An energy efficient and optimized load balanced localization method using CDS with one-hop neighbourhood and genetic algorithm in WSNs
Introduction
Localization defines the estimation of location of unknown nodes deployed in Wireless Sensor Networks (WSNs) (Kuriakose et al., 2014). Variety of applications of wireless sensor network ranges from environment conditions observation, agriculture, smart home environment, defence environment etc. The dynamic need of the technology and human requirement has made the WSNs from static to mobile. The mobility is incorporated in the environment for the robust output of an observation. In mobile environment estimating the location of unknown nodes is more critical. The accuracy of the location estimation in such networks depends on two basic processes (Han et al., 2013). The first one is location estimation (Ali et al., 2013) that deals with the calculation of unknown nodes' location and the second process is location verification (Ali et al., 2013), which verifies the calculated location with the actual location. Numerous algorithms have been introduced to improve the accuracy level of the localization process.
The wireless sensors are resource constrained. The anchor nodes are privileged in this regard with more power resources, but the unknown nodes are not having such resources to utilize. To perform a heavy computational task, the sensor nodes become exhausted and therefore either it has to be replaced or the communication at that part of the sensor node will be untouched further. But, the replacement of such sensors becomes vital and infeasible in some applications due to the dynamic configuration. Thus, the localization process should be efficient in the terms of resource utilization. To dynamically allocate the load among anchor nodes load balancing has been introduced in the localization process (Ko, 2015). The objective behind distributing the load among anchor nodes for location estimation is to increase network lifetime (Dietrich and Dressler, 2009).
In this paper, we have introduced a load balancing mechanism for location estimation in WSNs using Connected Dominating Set(CDS)as a backbone and Genetic Algorithm (GA). Section 2 deals with related work in this field. The proposed network model, shown in Section 3, is completed in two phases (i) to optimize the backbone by using genetic algorithm in Section 3.1. (ii) to estimate the location of unknown nodes with the help of optimized backbone in Section 3.2. The results and analysis are provided in Section 4. Finally Section 5 concludes the paper.
Section snippets
Related work
There are various approaches available to solve localization problem which differ in the assumptions including device hardware, signal propagation models, timing and energy requirements, composition of network, operational environment, beacon density, time synchronization, communication costs, error requirements, and node mobility. Localization techniques (Kuriakose et al., 2014) are broadly classified into two categories: Direct approaches and Indirect approaches. The direct approach can be
Proposed network model
The first consideration of the network model is to be self-organizing that is having no central control of deploying the sensor nodes in the network. The sensor network model has been considered to be in Two-Dimensional (2-D) and is represented by a graph consists of V, a set of vertices and E, a set of edges. The connectivity in such a network is an important parameter to analyse the overall performance metrics in localization algorithm. The size of the network can be defined as:
Result and discussion
The proposed algorithm has been simulated in Network Simulator-2 with parameters enlisted in the Table 2.
The performance of the algorithm has been analysed in the terms of, localization error and network lifetime.
Localization error is defined as squared error of the estimation given in the Eq. (28) and the average localization error as:The proposed algorithm uses linear energy model for each anchor node given as:where Ra is transmission range of anchor node aj and c1
Conclusion
In this paper, we have proposed a novel algorithm of optimizing a CDS which is based upon anchor nodes. Another major consideration is, the proposed algorithm applies random mobility for both the anchor nodes and the unknown nodes. The optimization process uses the genetic algorithm with elitism strategy so that the fittest solution can be retained accordingly for a fast convergence of the global solution. Obtaining the optimized CDS, the localization process executed that depends upon
Gulshan Kumar working as an Assistant Professor in Lovely Professional University, Punjab India and did his B.Tech from Amritsar College of Engineering, Amritsar (2009) in Computer Science Engineering, M.Tech from Lovely Professional University, Punjab India with area of specialization in Mobile Ad hoc and Sensor Networks. He has many publication in well renowned International journals and Conferences Mailing Address: House No-3, Gali No 3, Kamla Devi Avenue, Amritsar (Punjab), India.
References (29)
Cooperative energy-efficient localization with node lifetime extension in mobile long-thin networks
J. Netw. Comput. Appl.
(2016)- et al.
A machine learning approach for user localization exploiting connectivity data
Eng. Appl. Artif. Intell.
(2016) - et al.
Distributed efficient localization in swarm robotics using min-max and particle swarm optimization
Expert Syst. Appl.
(2016) - et al.
Approximation algorithms for load-balanced virtual backbone construction in wireless sensor networks
Theor. Comput. Sci.
(2013) - et al.
Energy efficient load-balanced clustering algorithm for wireless sensor networks
Procedia Technol.
(2012) - et al.
LIS: localization based on an intelligent distributed fuzzy system applied to a WSN
Ad Hoc Netw.
(2012) - et al.
