Informatics Publishing Limited

R. Abinaya ¹, S. Kamakshi ¹

Affiliations
1 School of Computing, SASTRA University, Thanjavur, Tamilnadu, IN

Abstract

The service provided by the Wireless Sensor Networks (WSN) should provide better quality as Quality of the network plays an important role in the improving the performance of the system. The dynamic topology and resource constraints of WSN are highly challenging in achieving QoS. Thus in this paper Quality of Service (QoS) is improved for some of the QoS parameters like packet loss and congestion. Categorization of nodes as qualified and unqualified is done using the parameters. The quality node then dynamically forms a network. A novel method to improve the QoS based on Artificial Neural Network (ANN) is used to train the unqualified nodes to make them as quality nodes. The structure of Artificial Neural Networks provides less complexity compared to other Computational Intelligence (CI) tools as wireless sensor networks are prone to many constraints like memory and energy it reduces the computation cost and time. Our Simulation results shows that the proposed system has better performance in improving the QoS by increasing the network lifetime and reducing the packet loss ratio.

Keywords

Artificial Neural Networks, Congestion Rate, Qualified Nodes, Quality of Service, Packet Loss Ratio, Unqualified Nodes, Wireless Sensor Networks

Full Text

   

J. Janani Deekshidha ¹, S. Kamakshi ¹

Affiliations
1 School of Computing, SASTRA University, Thanjavur - 613401, Tamil Nadu, IN

Abstract

Objectives: Congestion is one of the major threats that we experience while transferring the data in WSN (Wireless Sensor Network). It occurs when the incoming rate of the packets to the sensor exceeds its outgoing rate which leads to queuing delay and packet loss. Since the energy consumed is proportional to the amount of data being transmitted, energy consumption can be reduced by minimizing the unnecessary transmission of data. Methods: During congestion it leads to exhaustion of energy and decreases the efficiency of the node. In the existing method DAlPaS algorithm is used to increase the network life time by reducing energy depletion at the sensors along a single path by finding alternative path when more than one flow is initiated through the same path. Alternate path selection could be done either by soft stage scheme or by hard stage scheme. In soft stage, the node that receives more than one flow and has the risk of buffer overflow or low power status, advises one of the sending nodes to change the path and it continues to forward the packets for that flow till the sending node finds alternate path and changes the path. But in hard stage, the availability status of this neighbor node is set as false in the sending node’s neighbor table, so that the sending node is forced to find an alternate path before it sends the next packet. When more than one path exists, the sender will select the neighbor based on priority. The priority of a node is based on the distance from the sender and residual energy in the node. Findings: In the proposed method priority based dynamic alternative path selection algorithm is used to increase the network lifetime, to avoid network partition and packet loss due to energy depletion and to utilize the node’s energy efficiently. Applications: Some of the major applications of Wireless Sensor Networks are security surveillance, environmental data compilation and tracking of sensor node. Sensor network applications are wide in range and they vary significantly in the mode of deployment, modality of sensing or the power supply.

Keywords

Congestion Control, Delay, Dynamic Alternate Path, Energy Efficiency, Packet Delivery Ratio, Wireless Sensor Networks.

Full Text

   

G. Ravindran ¹, S. Kamakshi ¹

Affiliations
1 School of Computing, SASTRA University, Thirumalaisamudram, Thanjavur - 613401Tamilnadu, IN

Abstract

Background/Objectives: Security is an important feature for the communication of data in the mobile network. But, however the existing work does not satisfy the security feature of the communication of data in the network. Methods: Due to lack of security in mobile agents, there are many chances of security breach through intruders. To overcome these disadvantages, a general secured aglets system has been proposed and load balancing is achieved effectively through various models. Findings: The evolved aglet system exhibits high security under the external attackers. Hence, it distributes workload for all intermediate client users and determines various time calculations like Total run-time and Elapsed-time. Applications: Some mobile agent applications are Mobile Computing, Personal Assistant, E-commerce, Data Collection, Secure Brokering, Information Distribution, Network Security testing, Database searches and Parallel Computing.

Keywords

Aglets, Load Balancing, Mobile Agents, Security.

Full Text