Wireless Sensor and Actor Network (WSAN) derived from Wireless Sensor Network
(WSN) refers to a group of sensors and actors linked by wireless medium to perform
distributed sensing and actuation tasks. In the network, sensors gather information
about the physical world, while actors coordinate and make decisions to perform
appropriate actions upon the environment, which allows remote, automated interaction
with the environment. The WSAN has wide applications in both civil and military
fields such as microclimate control in buildings, home automation and environment
monitoring, biological and chemical attack detection and battlefield surveillance
(Akyildiz et al., 2000; Akyildiz
and Kasimoglu, 2004).
In WSAN, the energy efficiency of network communication is also crucial, since
sensors are resource-constrained nodes with a limited battery lifetime. The
energy-aware routing protocol should take advantage of actor nodes and use their
resources when possible.
In this study, EACBR protocol is proposed that enables low packet delay and energy consumption, which constructs the Shortest Path Tree (SPT) among actors and sensors by making use of sufficient energy capabilities, larger memory, better processing and communication capabilities of the actors.
In WSN, many studies on routing protocol are mostly focused on solutions that
try to prolong the lifetime of the network. A protocol architecture for sensor
network named Low-Energy Adaptive Clustering Hierarchy (LEACH) is developed
and analyzed that combines the ideas of energy-efficient cluster-based routing
and media access together with application-specific data aggregation to achieve
good performance in terms of system lifetime, latency and application-perceived
quality (Heinzelman et al., 2002). Power-efficient
gathering in sensor information systems (PEGASIS) is an improvement of the LEACH
protocol (Lindsey and Raghavendra, 2002). Rather than
forming multiple clusters, PEGASIS forms chains from sensors so that each node
transmits and receives from a neighbor and only one node is selected from that
chain to transmit to sink. The LEACH has a drawback that the cluster is not
evenly distributed due to its randomized rotation of local cluster-head. For
the drawback, MECH (Maximum Energy Cluster Head) routing protocol that has self-configuration
and hierarchical tree routing properties is presented by Ruay-Shiung
and Chia-Jou (2006). The MECH constructs clusters based on radio range and
the number of cluster members.
Shah and Rabaey (2002) proposed to use a set of sub-optimal
paths occasionally to increase the lifetime of the network. These paths are
chosen by means of a probability function, which depends on the energy consumption
of each path. To save energy, maximum Energy routing protocol based on strong
head (MESH) is introduced by Lim et al. (2007).
The collected data from cluster head is transmitted to sink by node defined
as a strong head. Tan and Ibrahim (2003) proposed PEDAP
(Power Efficient Data gathering and Aggregation Protocol), which was a near
optimal minimum spanning tree based routing scheme, where one of them was the
power-aware version of the other. Boukerche et al.
(2005) proposed EDA (Energy-Aware Data-Centric Routing algorithm for wireless
sensor networks), which represented an efficient energy-aware distributed protocol
to build a rooted broadcast tree with many leaves. All the leaf nodes are turned
off and all the non-leaf nodes are in charge of data aggregation and relaying
tasks. A cluster-based cooperative Multiple Input Multiple Output (MIMO) scheme
is provided by Yuan et al. (2006). The MIMO reduces
the adverse impacts caused by radio irregularity and fading in multi-hop wireless
Due to the coexistence of sensors and actors, there exists distinct difference
between WSAN and traditional WSN such as node heterogeneity, deployment, coordination.
So many protocols proposed for WSN cannot be well suited for the unique features
and application requirements of WSAN. In WSAN, Haidong et
al. (2006) proposed a novel three-level coordination model. In different
levels, sensor-sensor, sensor-actor, actor-actor coordination mechanisms are
addressed, respectively. But, the realization mechanism is not further discussed.
To analyze and solve the coordination and communication problems in WSAN, a
sensor-actor coordination model is proposed based on an event-driven partitioning
paradigm (Melodia et al., 2007). Sensors are
partitioned into different clusters and each cluster is constituted by a data-delivery
tree associated with a different actor. The optimal solution for the partitioning
strategy is determined by mathematical programming and a distributed solution
Here, we present the basic features of EACBR protocol, referring its analysis and simulation.
Routing protocol based on shortest path tree: Figure 1a,
is an undirected graph, while the Fig. 1b is the Shortest
Path Tree (SPT) that is from node N1 to other nodes using Dijkstra
algorithm. Figure 1c is the Minimum Spanning Tree (MST) using
Prim algorithm, which is suitable for dense graph. Since, WSN can be regarded
as an undirected dense graph, the routing tree of the network is constructed
adopting SPT and MST. By this way, sensor data is transmitted to root node by
the routing tree. As shown in Fig. 1, the edge number from
left node and intermediate node to root node in MST is more than the one in
||(a-c) SPT and MST using N1 as the root
In other words, the routing tree rooted at actor node is constructed using
SPT, which can reduce the hop count that the data is transmitted from sensors
to actors. The transmitting time reduce to some extent. That is to say, the
scheme adopting SPT can reduce delay.
