1. Please check and comment entries here.
Table of Contents

    Topic review

    Green Wireless Sensor Networks

    Definition

    The issue of energy balancing in Wireless Sensor Networks is a pivotal one, crucial in their deployment. This problem can be subdivided in three areas: (i) energy conservation techniques, usually implying minimizing the cost of communication at the nodes since it is known that the radio is the biggest consumer of the available energy; (ii) energy-harvesting techniques, converting energy from not full-time available environmental sources and usually storing it; and (iii) energy transfer techniques, sharing energy resources from one node (either specialized or not) to another one.

    1. Introduction

    Wireless Sensor Networks (WSN) is one of the most active research areas with applications in numerous fields such as transport, health, military, agriculture, environment monitoring and control, etc. Within this research area, the problem of energy balancing in WSNs (harvesting, transfer and conservation) has been always the one that has attracted the most interest.
    A WSN is a set of sensor nodes wirelessly interconnected with at least one central node, called base station (BS) or sink node. The base station can both control the network in a centralized way and communicate with end users or/and other networks.
    A sensor node is an electronic device that essentially has four main components: a sensing unit, a processing unit, a communications unit and a power unit. Other equipment such as a mobility unit or a position tracking unit are optional. The sensing unit allows the node to collect data related to its ambient conditions. These data are handled by the processing unit, and its communication unit is used to exchange data with other nodes and the base station. The power unit provides the necessary energy to the node and it is typically a chemical battery.
    A WSN may contain one or more BSs together with a lot of sensor nodes, static or mobile. Using many nodes enables the simultaneous acquisition of data related with the environment conditions in very wide areas. This makes WSNs ideal for an increasingly wide range of tasks such as fire detection, weather forecasting, energy management, biomedical applications, environmental and habitat monitoring, surveillance and reconnaissance, home automation, object tracking, traffic control, inventory control, agriculture, machinery failure diagnosis and various military applications.
    In a wide sense, the objective of a WSN is to feed some application with data dynamically generated by sensor sources distributed in a certain region and, as a result of the analysis of such information, other devices may be commanded remotely for some action.
    Usually, sensor nodes have a limited power source. Since in many cases the nodes are deployed in unreachable areas, it is not possible to recharge or replace the battery of the sensor nodes. Therefore, it is necessary to find appropriate techniques to save energy by reducing the power consumption of the nodes and so improving the lifetime of the WSN. It is widely recognized that sensors use more energy during communications than during sensing and preprocessing the data they must send. Thus, the radio is the main problem when it comes to extending the life of devices that depend on batteries to perform their mission and, therefore, the communications unit, which involves both transmission and reception of data, uses a significantly higher proportion of the available energy [1].
    There are several aspects of a WSN dealing with sensible energy strategies: conservation, harvesting and transfer.
    Energy conservation
    techniques simply aim to extend the lifetime of the network by reducing the energy used while the WSN continues to operate as required [2]. Energy saving usually implies minimizing the cost of communication at the nodes since it is known that the radio is the biggest consumer of the available energy [3]. Different energy-saving schemes were recently surveyed in [4][5].
    Energy-harvesting 
    techniques seek to increase the energy available to the nodes. The energy can be obtained from the external environment such as solar, wind, vibrations, radio frequency, thermal, etc. These techniques convert energy from the environment into electrical energy for the nodes. In [6] it is presented a recent survey. A major constraint of these techniques is that energy sources are not always accessible and therefore it is necessary to store the harvested energy using rechargeable batteries or low-powered supercapacitors.
    Energy transfer
    is another emerging technique applied to extend the lifetime of the network. The idea is that energy-rich nodes transfer energy to energy-deficient nodes. This transfer can be done wirelessly from a node specialized in energy harvesting or a node with enough energy resources to a node in need of energy in the same network. Engmann et al. [7] review several mechanisms for energy transfer such as inductive and magnetic coupling, or electromagnetic radiation.
    Usually, many energy balance approaches involve using a combination of several schemes to address the efficient use of energy of the WSN [7].
    This paper aims to survey the main contributions of these three types of techniques for extending the lifetime of WSNs (conservation, harvesting and energy transfer), also identifying the main research topics that are trending presently. There are many surveys addressing one of these three categories, but there are only a few surveys that covers all the types, one published in 2017 by Yetgin et al. [8] and another one published in 2018 by Engmann et al. [7]. In this survey, we will intend to classify and summarize the different techniques proposed in the literature, since at least 2018, for conservation, harvesting and transfer of energy in WSNs, focusing on the most recent surveys and, mainly, trying to identify the trending research topics and the most active areas at present, commenting the most relevant and interesting works in each case. The bibliographical revision was closed on November 2020.

    2. Energy Conservation

    Energy conservation methods try to reduce energy consumption of the sensor nodes to extend the lifetime of the nodes and, therefore, the lifetime of the network. Since radio communication is the main consumer of energy in a node, the design of communications protocols (MAC and routing software) is a key element. In the past, without rechargeable batteries, prolonging the life of the network was the major focus in the communication unit design, leaving network latency and throughput optimization at a second place. Since energy-harvesting capabilities has been added to nodes, Quality of Service (QoS) has come back to be its main focus, though energy constraints remain. Energy-Neutral Operation (ENO) arises as a new paradigm, and achieving the best performance of the application/network (in terms of throughput and time response) balancing the energy expenditure caused by sensor activity with the energy flow being able to harvest has attached attention in recent times.
    There is abundant literature relating to energy-saving in WSNs in recent years that have been collected and classified in different works. The survey in [9] uses a simple classification based on the network structure (flat or hierarchical) for data aggregation and routing schemes. This year two surveys have been published. In [4] energy-saving schemes such as duty-cycling, efficient routing, efficient MAC, data aggregation, cross-layer design and error control code are addressed. This survey includes works until 2019 and it is noticeable that the most active year was 2017.
    Another recent review by Singh et al. [5] uses an uncommon classification, considering battery management schemes (nodal management and energy balancing), transmission power-management schemes (MAC protocols, routing protocols and transmission policy), system power-management schemes (processor power and device management) and other miscellaneous schemes like load balancing, duty-cycling or cross-layer optimization. In this survey we decide to use the four main categories proposed in [10] (see Figure 1). In every category we explain the fundamentals of each type of solution, we consider some subcategories and when we find a particularly active area of research in the last two years, we review the most relevant works found in the literature.
    Figure 1. Taxonomy of energy-efficient mechanisms proposed in [10].

    2.1. Radio Optimization

    Since the radio module is the main component that causes the node battery to be exhausted, many researchers have addressed several mechanisms in wireless communications such as optimizing modulation or coding, cooperative communications, power transmission control, directional antennas or cognitive radio.

    2.1.1. Modulation Optimization

    The goal of modulation optimization is to find the optimal modulation parameters to minimize the energy consumed by the radio. Existing research studies try to find a good trade-off between the information rate, the transmission delay, the constellation size, the distance between the nodes and the signal-to-noise ratio. Cui et al. [11] showed that to minimize the total energy consumption, for a given BER and delay requirement, the transmission time needs to be optimized, showing that it is possible to achieve up to 80% energy savings. The literature provides some evidence that low-order modulations such as BPSK are suboptimal for short transmission distances. In [12] it is shown that each modulation scheme has a single optimal SNR at which the energy consumption is minimized. The optimal SNR and the minimal energy consumption are larger for higher values of BER. Therefore, they found that operated at its optimal SNR, BPSK and QPSK are the optimal choices for long transmission distances, but as the transmission distance shortens, the optimal modulation size grows to 16-QAM and even to 64-QAM.
    In recent years little interest has been shown in this research topic and only the work in [13] is noteworthy, where DMS (Dynamic Modulation Scaling), a technique which manages the constellation size to change the transmission time and energy consumption, is used in combination with topology control (see Section 2.3.4).

    2.1.2. Cooperative Communications

    Cooperative communications strategies aim to enhance the quality of the received signal by operating several single antennas that work together to form a virtual multi-antenna transmitter, taking advantage of the fact that data are often overheard by neighboring nodes because of the broadcast nature of wireless networks, therefore gaining the benefits of multiple-input multiple-output (MIMO) systems while overcoming their challenges [14][15].
    On the other hand, although a MIMO scheme is an effective way for reducing energy consumption in WSNs, the physical size and the energy available for the sensor nodes is limited, and therefore, installing multiple antennas at the sensor nodes is not ideal. Thus, cooperative MIMO (CMIMO) can be achieved by orchestrating collaboration among the single antennas installed on each sensor node. Therefore, spatial multiplexing can enhance the data rates, whereas spatial diversity can improve the bit-error rate (BER) using the CMIMO technique. Moreover, CMIMO minimizes the energy consumed in transmission, especially in long range transmission [16].
    CMIMO is an area where recent research activity can be found in recent years. In [17], a novel protocol is proposed in which mobile terminals form a virtual MIMO uplink by means of device relaying on Device to Device (D2D) tier in 5G Cellular Networks. Its focus is to design an incentive for terminals to form the virtual MIMO and cooperate in relaying other data.
    The work in [18] addresses how to extend the average battery capacity among the whole network through the selection of the cooperative coalition for CMIMO, proposing to apply the quantum-inspired particle swarm optimization (QPSO) algorithm to select the optimum cooperative coalitions of each hop in the routing path.
    CMIMO scheme reduces the energy consumption of sensor nodes quite effectively by using the space-time block coding scheme. However, in [19] it is shown that in networks with high node density, the scheme is ineffective due to the high degree of correlated data. Therefore, to enhance the energy efficiency in these cases, the authors propose to use the distributed source coding (DSC) with the virtual MIMO data transmission technique. The DSC-MIMO first compresses redundant source data using the DSC and then sends it to a virtual MIMO link. The results reveal that energy consumption is lower than that in the CMIMO technique.
    Aside from the pure CMIMO technique, CMIMO with spatial modulation (CMIMO-SM) [20][21] is another transmission approach recently proposed to achieve the spatial diversity of MIMO systems. The basic idea of CMIMO-SM is to give each individual antenna a preassigned index and let them cooperate with each other in a cluster to form a cooperative frame.
    In [22], applications of CMIMO are proposed in networks of intelligent transportation systems (ITS) for reducing the total energy consumption. The scheme proposed is based on CMIMO-SM and its detailed energy consumption is compared with the traditional single-input-single-output (SISO) scheme.

    2.1.3. Transmission Power Control

    Transmission Power Control (TPC) has been proposed in the literature to improve energy efficiency by adjusting the transmission power of the radio [23]. However, the TPC mechanism has an effect not only on energy but also on delays, interference and connectivity. In fact, when the transmission power is reduced, the risk of interference also diminishes. Conversely, the delay increases, because more hops will be needed to forward a packet. TPC has also been an active research area in recent years, essentially in the scope of Wireless Body Area Networks (WBANs) with medical applications, as shown in these two recent surveys [24][25]. Important work in this area is due to Sohdro et al. [26][27][28].
    The aim of TPC mechanisms in WBANs, which can be combined with other energy-saving mechanisms [29][30][31], is to reduce the energy consumption and external interferences by dynamically adjusting the transmission power output, with the minimum effect on other performance aspects, such as reliability and latency. Other recent works in this field are adaptive algorithms based on variations in body conditions [32] or on human motion [33].
    More recently, Qolami et al. [34] investigated TPC in 802.15.4 + RPL WSNs and proposed a mechanism where each node dynamically adjusts its transmission power based on channel conditions before sending every data and ACK packet. Their results confirm that this power control method improves network lifetime. A TPC mechanism is also proposed for Underwater Wireless Sensor Networks (UWSNs) in [35], where source nodes adjust transmission power according to the location of the destination node.

