Temperature State Awareness-Based Energy-Saving Routing Protocol for Wireless Body Area Network

Abstract

As an emerging information technology, Wireless Body Area Networks (WBANs) provide a lot of convenience for the development of the medical field. A WBAN is composed of many miniature sensor nodes in the form of an ad hoc network, which can realize remote medical monitoring. However, the data transmission between sensor nodes in the WBAN not only consumes the energy of the node but also causes the temperature of the node to rise, thereby causing human tissue damage. Therefore, in response to the energy consumption problem in the Wireless Body Area Network and the hot node problem in the transmission path, this paper proposes a temperature state awareness-based energy-saving routing protocol (TSAER). The protocol senses the temperature state of nodes and then calculates the data receiving probability of nodes in different temperature state intervals. A benefit function based on several parameters such as the residual energy of the node, the distance to sink, and the probability of receiving data was constructed. The neighbor node with the maximum benefit function was selected as the best forwarding node, and the data was forwarded. The simulation results show that compared with the existing M-ATTEPMT and iM-SIMPLE protocols, TSAER effectively prolongs the network lifetime and controls the formation of hot nodes in the network.

1. Introduction

The Wireless Body Area Network (WBAN) is an important branch of Wireless Sensor Networks (WSNs). The WBAN is mainly used in the medical and health field. It can realize early identification and prevent patients’ potential diseases through remote medical monitoring [1]. A WBAN is composed of miniature sensor nodes implanted in the human body or worn on the surface of the human body through an ad hoc network. These sensor nodes sense various physiological data of the human body and forward the sensed data to a sink. The sink further sends the data to a medical service terminal wirelessly for corresponding processing. Figure 1 presents the communication structure of a WBAN. The WBAN can realize the long-term and real-time monitoring of patients’ vital signs by medical terminals, and patients can grasp their health status more intuitively and obtain the emergency response capability for emergencies [2,3,4,5,6].

Although WBAN belongs to an important branch of WSN, WBAN has its unique network characteristics, which makes it difficult to establish communication paths between sensor nodes [7,8,9], such as human mobility and frequent data transmission between sensor nodes. The node temperature rises and the dead node caused by continuous energy consumption. Therefore, the routing protocol proposed for WSN is not suitable for WBAN [7].

Taking into account the small size of the sensor node in a WBAN, low battery capacity, and the difficulty of replacing the battery for implanted sensors, the power consumption of WBANs needs to be limited [10]. In order to ensure that WBANs can effectively transmit data under limited energy conditions, in addition to the circuit design of wireless sensor nodes to save energy, efficient energy-saving routing plays a vital role in reducing network energy consumption. The energy-saving routing of a WBAN often introduces the concept of multi-hop transmission into it [11,12], using forwarding nodes to forward data between the source node and the target node to reduce the energy consumption of long-distance transmission between the source node and the target node.

Furthermore, data transmission between sensor nodes not only consumes the energy of the node but also causes the temperature of the node to rise. The heat dissipation of nodes is another important challenge faced by the WBAN routing protocol. In WBANs, the reason for the increase in node temperature is that, on the one hand, the sensor nodes use radio frequency (RF) signal communication, and the radio frequency transmits the heat dissipated; on the other hand, it is the heat generated by the internal circuit of the sensor when processing the collected data [8]. When the temperature of a node increases to destroy human tissues or organs, it is called a hot node. The generation of hot nodes makes WBANs a security risk, and it is also not conducive to the normal transmission of data, which may cause data loss or delay. Therefore, the heat generated by sensor nodes is controlled to effectively transmit the data of WBANs in the routing process.

This paper proposes a temperature state awareness-based energy-saving routing protocol to solve the problem of energy consumption and hot nodes in WBANs. The main contributions are as follows:

• Two thresholds are set for the temperature state of the node, namely a high-temperature threshold and heating threshold. The heating threshold is set before the node reaches the high-temperature threshold. Use these two thresholds to divide the node temperature state into a normal interval, a heating interval, and a high-temperature interval, calculate the data reception probability of the node in the three temperature intervals, respectively, and delay the node temperature from reaching the high-temperature state.
• The efficiency function composed of the remaining node energy, the distance to sink, and the probability of receiving data is constructed; it compares the benefit function value of each neighbor node and selects the maximum value as the best forwarding node to transmit data to.
• The data are divided into emergency data and normal data; the corresponding forwarding node is selected according to the data type to forward the data.

