Data Collection through Device- to-Device Communications for Mobile - PowerPoint PPT Presentation

Data Collection through Device- to-Device Communications for Mobile Big Data Sensing Hanshang Li, Ting Li, Xinghua Shi and Yu Wang 
 College of Computing and Informatics University of North Carolina at Charlotte May 17 , 2016 @ The First Workshop of Mission-Critical Big Data Analytics (MCBDA 2016) 2

OUTLINE ➤ Introduction ➤ Mobile Data Collection ➤ Relay Selection Problem ➤ Our Solutions ➤ Simulations ➤ Conclusions 3

OUTLINE ➤ Introduction ➤ Mobile Data Collection ➤ Relay Selection Problem ➤ Our Solutions ➤ Simulations ➤ Conclusions 4

MOBILE DEVICES ➤ Nowadays, more and more smart mobile devices are utilized by humans as the primary personal devices, which have the functions of computing, sensing, communicating and so on. 5

MOBILE DEVICES AND USERS An Introduction to Mobile Marketing: The Past, Present, 
 Cisco VNI Global Mobile Data Traffic Forecast, and Future, Marketo, 2015 2015 - 2020, Cisco, 2016

MOBILE DATA EXPLOSION ➤ Mobile data tra ffi c grows! grew 74% in 2015, reached 3.7 exabytes/month, 4,000 times of the one in 2005 will surpass 30.6 exabytes per month in 2020 ➤ Mainly came from smart devices though smart devices 
 only represent 36% of 
 devices/connections, 
 they account for 89% 
 of all mobile tra ffi cs Source: Cisco VNI Mobile, 2016 Cisco VNI Global Mobile Data Traffic Forecast, 2015 - 2020, Cisco, 2016

MOBILE CROWD SENSING — “POWER OF THE CROWD” ➤ Individuals with sensing and computing devices collectively share data and extract information to measure and map phenomena of common interests ➤ Widely used in many applications - human as sensors 8

ADVANTAGES OF MOBILE CROWD SENSING ➤ Leverages existing sensing and communication 
 infrastructures with less additional costs ; ➤ Provides unprecedented spatial-temporal coverage, 
 especially for observing unpredictable events ; ➤ Integrates human intelligence into the sensing 
 and data processing. 9

GENERAL FRAMEWORK OF MOBILE CROWD SENSING ➤ A large number of Reward mobile participants ➤ A set of crowd sensing tasks Sensing Data Sensing Tasks ➤ Participant selection Tasks User Traces mechanism - the focus Participants of most current works Coverage Task Assignment Cost Incentive Selection Mechanism 10

GENERAL FRAMEWORK OF MOBILE CROWD SENSING ➤ A large number of Reward mobile participants ➤ A set of crowd sensing tasks Sensing Data Sensing Tasks ➤ Participant selection Tasks User Traces mechanism - the focus Participants of most current works Coverage Task Assignment Cost Incentive Selection Mechanism 10

CHALLENGE TO CURRENT NETWORK INFRASTRUCTURE ➤ Current cellular network do not have enough capacity to support all of the fast growing mobile big data from smart devices and mobile sensing

OUTLINE ➤ Introduction ➤ Mobile Data Collection ➤ Relay Selection Problem ➤ Our Solutions ➤ Simulations ➤ Conclusions 12

DATA COLLECTION IN MOBILE CROWD SENSING ➤ How to transfer sensing data back? Rewards cellular network (piggyback) WiFi or femtocell o ffl oading Sensing Data D2D/DTN relays 
 Sensing Tasks Tasks User Traces Participants Coverage Task Assignment Cost Incentive Selection Mechanism D2D: Device-to-Device 
 DTN: Delay Tolerant Networks

DATA COLLECTION IN MOBILE CROWD SENSING ➤ How to transfer sensing data back? Rewards cellular network (piggyback) WiFi or femtocell o ffl oading Sensing Data D2D/DTN relays 
 Sensing Tasks Tasks + low cost and easy to deploy User Traces Participants Coverage Task Assignment Cost Incentive Selection Mechanism D2D: Device-to-Device 
 DTN: Delay Tolerant Networks

DATA COLLECTION IN MOBILE CROWD SENSING ➤ How to transfer sensing data back? Rewards cellular network (piggyback) WiFi or femtocell o ffl oading Sensing Data D2D/DTN relays 
 Sensing Tasks Tasks + low cost and easy to deploy User Traces Participants - longer delay and low deliver ratio Coverage Task Assignment Cost Incentive Selection Mechanism D2D: Device-to-Device 
 DTN: Delay Tolerant Networks


 MOBILE DATA COLLECTION VIA D2D RELAYS ➤ Leverage user mobility to delivery the sensing data from the source to the sink(s) 
 14