    2.1.4. Directional Antennas

    Directional antennas are used to allow signals to be sent and received in more than one direction simultaneously, thus enhancing transmission range and throughput. Directional antennas may require positioning techniques to be oriented, but multiple communications can take place in close proximity, which results in spatial reuse of bandwidth. Many researchers have proposed the use of directional antennas to improve the energy efficiency of the network [36].
    Given that omnidirectional antennas transmit a signal equally to all directions and, in most of the applications, communication is unicast, we can infer that significant amount of energy can be saved using directional antennas. In [37] the authors study the effect of directional antennas on energy efficiency of the IEEE 802.15.4 standard in outdoor WSNs. In [38] we can read a review of energy efficiency using directional antennas in mobile ad hoc networks and [39] is a very recent review of directional antennas for WSN applications, where a little section is dedicated to contention reduction and energy efficiency.
    The use of directional antennas can also reduce the interference between radio streams and improve the SNR. Therefore, directional antennas can improve the reliability and reduce the number of retransmissions. However, use of directional antennas necessitates to find the right direction and parameters correctly and quickly. A recent work in this area [40] presents an energy-efficient localization of sensor nodes in WSNs using beacons from a rotating directional antenna.
    Finally, we must note that the most active research area in recent years in order to take advantage of the properties of directional antennas is the proposal of new MAC protocols. The work in [41] contains a review of this topic but there are interesting subsequent works [42][43][44][45][46][47].

    2.1.5. Energy-Efficient Cognitive Radio

    A cognitive radio (CR) is an intelligent radio that can dynamically choose a channel in the wireless spectrum and can set its transmission and reception parameters according to the channel status. The subjacent Software-Defined Radio (SDR) technology is intended to build fully reconfigurable wireless transceivers that automatically tune their communication parameters to the requirements of the network, resulting in improved context awareness. Nevertheless, SDR requires substantial power usage in comparison to conventional devices because of the greater complexity inherent to the sophisticated new functionalities. In this scenario, the design of energy-efficient CR sensor networks is a key challenge for the intelligent use of battery power.
    Recent cognitive radio studies are interested in the development of cross-layer approaches for MAC [48][49][50] and routing [51][52]. The field of clustering protocols have been especially active. For example, the work in [53] surveys 12 papers until 2017 that use clustering topologies in CR sensor networks with a focus on energy consumption. Several subsequent works also use CR together with clustering protocols. In [54] the Learning Automata-based Multilevel Heterogeneous Routing (LA-MHR) scheme for WSNs is proposed, and [55] presents the novel approach to integrate the sensor nodes with CR nodes to forward data towards the sink using opportunistically licensed channels.

    2.2. Data Reduction

    It is clear that another way to save energy is to reduce the volume of data to be transmitted to the sink. Several methods can be used for data reduction: aggregation, compression and prediction.
    It is known that the data collected at the sensors contains spatial correlation, especially in some types of measurements when the sensors are close together. If these redundant data are sent to the BS, then it causes a waste of bandwidth and an increase in the energy consumed by the nodes.
    Given that the spatial correlation is stronger among data from nodes located close to each other, Heinzelman et al. [56] proposed the Low Energy Adaptive Clustering Hierarchy (LEACH) as organization of the network that consists of grouping neighboring nodes in clusters, one of them acting as Cluster Head (CH). The nodes within the cluster send data to the CH, eventually after local processing at the node. Then, correlated data received from these nodes can be processed for data reduction at the CH before being routed towards the BS. In Section 2.4.1, we address hierarchical routing and LEACH in more detail.

    2.2.1. Data Aggregation

    The main idea under data aggregation techniques is to remove redundancies in the received data from the neighboring nodes, extracting the useful information by means of aggregation functions (maximum, minimum, average, etc.) before forwarding the final data to the BS, CH or any central node [57].
    Several data aggregation schemes have been proposed that, according to the aggregation methodology, can be classified [58] in: (a) centralized (all nodes send data to an energy-rich central node, responsible for aggregation); (b) in-network (aggregation is performed in all the intermediate nodes); (c) tree-based (a minimum spanning tree is built, where root node acts as base station, source nodes are the leaves that send the data to their parent intermediate node) and (d) cluster-based (CH is responsible for aggregating the data received from the cluster nodes and then sending the data to the base station).
    Data aggregation techniques is one of the most researched areas for energy efficiency in WSNs. It is noticeable that only in the last two years several surveys can be found in the literature in its field [59][60][61][62].
    When sensitive data must be aggregated and transmitted, assuring privacy is also an important issue. With this focus on mind, the work in [63] surveys various existing solutions for secure data aggregation, classifying them based on the node topology and mechanisms employed for ensuring privacy.
    Another interesting approach for aggregating data is to exploit machine learning to select efficient cluster heads or for dimensionality reduction of the data at sensor nodes or cluster heads. The recent survey in [64] focuses on the application of machine learning for energy efficiency in WSNs, and devotes a section to data aggregation.

    2.2.2. Data Compression

    Another approach to reduce the amount of data to transmit is to compress these data by minimizing the number of bits required to represent each data block. All the compression algorithms are based on taking advantage of the correlation of collected data. This correlation can be temporal, among the sensed data in each node, or spatial, among the data collected in the neighboring nodes, typically inside a cluster. According to these two types of correlation, data compression mechanisms can be classified in two categories: local compression and distributed compression. In sparse sensor networks, local data compression is a good approach, but in dense networks, distributed data compression should be used [65].
    Distributed Data Compression
    (also called in-network compression) is an asymmetric coding that uses the spatial correlation among the sensor nodes. In this case, the sensor nodes belonging to an area encode its data before sending it to the BS (or CH if clustering is used), where all the correlated flows are decoded jointly.Images and videos are usually compressed using a transform-based compression such as cosine transform or wavelets transform, where a sparse representation of the data is used to recover the original data at the decoding point with minimal loss. However, distributed transform coding (DTC) is difficult to implement in WSNs because they often need the knowledge of all measurements in the network at each node. Therefore, most articles in this field aim to modify DTC algorithms in order to make them suitable to WSNs.
    This problem can be avoided by means of distributed source coding (DSC) [66][67], based on the Slepian-Wolf theorem, since it does not require inter-communication among sensor nodes. The main problem of DSC is that it requires previous knowledge of the correlations in the data, and so its performance depends on specific assumptions.
    In contrast, compressive sensing (or compressed sensing) techniques, a relatively new idea in the field of WSNs, do not need any previous knowledge or assumption on data correlations. In compressed sensing (CS) [68] a small number of samples of a sparse signal contains enough information to successfully recover the original signal with almost no data loss. The great advantages in terms of limitations and data reduction (part of the redundant data is never acquired) make CS the most widely used technique presently in WSNs and IoT [69], even in the case of multimedia sensor networks [70].
    Local Data Compression
    has not the limitations mentioned above and it is a universal and robust compression technique. Huffman coding, LZW (Lempel-Ziv-Welch) and RLE (Run-length encoding) are common compression techniques used in WSN for local data compression. Moreover, it can be used in conjunction with distributed data compression in WSNs to exploit both temporal and spatial data correlations.
    There are several surveys about data compression techniques for energy efficiency in WSNs [71][72]. The survey in [73] focuses on in-network compression techniques, devoting special attention to CS, which is undoubtedly the most active research topic in this area in recent years. In addition, within CS schemes, the focus is currently on its use in conjunction with clustering, where interesting works have been published [74][75][76][77][78][79][80]. More recently we must mention the works in [81][82][83].

    2.2.3. Data Prediction

    An alternative approach to save energy is to predict part of the sensed data without any transmission. Many different prediction-based data reduction mechanisms are proposed in the literature, also using different prediction approaches such as regression, neural networks, or machine learning. In [84] we can find a useful survey that also contains a systematic procedure for selecting a suitable scheme to make predictions. This work classifies the mechanisms in two groups:
    • In the Single Prediction Schemes (SPSs), predictions are made in a single point in the network, either in the cluster head (CH), or in a sensor node when it is more expensive (in energy terms) to obtain a sample than to predict one. This scheme has been applied in conjunction with adaptive sampling [85], topology control [86] and clustering [87][88][89].
    • In Dual Prediction Schemes (DPSs), where clustering is used, the predictions are simultaneously made in CHs and nodes. The idea is that nodes and CH obtain the same prediction, but the node can check the precision of the prediction by comparing this with the real measure, and only send the measure to CH if the prediction is poor. In other case, CH uses the prediction.
    DPSs are potential candidates to optimize the data transmissions because they avoid unnecessary transmissions without affecting the quality of their measurements. Simulation results in [90] show that the number of transmissions can be reduced by almost 98% in the nodes with the highest load.
    Wu et al. [91] propose the combined use of Dual prediction (DP) and Data Compression (DC) schemes in clustered networks, where DP is used for data reduction between nodes and CH by exploiting the temporal correlation and DC is used between the CHs and the BS by exploiting the spatial correlation.
    This combination has received special attention in the literature. Jarwan et al. [92] present a comparative analysis using combinations of different mechanisms of each type. For the DP scheme, the work focuses on neural networks to perform predictions, in comparison to popular LMS approaches. Regarding data compression, principal component analysis, non-negative matrix factorization, truncated-singular value decomposition, and discrete wavelet are discussed and compared. The results show significant reduction in the number of transmissions when using both schemes while preserving the QoS requirements. More recently, the work in [93] proposes a gradient-based adaptive model that also uses a combination of DP and DC schemes.

    2.3. Sleep/Wake-Up Schemes

    Communications in WSN accounts for most of the energy consumed by network nodes, quite higher than that used for sensing and preprocessing the data they must send. Therefore, designing energy-efficient (EE) communication protocols (MAC and routing) is key to the proper functioning of the underlying application.
    In any case, to choose/design the right energy-aware communication protocols, other factors such as topology and the functioning application itself must be taken into account. For instance, in WSNs deployed for critical missions (perimeter surveillance, tsunami/earthquake early detection, fire control, etc.) data must be transmitted immediately, so sensors with new information should have the highest priority to access the channel, avoiding whichever delay and contention to do so. Same way, the limited transmission range and the distance to the sink determine if a sensor must also function as a relay to help others or must lean on other itself.

    2.3.1. Energy-Efficient MAC Protocols

    Due to its position in the protocol stack, MAC entity is ultimately the responsible for introducing and extracting packets in/from the wireless medium; and so, it is the one who commands the radio transceiver, which works in half-duplex regimen switching between transmitting and receiving modes as commanded.
    From the point of view of energy efficiency, MAC protocols for WSN must minimize wasting energy in processes such as idle listening (checking actively packet arrivals), overhearing (receiving packets not intended for this node), collisions (frames lost caused by overlapping transmissions), overtransmitting (transmitting when receiver is not ready) and protocol overhead (transmitting/receiving protocol control packets/bytes). In Energy-Harvesting (EH) aware MAC protocols; however, some of these basic principles may be dynamically relaxed to achieve some kind of QoS if the external energy contribution allows it.
    Almost by consensus, WSN MAC protocols are classified as Contention Free (TDMA/ FDMA/CDMA based), Contention-Based (Synchronous and Asynchronous) and Hybrid schemes (CDMA/TDMA + CSMA). However, from the point of view of energy efficiency the key concepts are Duty-cycling and Wake-up Radios.