A dynamic neural network approach for solving nonlinear inequalities defined on a graph and its application to distributed, routing-free, range-free localization of WSNs
Neurocomputing
(2013) - et al.
Minimum connected dominating sets and maximal independent sets in unit disk graphs
Theor. Comput. Sci.
(2006) - Ali, Nabil, Alrajeh, MB, BS., 2013. Localization techniques in wireless sensor networks. Int. J. Distrib. Sens. Netw.,...
- et al.
Localized protocols for ad hoc clustering and backbone formation: a performance comparison
IEEE Trans. Parallel Distrib. Syst.
(2006)
On the lifetime of wireless sensor networks
ACM Trans. Sens Netw.
Localization algorithms of wireless sensor networks: a survey
Telecom. Syst.
Cited by (30)
GREPHRO: Nature-inspired optimization duo for Internet-of-Things
2024, Internet of Things (Netherlands)ACRDA: An adaptive combined relay based dynamic data aggregation technique for wireless sensor networks
2022, Measurement: SensorsCitation Excerpt :The algorithms of HEED and LEACH were compared with the proposed energy-efficient clustering algorithm for obtaining the improved energy-efficiency. Based on the each node's residual energy and probability of the cluster head determination, the performance of the proposed technique has been improved that leads to the increased network lifetime and decreased energy consumption [26]. For increasing the WSNs lifetime and reducing the energy consumption, the novel algorithm of IEE-LEACH was considered in the paper [27].
A balanced energy efficient virtual backbone construction algorithm in wireless sensor networks
2019, AEU - International Journal of Electronics and CommunicationsCitation Excerpt :Kumar et al. proposed an optimized load balanced Connected Dominating Set using Genetic Algorithm. This involves the formation of CDS using anchor nodes the localization process [35]. Some researchers have exploited the familiar Genetic Algorithm in WSN for achieving network longevity [36] and maximum coverage [37].
Parallel memetic algorithm for training recurrent neural networks for the energy efficiency problem
2019, Applied Soft Computing JournalCitation Excerpt :Nevertheless, these approaches are ad hoc solutions to EE problems and are based on classical techniques which require high computational costs if we want to achieve optimal solutions. Different soft computing paradigms such as artificial neural networks (ANN) [8–10] or evolutionary computation (EC) [6,11,12] can be used to find predictable energy management systems that reduce energy waste in buildings: ANN-based systems have proved to be successful for models for energy consumption prediction [9,10,13], energy inefficiency diagnosis and fault detection [14,15]; EC-based ones support multi-objective applications that can be combined with Data Mining (DM). There are several EC applications worth mentioning and these include ones that deal with cost-optimal analysis [16], classification of new electricity customers [17], or selection of the most relevant features [18] and detection of outliers [14,19,20].
Differential evolution algorithm applied to wireless sensor distribution on different geometric shapes with area and energy optimization
2018, Journal of Network and Computer ApplicationsCitation Excerpt :The feasibility of DEA with these constraints is tested on different scenarios, one of them with obstacles. Recently, genetic algorithms have been applied to issues like load balancing (Kumar and Rai, 2017). It is done to reduce the amount of energy consumed by each node when it tries to locate the possible unknown nodes that surrounds it.
A framework for enhancing the performance of Internet of Things applications based on RFID and WSNs
2018, Journal of Network and Computer ApplicationsCitation Excerpt :One way to minimize the energy consumption and, thus, increasing the network lifetime is by deploying new or adapting available routing protocols (AlShawi et al., 2012). In this context, various studies have proposed new routing solutions in WSN applications (Machado et al., 2013; Tunca et al., 2015; Kumar and Rai, 2017; Abdul-Salaam et al., 2017; Mann and Singh, 2017). In this sense, the FSBRC component uses a fuzzy system to classify routes and assist routing protocols in IoT scenarios that integrate WSNs and RFID.
Gulshan Kumar working as an Assistant Professor in Lovely Professional University, Punjab India and did his B.Tech from Amritsar College of Engineering, Amritsar (2009) in Computer Science Engineering, M.Tech from Lovely Professional University, Punjab India with area of specialization in Mobile Ad hoc and Sensor Networks. He has many publication in well renowned International journals and Conferences Mailing Address: House No-3, Gali No 3, Kamla Devi Avenue, Amritsar (Punjab), India.
Mritunjay Kumar Rai received his Doctorate degree from Indian Institute of Information Technology and Management, Gwalior, India, after the completion the Master of Engineering degree in Digital system from Motilal Nehru National Institute of Technology, Allahabad, India. Presently he is working as an Associate Professor in Lovely Professional University, Phagwara, India. His research Interest include Wireless Networks, Network Security and Cognitive Radio Networks. He has published more than 40 research articles in reputed International Conferences and International Journals. Mailing Address: Room No 801, 41-C, Lovely Professional University, Phagwara (Punjab), India