Introduction: Wireless Sensor Networks (WSNs) consisting of spatially distributed devices (motes) using sensor nodes to cooperatively monitor

physical or environmental conditions at different locations. Devices in a WSN are resource constrained; they have low processing speed, storage capacity, and communication bandwidth. In most settings, the network must operate for long periods of time, but the nodes are battery powered, so the available energy resources limit their overall operation. To minimize energy consumption, most of the device components, including the radio, should be switched off most of the time. Another important characteristic is that sensor nodes have significant processing capability in the ensemble, but not individually. Nodes have to organize themselves, administering and managing the network all together, and it is much harder than controlling individual devices. Since the sensor nodes has been deployed in order to sense the physical environment for a long duration of time without any human intervention, so it is desired to effectively utilize the scares resources (such as battery power). The sensor nodes are deployed in large density in a given area so if every individual node send it sensed data to the base station then it creates congestion and redundancy of the correlated data at the base station which in turn leads to wastage of resources, bandwidth, large control over heads and increased processing time. WSNs are designed to gather information about the state of physical world and transmit sensed data to interested users. In most cases, they are unable to effect on the physical environment. In many applications, however, only observing the state of the physical system is not sufficient; it is also expected to respond to the sensed events/data by performing corresponding actions upon the system. This need for actuation heralds the emergence of wireless sensor/actuator networks (WSANs), a substantial extension of sensor networks that feature coexistence of sensors and actuators. WSANs enable the application systems to sense, interact, and change the physical world parameters. Sensors gather information about the state of physical world and transmit the collected data to actuators through single-hop or multi-hop communications. In this paper we propose a new communication and coordination frame work for effective

communication between sensors, between sensor and actuator and between actuator nodes. Related work: Since the energy of the sensor node is a scares resource so it needs to be utilized effectively. Sensor nodes collect data from the environment and send it to the base station for further analyses. But since sensor nodes deployed in the coverage area is very large in number so if the entire sensor send their individual readings it will leads to wastage of resources. Sometime it will leads to congestion in the network and resulting in packets loss and to compensate for packet loss sensor nodes have to retransmit their data which will shorten the network life time. So in order to effectively utilize the scares resources the sensor nodes has to communicate and coordinate with each other in order to schedule their duty cycle of operation i.e. only one sensor node wake up at a time and send the aggregated data to the base station while all other sensor nodes in a region are put in to sleep mode of operation. This process is called clustering in which only sensor node is selected as cluster head and the other nodes aggregate their data in the cluster head and goes in sleep mode. The cluster architecture allows for better resource allocation and enhances the power control. It balances with different network sizes and node densities under energy constraints. Clustering helps in reducing the number of exchanged communications in wireless sensor network resulting in low consumption of battery power of individual sensor nodes. This increases the life span of the wireless sensor network. The basic clustering protocol is LEACH (Low Energy Adaptive Clustering Hierarchy) has been proposed in [1] to maximize the network lifetime. It selects the cluster head based on the probability of selecting a random number between 0 and 1. But this protocol does not guarantee about equal number of CHs in each round and number of member nodes in each cluster. Clusters formed of uneven size makes network load unbalanced. Since node has different communication distance and remainder energy, LEACH cannot have chance to cluster head and consume their energy evenly with only probability. In addition, when the cluster head select, the cluster head causes on additional overhead to estimate the sum of the

current energy. Unlike LEACH, where nodes self-configure themselves into clusters, LEACH-C (LEACH-Centralized) [2] utilizes the base station for the cluster formation. During the setup phase of LEACH-C, the base station receives information regarding the location and energy level of each node in the network. Using this information, the base station finds a predetermined number of cluster heads and configures the network into clusters. The cluster groupings are chosen to minimize the energy required for non-cluster head nodes to transmit their data to their respective cluster heads. But, at the beginning of each round, LEACH-C causes on additional overhead to receive information. PEGASIS (power-efficient gathering in sensor information systems) [4],[3] is an improvement over LEACH by making only one node transmit data to the base station in this protocol every node transmits it's data only to its nearest/neighbour node in the data fusion phase. Xianghui Cao et al [5] focuses on the problem of designing control methods for actuators to control the environment and meet user requirements based only on local sensory information using distribute control strategy. But because of overlapping, actuators actuation will be doubled or even multiplied on the sensors places whose overlapping degree is larger. In the running phase, actuators should exchange their information and cooperate with each other and take their decision based on the overlapping extent of the neighbouring nodes. Alie El-Din Mady et al [6] enhance the functionality of a BAS through embedding advanced control strategies within a Wireless Sensor/Actuator Network (WSAN), and also make retrofit applications possible. Meng-Shiuan Pan et al [7] propose a WSN-based intelligent light control system for indoor environments. Wireless sensors are responsible for measuring current illuminations. Two kinds of lighting devices, namely, whole lighting and local lighting devices, are used to provide background and concentrated illuminations, respectively. Proposed communication and coordination framework: In this paper we propose a new clustering scheme based on the knowledge of the residual energy left in the nodes as well as the distance measurements with the base station and with the other nodes in the

