sensors

Review
A Survey on Data-Driven Approaches for Reliability, Robustness,
and Energy Efficiency in Wireless Body Area Networks
Pulak Majumdar 1 , Satyaki Roy 2 , Sudipta Sikdar 3 , Preetam Ghosh 4 and Nirnay Ghosh 1, *

1 Department of Computer Science and Technology, Indian Institute of Engineering Science and Technology,
Shibpur, Howrah 711103, India; 2022csm007.pulak@students.iiests.ac.in
2 Department of Mathematics, University of Alabama, Huntsville, AL 35899, USA; sr0215@uah.edu
3 Department of Computer Science and Technology & Computer Science and Information Technology,
University of Engineering & Management, Kolkata 700160, India; sudipta.sikdar@uem.edu.in
4 Department of Computer Science, Virginia Commonwealth University, Richmond, VA 23284, USA;
pghosh@vcu.edu
* Correspondence: nirnay@cs.iiests.ac.in

Abstract: Wireless Body Area Networks (WBANs) are pivotal in health care and wearable tech-
nologies, enabling seamless communication between miniature sensors and devices on or within
the human body. These biosensors capture critical physiological parameters, ranging from body
temperature and blood oxygen levels to real-time electrocardiogram readings. However, WBANs
face significant challenges during and after deployment, including energy conservation, security,
reliability, and failure vulnerability. Sensor nodes, which are often battery-operated, expend consider-
able energy during sensing and transmission due to inherent spatiotemporal patterns in biomedical
data streams. This paper provides a comprehensive survey of data-driven approaches that address
these challenges, focusing on device placement and routing, sampling rate calibration, and the
application of machine learning (ML) and statistical learning techniques to enhance network perfor-
mance. Additionally, we validate three existing models (statistical, ML, and coding-based models)
using two real datasets, namely the MIMIC clinical database and biomarkers collected from six
subjects with a prototype biosensing device developed by our team. Our findings offer insights
into strategies for optimizing energy efficiency while ensuring security and reliability in WBANs.
Citation: Majumdar, P.; Roy, S.; Sikdar,
We conclude by outlining future directions to leverage approaches to meet the evolving demands of
S.; Ghosh, P.; Ghosh, N. A Survey on
healthcare applications.
Data-Driven Approaches for
Reliability, Robustness, and Energy
Keywords: wireless body area networks; wearable biosensors; photoplethysmography; redundancy;
Efficiency in Wireless Body Area
Networks. Sensors 2024, 24, 6531.
energy efficiency; robustness; machine learning
https://doi.org/10.3390/s24206531

Academic Editors: Jesús Fontecha

and Iván González
1. Introduction
Received: 31 August 2024 Wireless Body Area Networks (WBANs) represent wearable healthcare technology,
Revised: 30 September 2024 facilitating seamless communication among miniature sensors and devices strategically
Accepted: 5 October 2024 positioned on or implanted within the human body. These networks hold immense impor-
Published: 10 October 2024 tance, as they enable real-time monitoring of physiological parameters, providing valuable
insights into an individual’s health status [1]. WBANs play a pivotal role in various medical
applications, ranging from continuous health monitoring and early disease detection to
Copyright: © 2024 by the authors.
the facilitation of personalized healthcare solutions. Through the wireless connections of
Licensee MDPI, Basel, Switzerland. biomedical sensors, actuators, and wearable devices, WBANs empower individuals and
This article is an open access article healthcare professionals with instantaneous access to critical patient data, paving the way
distributed under the terms and for targeted diagnosis, timely interventions, and enhanced patient care [2]. Beyond health-
conditions of the Creative Commons care, the versatility of WBANs extends to diverse domains such as sports monitoring,
Attribution (CC BY) license (https:// wellness tracking, and human–computer interaction, underscoring their significance in
creativecommons.org/licenses/by/ shaping the future of connected and personalized healthcare monitoring ecosystems [3–6].
4.0/).

