Presentation is loading. Please wait.

Presentation is loading. Please wait.

Reza Rejaie AT&T Labs - Research1 Reza Rejaie AT&T Labs – Research Menlo Park, CA Jussi Kangasharju Institut Eurocom France NOSSDAV 2001, New York June.

Similar presentations


Presentation on theme: "Reza Rejaie AT&T Labs - Research1 Reza Rejaie AT&T Labs – Research Menlo Park, CA Jussi Kangasharju Institut Eurocom France NOSSDAV 2001, New York June."— Presentation transcript:

1 Reza Rejaie AT&T Labs - Research1 Reza Rejaie AT&T Labs – Research Menlo Park, CA Jussi Kangasharju Institut Eurocom France NOSSDAV 2001, New York June 25, 2001 http://www.research.att.com/~reza Mocha: A quality Adaptive Multimedia Proxy Cache for Internet Streaming

2 Reza Rejaie AT&T Labs - Research2 Motivation Rapid growth in client-server multimedia streaming over the Internet Client-server architecture has two major limitations: Limited and unstable quality Limited scalability

3 Reza Rejaie AT&T Labs - Research3 Unicast Internet Streaming Multimedia streams are pipelined through the network Quality of stream should match available bandwidth  Internet streaming applications should be quality adaptive Cong. Ctrl Buffer Decoder Server Display Encoder Source TCP Internet QualityAdapt

4 Reza Rejaie AT&T Labs - Research4 Issues with Client-Server Streaming Limited & unstable quality Limited scalability Asynchronous access could be inefficient RTT could be high High delay VCR-functions Large startup delay  Multimedia proxy caching can address all these issues Server Client Internet Server Proxy Client Internet

5 Reza Rejaie AT&T Labs - Research5 Multimedia Caches should be Quality Adaptive To maximize delivered quality to heterogeneous clients: Multiple encodings/versions Trans-coding Layered encoding Design issues: What streams to cache? Which quality to cache?  Notion of “quality” affects both design and evaluation Server Proxy Client Internet 56 Kbps2 Mbps

6 Reza Rejaie AT&T Labs - Research6 Previous Work Existing Web caches can not support Multimedia streams efficiently Atomic delivery & Atomic replacement Multimedia caches = web cache + media player Caching only selected portions of streams Prefix caching, Video Staging, etc Memory Caching, batching, etc Resource-based Caching  Previous work treat multimedia streams in an atomic fashion => Not Quality Adaptive

7 Reza Rejaie AT&T Labs - Research7 This Paper Design and implementation of Mocha, a quality adaptive proxy cache for multimedia streams, on top of Squid Preliminary evaluation (Sanity Checking)

8 Reza Rejaie AT&T Labs - Research8 Design Goal Assuming locality of reference exists Goal: Cache popular streams with appropriate quality Appropriate quality is determined by Client bandwidth Popularity of streams Max. deliverable quality is not guaranteed  Caching appropriate quality => Higher performance

9 Reza Rejaie AT&T Labs - Research9 The idea Server Proxy Client Internet Exploit layered organization Relay on a cache miss Pre-fetch on a cache hit  If higher quality is required Two key components: Fine-grained Prefetching Fine-grained Replacement

10 Reza Rejaie AT&T Labs - Research10 Client-Server Architecture

11 Reza Rejaie AT&T Labs - Research11 Internal Architecture Mocha

12 Reza Rejaie AT&T Labs - Research12 RTSP Signaling Mocha Cache Hit Cache Miss

13 Reza Rejaie AT&T Labs - Research13 Main Components Object Management Fine-grained Pre-fetching Fine-grained Replacement Mocha

14 Reza Rejaie AT&T Labs - Research14 Object Management Cache RTP packets instead of raw payload Challenge: store and access partially received layers of a single streams All layers should be collectively viewed as a single object by Squid Need to extend Squid’s data structures: One file per layer, Each layer consists of Chunks A chunk contains a group of contiguous pkts Chunks are treated in an atomic fashion Mocha/Main Components

15 Reza Rejaie AT&T Labs - Research15 Data Structure Mocha/Object Management

16 Reza Rejaie AT&T Labs - Research16 Fine-grained Online Pre-fetching Pre-fetching stream is congestion controlled Pre-fetching & playback should remain loosely sync.  Sliding-window Batch of missing segments Prioritized delivery Extending “Range” header field of RTSP Time L 0 L 1 L 2 L 3 L 4 Quality ( layers ) Pre-fetching Window Td t p Client ProxyServer Pre-fetchPlayback A Segment Mocha/Main Components 1 2 3 4 56

