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

• rks and Data

c Storage Data Centric Storage istance (SBD) Routing Protocol

• loring 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

• use management and supply chain monitoring

ry applications: surveillance and recognition

• rage

work diate pro-active reporting

ng Mechanism

l Query Routing ry Uni-casting vs. Flooding

Data Centric Storage (DCS)

rnal Storage (ES) [1] l Storage (LS) [2] Centric Storage (DCS) [3]

Base Station D D D D D D D D D D D D D D Base Station/ Gateway / Query Node lf estination of nown R D D D Q R

Disk Based Data Centric Storage

MAPPING SID = {Ti×n + Sj}

Tm T1 T0 S0 S1 Sn 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15

(1) (14)

(b) (c) (a)

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 ase

• des 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

Sector Based Distance Routing Protocol (SBD)

[ ] [ ] [ ]

} , | ) ( ) {(

1

j i M i t C i

ID S k

ψ ψ ψ ≠ == Δ × + ×

− −



1 ) ( + Ψ sizeof

Grid Coloring Algorithm

do = (j-1) × n to (j × n)-1 do if i < n × j then Assign slot C0 to SHi end if if i +1 < n × j then Assign slot C1 to SHi+1 end if if i + n < m × n then Assign slot C2 to SHi+n end if if i+n+1 < m × n then Assign slot C3 to SHi+n+1 end if i = i + HD d for = + HD

C0 C1 C2 C3 24 25 26 27 28 29 18 23 22 21 20 19 17 16 15 14 13 12 11 10 9 8 7 6 5 4 3 2 1

Member-SH Association Stage

mplemented in SH nodes which is triggered at 1.0 msec

• se. The SH node broadcasts a beacon packet (see Fig.

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

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

• t to each member node.

t, sectorMembers, slotArray e id from each incoming packet of type 4 //and insert into mber node id and assign it to MID is destined for this node i.e. SHID == SELF_NET_ADDR sh MID to sectorMembers = 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)

[ ] [ ]

, | ) ( ) {(

1

i M i t C i t

ID S k i

ψ ψ ψ ≠ == Δ × + × =

− −



) 1 ) ( ( + Ψ Δ = sizeof t 

Relaying Phase

t_Hop(SHi), 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 ve toward right */

• pCol = nextHopCol + 1

stcol ve toward left */

• pCol = nextHopCol - 1

mn so move toward up or down stCol rRow < destRow /*Move vertically up*/ nextHopRow = nextHopRow + 1 else if curRow> destRow /*Move vertically down*/ nextHopRow = nextHopRow – 1 end if r number*/ * n + nextHopCol

(1) (14)

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

Castalia-3.2’s Existing Structure 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} wirelessChannel Castalia-3.2’s Current Extended Structu

Member-SH association stage

BD Beacon Frame (Packet Type = 3)

Source = SH ID Destination ID = -1 Source = Member Node ID SH ID

Head Join Packet (Packet Type = 4)

• urce
• r ID

Destination ID = -1 Size of Slot Array Slot Array

Scheduling Packet (Packet Type =5)

Packet Type=1 Source Member ID Destination Sector ID A S

SBD Local Data Update (Packet Type = 1)

Packet Type=2 Source Sector ID Destination Sector ID Previous Hop ID

SBD Remote Data Packet (Packet Type = 2)

Simulation Parameters Setup

Parameter Setting F i e l d S i z e 60x60 m2, 90x90 m2, 120x120 m2, 150x150 m2 N u m b e r

• f

N

• d

e s ( n ) 8 0 ( 3 6 0 0 m 2 ) , 1 8 0 ( 8 1 0 0 m 2 ) , 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

• 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

• d

e / 2 2 5 m

2

R a d i

• R

a n g e ( N S H ) ~ 8 m R a d i

• R

a n g e ( S H ) ~ 2 m T r a n s m i s s i

• n

P

• 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 . 2 m J

• u

l e Data Rate, Modulation Type, Bits Per Symbol, Bandwidth, Noise Bandwidth, Noise Floor, Sensitivity 2 5 0 K b p s , P S K , 4 , 20 MHz, 194 MHz, -100 dBm, -95 dBm p a t h L

• s

s E x p

• 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 . m Gaussian Zero-Mean Random Variable (Xα) 4 .

rformance Comparison of SBD with TMAC and SMAC

Performance Comparison of SBD with TMAC and SMAC

(a) (b)

Energy Consumption

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

SBD Throughput Performance

ghput (Number of Application Packets Received Successfully) b) Number of Packets (includes RTS, CTS and

(a) (b)

Property Rate

• de’s Sampling Rate

10 / s by Member Nodes to Head Node 5/s te by Head Nodes 1/s 0.5/s SMAC

Production and Consumption Rate

SBD Querying Performance

a) Latency (sec) b) Total RTS sent (a) (b)

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

• , 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 !!