Satvir Singh

The Firefly Optimization Algorithm: Convergence Analysis and Parameter Selection

bio-inspired optimization techniques have obtained great attention in recent years due to its robustness, simplicity and efficiency to solve complex optimization problems. The firefly Optimization (FA or FFA) algorithm is an optimization method with these features. The algorithm is inspired by the flashing behavior of fireflies. In the algorithm, randomly generated solutions will be considered as fireflies, and brightness is assigned depending on their performance on the objective function. The algorithm is analyzed on basis of performance and success rate using five standard benchmark functions by which guidelines of parameter selection are derived. The tradeoff between exploration and exploitation is illustrated and discussed.

A conceptual comparison of firefly algorithm, bat algorithm and cuckoo search

There are various mathematical optimization problems that can be effectively solved by metaheuristic algorithms. The advantage of these algorithms is that they perform iterative search processes which efficiently perform exploration and exploitation in the domain space containing local and global optima. In this context, three types of metaheuristic algorithms called firefly algorithm, bat algorithm and cuckoo search algorithm were used to find optimal solutions. Firefly is inspired by behavior of flies, bat algorithm is based on the echolocation behavior of bats while in cuckoo search, a pattern corresponds to a nest and similarly each individual attribute of the pattern corresponds to a cuckoo-egg. A series of computational experiments using each algorithm were conducted. Experimental results were analyzed and it is observed that firefly algorithm seems to perform better than bat algorithm and cuckoo search.

Energy efficient clustering protocol based on improved metaheuristic in wireless sensor networks

Energy efficient clustering is a well accepted NP-hard optimization problem in Wireless sensor networks (WSNs). Diverse paradigm of Computational intelligence (CI) including Evolutionary algorithms (EAs), Reinforcement learning (RL), Artificial immune systems (AIS), and more recently, Artificial bee colony (ABC) metaheuristic have been used for energy efficient clustering in WSNs. Due to ease of use and adaptive nature, ABC arose much interest over other population-based metaheuristics for solving optimization problems in WSNs. However, its search equation, which is comparably poor at exploitation and require storage of certain control parameters, contributes to its insufficiency. Thus, we present an improved Artificial bee colony (iABC) metaheuristic with an improved solution search equation to improve its exploitation capabilities. Additionally, in order to increase the global convergence of the proposed metaheuristic, an improved population sampling technique is introduced through Studentst distribution, which require only one control parameter to compute and store, hence increase efficiency of proposed metaheuristic. The proposed metaheuristic maintain a good balance between exploration and exploitation search abilities with least memory requirements, moreover the use of first of its kind compact Studentst distribution, make it suitable for limited hardware requirements of WSNs. Further, an energy efficient clustering protocol based on iABC metaheuristic is introduced, which inherit the capabilities of the proposed metaheuristic to obtain optimal cluster heads (CHs) and improve energy efficiency in WSNs. Simulation results shows that the proposed clustering protocol outperforms other well known protocols on the basis of packet delivery, throughput, energy consumption, network lifetime and latency as performance metric.

An improved butterfly optimization algorithm with chaos

Butterfly Optimization Algorithm (BOA) is a new comer in the category of nature inspired metaheuristic algorithms, inspired from food foraging behavior of the butterflies. Similar to other metaheuristic algorithms, it encounters two probable problems; (1) entrapment in local optima and (2) slow convergence speed. Chaotic maps are one of the best methods to improve the performance of metaheuristic algorithms. In the present study, chaos is introduced into BOA which increases its performance in terms of both local optima avoidance and convergence speed. Ten chaotic maps are employed to enhance the performance of the BOA. The proposed chaotic BOAs are validated on unimodal and multimodal benchmark test functions as well as on engineering design problems. The results indicate that the chaotic maps are able to significantly boost the performance of BOA.

Mutated firefly algorithm

In the standard firefly algorithm, every firefly has same parameter settings and its value changes from iteration to iteration. The solutions keeps on changing as the optima are approaching which results that it may fall into local optimum. Furthermore, the underlying strength of the algorithm lies in the attractiveness of less brighter firefly towards the brighter firefly which has an impact on the convergence speed and precision. So to avoid the algorithm to fall into local optimum and reduce the impact of maximum of iteration, a mutated firefly algorithm is proposed in this paper. The proposed algorithm is based on monitoring the movement of fireflies by using different probability for each firefly and then perform mutation on each firefly according to its probability. Simulations are performed to show the performance of proposed algorithm with standard firefly algorithm, based on ten standard benchmark functions. The results reveals that proposed algorithm improves the convergence speed, accurateness and prevent the premature convergence.

