Resource Management in IoT-
Clustering
Wireless Sensor Network (WSN)
Wireless Sensor Networks
features
• No wired infrastructure
• Homogeneous devices
• Stationary nodes
• Dispersed Network
• Large Network size
• Self-organized
• All nodes acts as routers
• Potential multihop routes
2
Routing protocol
• Route the packets(from source to sink) along the optimal
path
Routing protocols
• Single hop - a packet is directly sent from the source to
target
• Multi hop - a packet is sent to the target through several
nodes.
– Multi hop protocols can be divided into flat and hierarchical
• Flat - all nodes have the same responsibility (excessive data
redundancy and faster energy consumption)
• Hierarchical - creates clusters inside the network.
Routing protocol(Contd..)
Clustering
• Hierarchical - the entire network is divided into several clusters.
Each cluster consists of some source nodes and a cluster head
Sensor nodes
• It referred as source nodes, can gather information from the
monitoring region and send the sensing information to their
corresponding cluster head.
Cluster head
• It is elected from all the sensor nodes in a cluster according to
some criteria, and is responsible for collecting sensing data from
source nods.
• After receiving data from source nodes, the cluster head performs
data aggregation to reduce the data size before sending data to the
sink, which further reduces the power expended for data transfer
Clustering
Cluster member
Clusterhead
Gateway node
Intra-Cluster link
Cross-cluster link
• Nodes divided in virtual group according to some rules
• Nodes belonging in a group can execute different functions
from other nodes.
5
With/Without Clustering
Contd…
• Grouping of similar objects or sensors in our context
distance or proximity
• Clustering is a method of organising wireless network
topologies, in which a few nodes, the cluster heads
(CH), are elected as representatives to route the
traffic originated in the entire network.
Clustering in WSN
• Involves grouping nodes into clusters and electing a CH
– Members of a cluster can communicate with their CH directly
– CH can forward the aggregated data to the central base station
through other CHs
• Clustering goals
– Allows aggregation - reducing number of nodes taking part in
transmission
– Limits data transmission
– Facilitate the reusability of the resources
– CHs and gateway nodes can form a virtual backbone for
intercluster routing
– Cluster structure gives the impression of a smaller and more stable
network
– Improve network lifetime
• Reduce network traffic and the contention for the channel
• Data aggregation and updates take place in CHs
8
Goals of clustering
• Less energy consumption - Data transmission is done by the CHs in
clustering routing scheme by which saving of energy consumption
• Data Aggregation - to eliminate the redundant transmission and provide
aggregated data to the BS
• Scalability - Responsibilities of CHs are aggregation of data, data
processing, dissemination of information and management of network. At
each sensor nodes the amount of size of the routing table stored can be
reduced by localizing the route set up with the help of clustering topology.
• Network lifetime maximization - Network lifetime is the term which
cannot be neglected in WSNs because sensor nodes are impel in supply of
power, bandwidth of transmission.
• Robustness
Cluster Head(CH) Election
• Main objective of clustering is to extend the object
network lifetime by electing a representative network
member which collects all the communication within
the network and forwards it to the outside.
• CH election mechanism that an Internet of Things
device, representing an object, would take as part of a
clustered network of devices.
• As an important and powerful feature, CHs are elected
according to their residual energy
CH Election(Contd..)
• Let 𝑇𝐶𝐻 be the duration that a node will have, once the
CH role has been elected.
• 𝑇𝐶𝐻 could be calculated as a function of the node’s
residual energy:
𝑇𝐶𝐻 = C x Residual Energy, where C is a constant
• In each CH, a node would calculate its own 𝑇𝐶𝐻
• Send its 𝑇𝐶𝐻 to the rest of the cluster members
• Factors that would elect a new CH could vary, straightforward
decision would select the proposal with the highest 𝑇𝐶𝐻 .
• This selection method minimise the number of CH election
procedures and also conserve more energy over time.
Contd…
A CH election procedure would start when any of the following situations
occur:
• 𝑇𝐶𝐻 expires
• The current CH cannot communicate with any infrastructure gateway
• A CH, whom did not participate in the previous election and has more
residual energy than the current CH, receives an advertisement packet
from a gateway
• A CH cannot communicate with the current CH before 𝑇𝐶𝐻 expires
• A new object is added to the CH’s network.
