CCD: Efficient Customized Content Dissemination in Distributed Publish/Subscribe H. Jafarpour, B. Hore, S. Mehrotra and N. Venkatasubramanian Information Systems Group Dept. of Computer Science UC Irvine 1
Customized content dissemination on distributed Pub/Sub (CCD) Motivation Problem definition and formulation CCD algorithm Heuristic CCD algorithm Experimental evaluation CCD: Efficient Customized Content Dissemination in Distributed Pub/Sub 2 2
Domain: Emergency Notification Systems One or a few generic messages sent to the entire impacted population Under response Goal: Customized Notifications are sent to the population using multiple modalities Over response 3
Motivation Leveraging pub/sub framework for dissemination of rich content formats, e.g., multimedia content. Same content format may not be consumable by all subscribers!!! CCD: Efficient Customized Content Dissemination in Distributed Pub/Sub 4 4
Customized delivery Customize content to the required formats before delivery! Español Español!!! CCD: Efficient Customized Content Dissemination in Distributed Pub/Sub 5 5
Subscriptions in CCD Subscription: • Team: USC How to specify required • Video: Touch Down formats? Context: PC, DSL, AVI Receiving context: Subscription: Receiving device capabilities • Team: USC • Video: Touch Down Display screen, available Context: Phone, 3G, FLV software,… Communication capabilities Subscription: Available bandwidth • Team: USC • Video: Touch Down User profile Context: Laptop, 3G, AVI, Location, language,… Spanish subtitle CCD: Efficient Customized Content Dissemination in Distributed Pub/Sub 6 6
Content customization How is content customization done? Adaptation operators Low resolution and small Original content Transcoder content suitable for Operator mobile clients Size: 28MB Size: 8MB CCD: Efficient Customized Content Dissemination in Distributed Pub/Sub 7 7
Challenges How do we perform content customization in distributed pub/sub infrastructures? CCD: Efficient Customized Content Dissemination in Distributed Pub/Sub 8 8
Challenges Option 1: Perform all the required customizations in the sender broker 28MB 28+12+8 = 48MB 28+12+8 = 48MB 8MB 15MB 8MB 12MB 8MB 12MB 28MB 15MB 28MB 8MB 8MB CCD: Efficient Customized Content Dissemination in Distributed Pub/Sub 9 9
Challenges Option 2: Perform all the required customization in the proxy brokers (leaves) 28MB 28MB 28MB Repeated Operator 8MB 15MB 28MB 8MB 12MB 28MB 15MB 28MB 8MB 8MB CCD: Efficient Customized Content Dissemination in Distributed Pub/Sub 10 10
Challenges Option 3: Perform all the required customization in the broker overlay network 28MB 8MB 15MB 8MB 12MB 28MB 15MB 28MB 8MB 8MB CCD: Efficient Customized Content Dissemination in Distributed Pub/Sub 11 11
Customized content dissemination on distributed Pub/Sub (CCD) Motivation Problem definition and formulation CCD algorithm Heuristic CCD algorithm Experimental evaluation CCD: Efficient Customized Content Dissemination in Distributed Pub/Sub 12 12
DHT-based pub/sub DHT-based routing schema, We use Tapestry [ZHS04] Rendezvous Point CCD: Efficient Customized Content Dissemination in Distributed Pub/Sub 13 13
Dissemination tree For a published content we can estimate the dissemination tree in the broker overlay network Using DHT-based routing properties The dissemination tree is rooted at the corresponding rendezvous broker Rendezvous Point CCD: Efficient Customized Content Dissemination in Distributed Pub/Sub 14 14
Content Adaptation Graph (CAG) All possible content formats in the system All available adaptation operators in the system Size: 28MB Frame size: 1280x720 Frame rate: 30 Size: 15MB Frame size: 704x576 Size: 8MB Frame rate: 30 Frame size: 128x96 Frame rate: 30 Size: 10MB Frame size: 352x288 Frame rate: 30 CCD: Efficient Customized Content Dissemination in Distributed Pub/Sub 15 15
Content Adaptation Graph (CAG) A transmission (communication) cost is associated with each format Sending content in format F i from a broker to another one has the transmission cost of A computation cost is associated with each operator Performing operator O (i,j) on content has the computation cost of F 1 /28 V={F 1 ,F 2 ,F 3 ,F 4 } 60 60 E={O (1,2) ,O (1,3) ,O (1,4) ,O (2,3) ,O (2,4) ,O (3,4) } 60 25 25 F 2 /15 F 3 /12 F 4 /8 25 CCD: Efficient Customized Content Dissemination in Distributed Pub/Sub 16 16
