MAC Layer Protocols for Sensor Networks
Leonardo Leiria Fernandes
Contents
Basic Concepts S-MAC T-MAC B-MAC P-MAC Z-MAC
Basic Concepts
Problem TDMA
CSMA RTS / CTS
Hidden Nodes
MAC Challenges
Traditionally
Fairness Latency Throughput
For Sensor Networks
Power
efficiency Scalability
S-MAC - Sensor MAC
Nodes periodically sleep Trades energy efficiency for lower throughput and higher latency Sleep during other nodes transmissions
Listen
Sleep
Listen
Sleep
S-MAC
Listen significantly longer than clock drift Neighboring nodes exchange SYNC msgs Exchanged timestamps are relative rather than absolute RTS/CTS avoids hidden terminal Message passing provided Packets contain expected duration of message Every packet must be acknowledged Adaptive listening can be used so that potential next hop nodes wake up in time for possible transmissions
S-MAC Results
Latency and throughput are problems, but adaptive listening improves it significantly
S-MAC Results
Energy savings significant compared to non-sleeping protocols
T-MAC - Timeout MAC
Transmit all messages in bursts of variable length and sleep between bursts RTS / CTS / ACK Scheme Synchronization similar to S-MAC
T-MAC Operation
T-MAC Results
T-MAC saves energy compared to S-MAC The early sleeping problem limits the maximum throughput Further testing on real sensors needed
B-MAC - Berkeley MAC
B-MACs Goals:
Low
power operation Effective collision avoidance Simple implementation (small code) Efficient at both low and high data rates Reconfigurable by upper layers Tolerant to changes on the network Scalable to large number of nodes
B-MACs Features
Clear Channel Assessment (CCA) Low Power Listening (LPL) using preamble sampling Hidden terminal and multi-packet mechanisms not provided, should be implemented, if needed, by higher layers
Sender
Receiver
Sleep
Sleep
Preamble
Sleep
Message
t
Receive t
B-MAC Interface
CCA on/off Acknowledgements on/off Initial and congestion backoff in a per packet basis Configurable check interval and preamble length
B-MAC Lifetime Model
E E rx E tx E listen E d E sleep E rx t rx c rxbV E tx t tx c txbV E listen E sample E d t d c dataV E sleep t sleepc sleepV 1 ti
E can be calculated if hardware constants, sample rate, number of neighboring nodes and check time/preamble are known Better: E can be minimized by varying check time/preamble if constants, sample rate and neighboring nodes are known
B-MAC Results
Performs better than the other studied protocols in most cases System model can be complicated for application and routing protocol developers Protocol widely used because has good results even with default parameters
P-MAC - Pattern MAC
Patterns are 0*1 strings with size 1-N Every node starts with 1 as pattern Number of 0s grow exponentially up to a threshold and then linearly up to N-1 TR = CW + RTS + CTS + DATA + ACK N = tradeoff between latency and energy
Patterns vs Schedules
Local Pattern Bit 1 1 Packet to Send 1 1 Receiver Pattern Bit 1 0 Local Schedule 1 1-
1
0 0 0
0
1 1 0
*
1 0 *
11 0 0
P-MAC Evaluation
Simulated results are better than SMAC Good for relatively stable traffic conditions Adaptation to changes on traffic might be slow Loose time synchronization required Needs more testing and comparison with other protocols besides S-MAC
Z-MAC - Zebra MAC
Runs on top of B-MAC Combines TDMA and CSMA features
TDMA Pros
CSMA Pros
Simple Scalable
Naturally avoids collisions
Cons
Cons
Collisions due to hidden terminals RTS/CTS is overhead
Complexity of scheduling Synchronization needed
Z-MAC Initialization
Neighborhood discovery through ping messages containing known neighbors Two-hop neighborhood used as input for a scheduling algorithm (DRAND) Running time and message complexity of DRAND is O(), where is the two-hop neighborhood size The idea is to compensate the initialization energy consumption during the protocol normal operation
Z-MAC Time Slot Assignment
a1 a
Fi 2 1
a
l2 si ( for: l 0,1,2...)
Z-MAC Transmission Control
The Transmission Rule: If owner of slot
Take
a random backoff within To Run CCA and, if channel is clear, transmit
Else
Wait
for To Take a random backoff within [To,Tno] Run CCA and, if channel is clear, transmit
Z-MAC HCL Mode
Nodes can be in High Contention Level (HCL) A node is in HCL only if it recently received an Explicit Contention Notification (ECN) from a two-hop neighbor Nodes in HCL are not allowed to contend for the channel on their two-hop neighbors time slots A node decides to send an ECN if it is losing too many messages (application ACKs) or based on noise measured through CCA
Z-MAC Receiving Schedule
B-MAC based Time slots should be large enough for contention, CCA and one B-MAC packet transmission Slot size choice, like in B-MAC, left to application
Z-MAC Results
Z-MAC performs better than B-MAC when load is high As expected, fairness increases with Z-MAC Complexity of the protocol can be a problem
Conclusions
Between the protocols studied, B-MAC still seems to be the best one for applications in general Application developers seem not to use B-MACs control interface Middleware service could make such optimizations according to network status
Thank You
Questions or comments? Thank you for coming!
