TMMAC: An Energy Efficient Multi-Channel MAC Protocol for Ad Hoc Networks

1
TMMAC An Energy Efficient Multi-Channel MAC
Protocol for Ad Hoc Networks
Jingbin Zhang, Gang Zhou, Chengdu Huang, Sang
H. Son, John A. Stankovic Department of
Computer Science, University of
Virginia Department of Computer Science,
University of Illinois
2
Motivation

• TMMAC A TDMA based multi-channel MAC protocol
  using a single half duplex radio transceiver.
• Why Multi-channel?
• Increase the bandwidth
• Most IEEE 802.11 devices can switch channels
  dynamically.
• Why a single radio transceiver?
• Using multiple radio transceivers increases both
  the cost and energy consumption
• Most IEEE 802.11 devices use a single half-duplex
  radio transceiver
• Why TDMA?
• Increase the life time of the mobile devices
• Improve the throughput

3
Contribution

• Novel multi-channel MAC
• Energy efficient 74 less per packet energy
• High throughput 113 higher throughput
• Supporting broadcast efficiently.
• Accurate analytical model.
• Dynamic ATIM window adjustment scheme.

4
Outline

• State of the Art
• TMMAC Design
• Analytical Model
• Dynamic ATIM Window Adjustment
• Performance Evaluation
• Conclusion

5
State of the Art (1)

• Special hardware support
• Multiple radio transceivers Wu et al. 2000
  Raniwala et al. 2005 Adya et al. 2004
• Busy tone Deng et al. 1998
• FHSS Tang et al. 1999 Tyamaloukas et al.
  2000

6
State of the Art (2)

• Single radio transceiver
• Frequency negotiation So et al. 2004 Fitzek
  et al. 2003 Li et al. 2003 Jain et al. 2001
• Random number generators Bahl et al. 2004
• MMAC So et al. 2004
• Time synchronization
• Beacon interval ATIM window Communication
  window
• ATIM window Frequency negotiation
• Communication window Data transmission

802.11 DCF
7
TMMAC Design Overview

• Similar to 802.11 PSM MMAC
• Time synchronization, Beacon interval (ATIM
  window Communication window)
• Different from MMAC
• Communication window is divided into time slots
• Both the frequency and the time are negotiated in
  the ATIM window
• ATIM window is dynamically adjusted

8
TMMAC Design Example (1)

• Assumption Two channels The communication
  window contains 5 time slots

At the start of an ATIM window
Suppose node B has two packets to be sent to node
C in this beacon interval.
Channel Usage Bitmaps (CUBs)
0 0 0 0 0
0 0 0 0 0
Combined CUBs
Channel Allocation Bitmaps (CABs)
0 1 0 0 0
0 0 0 1 0
OR
0 0 0 0 0
0 0 0 0 0
0 0 0 0 0
0 0 0 0 0
0 1 0 0 0
0 0 0 1 0
0 0 0 0 0
0 0 0 0 0
0 0 0 0 0
0 0 0 0 0
0 1 0 0 0
0 0 0 1 0
0 1 0 0 0
0 0 0 1 0
0 1 0 0 0
0 0 0 1 0
0 0 0 0 0
0 0 0 0 0
D
A
B
C
E
Slot 2 Rec. in channel 1
Slot 4 Rec. in channel 2
Slot 2 Send in channel 1
Slot 4 Send in channel 2
ATIM packet
ATIM-ACK packet
ATIM-RES packet
9
TMMAC Design Example (2)
Suppose node E has two packets to be sent to node
D in this beacon interval.
0 1 0 0 0
0 0 0 1 0
0 0 0 0 0
0 1 1 0 0
CABs
OR
0 1 0 0 0
0 0 0 1 0
0 1 0 0 0
0 0 0 1 0
0 1 0 0 0
0 0 0 1 0
0 1 0 0 0
0 0 0 1 0
0 0 0 0 0
0 0 0 0 0
0 1 0 0 0
0 1 1 1 0
0 1 0 0 0
0 1 1 1 0
0 0 0 0 0
0 1 1 0 0
D
A
B
C
E
Slot 2 Rec. in channel 1
Slot 4 Rec. in channel 2
Slot 2 Send in channel 1
Slot 4 Send in channel 2
Slot 2 Send in channel 1
Slot 3 Send in channel 2
Slot 2 Rec. in channel 1
Slot 3 Rec. in channel 2
ATIM packet
ATIM-RES packet
ATIM-ACK packet
10
TMMAC Design Example (3)
Suppose node C has one packets to broadcast to
its neighbors in this beacon interval.
0 0 0 0 0
0 0 0 0 1
CABs
0 1 0 0 0
0 0 0 1 0
0 1 0 0 0
0 0 0 1 0
0 1 0 0 0
0 1 1 1 0
0 1 0 0 0
0 1 1 1 0
0 0 0 0 0
0 1 1 0 0
0 1 0 0 0
0 1 1 1 1
0 1 0 0 0
0 0 0 1 1
1 1 0 0 0
0 1 1 1 1
D
A
B
C
E
Slot 2 Rec. in channel 1
Slot 4 Rec. in channel 2
Slot 2 Send in channel 1
Slot 4 Send in channel 2
Slot 2 Send in channel 1
Slot 3 Send in channel 2
Slot 2 Rec. in channel 1
Slot 4 Rec. in channel 2
Slot 5 Send in channel 2
Slot 2 Send in channel 1
Slot 4 Send in channel 2
Slot 5 Rec. in Channel 2
Slot 1 Rec. in channel 1
Slot 3 Rec. in channel 2
Slot 2 Rec. in channel 1
Slot 3 Rec. in channel 2
Slot 5 Rec. in channel 2
ATIM-BRD packet
11
Analytical Model

• Analyze the saturation throughput of TMMAC in
  wireless LANs.
• Built upon Bianchi 2000, which is used to
  analyze the saturated throughput of 802.11.
• Validated through simulations in GloMoSim.

