Furthermore, an energy-efficient clustering protocol based on iABC metaheuristic is introduced, which inherits the capabilities of the proposed metaheuristic to achieve optimal cluster heads (CHs) and improve energy efficiency in WSNs. Selecting appropriate CH with optimal capacities while balancing the energy efficiency ratio of the network is a well-defined NP-hard optimization problem in WSNs[9]. Further, to exploit the capabilities of the proposed metaheuristics, an improved artificial bee colony-based clustering protocol, Beecluster, is introduced, which selects optimal cluster heads (CHs) with energy-efficient approach in WSNs.
Here we present only the researchers' vital contribution based on classical as well as CI based metaheuristic approach;. Advice on an improved solution search equation that tries to strike a balance between the exploitation and exploration capabilities of metaheuristics. Further, an improved solution search equation called ABC/rand-to-opt/1 is proposed, which is motivated by existing Differential evolution (DE) family framework, and trains an optimal solution from the current best solutions, thus improving the convergence speed of the proposed metaheuristic.
Beecluster - an improved artificial bee colony-based clustering protocol: By exploiting the capabilities of the proposed metaheuristics, we introduce Beecluster, an improved artificial bee colony-based clustering protocol for optimal cluster head (CH) selection, which is a well-identified NP-hard optimization problem in WSNs. These equations are used to refine the sampling process, which ultimately increases the convergence rate of the proposed metaheuristic globally. The proposed solution search equationABC/rand-to-opt/1, which only uses the information of optimal solutions in the current population, can improve the convergence speed of the proposed metaheuristic.
Beecluster - proposed clustering protocol
The first component; Tque is the queue delay, the second component; Ttran is the transmission delay and the third Tack is the delay due to the acknowledgment packet. Placing a BS in an appropriate location is a very important task in a WSN, as it affects its performance by reducing energy consumption and increasing network lifetime. Therefore, we present an analytical evaluation to determine the optimal BS location based on critical energy analysis of WSNs.
This is the first time we incorporate the importance of energy consumption when exchanging an ACK in WSNs. Therefore, the energy consumption for transmission of l bit data is composed of three parts: the energy consumed by transmitterEtrans, by receiverErec and by ACK packet exchangeEack. If there will be n nodes uniformly distributed in an m∗m field with k clusters, then there will be n k nodes per cluster.
Of these, there will be one CH node and n will remain. now the energy consumed by a non-CH node is given by :. and the energy consumed by a CH node is given by. Now, to get the optimal BS location, we need to minimize Etotal(l, Dssin+1), which depends on the type of nodes dominating the residual energy. If nearby nodes dominate i.eEresa>> Eresthen the centroid nodes of {s1, s2, ..sa} S will be the optimal location for the BS, and if Eresb >> Eresa then the center of {s1, s2, ..sb} Nodes S will be the optimal location for BS.
CH selection is one of the crucial tasks for cluster formation in WSNs, as it affects the overall performance of the network. CH will be responsible for collecting data originating from several SNs and transmitting aggregated data to the BS. Now we construct a fitness function to evaluate the fitness of individual food source of the population.
The forager bee will then select a food source with a higher fitness and perform a local search on xij, if the new solution has a better fitness then it will replace xij with the optimal solutionxopt,j and is set as a CH, otherwise the old solution will be retained. Now, if the fitness cannot be improved further, after a number of trials, the corresponding employed bee becomes a scout to produce a new food source randomly using Eq. Then each non-CH node will join the nearest CH node based on the squared Euclidean distance between them (Eq. 24), through a short Join-Acknowledgment (J-ACK) message which will be transmitted using a MAC CSMA/CD protocol, in became a member of the group.
After receiving J-ACK messages from all surrounding nodes, each CH must maintain a cluster member table and create a TDMA schedule for each member node of the cluster for data transfer.
Simulation results and discussion
In the first scenario of WSN # 1, a network of sensor nodes ranging from 100 to 700 is randomly distributed in an area of 150∗ 150 m2 with the BS located at the optimal position calculated by the proposed estimation of the energy equations [7.2] within the network field, while in the second scenario, WSN # 2 BS will be placed at a position (100m, 200m) outside the network field. Beecluster delivers about 100% packets to 100 nodes with the BS located at the optimal position in WSN # 1 scenario. Also in WSN # 2 scenario, Fig. [7] shows that Beecluster has the highest PDR compared to BeeSensor, MRP, and ERP at nodes 100 to 700.