Sensors 2024, 24, 6531. https://doi.org/10.3390/s24206531 https://www.mdpi.com/journal/sensors

Sensors 2024, 24, 6531 2 of 19

WBANs have many applications, from hospital patient monitoring to enhancing

military performance. First, the healthcare industry offers one of the most promising appli-
cations of WBANs [7]. Biosensors deployed in the human body can monitor physiological
parameters, allowing physicians, family members, and diagnostic centers to access these sig-
nals remotely. Real-time observation of body signals can detect critical events such as heart
attacks and strokes. WBANs can improve living conditions through the use of devices like
hearing aids and implants. Second, athletes and sports personnel can benefit from WBANs
by monitoring vital body parameters such as SpO2 , heartbeat, blood pressure, and glucose
levels. WBANs can enhance realism in the entertainment sector by tracking posture, facial
expressions, and movements, offering a more immersive experience [8]. Third, WBAN
technology is also vital in military operations, contributing to network-enabled capability
efforts to enhance military effects through the use of information systems [9]. Sensors
monitor vital signs and environmental conditions, helping soldiers avoid threats, while
inter-WBAN communications and spatial localization protect sensitive information.
WBAN communication suffers from several challenges [10]. The primary aim of sup-
porting life-saving medical applications within wireless body area networks (WBANs) un-
derscores the critical importance of Quality-of-Service (QoS) standards, as measured in terms
of data loss and communication latency in a delay-sensitive environment [11]. The chal-
lenge lies in configuring QoS to suit application needs while ensuring fair bandwidth
sharing among co-located WBANs and enabling graceful service degradation. Limited
device memory further complicates matters, necessitating effective error detection and
correction schemes. Reliability and robustness, as determined in terms of the ability of the
network to meet the QoS standards despite failures, are particularly crucial for ensuring the
success of these networks in critical medical applications. The potential for lost or corrupt
alarm/alert packets due to unreliable wireless networks encompasses the need for efficient
acknowledgment and retransmission mechanisms. Integrating heterogeneous wireless net-
working technologies poses a significant challenge in realizing the full potential of WBANs.
Integrating low-cost, limited-range, high-capacity wireless local area network (WLAN) and
personal area network (WPAN) infrastructure for indoor connectivity, alongside lower-
capacity, longer-range cellular infrastructure for outdoor connectivity, requires issues of
seamless roaming and end-to-end QoS to be addressed. The gateway device, acting as a
bridge between WBANs and infrastructure networks, must facilitate smooth data transfer,
storage, and offloading during non-real-time applications. Security concerns in WBANs
extend beyond conventional considerations due to limited resources, lack of user interface,
unskilled users, and global roaming. Traditional security and privacy mechanisms are
often unsuitable for WBANs, necessitating the development of resource-efficient methods.
In WBANs, addressing the inherent spatial and temporal patterns in biomedical data
streams is crucial for enhancing the reliability and robustness of data transmission [12].
By carefully analyzing these correlations and the consequent redundancy, systems can
better mitigate the impact of potential disruptions, leading to improved performance.
Specifically, excessive redundancy can introduce energy overheads and increase the vul-
nerability surface, compromising security, while insufficient redundancy may jeopardize
quality of service (QoS) during component failures [13]. Therefore, it is essential to control
the spatiotemporal data correlation to enhance reliability and robustness without degrad-
ing QoS, compromising biomedical information security, or reducing energy efficiency,
demanding a nuanced approach that considers the specific requirements of the wireless
communication system for clinical applications, ensuring resilience to disruptions.
Spatiotemporal data correlation influences WBAN QoS in the following specific ways.
Firstly, it is crucial in addressing the challenges faced by WBANs, playing a significant role
in maintaining quality of service (QoS). By offering alternative routes for data transmission,
redundancy helps mitigate delays and losses, ensuring consistent performance. However,
excessive redundancy can reduce network lifetime and potentially lead to adverse clinical
outcomes due to decreased energy efficiency, which is paramount in WBANs. Secondly, it
enhances the reliability of communication. Backup paths and retransmission mechanisms
Sensors 2024, 24, 6531 3 of 19