12
Impact of Time Synchronization Error
2 at maximum
18 to 31
13
Motivation

• There is no fixed optimal ATIM window size when
  the network is saturated.
• A smaller ATIM window is preferred when the
  network is not saturated.
• The dynamic ATIM window scheme used in 802.11 PSM
  is not applicable. Jung et al. 2002

Dynamic ATIM Window Adjustment
14
Rules for Dynamic ATIM Window Adjustment (1)

• A finite set of ATIM window sizes are used
  ATIM1, , ATIMi, ATIMi1, , ATIMm and
  ATIMi1-ATIMilslot
• The default channel is never used for data
  communication in the time slots before ATIMm.
• The ATIM window size for the next beacon interval
  is piggybacked in the ATIM control packets.
• Node A wants to send the packet to node B
• A knows Bs ATIM window size
• A does not know Bs ATIM window size

Dynamic ATIM Window Adjustment
15
Rules for Dynamic ATIM Window Adjustment (2)

• Decide whether the network is saturated.
• If the network is saturated
• If the communication window is fully used ?
  decrease the ATIM window size by one level
• If not ? Increase the ATIM window size by one
  level
• If the network is not saturated, decrease the
  ATIM window size by one level

gt? Saturation threshold
Dynamic ATIM Window Adjustment
16
Simulation Settings
Number of channels 3
Bit rate 2Mbps
Packet size 512 bytes
Channel switch delay 80us
Time synchronization error 0.1ms
Beacon interval 100ms
Network size 1000m by 1000m
Node number 200
Application layer CBR
Routing layer GF
MAC layer TMMAC, MMAC, 802.11
Communication Range 250m
Carrier sense range 500m
Performance Evaluation
17
Evaluation Metrics

• Aggregated Throughput
• Total throughput of all the nodes in the network
• Per packet energy
• The value of total energy consumed by the whole
  network divided by the total number of data
  packets successfully transmitted.

Performance Evaluation
18
Evaluation of Dynamic ATIM Window Adjustment (1)
Traffic pattern
Performance Evaluation
19
Evaluation of Dynamic ATIM Window Adjustment (2)
Traffic pattern
Performance Evaluation
20
Performance vs. System Loads (1)

• Aggregate throughput vs. packet arrival rate

113 more aggregated throughput
Performance Evaluation
21
Performance vs. System Loads (2)

• Per packet energy vs. packet arrival rate

74 less per packet energy
Performance Evaluation
22
Performance vs. System Loads (3)

• Aggregate throughput vs. packet arrival rate (6
  channels)

84 more aggregated throughput
Performance Evaluation
23
Conclusion

• TMMAC exploits the advantage of both multiple
  channels and TDMA in an efficient way.
• TMMAC achieves high communication throughput and
  low energy consumption.
• 113 higher communication throughput
• 74 less per packet energy

24
Publication

• Jingbin Zhang, Gang Zhou, Sang H. Son, John A.
  Stankovic, Kamin Whitehouse, "Performance
  Analysis of Group Based Detection for Sparse
  Wireless Sensor Networks," in Submission.
• Jingbin Zhang, Gang Zhou, Chengdu Huang, Sang H.
  Son, John A. Stankovic, "TMMAC An Energy
  Efficient Multi-Channel MAC Protocol  for Ad Hoc
  Networks," 2007 IEEE International Conference on
  Communications (IEEE ICC'07), Glasgow, Scotland,
  2007.
• Jingbin Zhang, Ting Yan, John A. Stankovic, Sang
  H. Son, "Thunder Towards Practical, Zero Cost
  Acoustic Localization for Outdoor Wireless Sensor
  Networks," ACM SIGMOBILE Mobile Computing and
  Communications Review (ACM MC2R), Special Issue
  on Localization Technologies and Algorithms,
  2007.
• Jingbin Zhang, Ting Yan, Sang H. Son, "Deployment
  Strategies for Differentiated Detection in
  Wireless Sensor Networks," Third Annual IEEE
  International Conference on Sensor Mesh and Ad
  Hoc Communications and Networks (IEEE SECON'06),
  Reston, VA, 2006.
• Shan Lin, Jingbin Zhang, Gang Zhou, Lin Gu, Tian
  He, John A. Stankovic, "ATPC Adaptive
  Transmission Power Control for Wireless Sensor
  Networks," 4th ACM International Conference on
  Embedded Networked Sensor Systems (ACM
  SenSys'06), Boulder, Colorado, 2006.
• Jingbin Zhang, Gang Zhou, Sang H. Son, John A.
  Stankovic, "Ears on the Ground An Acoustic
  Streaming Service in Wireless Sensor Networks,"
  Fifth IEEE/ACM International Conference on
  Information Processing in Sensor Networks
  (IEEE/ACM IPSN'06, Demo Abstract), Nashville, TN,
  2006.
• Arsalan Avatoii, Jingbin Zhang, Sang H. Son,
  "Group-Based Event Detection in Undersea Sensor
  Networks," Second International Workshop on
  Networked Sensing Systems (INSS'05), San Diego,
  California, 2005.

25
Questions?