It is further analyzed from above Fig.[6] [7] that the PDR remains the highest, in the optimal position of the BS, i.e. in the WSN. 1 scenario which clearly shows the importance of placing the BS in the optimal position, since manual intervention is possible for the BS position during the placement of nodes in a WSN. The low performance of BeeSensor and ERP is due to the fact that they are non-cluster-based protocols, hence the lack of performance when sending data from nodes to the BS.
1 Fig.[8], witness the highest throughput compared to WSN # 2 Fig.[9], so placing BS at optimal position will increase data delivery per second in WSNs. Fig. [10] shows that in scenario WSN # 1, energy consumption of the proposed protocol is approximately less than BeeSensor, MRP and ERP protocols, respectively, which is attributed to the use of compact student-t distribution and improved solution search comparison to select optimal CHs, thus energy consumption in the network. Moreover, placement of BS in an optimal location will reduce the energy consumption, as shown in Fig.
Further Fig.[12] shows the average percentage reduction in energy consumption, when BS is placed at optimal position in a square field with nodes ranging from 0 to 1000. The energy thus saved will extend the network lifetime and the nodes will provide data for a longer duration can transfer . It is clearly visible that Beecluster delivers data packets with minimum latency in both the scenarios among other protocols which ultimately increases the reliability of the network.
In WSN #1, the average latency decreases sharply with increase in number of rounds in Beecluster, which is due to the fact that the proposed protocol delivers data packets to the BS with minimum relay after calculating the optimal possible distance for the next hop, moreover the CHs are placed at optimal distance to BS, thus maintaining a trade-off between transmission distance and hop count. In BeeSensor and MRP, data will be transmitted to BS using the maximum number of hop counts which eventually exhausts the network with unnecessary end-to-end delay.
Conclusions
In MRP, due to asymmetric data forwarding effects on the CHs, those close to the BS will die quickly, reducing the network lifetime. ERP has the smallest network lifetime among all its peers, due to the absence of a clear data aggregation and communication framework, especially for WSN # 2-like scenarios. It is further analyzed that every 1% increase in network lifetime for the proposed protocol will increase data delivery by 2.2%, thus increasing network robustness.
Even in WSN #2, when the BS is located at a large distance from the sensor nodes, the proposed protocol will be able to successfully deliver data packets with minimum delay. Finally, we compare the performance of the proposed protocol with other protocols to demonstrate its validity against various performance metrics. The significance of the optimal location of the BS is also analyzed by the percentage of energy consumption reduction at different distances with the number of sensor nodes.
Further, the proposed protocol should be implemented in the real testbed scenario of sensor nodes, which are set to work in a real-world application framework to judge its performance.
Acknowledgement
Fahmy, Heed: A hybrid, low-power, distributed clustering approach for ad hoc sensor networks, IEEE Transactions on Mobile Computing. Hong, Peach: Energy-efficient and adaptive clustering hierarchy protocol for wireless sensor networks, computer communications. Yang, Eemc: An energy-efficient multi-level clustering algorithm for large-scale wireless sensor networks, computer networks.
Xu, A multipath routing protocol based on clustering and ant colony optimization for wireless sensor networks, Sensors. Shen, Mobility-based clustering protocol for wireless sensor networks with mobile nodes, Wireless Sensor Systems, IET. Guan, A distributed energy-efficient clustering algorithm with improved coverage in wireless sensor networks, Future Generation Computer Systems.
Khalil, A novel evolution-based routing protocol for clustered heterogeneous wireless sensor networks, Applied Soft Computing (2012). Panda, A real-time implementation of a harmony search algorithm-based clustering protocol for energy-efficient wireless sensor networks, IEEE Transactions On Industrial Informatics (2014). Farooq, Beesensor: A bee-inspired power-aware routing protocol for wireless sensor networks, in: Applications of Evolutionary Computing, Springer, 2012, p.
Jana, Energy-efficient clustering and routing algorithms for wireless sensor networks: particle swarm optimization approach, engineering applications of artificial intelligence. Gaddafi Abdul-Salaam, A comparative analysis of energy saving approaches in hybrid wireless sensor networks, data collection protocols.,Telecommunication systems. Vishal Kumar Arora, Vishal Sharma, A study on leaching and other routing protocols in wireless sensor networks, Optik - International Journal for Light and Electron Optics (2016).
Preetha Thulasiraman, Topology Control of Tactical Wireless Sensor Networks Using Energy Efficient Zone Routing, Digital Communications and Networks 2 (2016) 1Ű14. Yongsheng Ding, A multi-path routing algorithm with dynamic immune clustering for event-driven wireless sensor networks, Neurocomputing (2016).