ensure the delivery of critical information, even in the presence of interference or packet loss,
thereby contributing to network robustness. Thirdly, in scenarios involving heterogeneous
wireless networking technologies, redundancy provides added reliability. Data replication
across multiple paths ensures seamless network functionality, even when individual links
fail. Lastly, redundant security measures, including multiple layers of authentication and
encryption, enhance data protection. Lastly, redundant data storage and transmission
mechanisms safeguard against potential security breaches or data corruption.
We provide a comprehensive survey, encompassing an in-depth review of data-driven
approaches aimed at enhancing energy efficiency, security, reliability, and robustness to
failures (see Figure 1a). We explore various device placement and routing approaches that
leverage these patterns to optimize network performance. Furthermore, we delve into exist-
ing strategies for calibrating sampling rates, as well as the application of machine learning
(ML) and statistical learning techniques in WBANs, highlighting their contributions to the
efficient, secure, and reliable operation of these time-critical WBAN applications.

(a) (b)

Figure 1. Spatiotemporal data correlation in wireless body area networks (WBANs). (a) Effect of
redundancy on security, reliability, robustness, and energy efficiency. (b) System model showing the
communication among sensors, coordinator nodes, and access points.

We implement a validation framework to conduct a detailed analysis of the pros

and cons of representative approaches within three distinct classes, namely ML, statistical
methods, and coding-based approaches. Our evaluation, based on real biomedical datasets
including the MIMIC dataset [14], as well as health markers such as heart rate and blood
oxygen levels collected from 6 subjects using the prototype device designed by our team,
provides insights into how each approach addresses data-related challenges in WBANs.
This comparative analysis offers a perspective on their practical applicability and perfor-
mance. We conclude with an outline of future directions for leveraging the spatiotemporal
data correlation to optimize WBAN communication. This includes potential advancements
in integrating redundancy with emerging technologies such as emerging ML models and
adaptive algorithms to further enhance the efficiency, security, and reliability of WBANs.

2. Communication Architecture of WBANs

The communication architecture of wireless body area networks (WBANs) can be
categorized into three tiers, namely intra-WBAN communication, inter-WBAN commu-
nication, and beyond-WBAN communication [15]. Each tier contributes to efficient data
transmission and system reliability. Figure 1b demonstrates these communication tiers
within a component-based WBAN system. In this figure, devices are dispersed across the
body in a centralized network architecture, with the exact location of each device tailored
to specific applications. Given that the body is often in motion (e.g., running or walking),
the ideal placement of sensor nodes can vary, making WBANs inherently dynamic.

2.1. Types of Nodes

WBAN nodes are classified into three types based on communication characteristics [16,17].
Sensors 2024, 24, 6531 4 of 19

• Coordinator or personal devices serve as a central hub for collecting signals from sensors
and forwarding information to recipients such as physicians, cloud services, or external
devices. For experimental validation (discussed in Section 4), we treat the coordination
node as the base station (BS).
• Sensors are the primary components responsible for acquiring physiological param-
eters from the body. They can measure various health parameters, environmental
conditions, and biokinetics. Sensors are deployed in diverse locations, such as wrist-
watches, smartphones, and various body parts.
• Actuators perform specific actions based on data received from sensor nodes, returning
feedback by executing tasks such as administering medicine or adjusting glucose
levels in response to sensor data.
IEEE 802.15.6 [18,19] presents another categorization based on sensor node placement
on or within the human body, distinguishing between implant, body-surface, and external
nodes. IEEE 802.15.6 defines key specifications for wireless body area networks (WBANs)
to ensure efficient and reliable communication in medical and non-medical applications.
The bit rate must range between 10 kbps and 10 Mbps, supporting a maximum of 256 nodes
per network. The packet error rate (PER) should remain under 10% in approximately
95% of the best-performing links, while node join and leave operations must occur in less
than 3 s. To maintain reliable communication, even with mobile nodes, the maximum
tolerable latency is 125 ms for medical applications and 250 ms for non-medical applications,
with jitter limited to 50 ms. Additionally, WBANs should comply with a Specific Absorption
Rate (SAR) of 1.6 W/kg in 1 g of body tissue, with devices transmitting at 0.1 mW and
a maximum transmission power of 1.0 mW. WBANs must also support interoperability
between devices following different standards and be capable of power-saving operations
while ensuring priority services and self-healing capabilities for quality of service (QoS).

