Taxonomy of Routing Protocols in WBAN
Over the years, a lot of WBAN routing protocols has been proposed in the literature. These protocols could be segregated into five categories, i.e., Thermal-aware routing protocols, Energy-Aware based routing protocols, QoS-aware based routing protocols, Postural-movement-based routing protocols, Cluster-based routing protocols and Cross-layered routing protocols [
46]. The overview of all these protocols is presented in
Figure 1.
The sensor nodes of WBAN are embedded and both produce electrical and magnetic fields during the transmission [
25]. This issue can result in absorption of radiation and increased heat in the human body that leads to harm of sensitive organs by the minute temperature rise. For instance, lens contraction gets effected due to lack of blood flow occurred because of temperature rise of a node [
16]. In [
45] a Specific Absorption Rate (SAR) is introduced as an indicator at which the energy is immersed from the body of human when it gets revealed to a radio frequency (RF) electromagnetic field. The formula to calculate SAR is presented in Equation (
1).
where
is electrical conductivity of body tissue,
E is induced electric field by radiation,
is density of body tissues,
W and
Kg represent the units of watts per kilogram. Primarily,
SAR defines the upper limit of the transmitting power. Therefore, WBAN routing protocols should be designed to reduce the radiation emission to transmit data. Considering safety, radiation absorption is required to be reduced either by controlling
SAR or using an appropriate routing protocol.
Bag et al. in [
27], a model is proposed to handle the upshot occurred during the communication by the bio-medical sensors on human body. In this study, a rate controlled recommendation model for WBAN transformation is presented based on the observation. Thermal-aware routing techniques are significant to discuss while designing the Intra-WBAN. To evade the temperature rise and radiation that cause harm to the human body, diversified routing protocols related to thermal-aware have been introduced in the literature. The contemporary thermal-aware routing protocols discussed in the literature include, TARA [
25], LTR [
26], ALTR [
26], LTRT [
27], HPR [
28], RAIN [
29], TSHR [
30], M-ATTEMPT [
31], TMQoS [
32], RE-Attempt [
33],
[
34], TLQoS [
35], DRC [
36], TTRP [
37], TRPL [
38].
The first temperature sensitive routing algorithm i.e., Thermal-Aware Routing Algorithm (TARA) is presented in [
25], Tang et al. The proposed algorithm aims to reduce the possibility of temperature rise of the implanted bio-medical sensor in the human body. TARA defined the overheated nodes as hotspots and routes away the data from these nodes during communication. The author defined the Temperature Increase Potential (TIP) to calculate the temperature rise of a node on the basis of SAR. During data transmission, each node observes the communication of neighboring nodes by counting the packets, calculating communication radiation and power consumption to evaluate the present temperature of nodes. This algorithm does not involve a node in data transmission when its temperature rises above the defined threshold value. TARA enforces the equal temperature dissemination among all the nodes within the network by considering temperature as a metric. As the temperature is the only metric, TARA has to face the issues such as network lifetime, packet drop and reliable transmission.
Least Temperature Rise (LTR) is an enhanced version of TARA to address the overheating of implanted sensors and delay. In LTR is proposed in [
26], the selection of the next hop is based on the temperature of neighboring nodes and the node with least temperature is selected as a next hop. To avoid the bandwidth consumption, predefined
MAX_HOP limit is defined in algorithm and when a packet exceeds that limit it is discarded. Previously visited lowest temperature node is saved in the list to avoid loop in the network. Loops are avoided by maintaining a list in the packet with the recently visited nodes.
Adaptive Least Temperature Routing (ALTR) protocol [
26] overcomes the problem faced in LTR. The approach of ALTR is similar to LTR with the variation changing
MAX_HOP by
MAX_HOP_ADAPTIVE. In ALTR, instead of discarding the packet after it exceeds the threshold value of
MAX_HOP_ADAPTIVE, it works as Shortest Hop Algorithm (SHA) to transfer the packet to the destination.