Energy-Efficient Hierarchical Routing for Wireless Sensor Networks: A Swarm Intelligence Approach

Energy efficient routing in wireless sensor networks (WSNs) require non-conventional paradigm for design and development of power aware protocols. Swarm intelligence (SI) based metaheuristic can be applied for optimal routing of data, in an energy constraint WSNs environment. In this paper, we present BeeSwarm, a SI based energy-efficient hierarchical routing protocol for WSNs. Our protocol consists of three phases: (1) Set-up phase-BeeCluster, (2) Route discovery phase-BeeSearch and (3) Data transmission phase-BeeCarrier. Integration of three phases for clustering, data routing and transmission, is the key aspect of our proposed protocol, which ultimately contributes to its robustness. Evaluation of simulation results show that BeeSwarm perform better in terms of packet delivery, energy consumption and throughput with increased network life compared to other SI based hierarchical routing protocols.

An Effective Hybrid Butterfly Optimization Algorithm with Artificial Bee Colony for Numerical Optimization

In this paper, a new hybrid optimization algorithm which combines the standard Butterfly Optimization Algorithm (BOA) with Artificial Bee Colony (ABC) algorithm is proposed. The proposed algorithm used the advantages of both the algorithms in order to balance the trade-off between exploration and exploitation. Experiments have been conducted on the proposed algorithm using ten benchmark problems having a broad range of dimensions and diverse complexities. The simulation results demonstrate that the convergence speed and accuracy of the proposed algorithm in finding optimal solutions is significantly better than BOA and ABC.

Artificial bee colony metaheuristic for energy-efficient clustering and routing in wireless sensor networks

Swarm intelligence (SI)-based metaheuristics are well applied to solve real-time optimization problems of efficient node clustering and energy-aware data routing in wireless sensor networks. This paper presents another superior approach for these optimization problems based on an artificial bee colony metaheuristic. The proposed clustering algorithm presents an efficient cluster formation mechanism with improved cluster head selection criteria based on a multi-objective fitness function, whereas the routing algorithm is devised to consume minimum energy with least hop-count for data transmission. Extensive evaluation and comparison of the proposed approach with existing well-known SI-based algorithms demonstrate its superiority over others in terms of packet delivery ratio, average energy consumed, average throughput and network life.

Butterfly algorithm with Lèvy Flights for global optimization

In recent years, various nature-inspired algorithms have been developed for solving complex real world problems with multiple constraints. These algorithms have shown remarkable performance which is the main reason of their preference over other conventional optimization algorithms. In this paper, a novel nature-inspired optimization algorithm based on the food foraging of butterflies is proposed. In the proposed algorithm, the butterflies act as search agents for finding food by sensing different fragrances from various flowers. The proposed algorithm is validated against various benchmark functions and then compared with some well-known nature-inspired algorithms. The results confirm superior performance of the proposed algorithm over other algorithms in solving global optimization problems.

Stochastic Algorithms for 3D Node Localization in Anisotropic Wireless Sensor Networks

This paper proposes two range based 3D node localization algorithms using application of Hybrid Particle Swarm Optimization (HPSO) and Biogeography Based Optimization (BBO) for anisotropic Wireless Sensor Networks (WSNs). Target nodes and anchor nodes are randomly deployed with constraints over three layer boundaries. The anchor nodes are randomly distributed over top layer only and target nodes over middle and bottom layers. Radio irregularity factor, i.e., an anisotropic property of propagation media and an heterogenous property (different battery backup statuses) of devices are considered. PSO models provide fast but less mature convergence whereas the proposed HPSO algorithm provides fast and mature convergence. Biogeography is based upon the collective learning of geographical allotment of biological organisms. BBO has a new comprehensive energy based on the science of biogeography and apply migration operator to share selective information between different habitats, i.e., problem solutions. Due to size and complexity of WSN, localization problem is articulated as an NP-hard optimization problem . In this work, an error model in a highly noisy environment is depicted for estimation of optimal node location to minimize the location error using HPSO and BBO algorithms. The simulation results establish the strength of the proposed algorithms by equating the performance in terms of the number of target nodes localized with accuracy, and computation time. It has been observed that existing sensor networks localization algorithms are not significant to support the rescue operations involving human lives. Proposed algorithms are beneficial for rescue operations too to find out the accurate location of target nodes in highly noisy environment.