Modified W-LEACH Protocol in Wireless Sensor Network

In this paper, a Modified Weighted Low Energy Adaptive Clustering Hierarchy (MW-LEACH) protocol is implemented to improve the Quality of Service (QoS) in Wireless Sensor Network (WSN) with mobile sink node. The Quality of Service is measured in terms of Throughput Ratio (TR), Packet Loss Ratio (PLR) and Energy Consumption (EC). The protocol is implemented based on Python simulation. Simulation Results showed that the proposed protocol provides better Quality of Service in comparison with Weighted Low Energy Cluster Hierarchy (WLEACH) protocol by 63%.


INTRODUCTION
A Wireless Sensor Network (WSN) contains sensor nodes and mobile sink node as shown in Fig. 1.A large number of sensor nodes are set up in the monitoring area.The monitoring data aggregated by the sensor nodes is transmitted to the sink node.The WSN's nodes are divided into groups each group has a central node which is called the Cluster Head(CH); that is responsible for storing the sensory data of all group nodes and sending them to the mobile sink Sensor nodes have limited energy and memory that is difficult to change the battery and to increase the memory; especially in places that are hard to reach, such as war field, polluted area Rad, et al., 2017.There are few different approaches that are applied in the literature to reduce EC in WSN and extended network lifetime.Some of these methods are: improving routing protocols, minimizing the amount of packets that travel through the network, making the sink node mobile which is another method for improving the lifetime of wireless sensor network and putting some sensor nodes into sleep mode and using only a necessary set of sensor nodes for sensing and communication Vinutha, et al., 2017, Soua, and Minet, 2011.Weighted Low Energy Adaptive Cluster Hierarchy (W-LEACH) Protocol is one of the efficient types of LEACH protocol, that is a centralized algorithm, similar to LEACH, W-LEACH consists of two phases, setup, and a steady phase.In the first phase, W-LEACH chooses a maximum of probability p% of alive sensor nodes to be cluster-heads(CHs).Every CH in the network is selected based on a weight value   , that is assigned to each sensor node   , without

Mobile sink node
taking into account thefree size of the buffer for each sensor in the cluster.Different from LEACH, each sensor node in cluster sends data to its CH during steady state phase, CH collects data from all, W-LEACH elects only x% of sensor nodes in each cluster for sending sensory data to their CH.Elected x% of sensor nodes based on the proximity between sensor nodes, as an example, if there are two sensors, and the distance between them is close, one of them is chosen to send the data and the other to put in a sleep state.The candidates for sending data to CHs are also chosen based on their weights, Snigdh, and Gosain, 2015.In general, W-LEACH works well.However, when the number of nodes in each cluster is larger; the performance of the protocol will be severely degraded.This is because W-LEACH uses x% of sensor nodes in each cluster.Which is easy to cause network congestion when the number of nodes in each cluster is large and the elected x% is also large.Therefore, it is the key to improve network performance by changing the number of the active sensors for each cluster.Solving the problems, via modifying W-LEACH protocol that will be done by coordinating x% with CH's capacity of the buffer to avoid buffer overflow.The aim of this paper is how to modify routing protocol, which is able to enhance QoS network by minimizing energy consumption this will be achieved by increasing Throughput Ratio and reducing PLR of the network.Finally, the results will save the energy and extend the network lifetime.

W-LEACH PROTOCOL
Supposing that a WSN consist of n alive sensors, for each round, W-LEACH assigns a weight   for each sensor node; based on two factors, the residual energy   , and the density   .The   is the energy remaining in the sensor node   after some round t.   is the density refers to the ratio of the number of alive sensors that are within a range r of   to the total number of alive sensors in the network n, at some round t,  ℎ is the minimum density threshold, represented in the Eq. ( 1) and Eq. ( 2): Thus, the value of density for every sensor node is updated during rounds depending on whether or not other sensor nodes in their ranges die.Choosing a CH has a high-density area leads us to save sensors' energy for its cluster.Since the distance between the sensor nodes and the CH is relatively short, energy consumption will be reduced.However, choosing a CH in low-density regions would result in that the cluster members lose more energy while sending their data when it is compared to long distances.Accordingly, W-LEACH efficiency is not affected by the remaining memory in the sensor that causes dissipated energy by sending sensory data.

MW-LEACH PROTOCOL
Buffer-overflow is a challenge that occurs in the CH.It is noted from the above that W-LEACH protocol does not take into account the remaining memory in the sensor as a factor to select CH, which leads to loss of sensory-data because it does not choose the suitable buffer size from the sensor nodes.Here it is proposed by this work to include an additional factor to the weight that chooses the CH and the size of the memory to reduce the Packet Loss Ratio (PLR) as it is demonstrated in the Eq. ( 3) and Eq. ( 4) : is the factor of buffer size capacity for every sensor to determine the ratio of    which refer to the maximum capacity of buffer size in sensor   to the number of alive sensor nodes in range r for   .The term density of one node is the amount of aggregated neighboring nodes in a place in range r.In a small size CH buffer, the communication between the boundary cluster node and the CH consumes too much energy and buffer.Additionally, too many group members causing more collision.In contrast, a big size CH buffer has a large detection range which also would cause additional power consumption.Thus, it requires to coordinate the buffer size of Cluster-Head with a number of active sensors in range nr, to enhance incoming traffic rate from sensors to CHs; in order to ensure less buffer-overflow which leads to saving energy.The above mentioned will be presented mathematically by the following Eq.( 5)

𝑛𝑟 𝛼 𝐵𝑢𝑓𝑓𝑒𝑟 𝑠𝑖𝑧𝑒 𝑓𝑜𝑟 𝐶𝐻
(5) 07 Changing the state of sensors that are out of dynamic elected CH range to sleep state; in this way, the real profit will be in energy saving and minimizing the redundant data that is sent to CH and sink node.Since the selected sensor as a CH has a low energy, the CH will die before sending the entire data and this will result in data loss.To overcome this drawback, the energy threshold must be checked to select CH from sensor nodes, which is showing in Fig. 2.

