300 likes | 444 Views
Enhua Tan 1 , Lei Guo 1 , Songqing Chen 2 , Xiaodong Zhang 1 1 The Ohio State University 2 George Mason University. SCAP: Smart Caching in Wireless Access Points to Improve P2P Streaming. Background. Wireless access to Internet is pervasive:
E N D
Enhua Tan1, Lei Guo1, Songqing Chen2, Xiaodong Zhang1 1The Ohio State University 2George Mason University SCAP: Smart Caching in Wireless Access Points to Improve P2P Streaming ICDCS’07, Toronto, Canada
Background • Wireless access to Internet is pervasive: • On campus, in offices, at home, and public utilities • Most are supported by Wireless LANs • Peer-to-Peer applications are widely used: • Streaming: PPLive, Joost, etc … • VoIP: Skype, etc … • Large file distribution: BitTorrent, etc … Our Focus: Interaction between wireless users and P2P streaming applications
Wired/wireless Communications Internet WLAN Access Point (AP) Wired users Wireless users
P2P Streaming for Wired/wireless Users:Workflow Internet Source Peer Access Point Viewing Peer Wireless Peer WLAN
P2P Streaming for Wired/wireless Users: Problems Internet Downstream traffic for other wireless users AFFECTED Source Peer WLAN Generating upstream traffic Viewing Peer Streaming content Streaming quality degraded Wireless Peer (Relay/Viewing) Other packets
Problem Summary • Peers in WLAN may relay streaming content by uploading a lot of traffic: • Congest the WLAN due to channel competitions • Provide low quality of service to the Internet peers • Downstreams have lower priority due to upstreams • Extra upstream traffic: • further increase the number of transmission errors • increase the cost of contention window back-off Major problem source:upstream relay traffic Can we minimize upstream traffic with low overhead? • to improve WLAN throughput • to improve service quality for Internet peers
P2P Streaming for Wired/wireless Users:Workflow The same content is transferred twice in the WLAN! Duplicated traffic Internet Source Peer Access Point Viewing Peer Wireless Peer WLAN
Contributions • Our measurements show that > 75% upstream traffic is duplicated with the downstream traffic for three representative applications • SCAP: Smart Caching in the Access Point for minimizing upstream traffic: design & prototype implementation • Evaluation results show SCAP can improve the throughput of the WLAN by up to 88%: • SCAP also reduces the delay to Internet peers
Outline • Problem Summary and Contributions • Measurement & Analysis of P2P Streaming Traffic • SCAP Design & Implementation • Evaluation • Summary
Measurement & Analysis of P2P Streaming Traffic • Aim to answer two questions: • How much duplicated traffic in practice? • How much overhead in identifying such duplications? • Measurement: • Collect traces of three representative P2P live streaming applications: PPLive, ESM, and TVAnts • In LAN (100Mbps) and WLAN (802.11b)
Workload Statistics • Downstream throughput is typically 300~400Kbps • Upstream traffic to downstream traffic: • Can be as large as10times for PPLive due to its popularity • Between 2 to 4 times for TVAnts • Not too much for ESM • PPLive and ESM: most in TCP • TVAnts: 74% in UDP for WLAN
Duplication Detection Methods:Fixed Hashing • Offline workload analysis: • Fixed Hashing (FH) • Compute only 1 fingerprint (hash value) for a downstream packet; store this fingerprint in a hash table, and cached the packet in FIFO buffer • For each upstream packet, also compute the fingerprint, and look it up in the hash table to locate the duplicated downstream packet; If found the same fingerprint, do further byte-to-byte comparison Lookup Downstream packet fingerprint hash table Upstream packet Downstream packet FIFO buffer Upstream packet fingerprint
Duplication Detection Methods:Rabin Fingerprinting • Rabin Fingerprinting (RF) • A unique hash function: produce fingerprints for a continuous data stream quickly (NSDI’07 BitTyrant) • We scan the whole packet and only store fingerprints ending with 8 zeros over 64 bytes content • averagely 5 fingerprints for a 1400 bytes packet (1/28) • FIFO Buffer: stores latest 50,000 downstream packets • Buffer + hash table: need about 75MB memory totally
Dup Ratio & Tput • Offline analysis processing throughput of RF is less than FH: • Still large enough (> 90Mbps) for process P2P streaming (400 Kbps) • RF can detect more duplications than FH • All the duplication ratios are larger than 75%
Duplication Beginning Offset • FH can only detect the duplication when the offsets for up/downstream packets are the same (no re-packetizing) • ESM does not have any offset differences FH performs well • TVAnts has a lot of re-packetizing FH performs the worst