The remaining work of this paper is arranged as follows: Section 2 describes the related work, and Section 3 introduces the network model, including the system model, the energy model, and the heat calculation model. Section 4 describes the specific method and routing process of the TSAER protocol, Section 5 simulates and analyzes the performance of TSAER, and Section 6 presents a summary and future prospects.

2. Related Work

In the past decade, many scholars have proposed corresponding routing protocols for WBANs to solve the problems of energy consumption and hot nodes.

In improving energy efficiency, energy-saving routing protocols use multi-hop transmission to reduce energy consumption. Nadeem et al. [13] proposed a routing protocol (SIMPLE) to improve throughput and energy efficiency, which considered parameters such as the remaining node energy and the distance to sink, used the cost function constructed by these parameters to select the forwarding node in the path to transmit data, and finally reduced the energy consumption of the network. The iM-SIMPLE protocol proposed by Javaid et al. [14] is an extension of SIMPLE. The protocol also uses the cost function to select forwarding nodes, but the difference between iM-SIMPLE and SIMPLE is that iM-SIMPLE takes into account the dynamic characteristics of the WBAN, and the mobility of nodes may lead to reduced throughput and high path loss. Therefore, iM-SIMPLE attributed the problem of minimizing energy consumption and maximizing throughput to an integer linear program for analysis, and finally, the protocol extended the network lifetime and increased throughput. The sentence was amended to read:

Anwar et al [15] proposed an energy-aware link efficient routing protocol (ELR-W), which integrates the residual energy of the node, link efficiency, distance to the coordinator, and number of hops in order to construct a path cost model. Using the path cost model to select the appropriate forwarding node to forward data, this protocol can effectively extend the lifetime of the network, but it is not suitable for different types of data transmission and multiple QoS requirements.

In controlling the formation of hot nodes [16], the thermal-sensing routing protocol aims to minimize the temperature rise in sensor nodes, the methods of sensing node temperature and setting temperature thresholds are often used to control the generation of hot nodes. TARA [17], proposed by Tang et al., is the first heat-aware routing protocol. The protocol mainly focuses on research on the temperature rise in sensors implanted in the human body. TARA uses the specific absorption rate (SAR) to calculate the temperature rise in the node and selects the lowest temperature node to transmit data during the transmission process. If a node receives data with a temperature above a threshold, it does not forward the data to the next hop node, but returns the data to the previous node. Using this backoff hot node strategy can effectively control the generation of hot nodes in the network. The hot spot prevention routing (HPR) proposed by Bag et al. [18] also solves the problem of data transmission delay. The protocol adopts the shortest route method and selects the next hop node to transmit data in the shortest path according to the temperature (equal to or less than the threshold). Kim et al. [19] proposed a multi-criteria decision-making enhanced mobility and temperature-aware routing protocol. The protocol uses Analytic Hierarchy Process (AHP) to logically determine the weights of parameters such as temperature, link quality, and number of hops; based on these parameter values and their weights, a simple weighting method is used to compute the cost of each node value, and the node with the smallest cost value is selected as the best oriented node to transmit data. The protocol uses the storage and carrying mechanism of mobile nodes to effectively reduce the number of hot nodes, but this mechanism will increase the end-to-end delay.

