink 60.00 60.01 60.02 60.03 Time (sec) 17
  • 18. Proxying-blackout-freeze Analysis • Minimizing the impact from blackouts Proxying-blackout-freeze – Same blackout period 25 tput throughput 120 Cwnd and ssthresh (mss) cwnd 100 introduced 20 cwnd ssthresh Throughput (Mbps) ssthresh 80 – Avoiding slow start 15 60 10 – Quick resumption after link blackout 40 recovery 5 20 0 0 – Maintaining cwnd and 0 2 4 6 8 10 Time (sec) ssthresh 18
  • 19. Mirroring-loss-fetching Analysis • Recovering lost packets efficiently – Recovering 393 packets 500 out of 100k packets (0.3% 400 Sequence number loss on Wi-Fi) 300 guard )atad ( gnihcte f tsa f – Fast fetching recovers lost 200 )kca ( gnihcte f tsa f )atad( gnihcte f lamron )atad ( gnihcte f lam ron segments 36x faster 100 )kca( gnihcte f lamron )kca ( gnihcte f lam ron 0 – Guard time makes sure 0.4 0.5 0.6 0.7 0.8 0.9 1.0 packet delivery on 3G Time (sec) 19
  • 20. Integrated Operation Performance • Integrated operations – Evaluated with a scenario with blackouts and random losses • Enter 3G when t=5 30 • 2-sec blackouts when 25 PCT tluafed PCT tluafed noitagergga-elpmis noitagergga-elpmis t=50 & t=62.5 noitage rgga-repus Throughput (Mbps) 20 • 1% packet loss at AP3 15 – Improving throughput by 10 169% in the scenario 5 0 0 10 20 30 40 50 60 70 80 90 100 Time (sec) 20
  • 21. Related Works • Simple aggregation – pTCP [ICNP`02], WAMP [Globecom`99], RMTP [ICNP`01], MC2 [ToMC`07], MAR [MobiSys`04] – Wireless specific: R2CP [MobiCom`05] and PRISM [ToMC`07] – Requiring two-point deployment • TCP enhancement over a single wireless network – Random losses: Snoop [MobiCom`95], WTCP [MobiCom`99] – Blackout: Freeze-TCP [Infocom`00] • Multi-interface mechanisms for energy efficiency – CoolSpots [MobiSys`06], Cell2Notify [MobiSys`07], Context-for- Wireless [MobiCom`06] 21
  • 22. Concluding Remarks • Study super-aggregation of heterogeneous wireless interfaces • Propose super-aggregation principles – Offloading-ACK – Proxying-blackout-freeze – Mirroring-loss-fetching – Generalization to rate-adaptive video streaming and more other wireless technologies • Design and prototype the integrated architecture • Evaluate on laptop/smartphone in testbed 22
  • 23. Thank you!