Contd…
Undesirable Situations,
• if a network loses its CH and no CH receives an
advertisement packet from a gateway, the election
process would not start again to choose a new CH.
To avoid this problem,
• When the connection with the information
infrastructure is re-established, networks with a CH
selected in this way would start a regular CH election
procedure again.
Direct v. Minimum Transmission
• The amount of energy
used in figure (a) can
be modeled by this
formula:
– e k(3d + d )2
amp 1 2
• Whereas the amount
of energy used in
figure (b) uses this
formula:
– e k(3d 2 + d 2)
amp 1 2
The Amount of Energy Depletion
• This is the formula for the amount of
energy depletion by data transfer:
Cluster-Based Routing Protocol
• In clustered based network nodes can be partitioned into
a number of small groups called clusters. Each cluster has
a coordinator referred to as cluster head (CH) and a
number of sensor nodes (SNs). Clustering results in a
two-tier hierarchy in which CHs form the higher tier
while SNs form the lower tier.
• Cluster members sense the data from the physical
environment and send it to their respective CHs. CHs fuse
the data and transmit it to the sink either directly or in
multi-hop manner.
Cluster-Based Routing Protocol
• LEACH-Low-Energy Adaptive Clustering Hierarchy for WSN
• HEED: A hybrid energy efficient distributed clustering approach for
ad-hoc sensor networks
• MRECA: Mobility resistant efficient clustering approach for ad-hoc
sensor networks
• EEDC- Energy efficient Dynamic clustering and energy efficient
routing technique for WSN
LEACH
• Low Energy Adaptive Clustering Hierarchy (LEACH) selects the sensor
nodes as CHs randomly.
• LEACH is a clustering algorithm in which selection of cluster head is
random and its implementation is cyclical. Every rounds of LEACH consists
of two phases such as establishment of cluster and stable data
communication.
• In establishment stage,
– The cluster head selected randomly in cyclic way.
– Every node generates a random number between 0 to 1.If the number is less
than the threshold value T(n), then the node becomes the cluster head.
– After that it broadcast the message, if any node wants to join the nearest
cluster, it totally depends on the strength of the received signal.
• In data communication stage,
– The cluster member transmits the sensing data directly to the cluster head.
– Then the cluster head aggregates the data received form the cluster nodes
and send it to the base station.
Contd…
Advantages
• The nodes in LEACH dies randomly and the dynamic clustering process increases
the network lifetime.
• As LEACH is distributed, it does not require any global knowledge of network.
• It does not require location information to form the clusters.
• It follows single hop routing from node to cluster head so that it saves energy.
Disadvantages
• It assumes that nodes have always data to send and initially nodes along with
Cluster head have same initial energy.
• LEACH is not applicable to the networks deployed in the large region as Cluster
head sends the data to the base station in single hop manner.
• Cluster heads are randomly selected rotationally.
• It consists of different rounds while each round every nodes are taking part for
reconstructing the cluster heads, so it takes large energy.
• Cluster head in the network is not uniformly distributed.
HEED
• Hybrid Energy-Efficient Distributed clustering (HEED) is
a multi-hop clustering algorithm.
• It gets an energy-efficient clustering routing with
unambiguous concern of energy.
• HEED is different from LEACH because it does not
selects CH randomly.
• The way of construction of cluster is achieved on the
fusion combination of two factors, residual energy and
communication cost of intra-cluster. In HEED, chosen
CH have comparatively high average residual energy
(evenly distributed CHs all over the networks)
Contd…
Advantages of HEED:
1) It contains fully distributed clustering method by using two
parameters for the CH selection.
2) Provides CH distribution and load balancing through the network.
3) It provides multi-hop communication between CHs and BS to
support more energy conservation and scalability difference from
single-hop.
Disadvantages of HEED:
1) HEED undergoes a consequent overhead meanwhile it needs some
iteration to form cluster.
2) The CHs which are near to the sink node dies earlier because of it
has extra work load.
3) HEED causes visible energy dissipation which causes to decrease
the lifetime of network.