2.2. Hierarchical Communication System

The communication architecture of WBANs is a multi-tiered system designed to ensure
seamless and efficient data transmission across different levels of operation. Each tier,
from intra-WBAN communication to beyond-WBAN communication, plays a critical role
in enhancing the functionality, reliability, and scalability of WBANs in various applications,
particularly in health care and medical monitoring systems.

2.2.1. Tier 1: Intra-WBAN Communication

Tier 1 focuses on communication within a single WBAN, covering interactions among
sensor nodes located in and around the human body within a transmission range of
approximately 2 m. The sensors capture physiological signals and transmit them to an
access point (AP) in tier 2.

2.2.2. Tier 2: Inter-WBAN Communication

Tier 2 deals with communication between the BS and multiple access points (APs).
These APs can be part of the infrastructure or strategically positioned in dynamic environ-
ments to manage emergencies. The primary goal is to interconnect WBANs with other
networks, such as cellular networks and the Internet, facilitating easy access in everyday
life. This tier’s communication is divided into the following two subcategories:
• Infrastructure-based Architecture: This common architecture is used in many WBAN
applications. It supports dynamic deployment within confined spaces like hospitals
and offers centralized management and security control, functioning as an application-
specific database server.
• Ad hoc Architecture: In this architecture, multiple APs transmit information within
medical centers, forming a mesh network. This configuration allows for flexible and
rapid deployment and extended radio coverage through multi-hop dissemination
and supports patient mobility. The coverage range extends up to 100 m, making it
suitable for both short-term and long-term setups.
Sensors 2024, 24, 6531 5 of 19

2.2.3. Tier 3: Beyond-WBAN Communication

Tier 3 encompasses communication beyond the WBAN, particularly in metropolitan
areas. A gateway device, such as a personal digital assistant (PDA), bridges the connection
between tiers 2 and 3, linking the Internet to a Medical Server (MS) in specific applications.
The design of this tier is highly application-specific. In medical environments, a database
forms a critical component of tier 3, storing clinical profiles and enabling notifications of
emergency statuses to doctors or patients via the Internet or SMS. Tier 3 also facilitates
the retrieval of essential patient information for treatment purposes. Depending on the
application, the BS in tier 1 may utilize GPRS/3G/4G technology instead of directly
communicating with an AP.

3. Methodologies
The surveyed methodologies (summarized in Figure 2) utilize approaches such as
(1) energy saving and robustness maximization, (2) intelligent node placement and route
selection, (3) sampling-rate calibration, and (4) coding-based techniques to achieve the
following optimization goals.

Figure 2. Summary of methodologies surveyed based on different optimization goals and approaches.

• Reliability, Security, and Trust. Security is paramount in WBANs, where transmitting

sensitive health data demands protection against unauthorized access and attacks.
Data-driven approaches have been employed to ensure the integrity and confidential-
ity of health information by detecting and mitigating security breaches in real-time.
This category encompasses studies that explore how such techniques can mitigate
security threats, detect and recover from attacks, and establish trust in WBAN com-
munication. Examples include secure key management, authentication protocols,
and intrusion detection systems driven by data insights.
• Load Balancing. Load balancing is critical in WBANs, where multiple sensors collabo-
rate to monitor physiological parameters, optimizing the distribution of computational
Sensors 2024, 24, 6531 6 of 19

and communication load across the network, ensuring resource efficiency, and pre-
venting overload on specific components. The studies in this category investigate
how data-driven mechanisms contribute to effective load balancing in WBANs. Top-
ics include dynamic resource allocation, task offloading, and adaptive strategies to
maintain an even distribution of network resources based on real-time data.
• QoS (Energy Efficiency and Robustness). Quality of service (QoS) is essential for the suc-
cess of WBANs, particularly in healthcare applications requiring real-time monitoring.
State-of-the-art approaches significantly enhance QoS by improving the robustness of
data transmission and ensuring overall network reliability. Reliability, in this context,
refers to the network’s ability to withstand and overcome faults, including energy de-
pletion of nodes and network failures arising from human mobility. These approaches
safeguard against potential disruptions and facilitate patient mobility, particularly
when patients move beyond Wi-Fi network coverage [20].

