Review Article

A Comprehensive Survey on Hierarchical-Based Routing Protocols for Mobile Wireless Sensor Networks: Review, Taxonomy, and Future Directions

Table 2

Comparison of the hierarchical-based routing protocols for MWSNs.

ProtocolDelayNetwork sizeEnergy-efficiencyScalabilityAdvantagesDisadvantages

LEACH-Mobile 2006 [27]HSmall scaleLLtd(i) Supports applications that require mobile sensor nodes.
(ii) It does not require the knowledge of the global network.
(i) Assumes that CHs are static.
(ii) It is not efficient in terms of energy consumption and data delivery rate.
(iii) Consumes more energy in idle listing and overhearing.
LEACH-ME 2008 [28]HSmall scaleLLtd(i) Selects CHs that have less mobility factor.
(ii) Solves the problem of LEACH-Mobile protocol.
(i) Consumes much energy for determining mobility factor of each node.
(ii) Requires extra slot for calculating the mobility factor and sending it to CH.
MSRP 2010 [29]HSmall/large scaleMG(i) Alleviates the hotspot problem.
(ii) Controls the movement of the sink based on the residual energy of CHs.
(i) It does not guarantee that all CHs in the network can communicate with sink during a round.
(ii) Requires more overheads for registration of CHs with a sink in each round.
CBR-Mobile 2011 [30]MSmall scaleLLtd(i) Supports the mobility of WSNs in efficient manner.
(ii) Maximizes delivery ratio and minimizes average delay.
(i) Requires more overhead for forming the two database tables.
(ii) Consumes more energy in control overheads.
MBC 2011 [31]MSmall/large scaleMG(i) Uses residual energy and mobility in CHs selection.
(ii) Considers connection time in construction of clusters.
(i) Fails to address the critical node occurrence problem, which causes link breakage, packet dropping, and reducing of network utilization.
EECC 2012 [32]HSmall scaleLLtd(i) Mitigates the hotspot problem.
(ii) Improves the energy utilization and the network lifetime.
(i) Ignores the transmission delay.
(ii) The sojourn locations of the mobile sink are fixed in each round, which drains the energy of neighbors to these locations.
MIEEPB 2013 [33]MSmall scaleMLtd(i) Avoids formation of long link.
(ii) Selects leader of each chain based on residual energy and distance from BS.
(i) Uses fixed sojourn locations for the mobile sink in each round, which causes hotspots.
(ii) Requires location information.
ECBR-MWSN 2013 [34]HSmall scaleMLtd(i) Prolongs the lifetime of MWSNs.
(ii) Balances the energy consumption among the nodes.
(i) Its scalability is limited.
(ii) Its overhead is high.
MACRO 2014 [35]MSmall scaleLLtd(i) Discovers the best route based on the link quality and mobility of nodes.
(ii) Sustains the reliability of a route.
(i) Causes a significant delay in discovering a route due to the higher number of nodes and the frequent topology change of MWSN.
CIDT 2014 [36]MLarge scaleMVG(i) Minimizes the energy exploitation, end-to-end delay, and traffic of CH due to transfer of data with DCT.(i) Fails to maintain the data rate, traffic, tree intensity, and coverage distance of the tree structure.
VELCT 2015 [37]LLarge scaleMVG(i) Overcomes the drawbacks of CIDT protocol.(i) Requires fault tolerance to maintain the network connectivity.
PHASeR 2015 [38]LSmall scaleMLtd(i) Uses the blind forwarding technique to pass messages through the network in a multipath manner.
(ii) Uses the hop-count gradient in forwarding messages.
(i) It suffers from the stability problem.
(ii) Authors do not study the effect of channel fading on the performance of PHASeR.
Optimizing LEACH 2015 [39]MSmall/large scaleMLtd(i) Combines the concepts of a mobile sink and rendezvous nodes to reduce the consumption energy.
(ii) Considers the residual energy in CH selection.
(i) Requires the locations of sensor nodes to select the rendezvous nodes.
(ii) The mobile sink moves along predefined path.
Ring protocol 2015 [40]LSmall/large scaleMLtd(i) Minimizes the overheads of broadcasting the sink position.
(ii) Accelerates the delivery of the sensory data to the sink.