The temperature rise in the nodes in the network will not only consume a lot of energy but also cause damage to human tissues when the temperature is too high. Therefore, some scholars have optimized the design of routes related to energy saving and controlling the formation of hot nodes, so that network performance can be fully utilized [20,21,22,23,24]. This type of routing protocol has the characteristics of multi-hop transmission, sensing node temperature, and setting temperature thresholds. Javaid et al [25] proposed a Mobility Based Thermal Aware Energy Efficient Multi-Hop Protocol (M-ATTEMPT), this protocol supports WBANs to transmit sensitive data, on-demand data, and normal data using single hops. For multi-hop transmission, the selected best forwarding node has the characteristics of fewer hops and high residual energy. Moreover, M-ATTEMPT uses temperature thresholds to control the hot node problem in WBANs [26,27]. However, when the temperature of a node reaches the threshold, the node disconnects all connections with its neighboring nodes to prevent the temperature from rising, but the disconnection leads to data retransmission and reduced network reliability. Bhangwar et al. [28] proposed a weighted temperature and energy-aware QoS routing protocol (WETRP). In the routing decision process, the protocol uses the node’s temperature, remaining energy, and link delay estimation to construct a cost function and selects the node with the minimum cost function to transmit data. This protocol not only reduces energy consumption but also reduces the generation of hot nodes. However, this protocol cannot provide corresponding QoS requirements for multiple data types. A temperature and energy optimization routing protocol (TAEO) was proposed by Ref [26,27,29,30,31]. The protocol selects nodes with low temperature, a short path to sink, and high residual energy as the forwarding nodes to transmit data, and the detection of hot nodes was performed by estimating the temperature rise value of nodes in each round. When the temperature of the selected forwarding node is higher than the threshold, it will not forward data until the node returns to normal. TAEO can effectively reduce the generation of hot nodes and extend the life of the network.

3. System Model

3.1. Network Model

In this paper, we consider a WBAN consisting of a sensor node and a sink node, which is located at the waist of the human body. The sensor node is mainly responsible for forwarding the collected data to the sink. The sink node is mainly responsible for forwarding the data.

Model Assumptions and Features:

• All sensor nodes are placed at different locations of the human body to collect corresponding physiological information and have specific IDs.
• All sensor nodes have the same initial energy and transmission range.
• The sink node has strong information processing capability and receives data from sensors only, and the energy of the sink is not considered.

The sensor nodes classify the sensed physiological data into emergency data and common data based on their different levels of importance and then prioritize the data, and the priority of emergency data is higher than that of common data.

3.2. Radio Model

In the WBAN, each sensor node consumes energy during the processes of data sensing, processing, and transmission, but the data transmission process consumes the largest percentage of energy, so we use a first-order radio model for the calculation. The formula is as follows:

E_Tx(N, l) = E_Tx-elec × N + E_Amp × N × l²,

(1)

E_Rx(N) = E_Rx-elec × N.

(2)

where E_Tx(N, l) represents the energy consumed by sending data, E_Rx(N) represents the energy consumed by receiving data, N is the size of the data packet, l is the data transmission distance, and E_Tx-elec and E_Rx-elec represent the energy consumption of the circuit when the node sends data and the circuit when receiving data. E_Amp represents the energy consumed by the amplifying circuit.

In the WBAN, the human body as a communication medium attenuates the transmission signal. Therefore, the path loss parameter is added to the energy model, and the formula for calculating the energy consumption of the node is

E_Tx(N, l) = E_Tx-elec × N + E_Amp × N × c × l².

(3)

3.3. Thermal Model

In order to perceive the node temperature state, it is necessary to calculate the node temperature rise rate. The temperature rise rate can be calculated by the Pennis bioheat equation [19], as shown in Formula (4).

ρ T_s dT/dt = K∇²T − b(T − T_b) + ρSAR + P_c.

(4)

Here, T_s represents the specific heat of human tissue, dT/dt represents the rate of temperature increase, K∇²T indicates the temperature increase due to the thermal conductivity of human tissue, b(T − T_b) is the heat caused by blood perfusion in the human body, and Pc refers to the sensor heat caused by the circuit. In the analysis of this paper, consider the typical power consumption of a conventional sensor circuit. ρSAR represents the antenna radiation absorption, the specific absorption rate (SAR) measures the rate at which the human body absorbs heat [32], and its calculation formula is as follows.

SAR = σE²/ρ.

(5)

where the conductivity of human tissue is denoted using σ, the induced electric field is denoted using E, and ρ is used to represent tissue density.