    2.3.2. Duty-Cycling

    Contrary to transmission, which is a synchronous activity always fired by a previous event (timer expiration, physical threshold exceeded, packet arrival), reception requires being listening in the channel prematurely (idle listening), wasting as much energy as would be wasted in any of the other two productive states (tx/rx) [3]. So being actively waiting the reception of a packet for a long time would deplete the battery very quickly, and it is not an option. In duty-cycling schemes the radio is switched off to save energy, and is only turned on periodically—during a limited interval—to check if traffic is pending somewhere in the network for this device.
    Depending on the considered solution, packets ready for transmission would either make the transceiver abandon immediately the sleep state or wait for the next on period (duty cycle) to try to send them. To achieve the mandatory rendezvous between transmitter and receiver, as both must be awake simultaneously to exchange information, a wide variety of solutions have been proposed over time, and have been mostly classified as contention-based asynchronous protocols and scheduled-based with contention protocols. In [94] the reader can find a recent and updated review of duty-cycle-based MAC protocols gathering the most significative contributions till 2017, including some especially tuned for energy harvesting.
    In asynchronous protocols duty cycles of different nodes are not synchronized avoiding the technical issues that will imply so precise timing adjustment, and so they are a good option for non-clustered implementations. In these schemes, rendezvous between both ends of the link is the most significant challenge. Transmitter Initiated (TI) and Receiver Initiated (RI) strategies determine a particular intrinsic subclassification, depending on which end triggers the events, with RI alternatives being more energy-efficient. Idle listening and overhearing are their main drawbacks, and though partially mitigated still remain. Nevertheless, the wide range of deployments/applications where synchronization is not only difficult but impossible, maintains the interest of the research community on them.
    Scheduled-based duty-cycled protocols synchronize the beginning of the active periods of the neighboring nodes, with a not negligible cost in protocol overhead and so in energy expenditure. Internally, a leading broadcast beacon precedes a flexible virtual time structure headed by a contention period (CP) where transmitters try to multiplex their packets or organize transmission turns (reservation) in an optional glued contention free period (CFP) that would be so dynamically managed.
    In wireless MAC contention protocols, the lack of a centralized access control mechanism makes collisions an issue to consider, both from the point of view of energy efficiency (useless transmissions) and QoS degradation (higher latency and wasted bandwidth). Backoff intervals are used to try to avoid collisions and to resolve those that have taken place. This has been used in several academic proposals to favor the most depleted devices (raising their priority with shorter backoff intervals) to try to avoid their collision or to shorten their contention resolution stage. CSMA/CA—and its many variants—usually with RTS/CTS packet exchange to avoid hidden terminal issue, is the regular protocol in CP intervals (both in asynchronous and synchronous-based MAC alternatives); and its Network Allocation Vector functionality (NAV) can be used by nodes in CP to tune even more their on and off intervals when a node loses contention or is not the destination of the current transmission.
    Pure TDMA protocols are usually classified outside duty-cycling schemes, though really they are just a special case of contention free scheduled rendezvous where synchronized devices switch from sleep to active state following a fixed—and statically configured in advance—slot pattern, in a repetitive superframe structure. It is the best solution for those applications—industrial, critical mission, WBAN—that need deterministic (guaranteed delay and throughput) and reliable communications. Collisions, overhearing and overtransmissions are completely eliminated, and receivers, looking just in the individually assigned slots, minimize idle listening intervals. Network throughput is maximized if traffic load is high; otherwise, bandwidth is wasted as many slots will be empty but reserved. Hybrid schemes (TDMA-based with some kind of contention) look for implementing dynamic TDMA reservation, or some kind of adaptive behavior, to prevent nodes from waking up in preassigned but finally not used slots, and to adjust the contents (ownership) and length (periodicity) of the superframe structure.
    Needless to say, that synchronized schemes (TDMA-based or scheduled with CP) fit perfectly in clustered architectures were the base station or virtual cluster heads can facilitate to surrounding nodes the necessary stringent time reference. If such special entities are not line powered, its special role will compromise their batteries quicker, so MAC protocols also must contribute with some kind of toggle mechanism to balance this extra consumption.
    Pursuing the ideal where the receivers do not wake-up when there are no transmissions ready for them, and attending to other certainties such as that wake-up points should not be scheduled/placed without considering the characteristics of real network traffic—which can be heterogeneous and unpredictable—or that higher traffic load and smaller delay would benefit from longer duty cycles, the research community continues squeezing MAC protocol internals looking for a better duty cycle and rendezvous tuning. Following, in a very descriptive way, we introduce some of the recent contributions in the immense universe of EE (Energy-Efficient)/EH (Energy-Harvesting) WSN MAC protocols:
    • In ADP-MAC (Adaptive and Dynamic Polling-based MAC), Siddiqui et al. [95] propose modifying the polling interval distribution at the receivers dynamically, every TA
    • , based on the analysis of the coefficient of variation of the incoming traffic; and they resolve that the best results are achieved when the polling interval distribution is the same as that of interarrivals.
    • In AWR-PS-MAC (Adaptive wake-up interval to enhance Receiver-based Preamble Sampling MAC) [96], the transmitter reports on its particular traffic to the receiver—using piggybacking in DATA packets—and with such information the receiver adjust the next time to wake-up for this particular sender, which takes it from the corresponding ACK.
    • In ADMC-MAC (Adaptive MAC for Critical Missions) [97], a synchronous contention-based scheme, the authors design a protocol to adjust dynamically the duty-cycle of neighboring nodes, using a discrete set of values (40%, 20%, 10%), so adapting the cycle to traffic load: in a first stage, neighboring nodes determine which one will be the cluster head (each node broadcasts sync packets with the length of its transmission queue, the number of neighbors detected and its remaining energy, being such order what determines the priority in the election). Next, the selected node, using a single expression—determined after applying a regression technique to data obtained by a simulation study of S_MAC [98], its parent protocol—determines a duty-cycle-factor (a real number), which is then transformed—using some thresholds—in a duty-cycle value of a configured discrete set. Finally, this duty-cycle is broadcasted to the neighborhood and adopted as such.
    • In AP-MAC [99], to reduce the probability of collision among receiver awake intervals in a RI asynchronous scheme, the receiver add a random value to their precalculated next wake-up time before broadcasting it—via beacon frames—to the neighborhood, so transmitters can forecast the next potential rendezvous.
    • In [100], to minimize the latency in a multihop sink rooted tree topology, where data gathering is commanded by a request starting at then sink, Monica et al. [100] design a request-response protocol with extra wakeups generated dynamically to match the predicted arrival time of the response packet, as if the participating nodes were “hit” by (upward and downward) “waves”, and so sequentially awoken according to their depths.
    • In [101], the concept of duty-cycle is eliminated as a repetitive sequence of on-off periods, and is substituted by a slotted vision of time, where, self-adaptively, each node autonomously decides in every slot to sleep or wake-up. The authors claim that this way the trade off between energy-saving and packet delivery delay can be avoided. Unlike other prediction-based approaches where nodes must exchange information between each other, these enables nodes to approximate their neighbors’ situation without requesting information from them, thus saving the large amount of energy usually used for information exchange. In addition, to accomplish it, they propose an alternative approach based on game theory and a reinforcement learning technique (Q learning algorithm), where through trial-and-error interactions within the dynamic environment nodes are able to learn optimal actions.
    • Liu et al. [102] present QTSAC (Quorum Time Slot Adaptive Condensing-based MAC protocol), derived from previous QMAC family protocols [103], for achieving delay minimization and energy efficiency in a synchronized sink driven and rooted topology. As a particular rendezvous scheme, QTSAC condenses the Quorum Time Slots (QTS) into the period in which nodes transmit data—at the beginning of the polling cycle for the furthest nodes, and towards the end for those closest to the sink—which increases the number of intersection slots, improving network performance (shorter delay to meet a node in the right next-hop-group), and uses more QTS in the area that is far from the sink (according to their less or null energy expenditure acting as a relay for others).
    With energy-harvesting onboard, energy efficiency constraints become relaxed someway, so the asymmetric weights of sleep/awake periods—traditionally set looking only at the remaining energy of the nodes—may now take into consideration other issues related to the provision of the QoS that certain application demands. We are mainly talking about minimizing or guaranteeing end-to-end latency, paying differentiated attention to data with different priorities, or, to a lesser extent for most WSN applications, achieving a minimum throughput. Wireless Body Area Networks (WBAN) is a good example: beyond its special environment (with possible temporal fading produced by the individual’s own movements, very close distances between sensors and sink, and moving scene), QoS is essential as sensed signals (ECG, EMG, EEG, glucose, blood pressure, SpO2) have clearly different importance and transmission requirements (latency, throughput and reliability).
    8.
    Liu et al. [104] propose a QoS driven EE TDMA-based protocol that dynamically adjusts the transmission order and transmission duration of the nodes—frame slots reordering—and so their on/off scheduling, based on data priority, channel status (temporal fading of parts of the body) and application context (the individual’s activity, an emergency, etc.). To reduce the synchronization overhead their proposal uses guard time intervals (to deal with clock drifts), and clock tuning using data/ack exchange (piggybacking) with the personal server (onboard sink).
    9.
    In the same research field (WBAN), Rismanian Yazdi et al. [105] design ECTP-MAC (Energy Consumption Traffic Prioritization MAC) and modify the frame structure of IEEE 802.15.4 to include an extra phase for emergency data. They define three types of data (normal discontinuous, periodic continuous and emergency) and schedule transmissions in CP, CFP, and extra interval according to a priority value proportional to data type and inversely proportional to data length and frequency.
    10.
    In a more general context, in QPPD MAC (QoS MAC Protocol for Prioritized Data) [106], after broadcasting a beacon with QoS mode on (asynchronous contention-based RI protocol), the receiver uses a limited interval and listen for senders’ requests. The protocol implements a request-allocation scheme based on transmitter beacons and data priority tags: the highest priority transmitter beacon received at interval expiration, or, immediately, if received one tagged with the top level priority, determines which contender will obtain receiver assignment. From the point of view of EE/EH, the protocol adjusts the receiver duty cycle according to its current power level, working at 5% if remaining energy is below 10%, at full activity if battery state is above 90%, and with the value determined by a particular expression (continuous function) otherwise. The authors say the biggest advantage of their proposal over others (TDMA -based) is the energy savings due to the lack of synchronization.
    11.
    In EEQ-MAC (Energy-Efficient and QoS-aware MAC) [107], queue length and data priority are used to adapt the node’s duty cycle, increasing the length of its active period in the event of high traffic which provides less waiting time to support time-bounded delivery of priority packets. A Random Early Detection mechanism is also included to avoid starvation of low priority data.

