Improving the Slotted CSMA/CA MAC for Time-Critical Events in WSN

In this paper, four scenarios with different settings of the slotted CSMA/CA parameters are defined. Each scenario is evaluated for FIFO and Priority Queuing. It also analyze the impact of the hidden-node problem and the solution to mitigate hidden-node problem.

The slotted CSMA/CA algorithm is based on a basic time unit called Backoff Period (BP) which is equal to aUnitBackoffPeriod = 80 bits (0,32ms). The slotted CSMA/CA backoff algorithm based on three variables : 1) Backoff Exponent (BE) enables the computation of the backoff delay, 2) Contention Window (CW) represent the number of BPs during which channel must be sensed idle before channel access, 3) the Number of Backoff (NB) denotes the number of times the CSMA/CA was required to backoff while attempting to access the channel.


Service Differentiation Strategies for Slotted CSMA/CA :
Slotted CSMA/CA affected by four initialization parameters : 1) the minimum backoff exponent (macMinBE), 2) the maximum backoff exponent (aMaxBE), 3) the initial value of the CW (CWinit) and 4) the maximum number of backoffs (macMaxCSMABackoffs).
in this paper, their service differentiation mechanisms are the macMinBE, aMaxBE and CWinit parameters.

IEEE 802.15.4 define two frame types : 1) data traffic which represents sensory data broadcasted to the network (without using acknowledgments), 2) command traffic which comprises critical messages (alarm reports, PAN management messages and GTS requests) sent by sensor nodes to PAN coordinator.
In this paper, they consider command frames as the high priority service class and data frames as the low priority class.

ffeb27_1.jpg

Instead of having the same CSMA/CA parameters for both traffic types, they assign each class its own attributes. They denote [macMinBEhp, aMaxBEhp] and CWhp for high priority traffic and [macMinBElp, aMaxBElp] and CWlp for low priority traffic.

Evaluation :
using OPNET simulator.
performance of data frames is analyzed in terms of average delay and probability of success. the probability of success is measured by the throughput of data frames divided by the offered load of data frames generated by MAC layer.

  • Fully connected network (no hidden-node problem) : Success probability of command frames without hidden nodes, Average delay of command frames (ms), Command traffic sent by the MAC layer, Success probability of data frames, Average delay of Data Frames.
    ffeb27_2.jpg
  • Partially connected network (hidden-node problem) : Succes probability of command/data frames, Average delay of command frames, Average delay of data frames, command traffic sent by MAC layer.

    ffeb27_3.jpg

they have run the same simulation scenarios using NS-2. they said that the values of the average delays observed in NS-2 results are greater than those obtained with OPNET. Also, NS-2 produce lower throughput than those obtained used OPNET.

Reference :
Improving the IEEE 802.15.4 Slotted CSMA/CA MAC for Time-Critical Events in Wireless Sensor Networks
Anis KOUBAA et all – IPP-HURRAY

Note : This resume is created for self-learning only. Author and Publisher hold copyrights

Paper Predator Jutsu [Mode On] – Searching Research Idea

February 27 2008,
Taipei City
High Speed Network Lab

Udin Harun

Comments are closed.

%d bloggers like this: