Wireless Sensor Networks : Technology, Protocols, and Applications
Wireless Sensor Networks : Technology, Protocols, and Applications
Wireless Sensor Networks : Technology, Protocols, and Applications
Create successful ePaper yourself
Turn your PDF publications into a flip-book with our unique Google optimized e-Paper software.
146 MEDIUM ACCESS CONTROL PROTOCOLS FOR WIRELESS SENSOR NETWORKS<br />
of state transitions between message arrival <strong>and</strong> message transmission. Therefore,<br />
deterministic MAC schemes guarantee an upper bound for the access time. Determinism<br />
is a crucial requirement in a real-time environment, where the correctness<br />
of the application depends on the adherence of its underlying tasks to their specified<br />
execution deadline.<br />
Throughput Throughput is typically defined as the rate at which messages are<br />
serviced by a communication system. It is usually measured either in messages<br />
per second or bits per second. In wireless environments it represents the fraction<br />
of the channel capacity used for data transmission. Throughput increases as the<br />
load on the communication system increases initially. After the load reaches a certain<br />
threshold, the throughput ceases to increase, <strong>and</strong> in some cases, it may start to<br />
decrease. An important objective of a MAC protocol is to maximize the channel<br />
throughput while minimizing message delay.<br />
Robustness Robustness, defined as a combination of reliability, availability, <strong>and</strong><br />
dependability requirements, reflects the degree of the protocol insensitivity to errors<br />
<strong>and</strong> misinformation. Robustness is a multidimensional activity that must simultaneously<br />
address issues such as error confinement, error detection <strong>and</strong> masking,<br />
reconfiguration, <strong>and</strong> restart. Achieving robustness in a time-varying network such<br />
as a WSN is difficult, as it depends strongly on the failure models of both the links<br />
<strong>and</strong> the communicating nodes.<br />
Scalability Scalability refers to the ability of a communications system to meet its<br />
performance characteristics regardless of the size of the network or the number of<br />
competing nodes. In WSNs, the number of sensor nodes may be very large, exceeding<br />
thous<strong>and</strong>s <strong>and</strong> in some cases millions of nodes. In these networks, scalability<br />
becomes a critical factor. Achieving scalability is challenging, especially in timevarying<br />
environments such as wireless networks. A common approach to achieve<br />
scalability is to avoid relying on globally consistent network states. Another<br />
approach is to localize interactions among the communicating nodes, through the<br />
development of hierarchical structures <strong>and</strong> information aggregation strategies.<br />
Grouping sensor nodes into clusters, for example, allows the design of shared medium<br />
access protocols which are highly scalable. Similarly, aggregating information<br />
from different sensors allows the development of traffic patterns which can be<br />
exploited efficiently to scale the MAC protocol to a large number of sensor nodes.<br />
Stability Stability refers to the ability of a communications system to h<strong>and</strong>le fluctuations<br />
of the traffic load over sustained periods of time. A stable MAC protocol,<br />
for example, must be able to h<strong>and</strong>le instantaneous loads which exceed the maximum<br />
sustained load as long as the long-term load offered does not exceed the maximum<br />
capacity of the channel. Typically, the scalability of a MAC protocol is<br />
studied with respect to either delay or throughput. A MAC protocol is considered<br />
to be stable, with respect to delay, if the message waiting time is bounded. These<br />
systems can be characterized by a bounded backlog of messages in the transmission