| dc.description.abstract | Wireless sensor network (WSN) is envisioned for a wide range of applications ranging
from environmental surveillance, inventory tracking, health monitoring, home au-
tomation to networking in or around a human body. Much of the development in
recent years has focused on new sensor node hardware, integration of sensing and
radio circuitry as well as design of suitable networking protocols to meet the require-
ments of low cost and low power operation. Despite above advances in both sensor
hardware and development of suitable sensor networking protocols, lack of a suitable
WSN standard and associated commercial product has slowed the maturation pro-
cess of this technology. The situation is expected to change with the release of the
IEEE 802.15.4 Wireless MAC and PHY speci¯cations for low-rate, low-power wireless
personal area networks (WPANs) due to signi¯cant interest from companies that are
already beginning to ship products based on this standard.
Performance analysis of wireless sensor networks in terms of data transmission
capacity and lifetime of the networks are critical research issues towards the design of
optimal deployment strategies of the sensor networks. In contrast to other types of
wireless networks, nodes in wireless sensor networks are densely deployed and organize
themselves in an ad hoc fashion. Due to limited transmission ranges, any two nodes
cannot reach each other directly and rely on other sensor nodes to relay data between
them. Hence, the data packets between the source and destination are routed through
multi-hops. Besides, sensor nodes are normally small in size and have constrained
energy sources. Hence energy consumption analysis is an important performance
measure in wireless sensor networks. Since, IEEE 802.15.4 standard de¯nes medium access control protocol for the devices using low data rate, low power and short-range
transmissions, modeling and performance analysis of its MAC for the wireless sensor
network is worth to study.
In this thesis, the contention free and contention based channel access mechanism
in wireless sensor networks is discussed. Considering di®erent MAC mechanisms in
wireless sensor networks, mathematical models for the energy consumption issues are
analyzed. A hybrid channel access mechanism is proposed for the wireless sensor net-
work that considers the channel access procedure of IEEE 802.15.4 and combines the
binary exponential backo® mechanism of IEEE 802.11. Taking the backlogged nodes
due to collision after successful channel assessment, extended linear feedback model
and discrete time Markov chain model are designed to analyze the successful and
failure probabilities of the system model of the wireless sensor network. Besides, en-
ergy consumption model for the star topology of wireless sensor network is developed
based on the hybrid channel access mechanism.
Besides, analytical models for the beacon-enabled slotted CSMA-CA mechanism
of IEEE 802.15.4 is designed for the star topology of wireless sensor network taking
acknowledged transmissions. The current mechanism of IEEE 802.15.4 CSMA-CA
is extended to include the retransmission limit of the nodes with packet collision
probability. A three-dimensional Markov chain model for the uplink tra±c of IEEE
802.15.4 wireless sensor network is developed to analyze the energy consumption and
throughput of the nodes under unsaturated tra±c conditions. Extensive performance
analysis are made to analyze energy consumption and throughput of the system mod-
els and to study the impact of di®erent network and tra±c parameters such as the
packet arrival rate, packet size, node numbers and data rates. The e®ect of binary
exponential contention window on energy consumption of the nodes is veri¯ed for the
wireless sensor networks. The comparison with comprehensive simulations and vali-
dations of our analytical models shows that the proposed analytical frameworks are
totally correct and provides accurate performance predictions for energy consumption
and throughput analysis.
| en_US |