RELATED WORKS ➤ Data Collection in Mobile Sensing Wang et al. [UbiComp 2013] consider Bluetooth/Wifi o ffl oading (one-hop) to reduce energy consumption and data cost of data-plan users Karaliopoulos et al. [InfoCom 2015] consider a joint user recruitment with D2D data collection (multi-hop), however, the time complexity of proposed greedy algorithm is large due to search over all space-time paths ➤ DTN/D2D Routing Focus on point to point delivery over D2D relays, selecting relay node on ride ➤ Data O ffl oading WiFi [Lee et al. 2010, Dimatteo et al. 2011], FemtoCell [Chandrasekhar et al. 2008] D2D [Han et al. 2012, Li et al. 2014, Zhu et al., 2013], broadcasting or point-to-point

OUTLINE ➤ Introduction ➤ Mobile Data Collection ➤ Relay Selection Problem ➤ Our Solutions ➤ Simulations ➤ Conclusions 16

MODEL AND ASSUMPTIONS ➤ n mobile users, User= u 1 ,u 2 , …, u n ➤ m locations, Location= l 1 ,l 2 , …, l m ➤ T , time period for delivery ➤ Known probability p(i,j,t) , mobile user u i visits 
 location l j at time t (learn from historical data) ➤ T wo devices can transfer sensing data if they are 
 visiting the same location within a particular time slot ➤ C ollection task : sending the data from a source 
 node s to a sink node d (a mobile device or a location) ➤ Restricted flooding (Epidemic routing) is used within selected relay nodes U(s,d)

RELAY SELECTION PROBLEM ➤ Goal: minimize the number relay nodes U(s,d) while maximize the data delivery ➤ T wo versions of the optimization problem Minimum Relay Problem K Relay Problem

TWO CHALLENGES ➤ How to model the time-evolving D2D network and estimate the delivery probability? weighted space-time graph and reliability calculation ➤ How to identify a small set of relay nodes from a huge candidate pool to guarantee certain level of data delivery? greedy algorithm

OUTLINE ➤ Introduction ➤ Mobile Data Collection ➤ Relay Selection Problem ➤ Our Solutions ➤ Simulations ➤ Conclusions 20

SPACE-TIME GRAPH ➤ Space-time graph describes all characteristics among the selected relay nodes in both spacial and temporal spaces 1 5 0 5 u u 0 s= u =s 0 0 0 1 5 u u 0 u 1 1 1 1 5 u u 0 u 2 2 2 1 5 u u 0 u 3 3 3 0 5 1 5 u u 0 =d u d= 4 4 4 t=1 t=2 t=3 t=4 t=5 21


 
 
 
 
 DELIVERY PROBABILITY OVER SPACE-TIME GRAPH ➤ Each spacial link has a delivery probability 1 5 0 5 u u 0 s= u =s 0 0 0 m 1 5 − − − − → − − − − → 0 u u u 1 1 1 u t − 1 Y u t − 1 u t u t k ) = (1 − (1 − p ( j, i, t ) p ( k, i, t ))) · r ( p ( k ) , j j 1 5 u u 0 u 2 2 2 i =1 1 5 0 u u u 3 3 3 Q ➤ With flooding, the delivery probability can be calculated 
 0 5 1 5 u u 0 =d u d= 4 4 4 via the following dynamic programming 
 t=1 t=2 t=3 t=4 t=5 delivery probability based on the ws p ( U ( s, d ) , s, d ) = p G ( s 0 , d T ) Thus, 22


 
 
 
 
 DELIVERY PROBABILITY OVER SPACE-TIME GRAPH ➤ Each spacial link has a delivery probability 1 5 0 5 u u 0 s= u =s 0 0 0 m 1 5 − − − − → − − − − → 0 u u u 1 1 1 u t − 1 Y u t − 1 u t u t k ) = (1 − (1 − p ( j, i, t ) p ( k, i, t ))) · r ( p ( k ) , j j 1 5 u u 0 u 2 2 2 i =1 1 5 0 u u u 3 3 3 Q ➤ With flooding, the delivery probability can be calculated 
 0 5 1 5 u u 0 =d u d= 4 4 4 via the following dynamic programming 
 t=1 t=2 t=3 t=4 t=5 delivery probability based on the ws p ( U ( s, d ) , s, d ) = p G ( s 0 , d T ) Thus, 22

RELAY SELECTION ALGORITHM ➤ Greedy Algorithm Algorithm 1 Relay Selection Algorithm Input: potential user set User , call probability p ( i, j, t ) for each user in User , the source s and the sink d . in each step, greedily selects the user u 
 Output: selected relay nodes U ( s, d ) . which leads to maximal improvement 
 1: U ( s, d ) = ∅ 2: while G U ( s,d ) is connected do of p(U(s, d), s, d) into U(s, d) Choose the most active user and add it into U ( s, d ) 3: 4: while | U ( s, d ) | < K or p ( U ( s, d ) , s, d ) < γ (for K relay ➤ Cold Start Problem problem or minimum relay problem , respectively) do for all u i ∈ User and / ∈ U ( s, d ) do 5: Calculate the improvement of p ( U ( s, d ) , s, d ) by 6: initially, the space-time is not connected 
 adding u i in to U ( s, d ) Select the user u i with the largest reliability improve- 1, 7: at all, and adding a single user cannot 
 ment and add it into U ( s, d ) 8: return U ( s, d ) solve this arding the solution: simply pick the most active user