Contents lists available at ScienceDirect Journal of Network and Computer Applications journal homepage: www.elsevier.com/locate/jnca Review Cache-based transport protocols in wireless sensor networks: A survey and future directions Melchizedek Alipio a, ⁎ , Nestor Michael Tiglao a , Antonio Grilo b , Fawaz Bokhari c , Umair Chaudhry c , Shavez Qureshi c a Ubiquitous Computing Laboratory, Electrical and Electronics Engineering Institute, University of the Philippines Diliman, Philippines b INESC-ID/Instituto Superior Técnico, Technical University of Lisbon, Portugal c Punjab University College of Information Technology (P.U.C.I.T), University of the Punjab, Lahore, Pakistan ARTICLE INFO Keywords: Caching Intermediate nodes Internet of Things Machine type communications Wireless sensor networks ABSTRACT Sensor nodes in Wireless Sensor Networks (WSNs) are battery-powered devices that consume energy during data transmission and processing. One of the most critical tasks in a sensor network is dealing with reliable end- to-end transmissions and optimizing the power consumption. WSNs are error prone due to the constrained nature of the nodes and the interference with other wireless technologies such as Wifi and Bluetooth. One possible way to minimize this problem is caching the data. Data caching is one technique of improving the performance of a transport protocol. Since a typical data transmission consumes more energy than processing in a sensor network, the use of caching enables quick access to data. Therefore, caching, if used efficiently, could reduce overall network traffic and hence bandwidth can be optimally utilized. Unfortunately, a systematic analysis of caching in WSNs was until now lacking. Given that WSNs are expected to play an important role in Machine Type Communications (MTC) and Internet of Things (IoT), the authors believe that now is time to collect the results of years of research on this important topic. This paper presents a comprehensive survey on the state-of-the-art cache-based transport protocols in wireless sensor networks. We classify the transport protocols by presenting a thematic taxonomy of the current cache management mechanisms in wireless sensor networks. Moreover, the critical aspects of the existing cache-ware schemes in sensor networks are analyzed to determine the strengths and weaknesses of such protocols. The similarities and differences of the transport protocols based on the important parameters, such as cache insertion/replacement policy, cache size requirement, cache location, cache partition, and cache decision are investigated in this paper. In addition to that, we discuss open research issues and challenges of cache-based transport protocols in wireless sensor networks. We strongly believe that this study can serve as a basis so that future implementations can choose the combination of caching mechanisms that best fits their target application scenario. 1. Introduction Wireless sensor networks (WSNs) are increasingly used in a range of different applications from home and environmental monitoring to industrial automation. Such constrained networks will be an integral part of Machine Type Communications (MTC) and Internet of Things (IoT). A typical WSN consists of tiny nodes that are equipped with embedded computing devices interfacing with sensors or actuators. A sizable set of these nodes is dispersed over a wide geographical area to monitor a physical or environmental event. Therefore, packets gener- ated at source nodes that are outside the communication range of the sink have to be reached through by other intermediate nodes. Since a wide variety of applications running over WSNs need the optimization of communications among the sensors, the network shall experience small latency and less energy consumption (Akyildiz and Vuran, 2010). A reliable transport protocol is required for such networks. The traditional design of reliable transport protocol is the end-to-end delivery of data from source nodes to sink nodes. However these transport protocols should be modified in order to adapt to the ad hoc wireless and constrained network environment such as WSNs. This is the reason why a lot of new mechanisms have been developed over the past years to improve the transport layer in terms of packet loss detection and recovery of lost segments. This can be performed either end-to-end, where the endpoints (source and sink nodes) are respon- http://dx.doi.org/10.1016/j.jnca.2017.04.001 Received 30 October 2016; Received in revised form 25 February 2017; Accepted 4 April 2017 ⁎ Corresponding author. E-mail addresses: mialipio@upd.edu.ph (M. Alipio), nestor@eee.upd.edu.ph (N.M. Tiglao), antonio.grilo@inov.pt (A. Grilo), fawaz@pucit.edu.pk (F. Bokhari), mscsf13m034@pucit.edu.pk (U. Chaudhry), mscsf14m009@pucit.edu.pk (S. Qureshi). Journal of Network and Computer Applications 88 (2017) 29–49 Available online 07 April 2017 1084-8045/ © 2017 Elsevier Ltd. All rights reserved. MARK