3.1. Energy Efficiency and Robustness to Failures

Efficient energy utilization is a critical aspect of a heterogeneous wireless communica-
tion system, directly influencing its reliability and longevity. State-of-the-art approaches
focus on maximizing system lifetime by designing ultra-low-power radio transceivers,
enabling devices to sleep for extended periods and thereby reducing the duty cycle. Addi-
tionally, choosing low-power WBAN technology further contributes to energy conserva-
tion [21–25]. Addressing failures caused by energy depletion, technical faults, or attacks
is equally vital to maintain system robustness [26–28]. Kaleem et al. proposed a method
that uses Support Vector Machine (SVM) classifiers and a neuro-fuzzy inference system
(ANFIS) to detect interference and malicious sensor nodes in WBANs. By identifying and
correcting sensor failures, the system prevents energy wastage due to repeated transmission
errors, enhancing overall efficiency [29]. Similarly, Bedi et al. introduced the Thermal-
aware, Energy-efficient, Congestion-aware Routing Protocol (TECRP) to balance energy use,
reduce node temperature, and avoid congestion in both inter- and intra-WBAN communi-
cations. Their multi-objective approach to optimizing energy consumption and managing
node temperature ensures minimal energy depletion, extending network lifetime [30].
In the domain of WBAN communications, characterized by time-critical data delivery,
implanted devices have been developed to ensure reliable and energy-efficient medical data
transmission, particularly under conditions of path loss and deep fading. Implant WBANs,
as shown in Figure 3a, facilitate communication between an implanted device (source) and
an off-body access point (destination) through on-body devices (relays). Unlike on-body
devices, implanted devices encounter significant path loss in fading environments while
requiring high reliability, low power consumption, and extended transmission periods.
To address this, joint relay selection and power control strategies are often employed, prior-
itizing energy efficiency over latency [31]. To further secure and enhance energy efficiency
in WBANs, an inexpensive key agreement scheme has been proposed for cluster-based
intra-WBAN and inter-WBAN communications involving resource-constrained sensor
nodes [32]. Moreover, several studies have focused on optimizing network performance by
reducing traffic redundancy and its associated energy consumption. For instance, an ant
colony-based energy-efficient algorithm utilizes clustering to manage data packet trans-
mission, balance network load, and extend network lifetime [33]. Another study aimed to
improve WBAN throughput by minimizing packet drops and implementing a prioritization
scheme based on packet priority and source location [34].
Sensors 2024, 24, 6531 7 of 19

(a) (b)

Figure 3. Energy efficiency and strategic node placement in WBANs. (a) Communication of an
implant WBAN with the wireless access point. (b) Energy-efficient protocols for biosensor networks.
(This figure was redrawn from [35]).

3.2. Optimized Placement and Coordinated Routing Approaches

An autoregression-based prediction model was proposed to determine the optimal
data transmission rate from sensors placed at various locations on the human body to
coordinator nodes. This approach minimizes energy overhead while maintaining acceptable
prediction errors [36]. Additionally, the integration of multiple coordinators or gateway
nodes in wireless body area networks (WBANs) has been suggested [37]. These nodes are
designed to receive data from sensors, aggregate them to enhance efficiency, and transmit
them to the base station for centralized processing. The deployment of multiple coordinator
nodes effectively extends the network’s operational lifespan and enhances reliability by
providing redundancy and ensuring continuous data transmission. Moreover, intelligent
node placement has been highlighted to curb data overhead and improve network lifetime
by minimizing the distance between communicating nodes [38]. Another approach aimed
at minimizing communication overheads within WBANs was introduced in [35]. This
strategy comprises the following two routing methods to optimize energy efficiency:
• The cluster-based approach involves the strategic selection of cluster heads (CHs) based
on the density of biosensors within the network. CHs are identified before node
assignment and are responsible for aggregating and transmitting data to the base
station (BS). This clustering mechanism streamlines communication and facilitates the
efficient organization of biosensors, ensuring an optimal distribution of responsibilities
within the network (see Figure 3b(1)).
• In the tree-based approach, a hierarchical structure is proposed, where the BS serves
as the root of the tree, with the nearest biosensors acting as its children. Through
an iterative process, each biosensor is assigned to a designated parent within the
tree, creating a structured hierarchy for data transmission. This method leverages the
tree structure to systematically push data from biosensors to the BS, promoting an
organized and resource-efficient flow of information (see Figure 3b(2)).
Variants of the cluster-based approach have also been proposed. For instance, a clustering-
based energy-efficient WBAN focuses on reliable data transmission within a many-to-
one stream model, addressing the critical need for resilience in healthcare applications.
The proposed mechanism, termed the Network Coding-based Fault-tolerant Mechanism
(NCFM), employs a two-step approach [39]. First, a greedy grouping algorithm partitions
the network topology into smaller logical units. Next, a spanning tree is constructed
using random linear network coding, facilitating the creation of linearly independent
coding combinations. Another paper introduced a method for determining cluster heads
in WBANs, emphasizing effective data communication in healthcare monitoring systems.
The proposed method considers factors such as residual energy levels and node positions
Sensors 2024, 24, 6531 8 of 19