The nodes are in space, with some treatment of the bioheat equation. Finite difference time domain (FDTD) is used in heating applications [33], FDTD discretizes the modeling of time and space by discretizing the entire network space into small grids, each represented by a pair of coordinates, with the following results:

T^m+1(i, j) = [1 − (δ_tb/ρT_s) − (4δ_t K/ρT_sδ²)] T^m(i, j) + (δ_t/T_s) SAR + (δ_tb/ρT_s) T_b + (δ/ρT_s) P_c + (δ_t K/ρT_sδ²)[T^m(i + 1, j) + T^m(i, j + 1) + T^m (i − 1, j) + T^m (i, j − 1)].

(6)

where T^m+1 (i, j) is the temperature of grid (i, j) at time m + 1, δ_t represents a discrete time step, and δ is a step in a discrete space, that is, the size of the grid. It can be seen from Formula (6) that the temperature of T^m+1 (i, j) is calculated by the composition of T^m (i, j) and (T^m (i + 1, j), T^m (i, j + 1), T^m (i − 1, j), T^m (i, j − 1)).

4. TSAER: Protocol Description

4.1. Proposed Parameters

The main parameters considered in this paper are the node temperature, the remaining node energy, and the distance to sink, which are described in detail as follows.

Temperature T: Estimate the current node temperature using the Pennis bioheat equation. When the node forwards each data packet, it will cause its temperature to rise by one unit.

Remaining energy: the remaining energy refers to the difference between the initial energy of the node and the energy consumed. The calculation expression of the remaining energy of the node is as follows:

E_res(i) = E_initial(i) − ∑[E_tx(i) + E_rx(i) + E_ax(i)].

(7)

where E_initial(i) represents the initial energy of node i, E_tx(i) represents the energy consumption of node i to send data, E_rx(i) represents the energy consumption of node i to receive data, and E_ax(i) represents the energy consumed by node i to process and aggregate the data it has collected or received. This aggregation process, typically involving computational tasks, such as in reducing data volume and redundancy, consumes energy for processing and memory access. The forwarding node has the ability of data aggregation to calculate the energy consumption of data aggregation, while other nodes do not need the data aggregation function, so the energy consumption of data aggregation is not calculated.

Distance between node and sink: the expression of the distance between node i and the sink is as follows:

D(i, Sink) = ((X_i − X_Sink)² − (Y_i − Y_Sink)²)^1/2.

(8)

where D(i, Sink) represents the distance between node i and the sink, X_i and Y_i represent the position coordinates of node i, and X_Sink and Y_Sink represent the position coordinates of the sink.

4.2. Route Benefit Description

4.2.1. Performance Parameter Selection

Data reception and transmission between sensor nodes will cause the node temperature to rise, and the temperature rise not only affects the data transmission reliability but also causes damage to human tissues in severe cases. In order to reduce the problem of node temperature rise caused by frequent data transmission between sensor nodes, this paper proposes a method to adjust the probability of node data reception by sensing the temperature state of the node. The node dynamically adjusts the probability of receiving data according to its own temperature value, so that the data are more evenly distributed in the network, which is beneficial to delay the rise in the temperature of the node.

The node temperature state is divided into three intervals, namely the normal temperature interval, the heating interval, and the high-temperature interval. t is the current temperature value, as shown in Figure 2 as the temperature state interval.

Here, T_in is the initial value of the temperature, T₁ is the critical value of the heating range, and T_th is the critical value of the high-temperature range.

A node can adjust its reception probability according to different types of data and temperature state intervals. P is the data reception probability of the node, the probability of receiving urgent data is P1, and the probability of receiving normal data is P2. The detailed description is as follows.

When the temperature of the node is in the normal temperature interval, the node can forward and receive any type of data; therefore, P1 = P2 = 1.

When the temperature state of the node is in the heating interval, there are different receiving probability calculation methods for two different types of data to not affect the data transmission. Emergency data are still forwarded with probability 1, while normal data are forwarded with varying probability.

In the heating interval, in order to achieve a more obvious effect of suppressing the temperature rise in the node, the data reception probability P2 of the node at this time should be reduced as the temperature of the node rises. Therefore, the change trend of P2 should be a decreasing concave function. Consider the simple quadratic function f(t) as the change function describing P2; the maximum value is at T₁, f(T₁) = 1, and the minimum value is at T_th, f(T_th) = 0. Its formula is expressed as follows.