EACBR protocol: Similar to Leach, EACBR protocol configures clusters in every round. The process of EACBR is divided into two phases, which are cluster set-up phase and data transmission phase. The process of network operation is divided into rounds and each round consists of the two phases.
Cluster set-up phase: The operations in the cluster set-up phase are detailed in algorithm 1. Firstly, EACBR protocol is based on the assumption that sensors and actors are already deployed. Sensors and actors are divided into some clusters and every cluster including an actor and some sensors is considered as a different subnet. The cluster set-up phase is as follows: Actor Ai as the cluster head of subnets sends broadcast information in lines 2-4. Sensor Ni receives the information that all actors send because actors are equipped with better communication capabilities and greater transmitting power (line 5). Then sensor Ni is added into the subnet of the actor Aj with the shortest distance from node Ni (lines 6-13). The sensor Ni sends ACK information to the actor Aj with single-hop transmission (lines 14-16).
Algorithm 1: Cluster set-up phase
Data transmission phase: Network nodes are divided into different subnets including an actor and some sensors after cluster set-up phase. Every actor as a cluster head may determine all nodes of its subnet. The SPT from sensors to actor in subnet is calculated by Dijkstra algorithm with the product of the maximum consumption energy of the two nodes in link and the consumption required to send data package as a weight. Then actor sends the shortest path tree structure to all sensors in the subnet. Every sensor can transmit data package along the path of the shortest path tree in its subnet.
The operation of EACBR protocol is divided into rounds. In each round, SPTs are configured and data package is transmitted from sensors to the cluster-head. Repeat the process. The subnet lifetime is expired if the residual energy of a sensor in the subnet is exhausted.
The operations in the data transmission phase are detailed in algorithm 2.
Algorithm 2: Data transmission phase
ANALYSIS AND SIMULATION
Here, the analysis and simulation of EACBR protocol are presented.
Preparatory work: According to the study Heinzelman
et al. (2002), we assume a simple model for the radio hardware energy
dissipation where the transmitter dissipates energy to run the radio electronics
and the power amplifier and the receiver dissipates energy to run the radio
To receive k bits, the sensor expends:
To send k bits a distance d, the sensor expends:
The electronics energy, Eelec, depends on factors such as the digital coding, modulation, filtering and spreading of the signal, whereas the amplifier energy, εfsd2 or εempd4, depends on the distance to the receiver and the acceptable bit-error rate.
Energy consumption analysis and simulation: In some hierarchical protocols
such as LEACH, randomized rotation of the cluster head positions is used to
decrease energy consumption and prolong network lifetime. But in EACBR protocol,
actors equipped with longer battery life and better communication capabilities
are fixed as the cluster heads in each round. The method (Melodia
et al., 2007) defined how sensors communicate with actors is proposed,
which the objective is to minimize the overall energy consumption by integer
linear programming. However, the EACBR protocol is aimed to improve the energy
consumption balance of sensors and therefore can prolong network lifetime.
In EACBR protocol, network nodes including sensors and actors are divided into some clusters using location information that every cluster is composed of an actor as cluster-head and some sensors. Every cluster is considered as a different subnet. The shortest path tree from sensors as resource to actor as destination in every subnet is calculated by Dijkstra algorithm with the product of the maximum consumption energy of the two nodes in link and the consumption required to send data package as a weight. The network energy consumption is very low that sensors send data package among the SPT in its subnet.
It is assumed that sensor nodes are randomly deployed in a square area of 400x400 mm. There are four actors which the coordinates (x, y) are (400/3, 400/3), (800/3, 400/3), (400/3, 800/3) and (800/3, 800/3), respectively. The clusters and SPTs are shown in Fig. 2 which the dot denotes sensor and the triangle denotes actor.
In the simulation, the energy and packet parameters of network are shown in Table 1, respectively.
||The residual energy when first sensor dies
||The values of energy and packet parameters of network
We consider two different simulation scenarios. The SPTs are constant in each round in scenario 1 named C-EACBR, while they are dynamically changing in scenario 2 named EACBR. In EACBR protocol, the lifetime of each subnet is the time of first sensor dies. Figure 3 shows a comparison of the residual energy in a certain subnet when first sensor dies between C-EACBR and EACBR. Simulation results show that the energy consumption balance of the network is improved by EACBR compared with C-EACBR.
||The lifetime of four subnets
In the simulation, there are four actors which form four subnets. The lifetime of the four subnets in C-EACBR and EACBR is shown in Fig. 4. The simulation results verify that EACBR can significantly prolong the lifetime by 70% compared with C-EACBR.
In this study, an energy-aware cluster-based routing protocol named EACBR protocol is presented. Network nodes including sensors and actors are divided into some clusters using location information and every cluster is composed of an actor as cluster-head and some sensors. Every cluster is considered as a different subnet. The shortest path tree from sensors as resource to actor as destination in every subnet is calculated by Dijkstra algorithm. EACBR protocol can save energy and reduce delay that sensors send data package among the SPT in its subnet.
This project is sponsored by the National Natural Science Foundation of China under Grant No. 60773190.