Contents lists available at ScienceDirect Computer Communications journal homepage: www.elsevier.com/locate/comcom Distributed coverage hole detection and recovery scheme for heterogeneous wireless sensor networks Ahmed M. Khedr ⁎ ,a,c , Walid Osamy b,d , Ahmed Salim c,d a Computer Science Department, University of Sharjah, Sharjah, United Arab Emirates b Computer Science Department, Faculty of Computer and Information, Benha University, Egypt c Mathematics Department, Faculty of Science, Zagazig University, Egypt d Qassim University, Buridah 51931, Saudi Arabia ARTICLE INFO Keywords: Distributed algorithm Heterogeneous wireless sensor networks Hole detection Recovery scheme ABSTRACT In Wireless sensor networks (WSNs), the coverage problem is a fundamental issue. The coverage holes are generally caused by both, random deployment of the sensor nodes and node failures, and are hardly avoided in WSNs. It is impossible to recharge or replace the battery, and therefore, the collaborative detection and pre- diction of coverage holes and recovery of these holes has strategic importance in WSNs specially in hetero- geneous WSNs (HWSNs). In this paper, a new scheme is proposed which includes distributed algorithms for detecting holes caused by both random deployment and node failures in HWSNs, where nodes can collaborate to detect and predict the coverage holes. This scheme also presents the new coverage holes healing algorithm, which can efficiently leverage mobility to optimize the average coverage rate and the average movement dis- tance of the mobile nodes. The performance results of our scheme exceeds the existing algorithms in terms of coverage and energy saving. 1. Introduction WSNs have the potential to provide unique capabilities for mon- itoring battle field and environment through the large number of ran- domly deployed sensor nodes [2–6]. However, the energy of a sensor node is highly limited and replacing the battery of hundreds or thou- sands of sensor nodes is impossible; as a result, the energy efficiency is a major and challenging physical problem. Sensors vary in architecture and functionalities. In a WSN, if all sensors have identical characteristics it is called a homogeneous WSN. But different types of sensors may co-exist in a network. The sensor nodes are designed differently based on their applications. If the sensor nodes of a network are different in their attributes we refer to the network as heterogeneous WSN (Fig. 1). In comparison with homo- geneous WSNs, where all the devices possess the same communication and computing capabilities, HWSNs allow for a variety of operating environments, and hence HWSNs are useful in many civil or military applications. One example of HWSNs is the underwater WSNs, which usually consist of multiple types of sensor nodes deployed at different depths within the water as well as diverse onshore, surface, and un- derwater data sinks. Being equipped with different functionalities, these heterogeneous wireless sensor devices collaboratively form a scalable multipurpose WSN and facilitate many sensing tasks, such as oceano- graphic data collection, pollution monitoring, offshore exploration and tactical surveillance. WSNs can be deployed into a specific sensing field either in a pre-planned manner or randomly to monitor and collect data from the sensing field using either static sensors, mobile sensors, or hybrid (static and mobile) sensors. Once a WSN is deployed, sensor nodes can stop working at any time due to several reasons such as, depletion of energy, errors in the code or hardware failures which will cause coverage holes. Additionally, formation of coverage holes cannot be avoided in WSNs [25] and they will progressively increase in number and size if not dealt with on time [28]. The holes can cause network disconnection, disruption in transmission of data, and uncover the network. As a consequence, significant events may not be detected or incorrect information may be gathered due to lack of coverage. Given that, avoidance of coverage holes is nearly unfeasible, me- chanism that efficiently detects and repairs them immediately is of practical importance. Although coverage hole that occurs at preliminary deployment time has been under vigorous research [12,20–26], there is very little work done regarding the coverage hole problem that occurs during the WSN operation specially in heterogeneous WSNs. Existing approaches tend to be complex, involve the sink node, and presume the availability of https://doi.org/10.1016/j.comcom.2018.04.002 Received 5 November 2017; Received in revised form 28 March 2018; Accepted 7 April 2018 ⁎ Corresponding author at: Computer Science Department, University of Sharjah, Sharjah, United Arab Emirates. E-mail addresses: akhedr@sharjah.ac.ae, amkhedr@zu.edu.eg (A.M. Khedr), walid.osamy@fci.bu.edu.eg, w.elsherif@qu.edu.sa (W. Osamy), asalim@zu.edu.eg, a.salem@qu.edu.sa (A. Salim). Computer Communications 124 (2018) 61–75 Available online 14 April 2018 0140-3664/ © 2018 Elsevier B.V. All rights reserved. T