to designate cluster heads, which act as relay nodes for the transmission of messages
from surrounding sensor nodes to the base station. By utilizing probability distribution,
specific energy, network density, and distance from base stations, this method optimizes
the selection of cluster heads. Simulations revealed that this approach increases network
lifetime and decreases end-to-end delay compared to the LEACH protocol [40].
Shukla et al. presented an energy-efficient solution for hub node placement in WBANs,
using the Whale Optimization Algorithm (WOA) to find the optimal location for the central
hub. The WOA minimizes the energy spent by biosensor nodes by reducing the need for
multiple trial-and-error placements, speeding up the process compared to traditional meth-
ods. The algorithm leverages a population of candidate solutions, or “whale search agents”,
that iteratively converge on the best hub location with minimal network energy consump-
tion [41]. Patra et al. proposed a Free Search Krill Herd (FSKH) algorithm for optimization
of relay node placement in WBANs, focusing on energy efficiency. They complemented this
with a routing protocol based on the Harmony Search (HS) algorithm, further improving
the reliability and energy conservation of data transmission [42]. Lastly, other studies
have addressed the joint challenge of security and energy-efficient routing. One approach
aims to secure data transmission using particle swarm optimization for the selection of the
next hop [43]. Similarly, an optimization-based framework aims to improve throughput,
prolong node lifetime, and minimize redundancy through synchronized communication
and a handover mechanism among the coordinator nodes (CNs). Specifically, two-hop
communication was enforced to minimize packet drops under worst-case scenarios where
the communicating node is outside the range of the CNs [44].

3.3. Sampling Rate Calibration and Machine and Statistical Learning

Calibrating sampling rates is another key aspect of managing energy expenditure
in WBANs. A two-step energy-efficient health monitoring approach [45] first optimizes
the measured vital signs by eliminating redundancy, then calibrates sampling rates based
on risk scores derived from the measurements. Similarly, another model predicts sam-
pling rates based on patient activity and risk, utilizing an adaptive neuro-fuzzy inference
system and long short-term memory to forecast future sampling rates and health mea-
surements [46]. A hierarchical approach further reduces redundancy at edge sensors by
adapting sampling rates, while edge devices aggregate the information from these sensors.
This approach incorporates a support vector machine (SVM) to make informed decisions
about patient health [47]. In another two-step health monitoring system [48], the following
energy-saving protocol was proposed:
1. For any incoming value (x) in a stream of data with mean (m) and standard deviation
(σ), a Z score is calculated to determine its deviation from the mean, i.e.,
x−m
Z= (1)
σ
If Z is outside reasonable limits [lτ , hτ ], the data are discarded as faulty. It is considered
interesting if ∆Z = | Z − Z previous | is below a prespecified threshold; otherwise, it is
marked as uninteresting.
2. The isolation forest method is employed to detect any anomalies that could indicate a
potentially adverse medical condition requiring immediate clinical attention from the
filtered data.
Another statistical approach based on a one-way ANOVA model and Fisher test
operates in the following two phases: aggregation and transmission [49]. This method
first captures data in time slots and calculates the difference between consecutive values
(|mi − mi+1 |). If this difference exceeds a predefined threshold (tτ ), it is retained as a new
value; otherwise, the new value is considered redundant, and the frequency of the old
value is increased. During the transmission phase, the sensor node evaluates the similarity
between data captured over multiple periods using the ANOVA model and Fisher test.
Only dissimilar data groups are transmitted to the base station, effectively reducing the
Sensors 2024, 24, 6531 9 of 19

