MuNCC: Multi-hop Neighborhood Collaborative Caching in - PowerPoint PPT Presentation

MuNCC: Multi-hop Neighborhood Collaborative Caching in Information-Centric Networks Travis Mick , Reza Tourani, Satyajayant Misra New Mexico State University Dept. of Computer Science 28 Sept. 2016 ACM ICN, Kyoto Travis Mick, Reza Tourani, Satyajayant Misra MuNCC: Multi-hop Neighborhood Collaborative Caching

Outline Introduction Prior Work Solution Design Experimental Results Conclusion & Future Work 1 Travis Mick, Reza Tourani, Satyajayant Misra MuNCC: Multi-hop Neighborhood Collaborative Caching

Outline Introduction Prior Work Solution Design Experimental Results Conclusion & Future Work 1 Travis Mick, Reza Tourani, Satyajayant Misra MuNCC: Multi-hop Neighborhood Collaborative Caching

The amount of cacheable content is growing... 63% of all traffic Video Traffic 1.8 ZB per year 79% of traffic by 2020 2 2016 Cisco VNI Travis Mick, Reza Tourani, Satyajayant Misra MuNCC: Multi-hop Neighborhood Collaborative Caching

The ideal caching scheme has several requirements. High Hit Ratio Ideal Caching Reduced Cost Low Overhead Low Latency Improved QoE 3 Travis Mick, Reza Tourani, Satyajayant Misra MuNCC: Multi-hop Neighborhood Collaborative Caching

Caching encompasses several tightly-coupled problems. Replica Placement Admission Control State Redundancy Exchange Elimination Routing & Forwarding 4 Travis Mick, Reza Tourani, Satyajayant Misra MuNCC: Multi-hop Neighborhood Collaborative Caching

Outline Introduction Prior Work Solution Design Experimental Results Conclusion & Future Work 4 Travis Mick, Reza Tourani, Satyajayant Misra MuNCC: Multi-hop Neighborhood Collaborative Caching

Caching strategies can be broadly classified into four categories, each with their own shortcomings. Independent On-Path Neighborhood Global Caching Coordination Coordination Coordination 5 Travis Mick, Reza Tourani, Satyajayant Misra MuNCC: Multi-hop Neighborhood Collaborative Caching

Caching strategies can be broadly classified into four categories, each with their own shortcomings. Independent On-Path Neighborhood Global Caching Coordination Coordination Coordination High Limited High ? Redundancy Diversity Overhead 5 Travis Mick, Reza Tourani, Satyajayant Misra MuNCC: Multi-hop Neighborhood Collaborative Caching

On-path caching has minimal overhead. Content Provider On-path Replica Consumer 6 Travis Mick, Reza Tourani, Satyajayant Misra MuNCC: Multi-hop Neighborhood Collaborative Caching

On-path caching has minimal overhead; low diversity. Content Provider Off-path Replica Consumer 7 Travis Mick, Reza Tourani, Satyajayant Misra MuNCC: Multi-hop Neighborhood Collaborative Caching

On-path caching has minimal overhead; low diversity. Content Provider Off-path New Replica Replica Consumer 7 Travis Mick, Reza Tourani, Satyajayant Misra MuNCC: Multi-hop Neighborhood Collaborative Caching

Attempts have been made to improve on-path diversity. Leave Copy Everywhere (LCE) Leave Copy Down (LCD), Move Copy Down (MCD) 1 . ProbCache 2 , Prob-PD 3 . 1 Laoutaris, N., Che, H., & Stavrakakis, I. (2006). 2 Psaras, I., Chai, W. K., & Pavlou, G. (2012). 8 3 Ioannou, A., & Weber, S. (2014). Travis Mick, Reza Tourani, Satyajayant Misra MuNCC: Multi-hop Neighborhood Collaborative Caching

Hash routing maximizes diversity. Content Provider Designated Replica Consumer 9 Travis Mick, Reza Tourani, Satyajayant Misra MuNCC: Multi-hop Neighborhood Collaborative Caching

Hash routing maximizes diversity; increases path stretch. Designated Content Provider Replica Content Provider Designated Replica Consumer Consumer 9 Travis Mick, Reza Tourani, Satyajayant Misra MuNCC: Multi-hop Neighborhood Collaborative Caching

Variants of hash routing aim to reduce path stretch. Content Designated Content Designated Replica Replica Provider Provider Consumer Consumer Asymmetric Multicast 10 Saino, L., Psaras, I., & Pavlou, G. (2013). Travis Mick, Reza Tourani, Satyajayant Misra MuNCC: Multi-hop Neighborhood Collaborative Caching

The neighborhood can contain diverse content. Distant 3rd-hop 2nd-hop Immediate Content Neighbors Neighbors Neighbors Provider Searching Node Nearby Replica 11 Travis Mick, Reza Tourani, Satyajayant Misra MuNCC: Multi-hop Neighborhood Collaborative Caching

Neighborhood caching in ICN is not well-studied. Caching schemes utilizing Bloom filters. Radius Validation Applicability Tortelli, M., et al. 1-hop None CCN Wang, Y., et al. Arbitrary Theoretical General ICN Lee, M., et al. Arbitrary Simulation IP Wong, W., et al. Arbitrary Emulation CCN 12 Travis Mick, Reza Tourani, Satyajayant Misra MuNCC: Multi-hop Neighborhood Collaborative Caching

Outline Introduction Prior Work Solution Design Experimental Results Conclusion & Future Work 12 Travis Mick, Reza Tourani, Satyajayant Misra MuNCC: Multi-hop Neighborhood Collaborative Caching

MuNCC consists of four major components. Cache State Neighborhood Exchange Forwarding Error Diversity Reduction Enhancement 13 Travis Mick, Reza Tourani, Satyajayant Misra MuNCC: Multi-hop Neighborhood Collaborative Caching

Cached contents are advertised in Bloom filters. Attenuated Bloom Filter Construction: 1 Compress cache state into a Bloom filter. 2 Send Bloom filter to neighbors. 3 Aggregate received Bloom filters. 4 Repeat h times to establish h -hop neighborhood. 14 Travis Mick, Reza Tourani, Satyajayant Misra MuNCC: Multi-hop Neighborhood Collaborative Caching

Cache summary exchange for BF construction: Step 1. B v v B w B x w x B y B z y z 1 Compress cache state into a Bloom filter. 15 Travis Mick, Reza Tourani, Satyajayant Misra MuNCC: Multi-hop Neighborhood Collaborative Caching

Cache summary exchange for BF construction: Step 2. v B v B x B w B v w x B w B z B y B x y z 2 Send Bloom filter to neighbors. 15 Travis Mick, Reza Tourani, Satyajayant Misra MuNCC: Multi-hop Neighborhood Collaborative Caching

Cache summary exchange for BF construction: Step 3. B v ( w , 0) = B w B v ( x , 0) = B x v B w B x ( v , 0) = B v ( v , 0) = B v B w B x ( y , 0) = B y ( z , 0) = B z w x B y B z ( x , 0) = B x ( w , 0) = B w y z 3 Aggregate received Bloom filters. 15 Travis Mick, Reza Tourani, Satyajayant Misra MuNCC: Multi-hop Neighborhood Collaborative Caching

Cache summary exchange for BF construction: Step 4. v B w ∪ B x B v ∪ B z B v ∪ B y B w ∪ B x w x B v ∪ B y B x B w B v ∪ B z y z 4 Repeat. 15 Travis Mick, Reza Tourani, Satyajayant Misra MuNCC: Multi-hop Neighborhood Collaborative Caching

Cache summary exchange for BF construction: Step 4. B v ( w , 1) = B v ∪ B y B v ( x , 1) = B v ∪ B z v B w B x ( v , 1) = B w ∪ B x ( v , 1) = B w ∪ B x B w B x ( y , 1) = B w ( z , 1) = B x w x B y B z ( x , 1) = B v ∪ B z ( w , 1) = B v ∪ B y y z 4 Repeat. 15 Travis Mick, Reza Tourani, Satyajayant Misra MuNCC: Multi-hop Neighborhood Collaborative Caching

Forwarding decisions are made based on attenuated BFs. Implementation: Add distance flag D to interest, with default ∞ . If D � = ∞ , the interest is part of a neighborhood search. If D = ∞ : 1 Try to satisfy from cache. 2 Try to satisfy in neighborhood. 3 Forward toward content provider. If D � = ∞ : 1 Try to satisfy in level D of neighborhood. 2 Return NACK. 16 Travis Mick, Reza Tourani, Satyajayant Misra MuNCC: Multi-hop Neighborhood Collaborative Caching

Neighborhood search procedure: Scenario. Searching Node v False Positive w x True Positive y z v : Wants content i . v : i �∈ local cache 17 Travis Mick, Reza Tourani, Satyajayant Misra MuNCC: Multi-hop Neighborhood Collaborative Caching

Neighborhood search procedure: Level 1. Searching Node v False Positive w x True Positive y z v : i �∈ B v ( w , 0) v : i ∈ B v ( x , 0) 17 Travis Mick, Reza Tourani, Satyajayant Misra MuNCC: Multi-hop Neighborhood Collaborative Caching

Neighborhood search procedure: False positive. v w x y z v : Forward to x with tag D = 0 x : i �∈ local cache x : Return NACK to v 17 Travis Mick, Reza Tourani, Satyajayant Misra MuNCC: Multi-hop Neighborhood Collaborative Caching

Neighborhood search procedure: Level 2. v w x y z v : Process NACK from x v : i ∈ B v ( w , 1) v : Forward to w with tag D = 1 17 Travis Mick, Reza Tourani, Satyajayant Misra MuNCC: Multi-hop Neighborhood Collaborative Caching

Neighborhood search procedure: Next hop, level 1. v w x y z w : i ∈ B w ( y , 0) w : Forward to y with tag D = 0 17 Travis Mick, Reza Tourani, Satyajayant Misra MuNCC: Multi-hop Neighborhood Collaborative Caching

Neighborhood search procedure: Content delivery. v w x y z y : i ∈ local cache y : Deliver content to w w : Deliver content to v 17 Travis Mick, Reza Tourani, Satyajayant Misra MuNCC: Multi-hop Neighborhood Collaborative Caching

Cache churn undermines BF usefulness and is detrimental to performance. Increased False Eviction NACKs Latency Positive 18 Travis Mick, Reza Tourani, Satyajayant Misra MuNCC: Multi-hop Neighborhood Collaborative Caching