Reliability of data in energy-constrained sensor networks is limited due to low powered transmissions. Improvement in transmission reliability through channel-state based scheduling is proposed. In a cluster, a time-shared channel between sensor nodes to the cluster head is considered in a Rayleigh fading environment. A 2-state Markov model is presented to partition the received envelope into fade and non-fade states. The state-partitioning threshold is derived as a function of bit error probability and average duration of transmission, in addition to other channel and transmission parameters. It is shown that the channel-state based scheduling can guarantee bit error probability, arbitrarily close to zero at the cost of increasing idle channel time. Increasing the number of sensor nodes in the cluster can offset the increase in idle channel time.
展开▼