P2 = f(t) = (t² − 2 T_tht + T_th²)/(T₁ − T_th)².

(9)

When the temperature of the node is in the high-temperature interval, it is easy to cause damage to human tissues. At this time, the node no longer receives any data. Therefore, P1 = P2 = 0.

Sink nodes always receive data with probability 1, and the receiving probabilities of nodes in different temperature intervals are shown in

Table 1. Data receiving probability of nodes in different temperature intervals.

Data Receiving Probability P	Normal Temperature Interval	Heating Interval	High-Temperature Interval
Emergency data P1	1	1	0
Normal data P2	1	(t2 − 2 Ttht + Tth2)/(T1 − Tth)2	0

.

4.2.2. Benefit Function Calculation

When the source node cannot directly communicate with the sink, the forwarding node needs to be used to transfer the data. The selection of the best forwarding node uses the benefit function (BF) as a basis. The benefit function is composed of the remaining node energy, the distance to sink, and the data reception probability. The calculation formula is as follows:

BF(i) = (E_res(i)/D(i, Sink)) × P.

(10)

where BF(i) is the benefit function of node i, E_res(i) is the residual energy at node i, D(I, Sink) is the distance between node i and the sink, and P is the probability of the node receiving data.

Since the data are divided into two types in this paper, the benefit function of emergency data is calculated by the following formula:

BF(i) = (E_res(i)/D(i, Sink)) × P1.

(11)

The benefit function of normal data is calculated using the following formula:

BF(i) = (E_res(i)/D(i, Sink)) × P2.

(12)

After each node calculates the benefit function, the neighbor node with the maximum benefit function is selected as the best forwarding node of the path. The source node selects the corresponding forwarding node to transmit the data according to the data type, and the forwarding node receives, aggregates, and transmits the data to the sink. The BF table of sensor nodes is updated every 10 s once the system reaches stability.

4.3. TSAER: Operational Steps

In order to balance the energy consumption of the WBAN and control the hot node formation, this paper proposes a temperature state awareness-based energy-saving routing protocol. The protocol is mainly divided into three processes: initialization, routing, and data transmission. Figure 3 is the flow chart of the protocol.

4.3.1. Network Initialization

In the initialization phase, the sink broadcasts an information packet to other nodes, including the location of the sink. After receiving this information packet, the other sensor nodes will obtain the location information of the sink and broadcast a HELLO message packet (HM) to the neighbor nodes. The HELLO message packet includes the node ID, remaining energy, temperature, and distance to the sink. The inside of each sensor node will build or update the neighbor table (NT) according to the information in the HELLO packet. The neighbor table includes the information of the local node and the neighbor node.

The intercommunication of information between sensor nodes facilitates the routing of nodes based on the information in the neighbor table. Algorithm 1 presents the neighbor table generation process.

Algorithm 1 Neighbor table generation process

Input: HMj (HM from neighbor node j) Output: NTi (The neighbor table of node i) Start 1.   For each HM do 2.     If (HMj ≠ NTi) then 3.          Update NTi 4.          NTi(T, Eres, D) ← HMj(T, Eres, D) 5.        else 6.          Discard HMj 7.      End If 8.   End For End

4.3.2. Routing Process

When the source node has data to be transmitted, if there is a path to sink, they will be forwarded directly; if it does not exist, it will enter the routing process.

In the routing process, the source node sends a routing request packet to its neighbor nodes. After receiving it, the neighbor node calculates the routing benefit and confirms the reply to the source node. Then, the source node compares the routing benefits of each neighbor node and selects the node with the maximum benefit function as the forwarding node transmits the data.

The process of calculating the routing benefit: The node calculates the data reception probability corresponding to emergency data or normal data according to the information in the neighbor table and then calculates the corresponding benefit function value. As shown in Algorithm 2, the routing benefit calculation instructions for different types of data are shown. And further routing process is elucidated by Algorithm 3.