CCD plan A CCD plan for a content is the dissemination tree: Each node (broker) is annotated with the operator(s) that are performed on it Each link is annotated with the format(s) that are transmitted over it {O (1,2) ,O (2,4) } {F 2 } {F 2 } {F 4 } F 1 /28 60 60 60 {} {O (2,3) } {} 25 25 {F 2 } {F 3 } {F 4 } F 2 /15 F 3 /12 F 4 /8 25 {} {} {} CCD: Efficient Customized Content Dissemination in Distributed Pub/Sub 17 17
CCD plan cost Communication cost for a plan, Sum of all costs for the formats transmitted through all edges Computation cost for a plan, Sum of the costs for all operators in all plan nodes Total CCD plan cost CCD: Efficient Customized Content Dissemination in Distributed Pub/Sub 18 18
Problem definition For a given CAG and dissemination tree, , find CCD plan with minimum total cost. CCD: Efficient Customized Content Dissemination in Distributed Pub/Sub 19 19
Customized content dissemination on distributed Pub/Sub (CCD) Motivation Problem definition and formulation CCD algorithm Heuristic CCD algorithm Experimental evaluation CCD: Efficient Customized Content Dissemination in Distributed Pub/Sub 20 20
CCD algorithm Input: A dissemination tree A CAG The initial format Requested formats by each broker Output: The minimum cost CCD plan CCD: Efficient Customized Content Dissemination in Distributed Pub/Sub 21 21
CCD algorithm Based on dynamic programming Annotates the dissemination tree in a bottom-up fashion For each broker: Assume all the optimal sub plans are available for each child Find the optimal plan for the broker accordingly N i …. N k N j CCD: Efficient Customized Content Dissemination in Distributed Pub/Sub 22 22
CCD algorithm F 1 F 1 /28 F 2 F 4 60 60 60 25 25 F 2 /15 F 3 /12 F 4 /8 F 4 F 4 F 1 F 2 F 1 F 3 25 CCD: Efficient Customized Content Dissemination in Distributed Pub/Sub 23 23
CCD algorithm in leaf broker Input: All possible input format sets F 1 /28 Requested formats 60 60 60 Output: 25 25 F 2 /15 F 3 /12 F 4 /8 Optimal plan for each input format set 25 Plan cost: 28+60 = 86 Input format set {F 1 } {F 2 } {F 1 ,F 2 } {F 1 ,F 2 ,F 3 ,F 4 } …. …. {F 1 } {O(1,3)} Requested {F 1 , F 3 } {F 1 , F 3 } {F 1 , F 3 } {F 1 , F 3 } format set CCD: Efficient Customized Content Dissemination in Distributed Pub/Sub 24 24
CCD algorithm in for a non-leaf broker Input: All possible input format sets Optimal sub plan for child nodes for any given input format set Output: • Enumerate all Optimal plan for the given input format set combination of sub plans • Enumerate all possible output format sets {F 1 ,F 2 } N i 2 m sub plans 2 m sub plans …. N k N j Optimal sub plan for Optimal sub plan for input set: {F 2 } input set: {F 1 } Optimal sub plan for Optimal sub plan for input set: {F 1 } input set: {F 2 } CCD: Efficient Customized Content Dissemination in Distributed Pub/Sub 25 25
Complexity of CCD algorithm Algorithm complexity n : number of nodes in the tree k avg : average number of children for a node m : number of formats in the CAG : complexity of minimum conversion cost computation in CAG Exponential in m , CAG size CCD: Efficient Customized Content Dissemination in Distributed Pub/Sub 26 26
Customized content dissemination on distributed Pub/Sub (CCD) Motivation Problem definition and formulation CCD algorithm Heuristic CCD algorithm Experimental evaluation CCD: Efficient Customized Content Dissemination in Distributed Pub/Sub 27 27
CCD Problem is NP-hard Directed Steiner tree problem can be reduced to CCD Given a directed weighted graph G(V,E,w) , a specified root r and a subset of its vertices S, find a tree rooted at r of minimal weight which includes all vertices in S. CCD: Efficient Customized Content Dissemination in Distributed Pub/Sub 28 28
Multilayer graph representation Cartesian product of CAG and dissemination tree {F 1 } F 1 /10 5 7 F 2 /5 F 3 /8 3 4 {F 4 } {F 1 ,F 3 } {F 1 ,F 4 } F 4 /15 CCD: Efficient Customized Content Dissemination in Distributed Pub/Sub 29 29
Source F 1 /10 7 5 F 2 /5 Terminal 3 F 3 /8 4 F 4 /15 CCD: Efficient Customized Content Dissemination in Distributed Pub/Sub 30 30 30
Recommend
More recommend
