Optimized TDMA Based Distance Routing for Data Centric Storage - PowerPoint PPT Presentation

Optimized TDMA Based Distance Routing for Data Centric Storage Khandakar Ahmed (khandakar.ahmed@rmit.edu.au) Dr. Mark A. Gregory (mark.gregory@rmit.edu.au) School of Electrical and Computer Engineering RMIT University

Outline orks and Data c Storage Data Centric Storage istance (SBD) Routing Protocol oloring Algorithm r-Sector Head Association Stage Sector Slot Assignment Stage g Phase Result Castalia after implementation of SBD esult Analysis

Sensor Networks and Data Network nment and health monitoring ement of critical industrial areas ouse management and supply chain monitoring ry applications: surveillance and recognition orage work diate pro-active reporting ng Mechanism l Query Routing ry Uni-casting vs. Flooding

Data Centric Storage (DCS) D D rnal Storage (ES) [1] D D D D D D l Storage (LS) [2] D D Base Station D D D D Centric Storage (DCS) [3] R D Base Station/ Gateway / Query Node R D D lf estination of Q nown

Disk Based Data Centric Storage S 0 S 1 S n T m (14) 12 13 14 15 8 9 10 11 MAPPING 4 5 6 7 T 1 1 2 3 0 T 0 (1) S ID = { T i × n + S j } (a) (c) (b)

Sector Based Distance Routing Protocol (SBD) ase MA slot assignment among sectors using Grid Coloring Algorithm (GCA) mber-SH association stage, and BEACON Frame broadcasted by Sector Head node er node stores head node id in descending order based on RSSI er node sends joining request to the head node with highest RSSI MA sector slot assignment stage for member nodes managed by the SH creates Member node array ( Ψ ) d Ψ to child nodes mber nodes calculate its time slot based on its index position in the array [ ] [ ] [ ] {( i  ) ( C t ) | i M , i j } × + × Δ ψ == ψ ≠ ψ k 1 S ID − − 1 + sizeof ( ) Ψ ase odes send aged data received from application layer to SH node during their slot of s after collecting all data packets from member nodes send the aged data to the corre

Grid Coloring Algorithm do = ( j- 1) × n to ( j × n)-1 do if i < n × j then Assign slot C 0 to SH i 24 25 27 26 28 29 end if if i + 1 < n × j then 18 19 20 21 22 23 Assign slot C 1 to SH i+1 12 14 16 13 15 17 end if if i + n < m × n then 7 9 6 8 10 11 Assign slot C 2 to SH i+n 1 3 5 0 2 4 end if if i + n +1 < m × n then C 0 C 1 C 2 C 3 Assign slot C 3 to SH i+n+1 end if i = i + HD d for = + HD

Member-SH Association Stage Head_Selection() , implemented in member nodes, mplemented in SH nodes which is triggered at 1.0 msec se. The SH node broadcasts a beacon packet (see Fig. moment member nodes receive a beacon packet from member nodes in the vicinity will receive this beacon by the function fromMacLayer(). After sorting the so based on their received signal strength Indicator Head_Selection() unicasts a sector joining packet (Se twork layer packet, beaconCtrlPacket SH with highest RSSI. Packet Type and Source Input: beaconCtrlPacket, RSSI, SHInfo Type = 3 in beaconCtrlPacket 1: //Extracts source sector head id denote //received signal strength (RSS) from each ET_ADDR as source and -1 (broadcast) as //that a member node receives in beaconCtrlPacket 2. for each beacon frame it received do this packet 3. Extract SH S , RSSI from beaconCtrlPa 4. //Insert SH S to SHInfo list along with r(beaconCtrlPacket, BROADCAST_NET_ADD) 5: push SH S and RSSI into SHInfo 6: end for 7: Sort SHInfo in descending order based on RSSI 8: Create network layer packet joinCntrlPacket 9: SH D = Pop top element from SHInfo.SH S 10: Set SELF_NET_ADDR as source, SH D as Packet Type = 4 to joinCntrlPacket 11: //Unicast joining request to the closest head n 12: toMacLayer (joinCntrlPacket, SH D )

TDMA sector slot assignment stage SS() , implemented in SH nodes. By this algorithm, a SH will mber nodes, who send a join request and assign a conflict ot to each member node. t , sectorMembers, slotArray e id from each incoming packet of type 4 //and insert into [ ] [ ]  t {( i ) ( C t ) | i M , i = × + × Δ ψ == ψ ≠ ψ mber node id and assign it to M ID i k 1 S ID − − is destined for this node i.e. SH ID == SELF_NET_ADDR t Δ  = sh M ID to sectorMembers ( sizeof ( ) 1 ) Ψ + = size of sectorMembers rk layer packet TDMACntrlPacket ype = 5, SELF_NET_ADDR as source and -1 (broadcast) TDMACntrlPacket e i of sectorMembers do Assign sectorMembers[i] to slotArray[i] ySize and slotArray to TDMACntrlPacket DMACntrlPacket ( DMACntrlPacket, BROADCAST_NET_ADD )

