A Fair Multiple-Slot Assignment Protocol for TDMA Scheduling in Wireless Sensor Networks K. Banerjee, P. Basuchaudhuri, D. Sadhukhan and N. Das.

A Fair Multiple-Slot Assignment Protocol for TDMA Scheduling in Wireless Sensor Networks

K. Banerjee, P. Basuchaudhuri, D. Sadhukhan and N. Das

WOMS 2008

WSN : Wireless Sensor Networks

Organization

Collision Avoidance : TDMA

Scheduling : Frame length minimization problem

Distributed Protocol

Performance Evaluation

Conclusion

WOMS 2008

What is Sensor Network?

B

E

A

D

C

F

G

• A collection of sensor nodes• Engaged in data transmission, reception, aggregation and redirecting to a sink• An ad-hoc network

SINK

WOMS 2008

Major Applications

• Environmental Monitoring• Habitat Monitoring • Precision Agriculture• Disaster Recovery• Natural Calamity Prediction• Defense Applications• Assisted Living for aged & disabled• Health Care

WOMS 2008

Large number of nodes

Multi-hop network

Streaming data

No global knowledge about the network

Frequent node failure

Energy is the scarce resource

Limited memory

Autonomous

Unique Constraints

WOMS 2008

Need to shutdown the radio if possible

SENSORS

Power consumption of node subsystems

0

5

10

15

20

Po

wer

(m

W)

CPU TX RX IDLE SLEEP

RADIOSLEEPIDLERXTX EEEE

Energy Consumers

WOMS 2008

Communication uses a single channel over the same wireless medium

Communication in sensor network

.

A node broadcasts data packets and nodes within its transmission zone can receive those packets

Interference takes place when more than one transmission overlaps : Collision

Primary interference occurs due to exposed terminals

Primary Interference

YX

Z

Secondary interference occurs due to hidden terminals

Secondary Interference

YX

Z

W

Several collision avoidance methods are available while accessing the media-

• CSMA : listening also consumes energy

• FDMA : not suitable; generally single channel

• TDMA : best suited; nodes can sleep in idle times

Collision Avoidance

Collision causes retransmission : wastage of energy

Energy is the most scarce resource

WOMS 2008

Slot: Smallest time slice, in which a node can either transmit or receive

Frame: A minimal sequence of slots is a frame

A matrix is used to represent a schedule

Slot, Frame and Schedule

Node→

R--R-5-R---4--R-R3R--R-2--R-R154321Slot

TDMA

Time is slotted : each node is assigned at least one collision-free slot in a frame; frames are repeated

TDMA: Periodic listen and sleep

–Turn off radio when sleeping–Reduce duty cycle to ~10% (e.g. 200 ms on/2s off)–Increased latency for reduced energy

How to reduce the latency?

sleeplisten listen sleeptransmit

Nodes within 1 hop neighborhood creates primary interference

Nodes within 2 hop neighborhood (but not in 1 hop neighborhood) creates secondary interference

So no two nodes within 2 hop neighborhood can be given same time slot for transmission

Slots can be reused for nodes at more than 2-hop distance

Unique Slots

Problem Definition

How to find a TDMA schedule with minimum frame length that assigns at least one conflict-free slot to

each node?

Can be modeled as a graph-coloring Problem

NP-Complete Problem [Ephremides et al, 1990]

Distributed solution is needed

WOMS 2008

The Problem :

1

2

3

4

WOMS 2008

• WSN consists of N static nodes

• Each node is assigned a unique id i, 1< i<N

• No global knowledge about network topology: each node knows N, the total number of nodes in the network • A node can only be in one state at a time: broadcasting or receiving

• All the links are bi-directional

Assumptions Revisited

The easiest way to solve the problem is providing each node a particular time slot.But that leads to -

1. Frame length = Number of nodes.2. Wastage of time slots.

Assigning Slots

WOMS 2008

Ephremides & Truong (IEEE Tr. Comm., 1990)