    2.3.3. Wake-Up Radios

    Recently, low-power wake-up radio technology (WuR) appeared as a promising alternative, perhaps making duty-cycling unnecessary [108]. In general, duty-cycling saves energy at the cost of an increase in the end-to-end packet delivery delay. Moreover, such savings will not be optimal because of collateral effects as, for instance, protocol overhead in synchronized approaches and higher contention—so collisions and retransmissions—in collective awakenings. Finally, idle listening is not completely removed.
    To achieve higher energy efficiency without increasing latency, wake-up radios—a second radio onboard—have been proposed as a promising alternative for asynchronous protocols. A wake-up radio is a low-power hardware device—consuming three orders of magnitude less power than a main RF chip [108]—capable of immediately reacting to an external event, waking up the node—and so its main radio—that is in sleep mode. Moreover with recent advancement in micro-electronics it is even possible to perform destination address decoding without waking the microcontroller, which would avoid false wakeups and overhearing [109]. It is not vain to say the only disadvantage of WuR systems appears to be the extra hardware expenditure. Though, well, limited range is another one.
    Wake-up MAC protocols can be seen as on demand MAC protocols where sending nodes ask receivers to wake-up for attending an imminent packet arrival. The wake-up signal can be sent on an exclusive control channel or on the same used by the main radio; and, in the basic implementation, not even CCA is performed. In [110], state-of-the-art wake-up MAC protocols have been split into three categories that use different hardware technology: (i) duty-cycled wake-up MAC protocols, (ii) non-cycled wake-up MAC protocols, (iii) path reservation wake-up MAC protocols. The first group is a special case where duty cycling is applied to the wake-up radio, which has similar characteristics in terms of coverage range and power consumption as the main transceiver (actually something less but not enough). So neither it is attractive from the point of view of energy efficiency nor eliminates the extra delay associated with duty cycling technique.
    In the second group (non-cycled) we find two interesting alternatives: those based on an always-on low-power wake-up radio (active wake-up radio) that is able of transmitting and receiving wake-up messages, and those where the wake-up radio is passive (powered by the wake-up signal, reminding RFID technology) or low-power active (using just a few micro watts from the battery), being able to receive the wake-up signal but delegating on main radio its transmission. Both share high responsiveness; active wake-up radio has greater coverage—20 m vs. ≪10 m—though leaving it active all the time conducts to a non-negligible power consumption; and passive devices achieve the highest energy conservation. Finally, in path reservation wake-up MAC protocols nodes take advantage of the additional (wake-up) radio to perform at the same time the early warning reception and a wake-up message transmission to prepare a forwarding event in a multihop context, so eliminating path establishment waiting time. Other interesting taxonomies, a hardware review, and WuR-based MAC protocol survey can be found in [111].
    The clear advantage of RI schemes in duty-cycling asynchronous protocols disappears (RI vs. TI) in WuR, and, once more, the application and the scenario will determine the most suitable operation mode. In general, RI suits better for data collection whereas TI is more appropriate for event-triggered data reporting in WSNs. In the next lines we present the very last contributions in the field of EE/EH-MAC protocols:
    • In [112], FAWR (Fully Asynchronous Wake-up Radio-based MAC), a RI sink commanded scheme, is presented as a multihop WuR MAC able to overcome the short-range limitation (20 m) of this low power transceivers. Though sensor nodes are settled at one-hop distance from the BS (from the point of view of the main radio transmissions) they succeed to propagate WuR signals using a forwarding decision table built in an early topology discovery phase executed by the sink. This way, the covered area can be much larger. They also use sensor nodes’ state of charge information so that the BTS, knowledgeable about the application requirements, can properly poll deployed devices.
    • In RI-LD-WuR (RI Low Delay MAC), Singh and Sikdar [113] propose partitioning sensor nodes in k (almost) equal sized groups to reduce packet collisions in a broadcast-based RI WuR enabled WSN: the sink commands its cluster broadcasting a request in slot0—dynamic TDMA with contention—firing the data collection phase. N transmissions fit in each slot and only the sensor nodes of the respective group are able to contend for sending. They also contribute with a distributed algorithm for the initial composition of the groups.
    • In RI-CPT-WuR (RI Consecutive Packet Transmission MAC) [114], to cancel out the time wasted in collisions, the protocol enables transmitting multiple packets through a single competition.
    • In addition, precisely to avoid collisions, refs. [114][115] present different strategies to face such issue in typical asynchronous TI WuR implementations: The proposal of Guntupalli et al. [114] uses an early backoff at time to send the WuR signal, arguing that in the case of success the transmitter will not need any additional backoff to use the main radio. Ghose et al. [115], on the contrary, advocate to incorporate to the WuC (wake-up channel) a more complete MAC protocol using CCA (CCA-WuR), CSMA (CSMA-WuR) or both adaptively (ADP-WuR). In the last, a threshold determines switching between CCA and CSMA modes if CCA fails such number of consecutive attempts.
    • Finally, in SNW-MAC (Star Network WuRx MAC) [116], an Energy Management (EM) scheme is incorporated which uses the node’s residual energy to optimize the energy usage in truly EH-WSN scenarios. The protocol assumes that the time is divided into time slots of equal duration T, and the EM is executed at the end of each slot to set the throughput of the node for the next interval. First, the Energy Budget Computation module evaluates the energy that the node can consume in the next time slot k to remain sustainable; and then Throughput Computation module calculates the wake-up interval TWI[k] according to the energy budget eB[k], so determining the frequency at which the node performs sensing and sends the so-obtained data.

    2.3.4. Topology Control

    When sensors are redundantly distributed to provide good coverage, it is possible to turn off some nodes while maintaining network operation and connectivity. Topology control protocols take advantage of redundancy to dynamically adapt the network topology based on application needs to minimize the number of active nodes. In fact, nodes that are not needed to ensure coverage or connectivity can be powered down to prolong the life of the network.
    This area of research has been maintaining a moderate (tens of works per year) but continuous activity during the last ten years. In 2013 two surveys on topology control techniques were published. Li et al. [117] provide an overview of topology control techniques, classifying them into two categories: network coverage issues (blanket, barrier and sweep coverage) and network connectivity issues (power management and power control). The survey of Aziz et al. [118] presents a comprehensive study of topology control techniques for extending the lifetime of WSNs, classifying them according to the energy conservation approach they adopt.
    Singla and Munjal [119] published recently a new review of topology control algorithms found in the literature, classifying them into two categories (centralized and distributed algorithms).
    Next, we comment the most relevant research works published in recent years in topology control. Javadi et al. [120] propose the topology control protocol LBLATC, where every sensor has a learning automaton that chooses the most suitable transmission range using the reinforcement signal produced by neighboring nodes. As other typical problems related to WSNs, topology control can be also object of applying computational intelligence techniques. Primeau et al. [121] review the application of several computational intelligence methodologies based on fuzzy systems, neural networks, evolutionary computation, swarm intelligence, etc., to several problems in WSNs, and one of the addressed problems is topology control. Song et al. [122] propose a multihop topology control algorithm with double CH based on affinity propagation clustering (APDC-M). Exploiting the fact that combining topology control and network coding has more advantages than if we apply them separately, Khalily-Dermany et al. [123] propose a topology control algorithm where in addition to the transmission power, the consumed energy for reception is also considered. Khalily-Dermany and Nadjafi-Arani [124] also study an optimization and graph theory approach to propose a mathematical perspective for combining topology control with network coding.

    2.4. Energy-Efficient Routing

    In the network layer, the most important task in the WSN is to set up a route between sensor nodes and BS. Since routing is an additional energy-consuming task, especially in the nodes close to the sink because of forwarding much more traffic, we must use energy-aware routing protocols. For an extensive review, the most cited surveys are [9][125][126][127][128][129]. In the last two years, we can also find many surveys [130].
    The survey done by Maratha and Gupta [131] (as Ogundile and Alfa [132] had done before in [132]) classifies energy-efficient routing protocols according to the communication mode towards BS into single-hop or multihop, both cases with and without clustering. In the case of clustering, the mode refers to the communication between CH and BS. In the other hand, multihop routing protocols are classified between single-path or multipath (if traffic from source to BS is balanced between several paths).
    The survey by Mittal and Iwendi [133] uses a taxonomy of routing protocols based on the network structure, classifying them into flat network (or data-centric), hierarchical and location-based routing protocols. Ketshabetswe et al. [134] uses the same classification, but the QoS-aware routing protocols are also considered.
    Nakas et al. [135] made the most complete survey about EE routing because, in addition to the traditional classification category of network structure, they also consider the communication model, if location information is either used or not, and if QoS or multipath are or not supported, and moreover protocols are both described and compared in every category. This classification is shown in Figure 2.
    Figure 2. Classification of Energy-Efficient Routing Protocols used in [135].
    Regarding the communication model, the protocols can be:
    • Query-based: When BS needs new data, it broadcasts a query message to ask for these data. Next, the node which owns the requested data sends them to the BS.
    • Coherent or non-coherent: In coherent protocols, a node applies some processing to the collected data. However, in non-coherent algorithms, the collected data are preprocessed at the source nodes and then sent to a special node, called aggregator (usually the CH), where they are further processed for data reduction.
    • Negotiation-based: before real data transmission, negotiation messages are exchanged between a source node and their destination to prevent redundant data. These protocols use a naming scheme to advertise data to destination.
    Regarding whether or not location information is used, this latter survey considers the special case of using a mobile agent (a program that travels across the nodes to perform tasks based on environmental conditions in an autonomous way) and where the BS can move within the network to collect data from sensor nodes. Location-based routing is also included under this type, but it could be included under the network structure category as done in other surveys.
    Under the Reliable Routing category, protocols are classified depending upon their inclusion of QoS support or multiple paths to balance the load and tolerate path failures.
    Another recent work [130] makes a systematic review of energy efficient routing schemes examining the literature during 2016–2018. For this purpose, they consider different categories of protocols: location-based, heterogeneity, mobility-based, multipath, hierarchical, data-centric and QoS based.
    To sum up, most of the routing algorithms proposed in the literature can be classified in flat network, hierarchical and location-based, either using single-hop or multihop communication.
    In flat network routing, all nodes have the same responsibility and every node has all information, so that the user can send a query to any node to obtain information. We must note that it is not possible to use a global addressing scheme due to the huge number of nodes in the network. This fact makes classical IP-based routing inadequate. Therefore, the routing is data-centric, i.e., it is totally dependent on naming of desired data.
    The first data-centric protocol was SPIN (Sensor Protocol for Information via Negotiation) [136]. When a node has new data to share, it advertises this fact by transmitting an ADV message containing metadata (that identifies data). When the neighboring nodes receive this ADV message, if not repeated, they send a REQ message to the source node requesting complete sensor data. After receiving data, the process is repeated in the second node. This protocol is based on negotiation, but data-centric algorithms are mainly query-based, as is the case with Direct Diffusion [137] and Rumor Routing [138].
    In location-based routing, all the nodes calculate the distances to the neighboring nodes based on incoming signal strength. Another option is to use a GPS signal, but in this case, nodes should go to sleep mode whenever the communication is not active. Some major routing protocols of this type are Geographic Adaptive Fidelity (GAF) [139] or SPAN [140], that is integrated with IEEE 802.11 to improve transmission latency, and extend the network lifetime.
    As already explained in Section 2.2, since in most WSN applications data flows from sensor nodes to BS, and data from nearby nodes contains redundancy, clustering favors data reduction by exploiting the spatial correlation. This is the basis of hierarchical routing protocols, where nodes are grouped in clusters, and a node is selected as the CH, that is responsible for collecting data from the sensor nodes inside the cluster and routing it towards the sink, directly (single-hop) or through several CHs (multihop). Hierarchical protocols are the most popular and the preferred option for WSNs and they are clearly the most studied protocols presently. For this reason, we will focus on them.

