Image Processing Reference
In-Depth Information
The PEGASIS protocol is extended in [Lin], where not only energy efficiency but also delay is
targeted. he authors investigate data gathering schemes that balance the energy and delay costs, as
measured by the energy
delay metric. Two different solutions are proposed, a chain-based binary
scheme for sensor networks with CDMA nodes and a chain-based three-level scheme for sensor
networks with non-CDMA nodes.
The two approaches share the same basic idea, i.e., achieving multiple parallel communications by
different nodes. In order to improve energy and delay performance by simultaneous transmissions,
interferences between different communications should be minimized. The first solution proposed
to achieve multiple simultaneous communication with low interference is to use CDMA, so CDMA-
capable sensor nodes are used. Each pair of nodes can use a distinct code to minimize interference
withothertransmissionsandsoamaximumdegreeofparallelismcanbeimplemented.hismeans
that in an
N
-node WSN, after the chain has been built as in PEGASIS,
N
/ nodes can transmit at the
same time with a different code (e.g., all the nodes in odd positions can simultaneously transmit to
those in even positions). Then, at the second level of the hierarchy, only nodes that were receivers
at the first level are considered (i.e., the nodes in even positions). So
N
∗
nodes will be senders (i.e., once again the ones in even positions in the second-level set). Then the
set of receiving nodes will be considered as the third-level set, and so on, until only one-node set
remains. This is the leader node that has to perform direct transmission toward the sink. While in
PEGASIS delay is a linear function of the number of WSN nodes
N
,usingthisapproachdelayisa
logarithmic function.
However, CDMA may not be applicable for all sensor networks. he authors therefore investigate a
second approach to achieve a minimal energy
/
ofthissetcontaining
N
/
delay with non-CDMA nodes. In this case, in order to
minimize interferences, only distant nodes should transmit at the same time. A three-level hierarchy
for data gathering is therefore proposed, which allows simultaneous transmissions that are far apart
so as to minimize interference while achieving a reasonable delay cost.
Simultaneous data transmissions are carefully scheduled taking the position of every node into
consideration, so that at each level only the leaders of the previous level participate. he transmission
schedulecanbecalculatedonceatthebeginning,sothatallnodesknowwheretosenddataineach
communication round.
This approach is only based on experimental considerations and is not as efficient as the CDMA-
based approach. However, it performs much better than LEACH and PEGASIS in terms of the
energy
∗
delay metric. The energy model adopted here is the first-order radio model, as in LEACH
and in PEGASIS, so idle (and sleep) power consumption is not taken into account. Even these exten-
sions of the PEGASIS protocol do not directly decrease the duty cycle of nodes. Hence these results
may change if the physical and MAC layers feature high idle power consumption.
∗
7.5.2.7 Base Station-Controlled Dynamic Clustering (BCDPC) Protocol
The BCDPC presented in [Mur] is a centralized cluster-based protocol that tries to combine the
best features of LEACH and PEGASIS. BCDPC entrusts a high-energy BS with a large amount of
energy supply with the task of setting up clusters and routing paths, performing randomized rotation
of cluster heads and other energy-intensive tasks.
The key points in BCDCP are
•
Formation of balanced clusters, where each cluster head serves an approximately equal
number of member nodes to avoid cluster-head overload
•
Uniform placement of cluster heads throughout the whole WSN, and adoption of cluster-
head-to-cluster-head (CH-to-CH) routing to transfer the data to the BS
Like the PEGASIS protocol, BCDPC enables multi-hop forwarding of aggregated data toward a
leader node, which directly transmits to the BS. However, here the forwarding is only CH-to-CH,
through predefined paths that are computed and communicated by the BS.