Relaying Phase t_Hop( SH i ) , implemented at each sector head node. Node , where, m - number of tracks (rows) and n - number of sectors per track (columns) nd column position of destination //sector head and current head in the grid ; ; = ( SELF_NET_ADDR) % n w = ( SELF_NET_ADDR) /n t to the same column where //destination sector lies (14) ve toward right */ opCol = nextHopCol + 1 stcol ve toward left */ opCol = nextHopCol - 1 mn so move toward up or down stCol rRow < destRow (1) /*Move vertically up*/ nextHopRow = nextHopRow + 1 else if curRow> destRow /*Move vertically down*/ nextHopRow = nextHopRow – 1 end if r number*/ * n + nextHopCol

Castalia – A Simulator for Wireless Sensor Netw behind Choosing Castalia [4] [5] Open Source Available WSN Simulators were Falling Short of the Current State of the Art Modelling Done in Sensor Networks Models Remain Simplistic or Unsuitable for Short Range Low Power Communications Built on Top of OMNET++ and Hence Focus is on the Model and Overall Desig and not on the Event-Driven Simulation Engine. Realistic Node Behaviour Has Been Captured Beyond The Channel Open, Expandable and Reliable

Structure of Castalia s Do Not Connect to Each Other but Through ss Channel Module (s) The Modules and Their Connection q Castalia Offers Support for Building Se Protocols by Defining Appropriate Abstra q Existing Modules are also Highly Tune Different Parameters The Node Composite Module

SBD (V-1) in Castalia SN node[*] Application {BridgeTest, ConnectivityMa ValuePropagation, ValueReporting, D Communication MAC {BaselineBANMAC, Byp TMAC, TunableMAC} Radio Routing {ByPassRouting, Mul SBDRouting} MobilityManager {LineMobilityManage ResourceManager SensorManager physicalProcess[*] {CarsPhysicalProcess, Cu CustomPhysicalProcess} Castalia-3.2’s Existing Structure wirelessChannel Castalia-3.2’s Current Extended Structu

Member-SH association stage Packet Source Destination A Source = SH ID Destination ID = -1 Type=1 Member ID Sector ID S BD Beacon Frame (Packet Type = 3) SBD Local Data Update (Packet Type = 1) Source = Member Node ID SH ID Packet Source Destination Previous Head Join Packet (Packet Type = 4) Type=2 Sector ID Sector ID Hop ID SBD Remote Data Packet (Packet Type = 2) ource Destination Size of Slot Array Slot Array or ID ID = -1 Scheduling Packet (Packet Type =5)

Simulation Parameters Setup Parameter Setting F i e l d S i z e 60x60 m 2 , 90x90 m 2 , 120x120 m 2 , 150x150 m 2 8 0 ( 3 6 0 0 m 2 ) , 1 8 0 ( 8 1 0 0 m 2 ) , N u m b e r o f N o d e s ( n ) 3 2 0 ( 1 4 4 0 0 m 2 ) , 5 0 0 ( 2 2 5 0 0 m 2 ) M e m b e r N o d e D e n s i t y ( f m ) 1 n o d e / 5 6 . 2 5 m 2 S e c t o r H e a d N o d e ( S H ) D e n s i t y ( f S H ) 1 n o d e / 2 2 5 m 2 R a d i o R a n g e ( N S H ) ~ 8 m R a d i o R a n g e ( S H ) ~ 2 0 m T r a n s m i s s i o n P o w e r 0 dBm ( SH ), -5 dBm ( Member Node ) Power Consumption in Sending and Receiving Messages 57.42 mW ( SH ), 46.2 mW ( Member Node ) P o w e r C o n s u m p t i o n P e r S e n s i n g 0 . 0 2 m J o u l e Data Rate, Modulation Type, Bits Per Symbol, 2 5 0 K b p s , P S K , 4 , Bandwidth, Noise Bandwidth, Noise Floor, Sensitivity 20 MHz, 194 MHz, -100 dBm, -95 dBm p a t h L o s s E x p o n e n t 2 . 4 I n i t i a l A v e r a g e P a t h L o s s ( P L ( d 0 ) ) 5 5 R e f e r e n c e D i s t a n c e ( d 0 ) 1 . 0 m Gaussian Zero-Mean Random Variable ( X α ) 4 . 0

rformance Comparison of SBD with TMAC and SMAC (b) (a) Performance Comparison of SBD with TMAC and SMAC

Energy Consumption (a) (b) verage Energy Consumption Per Node (Joule) b) Total Number of Hops (Number of hops in storage and query routing

SBD Throughput Performance Production and Consumption Rate Property Rate ode ’ s Sampling Rate 10 / s by Member Nodes to Head Node 5/s te by Head Nodes 1/s 0.5/s SMAC (b) (a) ghput (Number of Application Packets Received Successfully) b) Number of Packets (includes RTS, CTS and

SBD Querying Performance (b) (a) a) Latency (sec) b) Total RTS sent

References and E.-P. Lim, "In-Network processing of nearest neighbor queries for wireless sensor networks," presented at t ional conference on Database Systems for Advanced Applications, Singapore, 2006 o, and T. L. Porta, "Data Dissemination with Ring-Based Index for Wireless Sensor Networks," IEEE Transactio 6, pp. 832-847, 2007. M. A. Gregory, "Techniques and Challenges of Data Centric Storage Scheme in Wireless Sensor Network," Journ rks, vol. 1, pp. 59-85, 2012. mulator for Wireless Sensor Network [Online]. Available: http://castalia.npc.nicta.com.au ia: A simulator for wireless sensor networks and body area networks: Version 3.2: User’s manual [Online]. Availa .com.au

Thank You !!