Forwarding Delay 200 seconds 200 seconds • PPLive and TVAnts: most upstream packets forwarded in 200 seconds • <20 seconds for 70% • ESM: within 10 ms • Implies the downstream buffer can be quite small 10 seconds 20 seconds 10 ms
Outline • Problem Summary and Contributions • Measurement & Analysis of P2P Streaming Traffic • SCAP Design & Implementation • Evaluation • Summary
SCAP (Smart Caching in Access Points) Overview Internet Access Point Metadata upstream packet (If duplications found in downstream buffer) Downstreams buffer Downstream buffer Relay/Viewing Peer Original upstream packet WLAN
Design Issues • Buffer size: • Need 7.5MB for storing recent 200 seconds traffic (in 300Kbps rate), which is affordable for a wireless station • But AP will need to buffer for multiple stations: • AP should dynamically adjust the buffer space for each station according to its duplication ratios in order to achieve highest traffic reduction with limited buffer space • Buffer synchronization between AP and station: • If a metadata upstream packet cannot be reassembled on AP due to a cache miss, TCP flow will be stalled • Wireless station caches several copies of recent sent upstream packets and resends the uncompressed packet when needed
Prototype Implementation • Modified HostAP driver in Linux kernel 2.6.16 for the AP and stations • Wireless card is based on Intersil Prism 2.5 chipset (802.11b) • Identification of the downstream packet • For AP to locate the packet in decompressing the upstream packet • Cannot use Sequence Control field (2 bytes) because it is filled by the firmware • Have to use the first fingerprint value (8 bytes)
Outline • Problem Summary and Contributions • Measurement & Analysis of P2P Streaming Traffic • SCAP Overview • Design & Implementation • Evaluation • Summary
Performance Evaluation: LAN Experiment 4.50 8.9 Mbps 4.43 Mbps • Station first receives a file from a server, then sends it back • RF: little overhead for the downstream throughput (1.5% decrease), and 88% improvement for the upstream throughput • FH: cannot have any improvement due to constant TCP re-packetizing 4.7
Performance Evaluation:Internet Experiment • Evaluate PPLive, TVAnts, and ESM • Run the applications in a VMWare-based Windows XP guest OS for HostAP driver to work • Measurement methods: • Because P2P Streaming is a Constant Bit Rate stream: • Upstream throughput will not change even if we reduces its traffic • Running iperf on another wireless station to observe the impact to WLAN TCP throughput • Running Ping to observe the impact to response time • Run multiple trials to get comparable P2P downstream throughput for comparison • Each trial runs for 600 seconds
Internet Experiment:Evaluation Results • RF/FH performs best for TVAnts since it has the largest volume of upstream traffic: • Increases TCP throughput by 0.95 Mbps (54% of upstream traffic) • Decrease Ping round-trip time by 83 ms (-26%) • Also performs well for PPLive/ESM
Summary • With the increasing popularity of P2P streaming applications and pervasive deployment of 802.11 WLANs, more peers will be connected by wireless • We study the impact of wireless peers to the performance of wireless and Internet users • Without a proper control of P2P traffic, the performance of both parties can be significantly affected • We designed and implemented SCAP (Smart Caching in Access Points) in order to reduce the upstream traffic for P2P live streaming applications • Our prototype based evaluation shows the effectiveness of SCAP: • SCAP improves the throughput of the WLAN by up to 88% • SCAP reduces the response delay to Internet peers as well
Thank you!Enhua Tan:etan@cse.ohio-state.eduhttp://www.cse.ohio-state.edu/hpcs/
SCAP (Smart Caching in Access Points) – Basic Idea • AP stores downstream data in buffer (1) • Station stores downstream data in buffer (2) • Compare upstream packet (3) with (2), upload difference (4) • AP will assemble upstream packet with data in (1) to the Internet
Rabin Fingerprinting • Rabin Fingerprinting (RF) can produce fingerprints for a continuous data stream quickly: • Advance the fingerprint only requires an addition, a multiplication, and a mask • Lack of this property for other hash functions like MD5/SHA (and they are also more complex)
Some Related Work • XORs in the Air: Practical Wireless Network Coding (Sigcomm’06) • Utilizing the broadcasting nature of wireless networks to improve throughput of multi-hop network (instead of application characteristics) • Our scheme is utilizing the traffic pattern of P2P applications • A Protocol-Independent Technique for Eliminating Redundant Network Traffic (Sigcomm’00) • reduces redundant traffic using Rabin Fingerprinting • A Low-bandwidth Network File System (SOSP’01) • Exploits similarities between different versions of a file to reduce update traffic