    2.4.1. Hierarchical Routing Protocols

    Heinzelman et al. [56] proposed the first hierarchical routing protocol referred to as LEACH (Low Energy Adaptive Clustering Hierarchy). LEACH proposes a random rotation method to select the node with maximum energy level as the CH, and so uniformly distribute the energy load among the sensors in the network. CHs send advertisement messages to the whole WSN using CSMA. Each sensor node joins the cluster from which it receives the strongest signal. Next, CH schedules TDMA slots for each member in the cluster to send data to it. CH uses aggregation techniques to combine the data received from sensor nodes to save energy and bandwidth, and then this aggregated information is forwarded directly to the BS, i.e., using only one hop, as shown in Figure 3.
    Figure 3. Different hierarchical routing strategies (BS base station, L leader, CH cluster head).
    The single-hop transmission is the simplest method, but usually it is not suitable for large networks, where multiple-hop transmission should be employed. In this case, data follows a multiple-hop route across several CHs towards the BS, and so it is essential to use an energy-aware routing protocol that avoids unnecessary transmissions and the overload in the nodes close to the BS.
    Clustering enhances energy efficiency in several ways: (i) it reduces the communication range within the cluster and so less transmission power is necessary, (ii) data reduction techniques can be performed by the CH, (iii) energy-intensive operations such as coordination or data reduction are only carried out by the CH, (iv) it enables the powering-off of some nodes, typically after sending data to the CH. On the other hand, hierarchical routing also improves network scalability by maintaining a hierarchical topology in the network.
    LEACH is still the most important and most used basic routing algorithm for WSNs. After 18 years of existence, much attention is still devoted to LEACH by the research community working in the area of routing in WSN. This itself shows its relevancy. In several recent works [141][142][143], the authors survey, classify and analyze different versions or improvements of LEACH, also using multihop transmission.
    Manjeshwar and Agrawal [144] proposed another popular cluster-based routing algorithm referred to as Threshold-sensitive Energy-Efficient sensor Network (TEEN) [144] that has been designed for time critical applications. TEEN combines the architecture based on clustering with the use of a data-centric mechanism. Adaptive Periodic TEEN (APTEEN) [145] is an enhancement of TEEN where CH broadcasts relevant parameters to the cluster members such as threshold values, TDMA schedule, and maximum time between consecutive reports.
    Another interesting cluster-based routing protocol is Hybrid Energy-Efficient Distributed (HEED) [146], where CH election is triggered in given intervals and it is based mainly on residual energy and other parameters as the number of neighbors or the distance to them. A survey recently published by Ullah [147] focus on HEED-based protocols.
    Since the relevancy of cluster-based routing, it is common to speak indistinctly of cluster-based and hierarchical routing, but strictly speaking, other types of hierarchical structures have been proposed in the literature. Recently, Chan et al. [148] survey and compare both LEACH-based clustering and these other hierarchical structures, classified into the following categories: (a) chain-based, (b) tree-based, (c) grid-based, and (d) area-based, also represented in the Figure 3.
    In chain-based hierarchical routing, the WSN is divided into chains; and a leader is chosen for every chain. Every node sends the data to the next node until the leader is reached. The main drawbacks are the delays suffered by the farthest nodes in long chains, the overload of the nodes close to the leader and the connectivity loss in a sub-chain when a node fails. The most relevant chain-based algorithm is PEGASIS (Power-Efficient Gathering in Sensor Information Systems) [149], where the leaders are rotated for energy reasons, and they send the aggregated data to the sink.
    In tree-based routing algorithms, a sink tree is created and there is a single path between each node and the sink. Unlike the chain-based case, there are no leaders, and a parent node can receive data from several children (or leaves), unlike the previous case, a node (parent) can have several children that send data to it, enabling the possibility of aggregation. The main drawbacks are similar to the chain-based case, i.e., the delays suffered by the farthest nodes in long trees, the overload of the nodes close to the sink and the connectivity loss in branches connected to a parent that fails. The most relevant tree-based algorithm is PEDAP (Power-Efficient Data gathering and Aggregation Protocol) [150] that uses the optimum sink tree based on data volume and transmission distance.
    In grid-based algorithms, the whole network is split into many grids (similar to clusters), based on the geographical location of the nodes. The leader selected for every grid is the responsible for routing the data through other leaders until reaching the sink, i.e., using multihop transmission. Each node only needs to know the location information about the leader of the grid. The most important proposal of this type is Two-tier data dissemination (TTDD) [151], where the mobile sink use flooding to send a data request to source nodes.
    In area-based mechanisms, the entire network is divided into multiple variable-sized areas. The BS also transmits a data request to the closest nodes that they forward via flooding until the data source is reached, which will send the data to the sink. A typical area-based algorithm is Line-based data dissemination (LBDD) [152], where a line of leaders is selected to split the whole network in two areas. The nodes send data to the closest leader on the line, and the leaders on the line store data from nodes and serve requests from sink if possible, and if not, send the request up and down the line. A little improvement was proposed in Ring Routing [153], using a ring instead of a line.
    In Figure 4, we show the most relevant algorithms of each type.
    Figure 4. Classification of hierarchical routing strategies.
    The main problems related to cluster-based routing are the cluster formation, the selection of CH in each cluster and the relay node placement. In addition to the classic approaches to address them, these problems have been object of optimization in hundreds of works in last years, and they are clearly one of the most active research areas presently. To solve this problem, the researchers have appealed to optimization (Swarm Intelligence Algorithms) and methodological approaches such as fuzzy logic or metaheuristic. The survey in [154] studies hierarchical energy-efficient routing protocols based on classical and swarm intelligence approaches.
    However, the subsequent survey in [155] focuses on methodological approaches in cluster-based multihop routing protocols that are classified into four categories: classical approaches, fuzzy-based approaches, metaheuristic-based approaches and hybrid metaheuristic- and fuzzy-based approaches.
    The very recent survey by Manuel et al. [156] is much more complete and the classification considers both swarm intelligence algorithms: ant colony optimization (ACO), artificial bee colony optimization (ABCO), fuzzy logic (FL), genetic algorithm (GA), whale algorithm (WA), or particle swarm optimization (PSO) and methodological approaches (fuzzy-based and metaheuristic-based approaches). See Figure 5 reproduced from this complete survey.
    Figure 5. Classification of clustering-based routing protocols reproduced from Manuel et al. [156].
    A review of the recent literature on these topics shows rapidly the high level of research activity in this area. The work in [157] studies the adoption of sink mobility to avoid the hot spot problem (CHs close to the BS). The mobile sink moves within the network and communicate directly with CHs without the need for routing. The ACO algorithm is studied for finding an optimal trajectory for the mobile sink. In [158] a special clustering method called Energy Centers searching using PSO (EC-PSO) is proposed for clustering and CH selection. In [159] a firefly algorithm is developed for selecting the CH optimally. Ezhilarasi and Krishnaveni [160] propose the evolutionary multipath energy-efficient routing protocol (EMEER) using a cuckoo search algorithm to optimally select the CH considering energy efficiency. Recently, Alghamdi [161] proposes an optimal CH selection by considering energy, delay, distance, and security using a new algorithm that hybridizes the concept of dragon fly and firefly algorithms. Additionally, recently, the Optimized QoS-based Clustering with Multipath Routing Protocol (OQoS-CMRP) has been proposed [162] by applying the Modified Particle Swarm Optimization to form clusters and select CHs.
    Ref. [163] introduces an algorithm that uses fuzzy logic for cluster construction and CH selection, and ACO for inter-cluster routing to mitigate the hot spot problem and extend network lifetime. In [164] an interesting PSO-based unequal and fault tolerant clustering protocol (PSO-UFC) is presented. In [165] the authors use a cuckoo optimization algorithm (COA) for clustering and selection of optimal CHs, considering four criteria such as the remaining energy of nodes, distance to the base station, within-cluster distances, and between cluster distances. In [166] a multihop LEACH protocol is optimized by means of an ACO algorithm, using a CH close to the BS. Other recent works that propose LEACH optimizations are the proposal in [167] using a Fuzzy C-means clustering (FCM) Algorithm, the work in [168] that uses a PSO algorithm or the optimization made in [169] by means of a Genetic Algorithm. Another interesting work is that of Jain and Goel [170] where fuzzy sets and fuzzy decision rules have been used for intelligent selection of CHs and to setup multihop routes to BS.
    Although LEACH is the preferred protocol for using as basis for optimization, other cluster-based protocols are also used. Therefore, several improvements of PEGASIS has been recently proposed. In [171] an Enhanced PEGASIS (EPEGASIS) protocol is proposed to mitigate the problem of hot spots from four directions. The work in [172] combines PEGASIS with Hamilton Loop algorithm, through a mixture of single-hop and multihop mechanisms, inserting a mobile agent node that is responsible for receiving and merging packets from the CHs. The authors in [173] also combines PEGASIS with a genetic algorithm to build the chain instead of the greedy algorithm.
    The problem of CH selection in APTEEN using artificial intelligence has also attracted the interest of researchers in recent years: using PSO [174], a combination of genetic algorithms and fruit fly optimization algorithm [175], or ACO [176][177].