Algorithm 2 Routing benefit calculation for different types of data

Input: Data Type DT, Temperature T, Residual Energy Eres, Distance from Sink D Output: Benefit function BF(i) Start 1. Obtain the temperature information of node and divide the temperature status interval. 2. Calculate the data receiving probability of node under different temperature states. 3.     If T = Normal temperature interval then 4.       P1 = P2 = 1 5.      else 6.        If T = Heating interval then 7.           P1 = (t2 − 2 Ttht − Tth2)/(T1 − Tth)2. 8.           P2 = 1 9.         else 10.           P1 = P2 = 0 11.       End if 12.    End if 13.  Calculate BF(i) 14.       If Data = Critical data then 15.         BF(i) = (Eres(i)/D(i, Sink)) × P1 16.        else 17.          BF(i) = (Eres(i)/D(i, Sink)) × P2 18.       End if End

Algorithm 3 Routing process

Input: Source node S, Target node Sink, Maximum benefit function BF(i), Neighbor table NTi Output: Best forwarding node Process: 1. For S has packets to transmit with Sink then 2.   Select the best forwarder node from NTi; 3.     For each record in NTi do 4.       Calculate BF(i); 5.        If Data = Critical data then 6.             BF(i) = (Eres(i)/D(i, Sink)) × P1 7.          else 8.             BF(i) = (Eres(i)/D(i, Sink)) × P2 9.        End if 10.      Forwarder Node = arg Max[BF(i)] 11.    End 12. End

4.3.3. Data Transmission Process

The source node sends the data to the selected forwarding node for transmission and repeats the above steps until the packet is transmitted to the sink. If the emergency data have the same forwarding node selected for normal data, the emergency data are forwarded first based on the priority principle.

5. Results and Discussion

In this study, Matlab R2018a was used to simulate the protocol proposed in this paper. Simulation refers to Nordic nRF2401 and tissue characteristic parameters [15,34]. Simulation parameters and their values are shown in

Table 2. Simulation parameters.

Parameters	Value
Initial energy	0.5 J
Transmission energy	16.7 nJ/bit
Receive energy	36.1 nJ/bit
Data aggregation energy	5 nJ/bit
Amplifier energy	1.97 × 10−9
Initial node temperature	37 °C
Node temperature threshold	43 °C
Node temperature low threshold	39.5 °C
Temperature drop at nodes after each round of sleep	0.02 °C
Specific heat Ts	3600 |J/kg°C|
Thermal conductivity K	0.498 |J/ms°C|
Blood perfusion constant b	2700 |J/m3s°C|
Fixed blood temperature Tb	37 °C
Time step δt	10 s

. The results are compared with those of two existing protocols, M-ATTMPT and iM-SIMPLE, to verify the performance of TSAER. It is worth noting that both M-ATTMPT and iM-SIMPLE are explicitly designed for wearable medical sensors, whose key metrics (energy efficiency, temperature) are in line with the focus of this paper, and secondly, as cluster-based layered protocols, they enable a fair assessment of routing efficiency. In the simulation, 10 sensor nodes were placed on different body parts, and 1 sink was deployed at the waist of the human body. The experiment mainly analyzed the network lifetime, stable period, residual energy, throughput, and average temperature rise in these protocols.

Performance introduction:

• Network lifetime: the time interval from the beginning of the network to the death of the last node;
• Network stability period: the time interval from the beginning of the network to the death of the first node;
• Residual energy: the average residual energy of nodes in the network;
• Throughput: the number of data packets effectively transmitted to the destination node;
• Average temperature rise: the calculated average temperature rise in all nodes.

5.1. Network Lifetime and Stable Period

Figure 4 displays the relationship between the number of dead nodes and rounds and reflects the lifetime and stable period of nodes in the WBAN. It can be seen from the figure that in TSAER, the first dead node appeared in round 4100, and the last node died in round 7050. In iM-SIMPLE, the first dead node appeared in round 3412, and the last dead node appeared in round 6160. In M-ATTEMPT, the first dead node appeared in round 3070, and the last dead node appeared in round 5740.