(i) Assumes that each node should know its position and position of its neighbors.
(ii) The procedure of ring construction is repeated until a closed loop is acquired and this consumes more energy in processing and overheads.
EMMS 2016 [41]HSmall scaleHG(i) Improves the energy- efficiency and data transmission quantity within the network significantly.
(ii) Uses multiple sinks to balance the load among the nodes.
(i) The sinks control process requires more overheads.
(ii) Its time complexity is high.
Anycast protocol 2016 [42]HLarge scaleMLtd(i) Uses unicast mode instead of broadcast mode.
(ii) Reduces the network traffic and extends the lifetime of nodes.
(i) Suffers from the scalability problem.
(ii) Requires high delay to establish a routing tree.
PSO-MBS 2011 [43]HSmall scaleMLtd(i) Improves the network lifetime and the data delivery.
(ii) Considers the distances between the sensor nodes and sink in the fitness function of PSO.
(i) Requires locations of sensor nodes.
(ii) Increases the packets delay due to waiting for CH when BS visits the cluster.
GAROUTE 2011 [44]HSmall scaleLLtd(i) Selects CHs based on the energy consumption and speed.
(ii) It does not need the location information of sensor nodes.
(i) It does not ensure that all mobile nodes can participate in the clustering process.
(ii) Requires the energy information and the list of neighbor of each node, which increases the overheads and consumes more energy.
Rendezvous algorithms 2012 [45]HSmall scaleLLtd(i) Finds the tour of BS and a set of RPs on the tour based on minimizing the transmission cost.(i) It does not ensure the maximization of the network lifetime.
(ii) Introduces some of delay during the data collection.
Sudarmani and Kumar 2013 [46]HSmall/large scaleLLtd(i) Solves the hotspot problem.
(ii) Considers the concept of adaptive transmission power control to balance the network load.
(i) It does not consider the transmission delay.
(ii) Requires locations of sensor nodes.
CAGM 2015 [47]HSmall scaleMLtd(i) Balances the energy consumption among the sensor nodes.
(ii) Utilizes the swarm optimization to find the best CHs.
(i) Increases the packets delay due to waiting for CH when BS visits the cluster.
(ii) Its overhead is high because it requires information of sensor nodes for optimization.
NSGAII-RP 2015 [48]MSmall scaleMLtd(i) Considers both the coverage and the routing problem.
(ii) Controls the mobile nodes to increase coverage and lifetime.
(i) Authors ignore the dissipated energy in movement of sensor nodes.
(ii) Requires the coordinates of all sensor nodes in the networks.
OZEEP 2015 [49]NASmall/large scaleHG(i) Utilizes the GFS to select the optimal CHs.
(ii) Considers energy, density, mobility, and distance in the CH selection.
(i) It does not consider the packets delay.
(ii) The disconnected nodes wait for the reclustering process of the next round to join with a new CH.
LOA-MSN 2015 [50]LSmall/large scaleHG(i) Considers the constraints of movement path, flow, energy consumption, and link transmission in finding the movement paths of the mobile sink.(i) Its cost is high due to using multiple sink nodes.
(ii) Requires the locations information of all nodes in the networks.
MLS 2015 [51]HLarge scaleHVG(i) Constructs a routing tree with avoiding burdening the nodes that already have heavier loads.
(ii) Maximizes the lifetime of large-scale network within a predefined delay tolerance level.
(i) Collecting data from high-density network increases the delay.
(ii) Using a fixed path for the mobile sink drains the energy of the closest nodes to this path, and as a result energy holes are produced.
Yue et al. 2016 [52]HSmall/large scaleMLtd(i) Uses the ABC algorithm to minimize the movement path of the mobile sink and maximize the data collection.
(ii) Selects the best CHs based on minimizing the consumption energy.
(i) Requires more overheads to find the best CHs and the sojourn location of the mobile sink.
(ii) Needs more delay to collect data from all CHs.

L: low, M: medium, H: high, VH: very high, Ltd: limited, G: good, VG: very good, and NA: not available.