17 Reza Rejaie AT&T Labs - Research17 Fine-grained popularity Assign popularity to individual layers Pipelining -> Hit could be any value within [0..1] This definition of popularity captures: Level of interest among clients Available bandwidth to interested clients Within a single stream, layer popularity monotonically decreases Fine-grained Replacement whit = weighted_hit = PlaybackTime(sec)/StreamLength(sec) Mocha/Main Components

18 Reza Rejaie AT&T Labs - Research18 Per-layer popularity  Victim layer Per-segment replacement Demand-driven Cached chunk Time Quality(Layer) Replacement Pattern Mocha

19 Reza Rejaie AT&T Labs - Research19 Data set: 6 layers per stream Layer BW = 6 Kbps random length within [30..180]sec Cache size: 30% of data set Popularity Win = Infinite 5000 requests Zipf-like popularity dist Proxy Server Client 1 Client 2 bw 1 2 sp Experiments (Sanity Check) Mocha

20 Reza Rejaie AT&T Labs - Research20 Fine-grained Replacement Single Client Pre-fetching Off 50 Streams bw_sp > 6 layers bw1 = 4 layers

21 Reza Rejaie AT&T Labs - Research21 Fine-grained Pre-fetching Two Clients Pre-fetching On 20 streams 70% high bw request bw_sp > 6 layers bw1 > 6 layers Bw2 = 2 layers

22 Reza Rejaie AT&T Labs - Research22 Average Delivered Quality Two Clients Pre-fetching On 20 Streams bw_sp > 6 layers bw1 > 6 layers Bw2 = 2 layers

23 Reza Rejaie AT&T Labs - Research23 Summary Mocha is a quality adaptive multimedia proxy cache Features: Able to manage layer-encoded streams Fine-grained pre-fetching Fine-grained replacement

24 Reza Rejaie AT&T Labs - Research24 Future Directions Developing a methodology for performance evaluation of multimedia proxy caches Examining various replacement and pre-fetching mechanisms Encoding/content specific replacement & pre-fetching

25 Reza Rejaie AT&T Labs - Research25 Reza Rejaie AT&T Labs – Research Menlo Park, CA Jussi Kangasharju Institut Eurocom France NOSSDAV 2001, New York June 25, 2001 http://www.research.att.com/~reza Mocha: A quality Adaptive Multimedia Proxy Cache for Internet Streaming

26 Reza Rejaie AT&T Labs - Research26 Evaluation Methodology Web caching evaluations are not sufficient Evaluation should be performed across Quality-Load plan Parameters of a request sequence Popularity distribution (similar to Web) Distribution of request among different classes of clients

27 Reza Rejaie AT&T Labs - Research27 Client-server Internet Streaming Internet streaming applications should be quality adaptive(QA) QA is often needed in a shared environment, e.g. Diff-serve without per-flow admission control Shared reservation Streaming over best-effort class Quality adaptation Adjust the quality with long-term changes in BW Limited quality & Limited scalability

28 Reza Rejaie AT&T Labs - Research28 Performance Evaluation Goal of Web caching: to maximize Byte-Hit-Ratio Goal of MCaching: to maximize Byte-Hit-Ratio, and to maximize delivered quality  MCaching performance should be evaluated across quality-BHR plan

29 Reza Rejaie AT&T Labs - Research29 Replication Client Server1 Internet Server2 Client Server1 ISP Campus

30 Reza Rejaie AT&T Labs - Research30 Proxy Caching Client Server1 Internet Server2 Client MCache ISP Campus

31 Reza Rejaie AT&T Labs - Research31 Replication vs Proxy Caching Location: Proxy is often closer to clients No bottleneck Min RTT Higher locality of reference! Content management (Push vs Pull) A proxy adaptively caches popular streams from different servers based on clients’ interest A mirror server statically replicates content of a specific group of servers Administration

32 Reza Rejaie AT&T Labs - Research32 Pre-fetching: An Example Missing pieces of the active layers are pre- fetched on-demand Required pieces are identified by QA Pre-fetching results in improvement of quality Pre-fetched data is always cached Time L 0 L 1 L 2 L 3 L 4 Quality ( no. active layers ) Pre-fetched data Stored stream Played back stream Mocha/Design Issues


Download ppt "Reza Rejaie AT&T Labs - Research1 Reza Rejaie AT&T Labs – Research Menlo Park, CA Jussi Kangasharju Institut Eurocom France NOSSDAV 2001, New York June."

Similar presentations


Ads by Google