Optimization by Genetic Algorithm in Wireless Sensor Networks Utilizing Multiple Sinks

WSN (Wireless Sensor Network) comprises of small-sized and constraint-capability SN (Sensor Nodes) which record, send and receive data, sensed to a sink. The network lifetime and energy usability are important challenges to be dealt with. During the working of the SN, the maximum amount of energy is consumed than sensing and processing of data. Therefore, an efficient transmission of the data is required so that the energy can be saved.

The main components of the WSNs are SNs, sink, user or database administrator. The sink is called BS (Base Station). Where, the SNs generates data through sensing the vicinity of the sensors and transmit it towards the sink. The transmission of the generated data totally depends upon the energy of the battery. Having the smallsized battery the SN gets depleted in a very short time [3][4]. The energy exhaustion of the nodes is greater in transmission than sensing and computation. Therefore, it is essential to route the data effectively and efficiently for the sake of the lifetime of the sensor nodes.
The sensors route the data through direct communication paradigm and multiple hops communication methods as illustrated in Fig. 1(a-c). Whereas, in Fig. 1(a) the trivial energy is consumed during the working of the sensors because each node routes the data towards the sinks is known as multiple hops communication method. In Fig. 1(b) the nodes transmit the data directly towards the sink, which is good for the nearer nodes but it is fatal for the farther SNs. This is called direct communication method [5]. Therefore, in Fig. 1

OVERVIEW OF GA
GA is the meta-heuristic evolutionary algorithm, which mimic, the natural selection process. The GA is utilized, mostly, for the optimization, where it finds the most optimal solution from the available solutions [6][7][8]. The

 
By applying the above criteria we get fit gene if the value obtained 1 and further that gene is assumed to undergo the crossover function. The research scholars of [15] proposed GA based optimization for deployment of SNs with respect to the area. The work of [16] focused upon optimization using multiple sinks. Another technique for localization SNs in a fully connected WSNs has been done utilizing GA.

Genetic Algorithms Optimization
Similarly, the localization and placement of multiple sinks with the help of GA for LAN (Local Area Network) traffic has been addressed in [17]. The lifespan enhancement of WSNs using GA is proposed by [18][19].

Simulation Results
The

Optimization by Genetic Algorithm in Wireless Sensor Networks Utilizing Multiple Sinks
For the comparison simulation, we considered a WSNs with 10 SNs deployed in an ad hoc manner as shown in Fig. 5(a) and the distance of all SNs from respective sinks as illustrated in Fig. 5(b). The direct distance of the SNs from the sinks helps in sending the data through minimum distant path. The path, furthermore, has been optimized through GA and data is routed through the best one from the available paths.
Notwithstanding, the direct distance is not the only option in WSNs for performing any routing scheme.

TABLE 2. DIRECT DISTANCE OF SENSORS TOWARDS EACH SINK
The multi-hop scheme as shown in Fig. 6, illustrates the path length towards the nearest sinks, which consumes minimum effort for routing and results trivial energy consumption. The multiple hops reduce the distance from/to the nearest sinks through the relay SNs and results in optimized path to route the data.
The data is routed through the optimal path towards the sink1 with the 94.8932 meters as illustrated in Fig.   6 (a-b) shows the total path distance towards sink4 is to the [20]. With equal number of rounds and same number of nodes the number of the dead nodes decreased as compared to the [20] as shown in Fig. 8, where the nodes die down slowly as compared to the [20]. The optimization through GA helps in reduction of the energy waste and resulted in improving the lifetime of the WSNs. The packet transmission of the proposed scheme advocates the better performance as compared to [20] as illustrated in Fig. 9, where the data packets received at the sinks are nearly 6000 while the data reception at the sinks in [20] remained nearly 2000 packets throughout the operation of the sensor nodes. The WSNs with 20 SNs deployed in an ad hoc manner are shown in Fig. 10(a) and the distance of all SNs from respective sinks as illustrated in Fig. 10 [20] as shown in Fig.   11. Simulation results show better performance of the proposed scheme as compared with [20]. With the proposed technique the number of Active SNs with respect to rounds are greater contrary to the [20] as illustrated in Fig. 12.
In Fig 11(a)

Optimization by Genetic Algorithm in Wireless Sensor Networks Utilizing Multiple Sinks
which shows the minimum path towards the sink 3 is 75.0206 while the SNs having minimum distance towards sink 4 over the optimal distance of 95.5024 as illustrated in Fig. 11(d).
The path length towards each sink reduces the distance for the sake of routing the data towards the respective sink and helps in trivial energy consumption. The energy expenditure has been reduced through the multiple hops routing scheme.
The network performance in terms of dead nodes has been depicted in Fig. 13. The results shows the proposed scheme performs well with dead nodes per round as compared to [20]. The packet transmission of proposed scheme outperforms the [20] as shown in