SIMULATION MODEL 4.1 Simulation Tool
The well-known simulation tool that is used is PYTHON 2.7.Different simulations results are presented with a different number of clusters in order to check the performance of the proposed modifying.The goal of the paper is to investigate the behavior of modified W-LEACH, conventional W-LEACH for Throughput, Energy Consumption, Lifetime and Packet Loss Ratio.

Simulation Setup
Extensive simulations are carried out to analyze the effectiveness of the proposed MW-LEACH.100 sensor nodes are randomly placed in a space of 100 x 100 m, with inner energy 2 joules for each sensor.Packet size is assumed to be fixed at 800 bytes and each node source generating data at a maximum rate of 5 packets per millisecond(msec.).In this simulation, the assumed sink buffer capacity is 150 packets and the initial presence of one sink located in the center of WSN.Simulation time for each scenario was set to 250 msec and repetitive simulations for each scenario were performed to verify the reliability of the results.

Performance Metrics for Evaluation  Throughput Ratio
The Throughput Ratio reflects the effective network capacity.It is defined as the total number of packets successfully delivered at the input sink buffer and a total number of packets generated by the active sensor nodes on the network in a given period of time.
Throughput Ratio shows the protocol's successful deliveries for a time: this means that the value of Throughput Ratio nearest to one is the better will be the protocol performance.

 Energy Consumption
The energy metric is taken as the average energy consumption for network calculated through simulation time.
 Lifetime Time spent until the death of all wireless sensor nodes in WSN.

 4. Packet Loss Ratio
It is defined as the ratio of the number of packets dropped at cluster head and sink buffer due to overflow and a total number of packets generated by the active sensor nodes on the network: this means that the value of Packet Loss Ratio nearest to zero is the better will be the protocol performance.

RESULTS & DISCUSSIONS
Fig. 3 displays the Throughput Ratio of the network.This figure shows that, maximizing the Throughput Ratio of the network is achieved by using MW-LEACH in comparison with W-LEACH; this feature is more benefit and it is achieved by decreasing traffic via making coordination between active sensor and maximum capacity of buffer in cluster head to send suitable amount of packets from sensors to the available buffer of the network, that lead to reducing packets drop, which is caused by overload on cluster heads limited buffer and sink node.It is noticeable that the starting chart of MW-LEACH Throughput Ratio is increasing after about 60 round t; because of the energy factor for each sensor   is high, therefore, will effect on the weight   , then the MW-LEACH work better because the energy is more consumed.The increasing of Throughput Ratio will decrease the energy consumption because of the dissipated energy for sending packets over the capacity of the buffer which is reduced.It is clear from Fig. 4 which show that EC, the MW-LEACH save power and prolong lifetime as represented in Fig. 5. Packet Loss Ratio was reduced in MW-LEACH protocol, therefore the buffer overflow will be reduced in cluster head and sink node by sending suitable amount of packets from sensors to the available buffer via making sensor nodes that are out of cluster head capacity change to inactive mode (sleep), that is clearly noticed in Fig. 6 .

Packet Loss Ratio
As it is shown in Fig. 7 the Throughput Ratio is increased when the number of clusters in the network is four while, the Throughput Ratio is decreased when the number of clusters is six and eight and that's because the M-WALLACH makes the number of active sensor nodes in every cluster suitable to the capacity of the cluster head buffer to avoid buffer overflow, which means increasing the number of clusters that will lead to increase the number of cluster heads that mean maximum the number of buffers that in turn increase the number of active sensor nodes in network.Traffic from six and eight clusters is greater than traffic that is coming from four clusters.Higher traffic causing more overflow on limited sink's buffer and decrease Throughput Ratio of the network by increasing Packet Loss Ratio because increasing number of packets dropped by overflow on sink buffer in high traffic in six clusters and higher in eight clusters as shown in Fig. 8.In Fig. 9 and 10 there is a clear difference in energy consumption and network lifetime when compared with four, six and eight clusters in network where the energy depletion in four clusters

CONCLUSIONS
This paper presented the Modified W-LEACH protocol to enhance the Quality of Service (QoS) in WSN.The QoS in terms of Throughput Ratio, Energy Consumption (EC), network lifetime and Packet Loss Ratio(PLR).The simulation results and discussion show that the Throughput Ratio is more increased in MW-LEACH and PLR is more reduced than the PLR when using W-LEACH protocol, reducing the PLR means minimize dissipated energy, which in turn saving network energy and prolong network lifetime when using MW-LEACH.

REFERENCES


Figure 6 .
Figure 6.Comparison between W-LEACH and MW-LEACH in PLR.
is decreased when compared with six and more decreased with eight clusters because the dissipated energy via PLR is increased with the increasing of the number of clusters in network.

Table 1
summarizes the simulation parameters.Evaluation metrics used performance test are Throughput, Energy Consumption (EC), Lifetime and Packet Loss Ratio(PLR).