Least Total Route Temperature (LTRT) [
27] is the combination of LTR and shortest path routing [
40]. In LTRT, the total route is selected from source to destination on the basis of temperature. LTRT selects a least temperature route to forward data instead of only considering the next hop. The temperature of sensors is transformed into graph weights and minimum temperature routes are obtained using Dijkstra’s algorithm. During the data transmission, the temperature of a sensor node rises by one unit upon receiving or transmitting data and decreases by one unit when it is inactive in transmission (defined time interval). In LTRT, hop count per packet and packet drop is decreased. It also aids in reducing overheating of sensors but the shortcoming in LTRT is that it requires to know the temperature of all sensors in the network. Moreover, energy consumption and overhead of acquiring the record of all temperatures is not measured.
Hotspot Preventing Routing (HPR) [
28] is another thermal-aware routing protocol that addresses delay during data transmission. It includes two phases i.e., setup phase and routing phase to avoid the delay and forming of hotspots in the network. In the first phase, all the sensor nodes exchange information about the shortest route and temperature of the nodes to build their routing tables. In routing phase, each node transfers data using the shortest available route in the routing table and maintains a hop-count. If the value exceeded than the threshold value of hop-count, the packet is discarded. If neighboring node is the destination node then the packet would be directly sent to it; otherwise, it chooses next hop from the shortest path based on temperature (equal or less than the threshold value). The threshold temperature of a sensor node is obtained by taking the average temperature of the neighboring nodes (including the temperature of sending node itself). In this approach, all nodes compute the threshold value of temperature on the basis of the temperature of its neighboring nodes and the temperature of the node itself, however, in other algorithms this value is predefined.
Bag et al. in [
29] presented the goal of Routing Algorithm for Network of Homogeneous and ID-Less Bio-Medical Sensor Nodes (RAIN) is to decrease the average temperature rise and average power consumption of sensor nodes. It constitutes three phases i.e., setup phase, routing phase and status update phase. In the setup phase, temporary IDs are random, usable for its operational life and ID ‘zero’ is kept for the sink node. All nodes broadcast their IDs using hello packet. In the routing phase, a packet ID [
N,
T and
R] is forwarded towards the destination, where
N is the ID of the sensor node,
T indicates the time at which the data packet is generated at the sending node and
R is any random number. Loops are avoided by maintaining a hop-count with every packet and if the hop-count of the packet exceeds the threshold value of
, then it is discarded. The nodes also avoid retransmission by maintaining the list of the packets IDs and a data packet is dropped if its ID is previously resides in that list. Every node knows the temperature of neighboring nodes by observing their communication actions. The packet is forwarded directly if the destination is one of the neighboring nodes otherwise it is forwarded to the neighbor node using probability which is inversely proportional to its temperature. In status update phase, the sink notifies all its neighboring nodes upon receiving any data packet by sending the packet’s ID to decrease the power consumption.
Shortest Hop Routing (SHR) is used to transfer the packet to the destination. However, it only considers average temperature metrics and having less Network lifetime and high end-to-end delay. Thermal-aware shortest hop routing protocol diminishes the overheating without influencing the packet delivery proportion, delay and power utilization of bio-restorative sensor hubs. Tabandeh et al. in [
30] claimed that TSHR is optimal for those applications that have high priority data to exchange the packets to the destination. The copy of transmission packet is additionally considered in TSHR. It incorporates two stages: setup stage and routing stage. In setup stage, routing tables are created while in routing stage every node transmits the packets to the destination on the basis of the shortest path.
Huang et al. in [
47] proposed a technique in which each sensor node is placed in different ring level. The ring level depicts the distance of each sensor node from the destination in the form of hop-count.
Abbas et al. in [
31], proposed an energy efficient and thermal-aware routing protocol for WBAN to diminish the temperature of nodes. In addition to diminish the delay for the critical data by utilizing heterogeneous bio-therapeutic sensor nodes MATTEMPT involves four stages that are initialization stage, routing stage, scheduling stage and data transmission stage. In initialization stage, all nodes broadcast the hello packet, while in routing stage, courses with fewer hop counts are chosen within the available routes based on the temperature and energy metrics. The sink nodes make a Time Division Multiple Access (TDMA) plan for all root nodes in the scheduling stage, while the root nodes send their data to the sink hub amid the data transmission stage.