Table 1

RR

1 5 4 3 2

Node→

-

R---5-R---4--R-3

--R2--R154321Slot

R

R

-

WOMS 2008

Improvements over Previous Works

Fairness – Even distribution of reserved slots

Compaction – Reduction of number of slots in the schedule matrix, wherever possible

WOMS 2008

Table 2: The initial-schedule-matrixNode→

Step – I : Initial-Schedule-Matrix

R---5-R---4--R-3

--R2--R154321Slot

1 5 4 3 2

WOMS 2008

Table 3: The contention matrix Node→

Step – II : Contention Matrix

Contention (Ci,j) = total number of 2-hop neighbors of nodei to which the slot Sj is available

XXX1X5XXXXX4XXXX132XXX22XX22X154321Slot

1 5 4 3 2

WOMS 2008

Table 4: The complete-schedule-matrix after Fair-ReservationNode→

Step – III : Complete-Schedule-Matrix

R--R-5-R---4--R-R3R--R-2--R-R154321Slot

X

X

Parallel Execution

WOMS 2008

Table 5: The compact-schedule-matrix Node→

Step – IV : Compact-Schedule-Matrix

-R---3R--R-2--R-R154321Slot

WOMS 2008

Simulation Environment

Random graph generation

• Graph generation algorithms have been used

• Number of nodes may vary from 50-250

• Randomly generated each time in Unix Environment

WOMS 2008

Performance Evaluation : Frame Length

Comparison based on frame length (L)

WOMS 2008

Comparison based on standard deviation of number of slots assigned to individual nodes

Performance Evaluation : Fairness

WOMS 2008

Tr = avg. # of slots reserved per node / frame length

Comparison based on transmission rates (Tr)

Performance Evaluation : Throughput

WOMS 2008

Conclusion

Proposed algorithm outperforms in terms of:

frame lengthfairness and throughput

Efficient for large networks with uniform trafficDistributed algorithm for compaction is to be

studied

References

1. A. Ephremides and T. V. Truong, “Scheduling Broadcasts in Multihop Radio Networks,” IEEE Transactions on Communications, Vol. 38, No. 4, April 1990, pp: 483-495.

2. I. F. Akyildiz, W. Su, Y. Sankarasubramaniam and E. Cayirci, “A Survey on Sensor Networks,” IEEE Communications Magazine, August 2002, pp: 102-114.

3. S. Ramanathan and E. L. Lloyd, “Scheduling Algorithms for Multihop Radio Networks,” IEEE/ACM Transactions on Networking, Vol. 1, No. 2, April 1993, pp: 166-177.

4. S. Ramanathan, “A Unified Framework and Algorithm for Channel Assignment in Wireless Sensor Networks,” Wireless Networks, Vol. 5, No. 2, 1999, pp: 81-94.

5. I. Rhee, A. Warrier, J. Min and L. Xu, “DRAND: Distributed Randomized TDMA Scheduling for Wireless Ad-hoc Networks,” Proc. of MobiHoc ’06, May 2006, pp: 190-201.

6. Y. Wang and I. Henning, “A Deterministic Distributed TDMA Scheduling Algorithm for Wireless Sensor Networks,” Proc. of International Conference on Wireless Communication, Networking and Mobile Computing, WiCOM 2007, pp: 2759-2762

7. S.  Gandham, M.  Dawande and R. Prakash, “Link scheduling in sensor networks: distributed edge coloring revisited,” Proc. of 24th Annual Joint Conference of the IEEE Computer and Communications Societies, INFOCOM 2005, pp: 2492- 2501.

8. S. Bhattacharjee and N. Das, “Distributed Time Slot Assignment in Wireless Ad Hoc Networks for STDMA,” Lecture Notes in Computer Science (Springer), No. 3618, Proc. of the 2nd International Conference on Distributed Computing and Internet Technology (ICDCIT 2005), Dec. 2005, pp. 93-104.

WOMS 2008