Game Theoretic Reward Based Adaptive Data Communication in Wireless Sensor Networks

Abstract

Various applications have been developed to realize smart homes with wirelessly connected devices, e-health care, and environmental and industrial monitoring by using wireless sensor networks (WSNs). Most of the sensing devices in WSNs operate on limited battery power as their only energy source. Therefore, any energy optimization scheme can lead to significant improvement in the lifetime of such networks. The sensor nodes consume their energies on sensing, processing, and transfer of data. Most of the energy is consumed by the radio communication used for transmission of data among the nodes. The pattern of data transfer in ad hoc networks, such as WSNs, is totally relied on multi-hop communication in which each node needs cooperation and coordination of other nodes. In multi-hop communication, most of the times nodes offer relaying services to each other, which in return severely increase their power consumption. Due to such situation, some nodes adopt a non-cooperative behavior by not offering any relay service to others. Non-cooperative or selfish nodes try to lengthen their lives by preserving energies for their own data transfer. Such behavior can reduce the level of collaboration among the nodes, which ultimately reduces the performance of the entire network. A WSN may face various issues such as increased end-to-end delays, unbalanced workload, non-availability of optimal routes and declined lifetime due to the existence of selfish nodes in it. Various techniques have been introduced to overcome the issue of selfish nodes in a network. The most effective and modern type of mechanism is to design an incentive-based framework. By using the incentive-based mechanisms, the nodes may be stimulated for cooperation with one another and the overall network performance can be improved. In most of the researches these incentives are some values referred to as virtual currency, money, points or scores. These values are exchanged for getting relaying services among the nodes. Each node tries to collect an adequate amount of such values so that it may easily transmit its own data towards the based station (BS). In this work, we propose an incentive-based mechanism which is based on the fundamental parameters of nodes and their placement in the network. The incentives, called as scores, are initially generated by the BS and then the nodes pay and collect these scores during the data transmission. The BS intelligently determines and assigns scores to nodes according to their features. Moreover, a node blocking mechanism is also introduced by using a card system. Each node is given a card according to its importance and participation level in the network. A novel technique for computing nodes’ individual importance is also designed by introducing a new term i.e. closed neighbors. A set of closed neighboring nodes can be considered those nodes which take relatively similar relay request due to their shorter distance with each other. This work has been simulated in MATLAB and NS2 for validation and comparison with other protocols. Results show that our proposed mechanism outperforms as compared with other experimented protocols.