    The entry is from 10.3390/s21134281

    References

    1. Akyildiz, I.F.; Weilian, S.; Sankarasubramaniam, Y.; Cayirci, E. A survey on sensor networks. IEEE Commun. Mag. 2002, 40, 102–114.
    2. Anastasi, G.; Conti, M.; Di Francesco, M.; Passarella, A. Energy conservation in wireless sensor networks: A survey. Ad Hoc Netw. 2009, 7, 537–568.
    3. Raghunathan, V.; Schurgers, C.; Park, S.; Srivastava, M. Energy-aware wireless microsensor networks. IEEE Signal Process. Mag. 2002, 19, 40–50.
    4. Chowdhury, S.; Hossain, A. Different Energy Saving Schemes in Wireless Sensor Networks: A Survey. Wirel. Pers. Commun. 2020, 114, 2043–2062.
    5. Singh, J.; Kaur, R.; Singh, D. A survey and taxonomy on energy management schemes in wireless sensor networks. J. Syst. Archit. 2020, 111.
    6. Singh, J.; Kaur, R.; Singh, D. Energy harvesting in wireless sensor networks: A taxonomic survey. Int. J. Energy Res. 2020.
    7. Engmann, F.; Katsriku, F.A.; Abdulai, J.; Adu-Manu, K.S.; Banaseka, F.K. Prolonging the Lifetime of Wireless Sensor Networks: A Review of Current Techniques. Wirel. Commun. Mob. Comput. 2018, 2018.
    8. Yetgin, H.; Cheung, K.; El-Hajjar, M.; Hanzo, L. A Survey of Network Lifetime Maximization Techniques in Wireless Sensor Networks. IEEE Commun. Surv. Tutor. 2017, 19, 828–854.
    9. Yadav, S.; Yadav, R. A review on energy efficient protocols in wireless sensor networks. Wirel. Netw. 2016, 22, 335–350.
    10. Rault, T.; Bouabdallah, A.; Challal, Y. Energy efficiency in wireless sensor networks: A top-down survey. Comput. Netw. 2014, 67, 104–122.
    11. Cui, S.; Goldsmith, A.; Bahai, A. Energy-constrained modulation optimization. IEEE Trans. Wirel. Commun. 2005, 4, 2349–2360.
    12. Rosas, F.; Oberli, C. Modulation and SNR optimization for achieving energy-efficient communications over short-range fading channels. IEEE Trans. Wirel. Commun. 2012, 11, 4286–4295.
    13. Gumusalan, A.; Simon, R.; Aydin, H. Dynamic modulation scaling enabled multi-hop topology control for time critical wireless sensor networks. Wirel. Netw. 2020, 26, 1203–1226.
    14. Nosratinia, A.; Hunter, T.E.; Hedayat, A. Cooperative communication in wireless networks. IEEE Commun. Mag. 2004, 42, 74–80.
    15. Li, Q.; Hu, R.Q.; Qian, Y.; Wu, G. Cooperative communications for wireless networks: Techniques and applications in LTE-advanced systems. IEEE Wirel. Commun. 2012, 19, 22–29.
    16. Shuguang, C.; Goldsmith, A.J.; Bahai, A. Energy-efficiency of MIMO and cooperative MIMO techniques in sensor networks. IEEE J. Sel. Areas Commun. 2004, 22, 1089–1098.
    17. Soorki, M.; Manshaei, M.; Maham, B.; Saidi, H. On uplink virtual MIMO with device relaying cooperation enforcement in 5G networks. IEEE Trans. Mob. Comput. 2018, 17, 155–168.
    18. Song, L.; Chai, K.; Chen, Y.; Loo, J.; Jimaa, S.; Iraqi, Y. Energy efficient cooperative coalition selection in cluster-based capillary networks for CMIMO IoT systems. Comput. Netw. 2019, 153, 92–102.
    19. Singh, M.; Amin, S. Energy-efficient data transmission technique for wireless sensor networks based on DSC and virtual MIMO. ETRI J. 2020, 42, 341–350.
    20. Peng, Y.; Al-Hazemi, F.; Kim, H.; Youn, C.H. Design and Optimization for Energy-Efficient Cooperative MIMO Transmission in Ad Hoc Networks. IEEE Trans. Veh. Technol. 2017, 66, 710–719.
    21. Peng, Y.; Al-Hazemi, F.; Boutaba, R.; Tong, F.; Hwang, I.; Youn, C. Enhancing Energy Efficiency via Cooperative MIMO in Wireless Sensor Networks: State of the Art and Future Research Directions. IEEE Commun. Mag. 2017, 55, 47–53.
    22. Peng, Y.; Li, J.; Park, S.; Zhu, K.; Hassan, M.; Alsanad, A. Energy-efficient cooperative transmission for intelligent transportation systems. Future Gener. Comput. Syst. 2019, 94, 634–640.
    23. Correia, L.H.; Macedo, D.F.; dos Santos, A.L.; Loureiro, A.A.; Nogueira, J.M.S. Transmission power control techniques for wireless sensor networks. Comput. Netw. 2007, 51, 4765–4779.
    24. Fernandes, D.; Ferreira, A.; Abrishambaf, R.; Mendes, J.; Cabral, J. Survey and taxonomy of transmissions power control mechanisms for wireless body area networks. IEEE Commun. Surv. Tutor. 2018, 20, 1292–1328.
    25. Lee, W.; Kim, H.; Hong, M.; Kang, M.G.; Jeong, S.; Kim, N. A survey of the transmission-power-control schemes in wireless body-sensor networks. KSII Trans. Internet Inf. Syst. 2018, 12, 1854–1868.
    26. Sodhro, A.; Chen, L.; Sekhari, A.; Ouzrout, Y.; Wu, W. Energy efficiency comparison between data rate control and transmission power control algorithms for wireless body sensor networks. Int. J. Distrib. Sens. Netw. 2018, 14.
    27. Sodhro, A.; Sangaiah, A.; Sodhro, G.; Lohano, S.; Pirbhulal, S. An energy-efficient algorithm for wearable electrocardiogram signal processing in ubiquitous healthcare applications. Sensors 2018, 18, 923.
    28. Sodhro, A.; Pirbhulal, S.; Qaraqe, M.; Lohano, S.; Sodhro, G.; Junejo, N.; Luo, Z. Power Control Algorithms for Media Transmission in Remote Healthcare Systems. IEEE Access 2018, 6, 42384–42393.
    29. Sodhro, A.; Pirbhulal, S.; Sodhro, G.; Gurtov, A.; Muzammal, M.; Luo, Z. A Joint Transmission Power Control and Duty-Cycle Approach for Smart Healthcare System. IEEE Sens. J. 2019, 19, 8479–8486.
    30. Zhang, H.; Safaei, F.; Tran, L. Joint transmission power control and relay cooperation for WBAN systems. Sensors 2018, 18, 4283.
    31. Tanessakulwattana, S.; Pornavalai, C. Multipath energy balancing for clustered wireless sensor networks. Wirel. Netw. 2019, 25, 2537–2558.
    32. Lee, W.; Kim, N.; Lee, B.D. An adaptive transmission power control algorithm for wearable healthcare systems based on variations in the body conditions. J. Inf. Process. Syst. 2019, 15, 593–603.
    33. Archasantisuk, S.; Aoyagi, T. Transmission power control using human motion classification for reliable and energy-efficient communication in WBAN. IEICE Trans. Commun. 2019, 102, 1104–1112.
    34. Qolami, A.; Nassiri, M.; Abdoli, H. A transmission power control mechanism for 802.15.4+rpl-operated wireless sensor network. Int. J. Sens. Wirel. Commun. Control 2020, 10, 197–206.
    35. Mohsan, S.; Amjad, H.; Mazinani, A.; Shahzad, S.; Khan, M.; Islam, A.; Mahmood, A.; Soban, A. Investigating transmission power control strategy for underwater wireless sensor networks. Int. J. Adv. Comput. Sci. Appl. 2020, 11, 281–285.
    36. Kranakis, E.; Krizanc, D.; Williams, E. Directional versus omnidirectional antennas for energy consumption and k-connectivity of networks of sensors. Lect. Notes Comput. Sci. 2005, 3544, 357–368.
    37. Michalopoulou, A.; Koxias, E.; Lazarakis, F.; Zervos, T.; Alexandridis, A.A. Investigation of directional antennas effect on energy efficiency and reliability of the IEEE 802.15.4 standard in outdoor wireless sensor networks. In Proceedings of the 2015 IEEE 15th Mediterranean Microwave Symposium (MMS), Lecce, Italy, 30 November–2 December 2015; pp. 1–4.
    38. Kumai, N.; Kumar, R.; Bajaj, R. Mobile ad hoc networks and energy efficiency using directional antennas: A review. In Proceedings of the 2017 International Conference on Intelligent Computing and Control Systems (ICICCS), Madurai, India, 15–16 June 2017; pp. 1213–1219.
    39. George, R.; Mary, T. Review on directional antenna for wireless sensor network applications. IET Commun. 2020, 14, 715–722.
    40. Gautam, P.; Kumar, S.; Verma, A.; Rashid, T.; Kumar, A. Energy-Efficient Localization of Sensor Nodes in WSNs Using Beacons from Rotating Directional Antenna. IEEE Trans. Ind. Inform. 2019, 15, 5827–5836.
    41. Ali, A.; Wang, H.; Lv, H.; Chen, X. A survey of MAC protocols design strategies and techniques in wireless Ad Hoc networks. J. Commun. 2014, 9, 30–38.
    42. Choudhury, R.; Yang, X.; Ramanathan, R.; Vaidya, N. On designing MAC protocols for wireless networks using directional antennas. IEEE Trans. Mob. Comput. 2006, 5, 477–491.
    43. Karapistoli, E.; Gragopoulos, I.; Tsetsinas, I.; Pavlidou, F.N. A MAC protocol for low-rate UWB wireless sensor networks using directional antennas. Comput. Netw. 2009, 53, 961–972.
    44. Dang, D.N.M.; Le, H.T.; Kang, H.S.; Hong, C.S.; Choe, J. Multi-channel MAC protocol with Directional Antennas in wireless ad hoc networks. In Proceedings of the 2015 International Conference on Information Networking (ICOIN), Siem Reap, Cambodia, 12–14 January 2015; pp. 81–86.
    45. Le, T.N.; Pegatoquet, A.; Le Huy, T.; Lizzi, L.; Ferrero, F. Improving Energy Efficiency of Mobile WSN Using Reconfigurable Directional Antennas. IEEE Commun. Lett. 2016, 20, 1243–1246.
    46. Chau, A.; Dawson, J.F.; Mitchell, P.D.; Loh, T.H. Medium access control protocol for wireless sensor networks in Harsh environments with directional antennas. In Proceedings of the Loughborough Antennas Propagation Conference (LAPC 2018), Loughborough, UK, 12–13 November 2018; pp. 1–5.
    47. Chau, A.; Dawson, J.; Mitchell, P.; Loh, T. Virtual sensing directional hub mac (Vsdh-mac) protocol with power control. Electronics 2020, 9, 1219.
    48. Luitel, S.; Moh, S. An Energy-Aware Robust MAC Protocol for Prolonging Network Lifetime in Cognitive Radio Sensor Networks. In Proceedings of the 2020 International Conference on Electronics, Information, and Communication (ICEIC), Barcelona, Spain, 19–22 January 2020; pp. 1–8.
    49. Rehan, W.; Fischer, S.; Chughtai, O.; Rehan, M.; Hail, M.; Saleem, S. A novel dynamic confidence interval based secure channel prediction approach for stream-based multichannel wireless sensor networks. Ad Hoc Netw. 2020, 108.
    50. Suguna, R.; Rathinasabapathy, V. An SoC architecture for energy detection based spectrum sensing using Low Latency Column Bit Compressed (LLCBC) MAC in cognitive radio wireless sensor networks. Microprocess. Microsyst. 2019, 69, 159–167.
    51. Rashid, A.; Tripathi, Y.; Prakash, A.; Tripathi, R. Load aware energy-balanced data gathering approach in CRSNs. IET Wirel. Sens. Syst. 2019, 9, 143–150.
    52. Bolla, D.; Shankar, S. A robust QSCTA-EDRA routing protocol for cognitive radio sensor networks. Int. J. Commun. Netw. Distrib. Syst. 2020, 25, 385–413.
    53. Ntshabele, K.; Isong, B.; Dladlu, N.; Abu-Mahfouz, A.M. Energy Consumption Challenges in Clustered Cognitive Radio Sensor Networks: A Review. In Proceedings of the 2019 IEEE 28th International Symposium on Industrial Electronics (ISIE), Vancouver, BC, Canada, 12–14 June 2019; pp. 1294–1299.
    54. Tanwar, S.; Tyagi, S.; Kumar, N.; Obaidat, M. LA-MHR: Learning automata based multilevel heterogeneous routing for opportunistic shared spectrum access to enhance lifetime of WSN. IEEE Syst. J. 2019, 13, 313–323.
    55. Carie, A.; Li, M.; Marapelli, B.; Reddy, P.; Dino, H.; Gohar, M. Cognitive radio assisted WSN with interference aware AODV routing protocol. J. Ambient. Intell. Humaniz. Comput. 2019, 10, 4033–4042.
    56. Heinzelman, W.; Chandrakasan, A.; Balakrishnan, H. An application-specific protocol architecture for wireless microsensor networks. IEEE Trans. Wirel. Commun. 2002, 1, 660–670.
    57. Zhang, P.; Wang, J.; Guo, K.; Wu, F.; Min, G. Multi-functional secure data aggregation schemes for WSNs. Ad Hoc Netw. 2018, 69, 86–99.
