Key performance factors of a P2P-VoD System

For more info about our services contact : help@bestpfe.com

Table of contents

1. Problem Statement
2. Overview of the proposed solution
3. Thesis outline
Literature Review
1. Content Streaming Networks
1.1. Server-based approaches
1.2. Peer-based approaches
2. Peer-to-Peer Video-on-Demand Systems
2.1. Tracker-based approaches
2.1.1. pcVoD
2.1.2. BitTorrent and BASS : BitTorrent Assisted Streaming System
2.2. Tree-based approaches
2.2.1. P2Cast and P2VoD
2.2.2. GridCast
2.3. Overlay
2.3.1. PROMISE
2.4. Hybrid CDN-P2P
2.4.1. LiveSky
3. Resource Allocation Strategies in P2P-VoD
3.1. User-based strategies
3.1.1. Random algorithm
3.1.2. Fastest Link
3.1.3. Greedy algorithm
3.1.4. Fit algorithm
3.2. Global-based strategies
3.2.1. Greedy/Fit algorithms
3.2.2. Batch Greedy/Batch Fit
3.3. Reputation System
3.3.1. Highest Reputation Strategy
3.3.2. Tit-for-tat Strategy
3.3.3. Black List Strategy
3.3.4. Comparable Reputation Strategy
3.3.5. FairTrust
3.4. Topology-based Strategies
3.4.1. Neighbor-based
3.4.2. Localization and Congestion-aware system
4. Summary
Analysis of a P2P-VoD System
1. Key performance factors of a P2P-VoD System
1.1. Number of sessions per peer
1.2. Number of peers
1.3. Number of titles
1.4. Popularity per title
1.5. Number of parts per title
1.6. Demand pattern and evolution
1.7. Original Content Dispatching: initial titles injection in the P2P network
2. Performance evaluation metrics of a P2P-VoD System
2.1. Rate of VoD sessions rejected
2.2. Peer Participation and streaming load distribution over active peers
2.3. VoD Provisioning Responsiveness
3. Summary
P2P-VoD System
1. Video Fragmentation
2. System Components
2.1. SuperNode
2.2. Peer
2.3. Cache
3. Summary
1. Popularity-to-Availability Translation
1.1. Impact of Video Content Fragmentation
1.2. Proportional Content Popularity to Content Availability Model
2. Content dispatching strategies
2.1. Serial content fragments dispatching
2.2. Random content fragments dispatching
2.3. Popularity-Weighted content fragments dispatching
3. Summary
Static Resource Allocation
1. Introduction to Peer resource allocation
1.1. Why resource allocation is a key performance factor
1.2. Characterizing the resource allocation process
2. Single-metric based resource allocation
2.1. Caracterisation of basic resource allocation strategies
2.2. Higher Available Uplink Capacity First (HUF)
2.3. Lowest Popularity Score (LPS)
2.4. Lowest Critical-Score (LCS)
3. Multi-criteria resource allocation
3.1. Hierarchically Combined Resource Allocation Strategy
3.2. Multi-objective Optimization
4. Summary
Dynamic Resource Allocation
1. Towards dynamic resource allocation
1.1. Why performances evolve over time
1.2. Why resource allocation is a key performance factor
2. Dynamic resource allocation
2.1. Dynamic Switching between different RA strategies
2.2. Resource Allocation Strategy Evaluation and Selection
2.2.1. Prior probabilities
2.2.2. Probability density
2.2.3. Posterior probabilities
2.2.4. Monte-carlo Simulations
3. Strategy selection based on Evidence Theory
3.1. From Bayes to Dempster-Schafer
3.2. Dempster-Schafer Theory
3.2.1. Basic Probability Assignment
3.2.2. Belief function
3.2.3. Plausibility function
3.3. Application to Resource Allocation Strategy selection
3.3.1. Plausibility applied to strategy evaluation
3.3.2. Problem formulation
3.3.3. Plausibility estimation
4. Summary
Simulation Platform..
1. Experimental platform overview
Content Injection
1. General Study: Random algorithm
2. Popularity-Weighted Content Dispatching (PWCD) Algorithm
Resource Allocation
1. Dynamic demand evolution
2. Resource Allocation Strategies evaluation
2.1. Dynamically Selected Strategies by LB-RA
2.2. Peer Saturation
2.3. Peer Participation
2.4. Rejection Rate
3. Dempster-Schafer Theory applied to LB-RA
3.1. Rejection Rate
3.2. Entropy
3.3. Overall results and discussion
4. Summary
5. Summary of Results and Contributions
6. Future Work
References