clustering zone. Since in the literature we have seen the cluster head is chosen based on the random probability or in some cases based on the energy model probability but it does not guarantee the energy efficient uniform and real time clustering approach. But clustering based on the residual energy model approach helps in energy efficient and load balancing of the network, while considering the distance with the base station and with other nodes help in efficient data aggregation and fusion in real time scenario. While it also lessens the problem with congestion in the network and large control overheads problems. In this paper the given simulation area is divided in equal number of zones and sensor nodes are distributed in these zones. The actuator nodes are distributed in the given area. Since actuator nodes are resource rich so it knows its position while it has also large memory and processing capability, so the sensor nodes send their data to the actuator nodes for further analysis and themselves goes in low power mode operation. The actuator node send a beacon signal to find out the number of zones in its vicinity i.e. zones organize themselves to the actuator nodes which is near to it in comparison to the other actuator nodes. Thus each actuator node is responsible for time scheduling of the zones its locality. Each actuator node send a beacon signal to each zone in it locality to form a variable window in order to allot variable time slot to each zone depending upon the distance between the zone and the actuator node. When a particular time slot came the actuator node send a alarm signal to the zone to wake up and send their aggregated data. Initially few nodes advertise themselves to become cluster head based on the random probability of selecting a number between 0 and 1 using basic leach protocol and are known as advanced nodes. Let A= {A1, A2, A3 AM} are the advanced nodes where m is the number of advanced nodes. Then the selected cluster head nodes calculate its residual energy. Since the sensor nodes knows the position of each other node in the zone so it calculate the distance between the selected cluster head and the actuator node. Also it calculates the distance between the selected cluster head and the other nodes in the zone and transmits their calculated values to the actuator node. Similar work is carried out with other advanced nodes and their values are

transmitted to the actuator node. Thus actuator node compares their probability of becoming final cluster head and the node that has maximum probability become the final cluster head of zone. A similar process is done with other zones in their allotted time slot. Proposed Clustering scheme: SELECT_CLUSTERHEAD (m, n) Begin: for i=1 to m do for j=1 to n do // S= {S1, S2, S3 Sn} where n is the number of nodes in a cluster Dij = distance from Si to Sj endfor endfor for i=1 to m do for j=1 to n do Di=Di + Dij endfor endfor for i=1 to m do for k = 1 to number of actuator nodes distributed in a given area DBSi = Distance from Ai to Xk // X = {X1, X2, X3Xk} number of actuator nodes Endfor for i =1 to m do NDBSi = DBSi + Di Endfor for i = 1 to m do Cali = Cali + Residual energy Ai Endfor Neti = [1/NDBSi] + Cali

Neti = Max (Net1 + Net2 + Net3+..+ NetM) End When the final cluster head is chosen based on the max probability, the actuator node then advertises it as cluster head to all the nodes in the zone and the nodes aggregate their data and fuses them in the cluster head which directly send their aggregated data to the nearest actuator node. Similar action is done cluster head of other zones that transmit their aggregated data to the nearest actuator node. Once each actuator node collects aggregated data from every nearest cluster head they form their individual weight matrix.
Weight factor=

Actuator nodes Weight matrix = Cluster

head
[ ]

Then each actuator communicates with each other and transmits their weight matrix to a single actuator node based on the difference of the distances with other actuator nodes. Thus this actuator node form a combined weight matrix and takes all the control decision to optimally adjust the actuation of each actuator based on the preference of each zone or based on the user preferences. In order to optimally adjust the actuation value of each actuator in order to save the energy while fulfilling the preference of each zone we use the linear programming simplex method of optimization. Linear programming (LP) is a method of allocating the resources in an optimal way. It has been used as a decision making device in all automation industries, in finance organization and service distribution organization. Pseudo code of distributed scenario 1: 2: 3: count = 1; for i = 1 to LENGTH for j = 1 to BREADTH

4: 5: 6: 7: 8: 9: 10: 11: 12: 13:

zone_id = count++; {X, Y} = CalcSensorLocation (); sensor_zone_id = GetZoneID (X, Y); Broadcast(X,Y); E = CalcResidualEnergy (); Broadcast (E); //Locate Actuators in the area. for i=1 to NO_OF_ACTUATORS TurnOnActuator (i); Delay (100);

29: {H1, H2..Hn} = ChooseInitialClusterHeads () This will be done with help of LEACH Protocol 30: H = ActuatorChooseFinalClusterHead (); Based on the max residual energy and the least distance from other nodes in same zone 31: data = CHAggregate (); 32: Amin = GetMinHopActuator (H); 33: Trans2MinHopActuator (Amin, data); 34: TurnOffSensors (); 35: 36: end for i = NO_OF_ACTUATOR

14: SenseQuantity () All sensors sense the level of actuation 15: 16: 17: 18: TurnOffActuator (i) end for i=1 to NO_OF_SENSORS CalcActuatorsImpactOnSensor (i) =

37: WM = ActuatorCreateWeightMatrix (); 38: Trans2MinHopActuator (Amin, WM); 39: end

40: Create Combine Weight Matrix (Amin); 41: ApplySimplexMethod (); Based on constraint variables defined by preference of each zone and final objective function Based on the combined weight matrix and preference of each zone the optimum value of each actuator actuation is decided by using simplex method algorithm. The constraint variable is taken and based on the constraints and the final objective function the value of the actuators is optimized.

19: {A1, A2An} GetOverlappingActuatorForSensor (i); 20: end

21: A = GetActuatorWithLeastHop (A1, A2An); 22: 23: ActuatorTransBeacon (A); ActuatorReceiveAck (A);

24: ActuatorCalcRecvTime(A); Calculate average time to receive from sensors of a particular zone 25: //Assign a time slot to each zone corresponding to calculated receive time. 26: for i = 1 to NO_OF_ZONES =

] [ ] [ ]

Therefore the constraint variables are C1A1U1+C1A2U2+C1A3U3++C1ANUN P1 C2A1U1+C2A2U2+C2A3U3++C2ANUN P2 : : : : : : : :

27: {zone_id, time_slot} ActuatorGetCurrentTimeSlot (A);

28: ActuatorTransAlarmSignal (A, zone_id, time_slot);

CNA1U1+CNA2U2+CNA3U3++CNANUN PN

The objective variable is either max or min. Y = U1+ U2+ U3++UN.