    58. Randhawa, S.; Jain, S. Data Aggregation in Wireless Sensor Networks: Previous Research, Current Status and Future Directions. Wirel. Pers. Commun. 2017, 97, 3355–3425.
    59. Yadav, S.; Yadav, R. Redundancy elimination during data aggregation in wireless sensor networks for IoT systems. Lect. Notes Electr. Eng. 2019, 524, 195–205.
    60. Kaur, M.; Munjal, A. Data aggregation algorithms for wireless sensor network: A review. Ad Hoc Netw. 2020, 100.
    61. Abbasian Dehkordi, S.; Farajzadeh, K.; Rezazadeh, J.; Farahbakhsh, R.; Sandrasegaran, K.; Abbasian Dehkordi, M. A survey on data aggregation techniques in IoT sensor networks. Wirel. Netw. 2020, 26, 1243–1263.
    62. Pushpalatha, S.; Shivaprakasha, K. Energy-Efficient Communication Using Data Aggregation and Data Compression Techniques in Wireless Sensor Networks: A Survey. Lect. Notes Electr. Eng. 2020, 614, 161–179.
    63. Vinodha, D.; Mary Anita, E. Secure Data Aggregation Techniques for Wireless Sensor Networks: A Review. Arch. Comput. Methods Eng. 2019, 26, 1007–1027.
    64. Praveen Kumar, D.; Amgoth, T.; Annavarapu, C. Machine learning algorithms for wireless sensor networks: A survey. Inf. Fusion 2019, 49, 1–25.
    65. Srisooksai, T.; Keamarungsi, K.; Lamsrichan, P.; Araki, K. Practical data compression in wireless sensor networks: A survey. J. Netw. Comput. Appl. 2012, 35, 37–59.
    66. Chou, J.; Petrovic, D.; Ramachandran, K. A distributed and adaptive signal processing approach to reducing energy consumption in sensor networks. In Proceedings of the IEEE INFOCOM 2003, Twenty-Second Annual Joint Conference of the IEEE Computer and Communications Societies (IEEE Cat. No. 03CH37428), San Francisco, CA, USA, 30 March–3 April 2003; Volume 2, pp. 1054–1062.
    67. Xiong, Z.; Liveris, A.D.; Cheng, S. Distributed source coding for sensor networks. IEEE Signal Process. Mag. 2004, 21, 80–94.
    68. Luo, C.; Wu, F.; Sun, J.; Chen, C.W. Compressive Data Gathering for Large-Scale Wireless Sensor Networks. In Proceedings of the 15th Annual International Conference on Mobile Computing and Networking, Beijing, China, 20–25 September 2009; Association for Computing Machinery: New York, NY, USA, 2009; pp. 145–156.
    69. Li, S.; Xu, L.; Wang, X. Compressed sensing signal and data acquisition in wireless sensor networks and internet of things. IEEE Trans. Ind. Inform. 2013, 9, 2177–2186.
    70. Tekin, N.; Gungor, V. Analysis of compressive sensing and energy harvesting for wireless multimedia sensor networks. Ad Hoc Netw. 2020, 103.
    71. Tuama, A.; Mohamed, M.; Muhammed, A.; Hanapi, Z.; Mohamed, R.; Abu Bakar, K. Recent advances of data compression in Wireless Sensor Network. J. Eng. Appl. Sci. 2018, 13, 9002–9015.
    72. Sheikh, S.; Dakhore, M. Data Compression Techniques for Wireless Sensor Network, (IJCSIT). Int. J. Comput. Sci. Inf. Technol. 2015, 6, 818–821.
    73. Singh, V.; Singh, V.; Kumar, M. In-Network Data Processing Based on Compressed Sensing in WSN: A Survey. Wirel. Pers. Commun. 2017, 96, 2087–2124.
    74. Zhang, D.G.; Zhang, T.; Zhang, J.; Dong, Y.; Zhang, X.D. A kind of effective data aggregating method based on compressive sensing for wireless sensor network. Eurasip J. Wirel. Commun. Netw. 2018, 2018.
    75. Ebrahimi, D.; Sharafeddine, S.; Ho, P.H.; Assi, C. UAV-Aided projection-based compressive data gathering in wireless sensor networks. IEEE Internet Things J. 2019, 6, 1893–1905.
    76. Osamy, W.; Khedr, A.; Aziz, A.; El-Sawy, A. Cluster-tree routing based entropy scheme for data gathering in wireless sensor networks. IEEE Access 2018, 6, 77372–77387.
    77. Qiao, J.; Zhang, X. Compressive Data Gathering Based on even Clustering for Wireless Sensor Networks. IEEE Access 2018, 6, 24391–24410.
    78. Pakdaman Tirani, S.; Avokh, A. On the performance of sink placement in WSNs considering energy-balanced compressive sensing-based data aggregation. J. Netw. Comput. Appl. 2018, 107, 38–55.
    79. Sun, P.; Wu, L.; Weng, Z.; Xiao, M.; Wang, Z. Sparsest random sampling for cluster-based compressive data gathering in wireless sensor networks. IEEE Access 2018, 6, 36383–36394.
    80. Li, G.; Chen, H.; Peng, S.; Li, X.; Wang, C.; Yu, S.; Yin, P. Collaborative data collection scheme based on optimal clustering for wireless sensor networks. Sensors 2018, 18, 2487.
    81. Pacharaney, U.; Gupta, R. Clustering and Compressive Data Gathering in Wireless Sensor Network. Wirel. Pers. Commun. 2019, 109, 1311–1331.
    82. Aziz, A.; Osamy, W.; Khedr, A.; El-Sawy, A.; Singh, K. Grey Wolf based compressive sensing scheme for data gathering in IoT based heterogeneous WSNs. Wirel. Netw. 2020, 26, 3395–3418.
    83. Manchanda, R.; Sharma, K. Energy efficient compression sensing-based clustering framework for IoT-based heterogeneous WSN. Telecommun. Syst. 2020, 74, 311–330.
    84. Dias, G.; Bellalta, B.; Oechsner, S. A survey about prediction-based data reduction in wireless sensor networks. ACM Comput. Surv. 2016, 49.
    85. Tayeh, G.; Makhoul, A.; Laiymani, D.; Demerjian, J. A distributed real-time data prediction and adaptive sensing approach for wireless sensor networks. Pervasive Mob. Comput. 2018, 49, 62–75.
    86. Hao, X.; Wang, L.; Yao, N.; Geng, D.; Chen, B. Topology control game algorithm based on Markov lifetime prediction model for wireless sensor network. Ad Hoc Netw. 2018, 78, 13–23.
    87. Diwakaran, S.; Perumal, B.; Vimala Devi, K. A cluster prediction model-based data collection for energy efficient wireless sensor network. J. Supercomput. 2019, 75, 3302–3316.
    88. Alduais, N.; Abdullah, J.; Jamil, A. Enhanced payload data reduction approach for Cluster Head (CH) nodes. Telkomnika (Telecommun. Comput. Electron. Control) 2017, 15, 1477–1484.
    89. Tayeh, G.; Makhoul, A.; Perera, C.; Demerjian, J. A Spatialoral Correlation Approach for Data Reduction in Cluster-Based Sensor Networks. IEEE Access 2019, 7, 50669–50680.
    90. Dias, G.; Bellalta, B.; Oechsner, S. The impact of dual prediction schemes on the reduction of the number of transmissions in sensor networks. Comput. Commun. 2017, 112, 58–72.
    91. Wu, M.; Tan, L.; Xiong, N. Data prediction, compression, and recovery in clustered wireless sensor networks for environmental monitoring applications. Inf. Sci. 2016, 329, 800–818.
    92. Jarwan, A.; Sabbah, A.; Ibnkahla, M. Data Transmission Reduction Schemes in WSNs for Efficient IoT Systems. IEEE J. Sel. Areas Commun. 2019, 37, 1307–1324.
    93. Li, P.; Youn, H. Gradient-based adaptive modeling for IoT data transmission reduction. Wirel. Netw. 2020, 26, 6175–6188.
    94. Quintero, V.L.; Estevez, C.; Orchard, M.E.; Pérez, A. Improvements of Energy-Efficient Techniques in WSNs: A MAC-Protocol Approach. IEEE Commun. Surv. Tutor. 2019, 21, 1188–1208.
    95. Siddiqui, S.; Ghani, S.; Khan, A.A. ADP-MAC: An Adaptive and Dynamic Polling-Based MAC Protocol for Wireless Sensor Networks. IEEE Sens. J. 2018, 18, 860–874.
    96. Adam, M.; Por, L.; Hussain, M.; Khan, N.; Ang, T.; Anisi, M.; Huang, Z.; Ali, I. An adaptive wake-up-interval to enhance receiver-based ps-mac protocol for wireless sensor networks. Sensors 2019, 19, 3732.
    97. Sakya, G.; Sharma, V. ADMC-MAC: Energy efficient adaptive MAC protocol for mission critical applications in WSN. Sustain. Comput. Inform. Syst. 2019, 23, 21–28.
    98. Ye, W.; Heidemann, J.; Estrin, D. An energy-efficient MAC protocol for wireless sensor networks. In Proceedings of the Twenty-First Annual Joint Conference of the IEEE Computer and Communications Societies, New York, USA, 23–27 June 2002; Volume 3, pp. 1567–1576.
    99. Zhang, D.G.; Zhou, S.; Tang, Y.M. A Low Duty Cycle Efficient MAC Protocol Based on Self-Adaption and Predictive Strategy. Mob. Netw. Appl. 2018, 23, 828–839.
    100. Monica, R.; Davoli, L.; Ferrari, G. A Wave-Based Request-Response Protocol for Latency Minimization in WSNs. IEEE Internet Things J. 2019, 6, 7971–7979.
    101. Ye, D.; Zhang, M. A Self-Adaptive Sleep/Wake-Up Scheduling Approach for Wireless Sensor Networks. IEEE Trans. Cybern. 2018, 48, 979–992.
    102. Liu, Y.; Ota, K.; Zhang, K.; Ma, M.; Xiong, N.; Liu, A.; Long, J. QTSAC: An Energy-Efficient MAC Protocol for Delay Minimization in Wireless Sensor Networks. IEEE Access 2018, 6, 8273–8291.
    103. Ekbatanifard, G.; Monsefi, R.; Yaghmaee M., M.; Hosseini S., S. Queen-MAC: A quorum-based energy-efficient medium access control protocol for wireless sensor networks. Comput. Netw. 2012, 56, 2221–2236.
    104. Liu, B.; Yan, Z.; Chen, C.W. Medium Access Control for Wireless Body Area Networks with QoS Provisioning and Energy Efficient Design. IEEE Trans. Mob. Comput. 2017, 16, 422–434.
    105. Rismanian Yazdi, F.; Hosseinzadeh, M.; Jabbehdari, S. A Priority-Based MAC Protocol for Energy Consumption and Delay Guaranteed in Wireless Body Area Networks. Wirel. Pers. Commun. 2019, 108, 1677–1696.
    106. Sarang, S.; Drieberg, M.; Awang, A.; Ahmad, R. A QoS MAC protocol for prioritized data in energy harvesting wireless sensor networks. Comput. Netw. 2018, 144, 141–153.
    107. Muzakkari, B.A.; Mohamed, M.A.; Kadir, M.F.A.; Mamat, M. Queue and Priority-Aware Adaptive Duty Cycle Scheme for Energy Efficient Wireless Sensor Networks. IEEE Access 2020, 8, 17231–17242.
    108. Oller, J.; Demirkol, I.; Casademont, J.; Paradells, J.; Gamm, G.U.; Reindl, L. Has Time Come to Switch from Duty-Cycled MAC Protocols to Wake-Up Radio for Wireless Sensor Networks? IEEE/ACM Trans. Netw. 2016, 24, 674–687.
    109. Khodr, H.; Kouzayha, N.; Abdallah, M.; Costantine, J.; Dawy, Z. Energy Efficient IoT Sensor With RF Wake-Up and Addressing Capability. IEEE Sens. Lett. 2017, 1, 1–4.
    110. Djiroun, F.Z.; Djenouri, D. MAC Protocols With Wake-Up Radio for Wireless Sensor Networks: A Review. IEEE Commun. Surv. Tutor. 2017, 19, 587–618.
    111. Piyare, R.; Murphy, A.L.; Kiraly, C.; Tosato, P.; Brunelli, D. Ultra Low Power Wake-Up Radios: A Hardware and Networking Survey. IEEE Commun. Surv. Tutor. 2017, 19, 2117–2157.
    112. Pegatoquet, A.; Le, T.; Magno, M. A Wake-Up Radio-Based MAC Protocol for Autonomous Wireless Sensor Networks. IEEE/ACM Trans. Netw. 2019, 27, 56–70.
    113. Singh, R.; Sikdar, B. A Receiver Initiated Low Delay MAC Protocol for Wake-Up Radio Enabled Wireless Sensor Networks. IEEE Sens. J. 2020, 20, 13796–13807.
    114. Guntupalli, L.; Ghose, D.; Li, F.Y.; Gidlund, M. Energy Efficient Consecutive Packet Transmissions in Receiver-Initiated Wake-Up Radio Enabled WSNs. IEEE Sens. J. 2018, 18, 4733–4745.
    115. Ghose, D.; Li, F.Y.; Pla, V. MAC Protocols for Wake-Up Radio: Principles, Modeling and Performance Analysis. IEEE Trans. Ind. Inform. 2018, 14, 2294–2306.
    116. Aoudia, F.; Gautier, M.; Magno, M.; Berder, O.; Benini, L. Leveraging energy harvesting and wake-up receivers for long-term wireless sensor networks. Sensors 2018, 18, 1578.
    117. Li, M.; Li, Z.; Vasilakos, A. A survey on topology control in wireless sensor networks: Taxonomy, comparative study, and open issues. Proc. IEEE 2013, 101, 2538–2557.