TMQoS is a thermal-aware multi-constrained intrabody routing protocol. The basic objective of this protocol is to keep the temperature of node at an adequate level. Hassan et al. in [
32], proposed TMQoS, which is a proactive routing protocol that builds ongoing routing table which includes multiple shortest path routes. Beacon packet is used to exchange the information between sensor nodes for the formation and maintenance of routing table. The beacon packet is received by neighboring nodes through mac receiver module. Routing table constructor module is used for formation and maintenance of routing table in term of hop-count, delay and temperature value. It also has hotspot avoidance mechanism that keeps the packet away from hotspot area.
Reliability Enhanced-Adaptive Threshold-based Thermal-aware Energy-Efficient Multi-hop Protocol (RE-ATTEMPT) is a thermal and energy efficient routing protocol. Javaid et al. [
33] presented that sink is placed between the nodes and nodes are sorted in descending order in term of data rate. The higher data rate node is placed where mobility is less and low data rate nodes are placed at mobility area of human body. This protocol uses direct communication for critical data and multi-hop communication for non-critical data. The protocol is divided into five phases: initialization phase, routing phase, scheduling phase and data transmission phase. The initialization phase is responsible for calculating the distance of nodes from sink node in the form of hop-count by broadcasting “Hello Message”. In this phase, all nodes are aware of their neighboring nodes, available routes and distance from sink. The routing phase is responsible for selecting an efficient route with less hop-count. In routing phase, critical data is directly sent to sink using direct communication while normal data is sent using multi-hop communication. In scheduling phase, time slot is assigned to sink and root nodes for communication of data. The root and sink nodes communicate within their assigned time slot. The last phase is data transmission phase wherein data is sent to sink node within their allocated time slot.
Rafatkhan et al. in [
34]
Multi-hop Routing Protocol is a thermal and energy efficient routing protocol. The protocol is divided into four phases including initialization, routing, scheduling and data transmission phase. The initialization phase is responsible for calculating distance of nodes from sink node by broadcasting “Hello Message”. The routing phase is responsible to send data to the medical server. If Home-Signal is received then one of the implanted nodes in the human will establish a link with routing table of the home node. In this way, a direct link is used to send the data from sink node to medical server. In the direct link, there is no energy restriction because critical data is sent over the selected link. If Home-Signal is not received then routing table formed from implanted nodes and multi-hop link get established for communication of data. This phase is also responsible for hotspot avoidance. In scheduling phase, time slot is assigned to sink and root nodes for communication of data. The root and sink nodes communicate within their assigned time slot. The last phase is data transmission phase wherein data is sent to sink node in their allocated time slot.
Thermal-aware localized QoS routing protocol (TLQoS) is a thermal-aware routing protocol which aims to address the required QoS demand of implanted bio-medical sensor in a human body. Monowar et al. in [
35] presented the temperature of the node at an acceptable level to avoid heating or damaging of tissue. The protocol also prevents loop formation during communication and avoid unnecessary routes with a large number of hop from sink to control delay in the network. The localized approach is used for selecting suitable forwarder nodes among all neighboring nodes in a routing table. Monowar et al. [
36] introduced an integrated rate control technique which ultimately aims to decrease the thermal effect of the implanted bio-medical sensor node and avoid congestion during communication between nodes.
Trust and Thermal-Aware Routing Protocol (TTRP) is another thermal-aware routing protocol which provides trusted and hotspot free network communication between the implanted bio-medical sensor nodes. Kumar et al. in [
37] stated that the communication is protected from malicious and faulty nodes within or outside the network. The TTRP is divided into three phases: trust estimation phase, route discovery phase and route maintenance phase. The trust estimation phase is further divided into two sub-phases: direct trust and indirect trust. The trust estimation phase is responsible for evaluating the trustiness of intermediate nodes. The route discovery phase is responsible for selecting trustworthy and hotspot free node route for communication in the network. In route maintenance phase, if the node becomes hotspot during communication inactive route then route maintenance phase re-initiates the route discovery phase.
The self-healing thermal-aware RPL routing protocol is a self-adaptive routing protocol. In this protocol, if the node is marked as a hotspot node, then the node itself decides the efficient path for communication of data. The selection of efficient path is based on low temperature and low power metrics. In [
38] the protocol work on IPv6 routing protocol for low power and lossy networks (RPL).
