Efficient connectivity in Wireless Sensor Network using adaptive strategy in evolutionary computation

— At present ,wireless sensor network (WSN) has reeived intense attention by reserchers because of its utilities in the various applications.In WSN ,sensors are battery operated devices and most of cases practicaly it is not possible to replace the battery once it loose the life .There are various reasons for energy consumption and among them one very significant factor is unefficient connectivity of sensors for sharing the relevent information under network.Connectivity situation can be more worse if there is a dynamic environment exist under network environment.In this paper ,evolutionary approach based on various form of genetic algorithm has proposed to handle this issue.Different strategy like ,redifiniton of agents,inclusion of flying agents and carrying the experience have included to enhance the qulaity of solution.

connectivity of networks and elevating network lifespan. The NP-complete problem is articulated by using combinatorial optimization method. They have suggested potential field deployment algorithm (PFDA) and multi-objective deployment algorithm (MODA). [6]This paper analyzes the metric relationship model with 3-D Clifford sensor network. The Clifford sensor network contains connection graph with independent of coordinate and is reliable with different targets in diverse dimensional space. [7] The authors planned a deliberate model with queuing to obtain an optimum solution to elevate energy depletion of the sensor node. [8] Has investigated the connectivity of random deployment nodes in WSN with Gaussian distribution. This method provides better results while simulating in nonlinear distribution of sensors. The important problem in WSNs is Coverage and connectivity problems, which have a abundant effect on the performance of WSNs. The Improved method with efficient node deployment approach is used. In [9], authors have categorized the problem of coverage in WSN with different angles, define the estimated metrics by using the suitable algorithms. In [10],the node deployment problem is articulated with multi-objective optimization (MO) problem where, the objective of the problem is to discover a deployed sensor node to get the best out of coverage of targets, minimum the network energy depletion, better network lifespan, and connectivity between source and destination node for accurate transmission of information with minimum number of sensor nodes.
[11] Presented the essential study on the connectivity between nodes in wireless sensor networks and efficient coverage of targets is obtained from mathematical modeling, theoretical study, and performance assessment perceptions. [12] In this paper the authors are presented, node deployment pattern with polygon shape to obtain optimum position of sensors in Wireless Sensor Networks for efficient coverage and connectivity. The important objective of this problem is to provide efficient connectivity and maximize the coverage with minimum number of sensors. To obtain better solution with minimum computational sources for node deployment problem with NP -hard is perplexing problem of research in WSN.
[13]has presented an outline of WSN and node deployment problem in wireless sensor networks , and deliberations on metaheuristics and demonstrates how to use the meta heuristics methods to resolve the node Deployment Problem in WSN. The efficient method is used for curtail by using Sleep Scheduling (SS) mechanism and enhance lifespan of wireless sensor networks. In [14], authors presented software based algorithm to cope the energy of the wireless sensor network with Sleep Scheduling of nodes. In [15], review has presented the issues built on archetypal of WSNs: structured and non-structured for data gathering and aggregation and also discussed the importance of clustering and routing in wireless sensor networks for better energy preservation and lifespan of the network. In [16] authors have presented a distributed Resource Constrained Recovery method is used to restructured a network subdivided into dismember segments by deliberately relocation of nodes. The cases in which relocation nodes are inadequate to form steady topology of inter segment then mobile data gatherers with elevated routes to reduce data delay III. PROPOSED SOLUTON To obtain minimum distance, we have implemented adaptive genetic algorithms like, Redefined agents genetic algorithm [RAGA], Flying agents genetic algorithms[FAGA], Experienced agents genetic algorithms [EAGA].These algorithms provide better connectivity in dynamic environments of wireless sensor networks. The dynamic environments obtain when the sensors are moved from place to another place with the help of animals, robotics and human beings (soldiers). The adaptive genetic algorithms provide better solutions when the sensors in dynamic environment.
In WSN there is a number of applications where dynamic topology exists for example in the case of a) Track animals b) Soldier strategy in war field etc. To handle Dynamic wireless sensor networks there is requirement of high level adaptability. In natural system evolutions can we consider as best example for adoptability, hence genetic algorithm platform has adapted, but the existing challenges are To find the optimal connectivity from one sensor to another for communication Detect change in topology With the change topology as shown as possible reestablish the optimal connectivity. To handle all these three different approaches are developed as shown in the Fig-1 F_fun (MPOP) 5. Next Generation population Tournament selecetion 〖(f〗_(vi )) 5. Topology change detection f(f_vi-f_(v(i-1)) ) ; 6. If change detected Current Generation = New Random Agent population ; Else Current Generation = Next Generation; 7. Go to step 1.

Flying agent's with time [FAGA]:
The new member are always incorporated with the populations in result there is a better diversity available and if there is dynamic condition appeared this diversity will help to find out solutions . With this approach it is possible to handle dynamic conditions faster.With initial definition of population the genetic operator applied to create offspring population and through tournament selection new population is created .The fitness value estimated and weak one is going to replace by newly created flying agents and this population will replace population and process is goes on. The psudo code for FAGA has given below.
Psudo code for FAGA

Experienced agent's with time: [ EAGA]
If there is possibility to place experienced agents in the next generation always there is very good chance to handle dynamic topology in efficient manner because they are having the knowledge to handle the change in the topology with the time in result the optimal solutions can be achieved with very less time and chances failure is minimum.In EAGA with initial population definition and applying the genetic operator through selection process new population is created among the new population whichever is having higher fitness value that is solution stored in memory. This process will goes on in every generation .The previously stored experienced solution will replace the weak solution available in current population to create find new population and process will continue. The psudo code has given below.
Fundamentally the two different approaches has applied to handle dynamic available in topology-(i) Detect and change in topology and take action (ii) Be adaptive all times, if there any change occur in topology it will handle by the same process RAGA concept applied to handle detedct and change approach while FAGA and EAGA applied to be adaptive always.
Psudo code for EAGA   It is observed with experiment that RAGA may take longer period to generate optimal solutions and there may be chance in between topology changes further which make the situation worst and obtained result are no way useful. If there is a high levels of change occur in topology it may difficult to find the optimal solutions within time span by FAGA.EAGA has shown the superior results in all cases in terms of high level of adaptive characteristics and faster exploration of global solution.this is possible because of exerience from past event help to handle the new event if it occurs.
V. CONCLUSION Objective to increase the Connectivity of WSN can be obtained by concept of adaptive genetic algorithms. In this present paper a concept to increase connectivity of WSN by using adoptive genetic algorithm has presented, which help in find the solution to obtain high connectivity between source and destination. For various different sensors experiments has done on Redefine Adoptive genetic algorithm (RAGA), Flying Agents Adoptive genetic algorithm (FAGA), Experienced Adoptive genetic algorithm (EAGA) and shown very clearly that the proposed solution provide high connectivity between the source and destination by choosing minimum distance. In future the hybrid adoptive genetic algorithms is developed to enhance connectivity in dynamic environment of wireless sensor networks