Compared with iM-SIMPLE and M-ATTEMPT, TSAER obviously has a longer network lifetime and stable period. This paper adopts the method of calculating the node receiving probability and forcing the hot node to sleep, which effectively saves energy consumption. Calculate the receiving node probability according to the temperature condition of each node, obtain the data receiving condition of the available nodes in the link, avoid the rapid generation of hot nodes, and balance the network energy. The transmission mode of M-ATTEMPT is single-hop and multi-hop hybrid transmission. In order to ensure the timeliness of data transmission, emergency data are transmitted in a single-hop mode, but for nodes far away from the sink, it will cause a large amount of energy consumption. Furthermore, M-ATTEMPT uses a backoff method to solve the generation of hot nodes. When the node receives a data packet and the temperature reaches the threshold, it will return the received data to the previous node, which alleviates the node temperature rise but increases the probability of data retransmission and consumes additional energy. In iM-SIMPLE, the temperature state of the node is not considered. When the temperature of the intermediate node increases due to repeated data transmission, there is no mechanism to reduce the high temperature of the node. At this time, continuous work will consume additional energy.

5.2. Residual Energy

Residual energy in the WBAN decreases as the number of rounds increases. Figure 5 describes the relationship between the residual energy of the network and the number of rounds.

The simulation results show that the energy consumption of TSAER is less than that of the existing two protocols. The reason is that TSAER considers the residual energy when selecting the appropriate forwarding node and pays attention to the residual energy of each node in the network, which can effectively balance the network energy. M-ATTEMPT is more inclined to choose a path with fewer hops for normal data transmission, rather than selecting a path to transmit data based on the residual energy. In iM-SIMPLE, the continuous data transmission by the intermediate node causes this node to form a hot node and continues to work without any treatment measures, which not only causes more energy consumption but also burns human tissues in severe cases. Moreover, the existing two protocols do not prioritize data, and congestion may occur when there is a large load in the network, and the data retransmission cause additional energy consumption.

5.3. Throughput

Compared with iM-SIMPLE and M-ATTEMPT, TSAER has a higher throughput. Figure 6 presents the throughput analysis. The throughput is related to the network stable period. A long stable period means that there are more active nodes, and the continuous work of these nodes continuously send data to the sink. Based on the previously analyzed network lifetime and stability period, the temperature state perception adopted by TSAER buffers the process of nodes forming hot nodes in the network, improves the utilization of available nodes, and improves the link quality. Moreover, TSAER has a relatively longer lifetime and stable period; therefore, the simulation results show that it has the highest throughput. The stable period of M-ATTEMPT is shorter, and its throughput decreases as the number of available nodes decreases. Therefore, M-ATTEMPT is lower than the other two protocols.

5.4. Average Temperature Rise

Figure 7 shows the average temperature rise performance of all nodes in different rounds. The communication process of each sensor node will cause the node temperature to rise, thereby causing the generation of hot nodes. TSAER is significantly better than the other two protocols, mainly because of its temperature perception and threshold control. M-ATTEMPT adopts a temperature sensing method to adopt a backoff mechanism for nodes that reach the temperature threshold. A large number of retransmissions also cause the nodes’ temperature to rise. Since iM-SIMPLE does not have any measures to prevent the formation of hot nodes, the nodes’ temperature continues to rise, and the number of nodes that eventually become hot nodes is more than the other two protocols.

6. Conclusions

TSAER focuses on improving energy consumption and hot node issues in WBANs. TSAER divides the data types into emergency data and normal data and designs corresponding routing paths. On the one hand, in order to control the formation of hot nodes in the network, the protocol sets a heating threshold before the node reaches the high temperature threshold, divides the temperature of the node into different temperature intervals through these two thresholds, calculates the probability of data reception of the node in different temperature intervals, and delays the node to become a hot node by controlling the probability of data reception in the heating interval. On the other hand, in order to balance the network energy consumption, the benefit function of the node residual energy, the distance to sink, and the data reception probability is constructed. Through the selection of the largest neighbor node in the benefit function as the best forwarding node, the optimal transmission path is finally determined. Simulation results show that compared with the existing M-ATTEPMT and iM-SIMPLE, TSAER has significant advantages in meeting network QoS requirements, improving energy utilization, and controlling the formation of hot nodes. In future research, the influence of human movement diversity on data transmission in WBANs can be considered.