    118. Aziz, A.; Şekercioǧlu, Y.; Fitzpatrick, P.; Ivanovich, M. A survey on distributed topology control techniques for extending the lifetime of battery powered wireless sensor networks. IEEE Commun. Surv. Tutor. 2013, 15, 121–144.
    119. Singla, P.; Munjal, A. Topology Control Algorithms for Wireless Sensor Networks: A Review. Wirel. Pers. Commun. 2020, 113, 2363–2385.
    120. Javadi, M.; Mostafaei, H.; Chowdhurry, M.; Abawajy, J. Learning automaton based topology control protocol for extending wireless sensor networks lifetime. J. Netw. Comput. Appl. 2018, 122, 128–136.
    121. Primeau, N.; Falcon, R.; Abielmona, R.; Petriu, E. A review of computational intelligence techniques in wireless sensor and actuator networks. IEEE Commun. Surv. Tutor. 2018, 20, 2822–2854.
    122. Song, L.; Song, Q.; Ye, J.; Chen, Y. A hierarchical topology control algorithm for WSN, considering node residual energy and lightening cluster head burden based on affinity propagation. Sensors 2019, 19, 2925.
    123. Khalily-Dermany, M.; Nadjafi-Arani, M.; Doostali, S. Combining topology control and network coding to optimize lifetime in wireless-sensor networks. Comput. Netw. 2019, 162.
    124. Khalily-Dermany, M.; Nadjafi-Arani, M. Mathematical Aspects in Combining Network Coding with Transmission Range Adjustment. IEEE Commun. Lett. 2019, 23, 1568–1571.
    125. Pantazis, N.; Nikolidakis, S.; Vergados, D. Energy-efficient routing protocols in wireless sensor networks: A survey. IEEE Commun. Surv. Tutor. 2013, 15, 551–591.
    126. Tanwar, S.; Kumar, N.; Rodrigues, J. A systematic review on heterogeneous routing protocols for wireless sensor network. J. Netw. Comput. Appl. 2015, 53, 39–56.
    127. Yan, J.; Zhou, M.; Ding, Z. Recent Advances in Energy-Efficient Routing Protocols for Wireless Sensor Networks: A Review. IEEE Access 2016, 4, 5673–5686.
    128. Shen, H.; Bai, G. Routing in wireless multimedia sensor networks: A survey and challenges ahead. J. Netw. Comput. Appl. 2016, 71, 30–49.
    129. Han, G.; Jiang, J.; Guizani, M.; Rodrigues, J. Green routing protocols for wireless multimedia sensor networks. IEEE Wirel. Commun. 2016, 23, 140–146.
    130. Shafiq, M.; Ashraf, H.; Ullah, A.; Tahira, S. Systematic Literature Review on Energy Efficient Routing Schemes in WSN—A Survey. Mob. Netw. Appl. 2020, 25, 882–895.
    131. Maratha, P.; Gupta, K. A comprehensive and systematized review of energy-efficient routing protocols in wireless sensor networks. Int. J. Comput. Appl. 2019.
    132. Ogundile, O.; Alfa, A. A survey on an energy-efficient and energy-balanced routing protocol for wireless sensor networks. Sensors 2017, 17, 1084.
    133. Mittal, M.; Iwendi, C. A survey on energy-aware wireless sensor routing protocols. EAI Endorsed Trans. Energy Web 2019, 6.
    134. Ketshabetswe, L.; Zungeru, A.; Mangwala, M.; Chuma, J.; Sigweni, B. Communication protocols for wireless sensor networks: A survey and comparison. Heliyon 2019, 5.
    135. Nakas, C.; Kandris, D.; Visvardis, G. Energy efficient routing in wireless sensor networks: A comprehensive survey. Algorithms 2020, 13, 72.
    136. Kulik, J.; Heinzelman, W.; Balakrishnan, H. Negotiation-based protocols for disseminating information in wireless sensor networks. Wirel. Netw. 2002, 8, 169–185.
    137. Intanagonwiwat, C.; Govindan, R.; Estrin, D. Directed Diffusion: A Scalable and Robust Communication Paradigm for Sensor Networks. In Proceedings of the 6th Annual International Conference on Mobile Computing and Networking, Boston, MA, USA, 6–11 August 2000; Association for Computing Machinery: New York, NY, USA, 2000; pp. 56–67.
    138. Braginsky, D.; Estrin, D. Rumor Routing Algorthim for Sensor Networks. In Proceedings of the 1st ACM International Workshop on Wireless Sensor Networks and Applications, Atlanta, GA, USA, 28 September 2002; Association for Computing Machinery: New York, NY, USA, 2002; pp. 22–31.
    139. Xu, Y.; Heidemann, J.; Estrin, D. Geography-Informed Energy Conservation for Ad Hoc Routing. In Proceedings of the 7th Annual International Conference on Mobile Computing and Networking, Rome, Italy, 16–21 July 2001; Association for Computing Machinery: New York, NY, USA, 2001; pp. 70–84.
    140. Chen, B.; Jamieson, K.; Balakrishnan, H.; Morris, R. Span: An energy-efficient coordination algorithm for topology maintenance in ad hoc wireless networks. Wirel. Netw. 2002, 8, 481–494.
    141. Arora, V.; Sharma, V.; Sachdeva, M. A survey on LEACH and other’s routing protocols in wireless sensor network. Optik 2016, 127, 6590–6600.
    142. Singh, S.K.; Kumar, P.; Singh, J.P. A Survey on Successors of LEACH Protocol. IEEE Access 2017, 5, 4298–4328.
    143. Al-Shalabi, M.; Anbar, M.; Wan, T.C.; Khasawneh, A. Variants of the low-energy adaptive clustering hierarchy protocol: Survey, issues and challenges. Electronics 2018, 7, 136.
    144. Manjeshwar, A.; Agrawal, D.P. TEEN: A routing protocol for enhanced efficiency in wireless sensor networks. In Proceedings of the 15th International Parallel and Distributed Processing Symposium, IPDPS 2001, San Francisco, CA, USA, 23–27 April 2001; pp. 2009–2015.
    145. Manjeshwar, A.; Agrawal, D.P. APTEEN: A hybrid protocol for efficient routing and comprehensive information retrieval in wireless. In Proceedings of the 16th International Parallel and Distributed Processing Symposium, Ft. Lauderdale, FL, USA, 15–19 April 2002; p. 8.
    146. Younis, O.; Fahmy, S. HEED: A hybrid, energy-efficient, distributed clustering approach for ad hoc sensor networks. IEEE Trans. Mob. Comput. 2004, 3, 366–379.
    147. Ullah, Z. A Survey on Hybrid, Energy Efficient and Distributed (HEED) Based Energy Efficient Clustering Protocols for Wireless Sensor Networks. Wirel. Pers. Commun. 2020, 112, 2685–2713.
    148. Chan, L.; Gomez Chavez, K.; Rudolph, H.; Hourani, A. Hierarchical routing protocols for wireless sensor network: A compressive survey. Wirel. Netw. 2020, 26, 3291–3314.
    149. Lindsey, S.; Raghavendra, C.S. PEGASIS: Power-efficient gathering in sensor information systems. In Proceedings of the IEEE Aerospace Conference, Big Sky, MT, USA, 9–16 March 2002; Volume 3, p. 3.
    150. Tan, H.; Körpeoǧlu, I. Power efficient data gathering and aggregation in wireless sensor networks. SIGMOD Rec. 2003, 32, 66–71.
    151. Luo, H.; Ye, F.; Cheng, J.; Lu, S.; Zhang, L. TTDD: Two-tier data dissemination in large-scale wireless sensor networks. Wirel. Netw. 2005, 11, 161–175.
    152. Hamida, E.B.; Chelius, G. A Line-Based Data Dissemination Protocol for Wireless Sensor Networks with Mobile Sink. In Proceedings of the 2008 IEEE International Conference on Communications, Hangzhou, China, 10–12 November 2008; pp. 2201–2205.
    153. Tunca, C.; Isik, S.; Donmez, M.; Ersoy, C. Ring Routing: An Energy-Efficient Routing Protocol for Wireless Sensor Networks with a Mobile Sink. IEEE Trans. Mob. Comput. 2015, 14, 1947–1960.
    154. Guleria, K.; Verma, A. Comprehensive review for energy efficient hierarchical routing protocols on wireless sensor networks. Wirel. Netw. 2019, 25, 1159–1183.
    155. Fanian, F.; Kuchaki Rafsanjani, M. Cluster-based routing protocols in wireless sensor networks: A survey based on methodology. J. Netw. Comput. Appl. 2019, 142, 111–142.
    156. Manuel, A.; Deverajan, G.; Patan, R.; Gandomi, A. Optimization of routing-based clustering approaches in wireless sensor network: Review and open research issues. Electronics 2020, 9, 1630.
    157. Wang, J.; Cao, J.; Sherratt, R.; Park, J. An improved ant colony optimization-based approach with mobile sink for wireless sensor networks. J. Supercomput. 2018, 74, 6633–6645.
    158. Wang, J.; Gao, Y.; Liu, W.; Sangaiah, A.; Kim, H.J. An improved routing schema with special clustering using PSO algorithm for heterogeneous wireless sensor network. Sensors 2019, 19, 671.
    159. Sarkar, A.; Senthil Murugan, T. Cluster head selection for energy efficient and delay-less routing in wireless sensor network. Wirel. Netw. 2019, 25, 303–320.
    160. Ezhilarasi, M.; Krishnaveni, V. An evolutionary multipath energy-efficient routing protocol (EMEER) for network lifetime enhancement in wireless sensor networks. Soft Comput. 2019, 23, 8367–8377.
    161. Alghamdi, T. Energy efficient protocol in wireless sensor network: Optimized cluster head selection model. Telecommun. Syst. 2020, 74, 331–345.
    162. Deepa, O.; Suguna, J. An optimized QoS-based clustering with multipath routing protocol for Wireless Sensor Networks. J. King Saud Univ. Comput. Inf. Sci. 2020, 32, 763–774.
    163. Arjunan, S.; Sujatha, P. Lifetime maximization of wireless sensor network using fuzzy based unequal clustering and ACO based routing hybrid protocol. Appl. Intell. 2018, 48, 2229–2246.
    164. Kaur, T.; Kumar, D. Particle Swarm Optimization-Based Unequal and Fault Tolerant Clustering Protocol for Wireless Sensor Networks. IEEE Sens. J. 2018, 18, 4614–4622.
    165. Khabiri, M.; Ghaffari, A. Energy-Aware Clustering-Based Routing in Wireless Sensor Networks Using Cuckoo Optimization Algorithm. Wirel. Pers. Commun. 2018, 98, 2473–2495.
    166. Liang, H.; Yang, S.; Li, L.; Gao, J. Research on routing optimization of WSNs based on improved LEACH protocol. Eurasip J. Wirel. Commun. Netw. 2019, 2019.
    167. Huang, W.; Ling, Y.; Zhou, W. An Improved LEACH Routing Algorithm for Wireless Sensor Network. Int. J. Wirel. Inf. Netw. 2018, 25, 323–331.
    168. Moorthi; Thiagarajan, R. Energy consumption and network connectivity based on Novel-LEACH-POS protocol networks. Comput. Commun. 2020, 149, 90–98.
    169. Bhola, J.; Soni, S.; Cheema, G. Genetic algorithm based optimized leach protocol for energy efficient wireless sensor networks. J. Ambient. Intell. Humaniz. Comput. 2020, 11, 1281–1288.
    170. Jain, A.; Goel, A. Energy Efficient Fuzzy Routing Protocol for Wireless Sensor Networks. Wirel. Pers. Commun. 2020, 110, 1459–1474.
    171. Wang, J.; Gao, Y.; Yin, X.; Li, F.; Kim, H.J. An Enhanced PEGASIS Algorithm with Mobile Sink Support for Wireless Sensor Networks. Wirel. Commun. Mob. Comput. 2018, 2018.
    172. Wang, J.; Gu, X.; Liu, W.; Sangaiah, A.; Kim, H.J. An empower hamilton loop based data collection algorithm with mobile agent for WSNs. Hum. Centric Comput. Inf. Sci. 2019, 9.
    173. Somauroo, A.; Bassoo, V. Energy-efficient genetic algorithm variants of PEGASIS for 3D Wireless Sensor Networks. Appl. Comput. Inform. 2019.
    174. Zhang, B.; Wang, S.; Wang, M. Area double cluster head APTEEN routing protocol-based particle swarm optimization for wireless sensor networks. Eurasip J. Wirel. Commun. Netw. 2020, 2020.
    175. Wang, M.; Wang, S.; Zhang, B. APTEEN routing protocol optimization in wireless sensor networks based on combination of genetic algorithms and fruit fly optimization algorithm. Ad Hoc Netw. 2020, 102.
    176. Wang, C.; Wang, S. Research on Uneven Clustering APTEEN in CWSN Based on Ant Colony Algorithm. IEEE Access 2019, 7, 163654–163664.
    177. Ma, J.; Wang, S.; Meng, C.; Ge, Y.; Du, J. Hybrid energy-efficient APTEEN protocol based on ant colony algorithm in wireless sensor network. Eurasip J. Wirel. Commun. Netw. 2018, 2018.
    More