A Modified Anchor Free Localization Technique for Wireless Sensor Network

INTRODUCTION: A novel anchor-free localization technique for a single-hop wireless sensor network is proposed, which is based on internode distances only. The average of the internode distances is mainly used to localize the reference node in the proposed algorithm. Average localization error is used as the main performance metric. The simulation result shows that the proposed algorithm significantly reduces localization error as compared with conventional localization techniques with respect to node density variation and network size. OBJECTIVES: To find the location of the randomly deployed sensor nodes. METHODS: An anchor free localization method is used to find the location of the deployed sensor nodes. RESULTS: Location error is comparatively less in the proposed strategy. CONCLUSION: For single-hop wireless sensor networks using inters node distance, a novel anchor free localization technique is presented


Introduction
Wireless Sensor Networks (WSNs) are autonomous networks for which there is no need of any pre-established network. It consists of small sensors, control units, transceivers, and batteries. These sensors are deployed over the areas of inaccessibility [1]- [3]. In this type of network, nodes are responsible mainly for routing data to the base station.
In order to compute the physical location of the deployed wireless sensor nodes, localization strategies are used. Finding the exact location of the data is one of the prime concerns in WSN. Erroneous location of the nodes results in enormous data, energy consumption, and degree on connectivity. The localization method of the sensor network is classified into two categories-(1) Range-based localization, and (2) Range free localization algorithms. In the range-based technique, geometrical computation is used. While in range free localization, average hop count between anchor and sensor node is the prime parameter for localization. The accuracy of localization can be improved by increasing the number of nodes in the region. However, anchor nodes are more costly than the ordinary nodes. In range based strategy for getting the higher accuracy additional hardware accessories are required. However, these accessories increase the cost of the hardware such as Received signal strength indicator (RSSI), Time of Arrival (ToA) Angle of Arrival (AoA) Comparison between Range based and Range Free localization Techniques are summarized in table 1, [4], [5].
In this paper, a range free node localization algorithm is proposed for single hope WSN. The proposed algorithm is based on the average internode distance.
There, RSSI value is used to determine the distance between the nodes. The rest of the paper is organized as follows, Section 2 presents previous works that have done in the field of localization in WSN. Section 3 is described as the metrics, which are used in the proposed algorithm. Section 4 depicts the proposed algorithms with flow chart. Section 5 presents simulation and performance evaluation of the proposed algorithm and Section 6 concluded the proposed work.

Literature Review
Many researchers propose various techniques of localization of WSN nodes. In this section, we present previous research work in the field of sensor node localization.
(Pandey and Varma, 2016) [6] In order to evaluate the impact of localization, a distributed cooperative approach based on hybrid extendable range have been proposed. (Tuba, Tuba and Simian, 2016) [7] Localization approach based on bat algorithm is proposed. In the proposed strategy, position of the end nodes in the regions is calculated. After that, using range based strategy positions of the other nodes are calculated. (

Metrics of Proposed Algorithm
Following parameters are used in the proposed algorithm-

(1) Received Signal Strength Indicator (RSSI):-The
Received Signal Strength Indication (RSSI) defines the relationship between wirelessly transmitted signal power (P t ), received signal power (P r ), and the distance between the nodes (d). This relationship is expressed as Eq. 1, in which n is the transmission factor whose value depends on the propagation environment.
(2) The average distance of a node (Davg):-If n number of sensor nodes are randomly deployed at (x 1 ,y 1 ), (x 2 ,y 2 ), (x 3 ,y 3 ) ….. (x n ,y n ) with inter-node distance as , , ∀( , ) ∈ , ≠ respectively. Therefore, d ij can be defined as - (3) The centroid of the nodes can be calculated as [21] - (3) (4) The average node distance (Davg) between all distributed nodes will be roughly equal to the distance from that node to the centroid (Dc In proposed anchor free localization this D avg and internodes distance are used to localize reference nodes first then for other nodes.

Proposed Algorithm
In the proposed algorithm we consider a square topology in which n sensor nodes are randomly distributed.
Step 1: Using allocated TDMA scheduling, each node transmits a beacon message with its node ID. The node uses RSSI to measure (n-1) distances (dij) from all other nodes using equation (1).
Step 3: Sharing of Davg among all the nodes and make the highest Davg node as the first reference node shown in Fig. 1(a).  Step 4: Next reference node is the farthest node from the first reference node shown in Fig. 1(b).
Step 5: Similarly, compute the Third and Fourth farthest node from the previous nodes as shown in Fig. 2 Step 6: Identify the Fifth node which has minimum Davg and make it as origin as shown in Fig. 3.

Fig. 3 Estimation of Fifth reference node and
Localization of Node 'j' By considering reference node 5 as origin and node 1 is considered on the positive x-axis, as shown in Fig. 3. For localization of any node j, compute the angle 'θ' as given in Eq. 6.
5j .d 53 2d The localization of node depends on the quadrant in which node is lying, For example, if the node j is lying in 1st quadrant By using Eq. 6 the x and y position of the 'j' node can be given as:

Simulation and Performance analysis
For analysis of the proposed algorithm, MATLAB software is used, in which a square field of length 100 m 2 to 400 m 2 with a step of 100 meters, and the number of nodes varied from 100 to 500 with a step size of 50 nodes are considered. For simulation following parameters are used- From above Fig 4, it is obvious that the average localization error decreases with an increase in node density due to the decreased (D avg ) which is used in the estimation of reference node localization. Initially proposed algorithm also shows a higher average error, but with increased node density in the topology algorithm shows less average error. Localization error for different sizes of the network would be different as shown in Fig. 5, Initially, in a comparatively small network, comparatively node density is high therefore less localization error occurs. On the other hand, with the same node density will be topology with increased size, high localization error exists, as all positioning estimation depends at D avg , Which uses node density.

Conclusion
In this paper, a novel anchor free localization technique for single-hop wireless sensor networks using inter node distance is presented. The proposed algorithm is investigated on the metric of node density and varying network size. Results have shown that the proposed strategy performs better than the existing classical localization techniques such as Centroid, Amorphous, DV-Hop, and DV-Hope Max.