In Multipath ring routing (MRRP) [
47], multiple paths are constructed to the destination to avoid congestion in the network. The protocol is divided into two phases: Multipath construction phase and Data transmission phase. In Multipath construction phase, a setup request packet is sent by the sink to all its neighboring nodes with ring level 0. Nodes which received packet will update its ring level by incrementing its value to 1 in it and broadcast the packet. The main aim of this phase is to organize the sensor nodes in the form of ring level. At the end of this phase, each node is assigned a ring level value and these nodes are separated according to their ring level value. In data transmission phase, each node is arranged according to their assigned ring level value. The sensor node sends the packet, broadcasts the packet to its neighboring node and the process continues until the packet is reached to the sink node.
Ordinarily, the little size of the battery in WBAN suggests that the capacity of the battery is low with little energy. Different schemes have been utilized to utilize power eruditely for the broadened life of remote systems (wireless network). Some power sparing schemes maintain a strategic distance from repetitive retransmissions, diminish the frequency of sending system control messages, lessening the size of message headers and utilizing the standby or sleep mode when conceivable [
48]. Among the various difficulties of WBAN, energy utilization is an essentially vital long-term arrangement of WBANs, which is elaborated in various studies (ESR [
14,
15,
16,
17,
23]). Here we shed a light on some of the energy-aware routing protocols including SIMPLE [
49], Co-CEStat [
50], RSSI [
51], MEPF [
52], DARE [
53] and ESR [
54].
The routing algorithm of QoS- aware are modular based and use various modules for different types of QoS metrics. The complexity and synchronization of the modules make it a challengeable task to design a QoS protocols. To provide QoS in WBAN, routing demands significant consideration as the patient’s data (normal or critical) should be transmitted to medical server efficiently and reliably. Broadly, the QoS routing protocols can be classified into reliability-based routing and delay-tolerant based protocols. The reliability-based routing does not consider delay and ensures throughput. The delay-tolerant based routing protocols guarantee in-time delivery of packets. Some of the important QoS-aware routing protocols are RL-QRP [
55], LOCALMOR [
56], DMQoS [
57], EPR [
53], QPRD [
58] and QPRR [
59] etc.
The third routing class of WBAN is cluster-based routing that is known as an appropriate protocol for increasing the lifespan of a network by reducing energy consumption. In this approach, WBAN network is divided into clusters and each cluster chooses Cluster Head (CH) that is responsible for communication to the base station to minimize the energy consumption of the other sensing nodes. The responsibilities of a CH include collecting, aggregating and forwarding of data. The two main cluster based WBAN routing protocols are Hybrid Indirect Transmissions (HIT) [
60] and Anybody [
61].
In WBAN topology the communication among the nodes frequently come across with the issues of disconnection and partitioning due to postural movements of the human body. Various studies have been conducted to resolve this issue and a cost function is defined in WBAN, that is updated at intervals. These protocols send packet with least cost on the basis of path selection from bio-medical sensor to base station. The routing protocols based on postural movements are postural-based routing protocol Probabilistic Routing (PRPLC) [
62], On-Body Store and Flood Routing (OBSFR) [
63], Dynamic Postural Partitioning (DVRPLC) [
64], opportunistic routing protocol [
65] and Energy Efficient Thermal and Power Aware Routing (ETPA) [
66].
A Cross-layered routing protocol is the fifth class of WBAN routing. The Cross-layered scheme improves the synchronization between the layers without disturbing the main functions of the layers. The research community remained interested in the areas related to WSN because of its effective implementation in WSN. These protocols contribute in improving the performance of the network by addressing the challenges and issues faced in network and MAC layers simultaneously. The routing protocols based on cross-layered routing protocols are Wireless Autonomous Spanning Tree Protocol (WASP) [
67], Cascading Information Retrieval by Controlling Access with Dynamic Slot Assignment (CICADA) [
68], Timezone Coordinated Sleeping Scheduling (TICOSS) [
69], Biocomm [
70] and Biocomm-D [
70].
Based on the comprehensive analysis of state-of-the-art approaches in the field, limitations of available thermal-aware techniques are described in
Table 1.