transmission rate by eliminating redundant data vectors. Mohammadi et al. addressed

the challenge of determining an optimal sampling rate in energy-harvesting WBANs to
maintain self-sustainability without compromising service quality. They highlighted the
unpredictability of harvestable energy rates and the variable nature of patient vital signs.
Their paper proposed dynamic sampling-rate technique using deep reinforcement learning,
where the Markov decision process framework captures both energy and data variability
aspects [50]. Li et al. focused on minimizing energy consumption in WBANs during
transmission sessions using amplify-and-forward (AF) relay through a hybrid supervised
and DRL-based method that optimizes transmission parameters like power levels and
block length, ensuring reliable, energy-efficient transmission under fluctuating channel
conditions [51].

3.4. Coding-Based Approaches

Several coding-based schemes have been proposed to enhance the reliability of
WBANs by adopting cooperative communication and network coding strategies. The first
approach minimizes channel impairment and body fading effects, thereby reducing faults,
bit error rates, and energy consumption. This method was demonstrated through a case
study focused on remote sepsis monitoring, showcasing its effectiveness [12]. Another net-
work coding approach was introduced to address load balancing in WBANs, dynamically
adjusting packet redundancy levels via a Markov Decision Process (MDP) algorithm. This
method significantly enhances data storage reliability and retrieval while optimizing energy
consumption in the network [52]. GROWN, a data compression algorithm designed to
reduce information redundancy in sensory data [53], operates in the following two stages:
lossy data acquisition and lossless compression.
1. During the lossy phase, the data within an acceptable range are discarded, minimiz-
ing the bits required for compression in the subsequent lossless phase. The initial
data point (di ) is stored and compared with subsequent points. If the difference
(∆Z = |di − d j |) for a later data point (d j ) exceeds a predefined threshold (tτ ), ∆Z is
sent for compression and di is updated to d j ; otherwise, d j is discarded by considering
it a repetition.
2. In the lossless compression stage, GROWN employs a modified exponential Golomb
code, a universal lossless encoding method suitable for coding non-negative values.
Decoding tables are used at the server or base station for signal decompression and
recovery of the original signal.
Another approach employs rateless channel coding to improve energy efficiency in Blue-
tooth networks for augmented reality [54]. Using a block-based rateless coding technique,
this study demonstrated reduced energy consumption and enhanced video quality com-
pared to traditional Bluetooth-based FEC schemes, especially under challenging channel
conditions. This approach mitigates wireless channel errors and minimizes network redun-
dancy, contributing to energy savings and improved video quality. Wang et al. focused
on the use of network coding in IoMT systems and proposed a lightweight identity-based
network coding scheme (IBNS) that enhances security while reducing computational
overhead in signature verification, making it feasible for resource-constrained IoMT envi-
ronments [55].
A logical XOR-based architecture [56] introduces a static approach for efficient wire-
less communication in WBANs while enhancing small-scale body area network reliability.
Grounded in linear block coding theory, the architecture involves sensor nodes and re-
lay nodes communicating with a destination. Each relay node strategically selects two
sensors and three relay nodes to reduce the effective error rate in noisy environments.
The XOR operation on data from these sources enhances data accuracy, with experiments
demonstrating effectiveness, particularly in Rician channels. Finally, another coding-based
approach, termed the Redundancy-Balanced Data Transmission Scheme (RDBT), empha-
sizes quality-of-service (QoS) parameters such as network lifetime, stability, throughput,
and latency [57]. RDBT leverages compressive sampling on medical data characterized by
Sensors 2024, 24, 6531 10 of 19

low variation and high redundancy, mitigating energy loss from excessive data transmis-
sion. It categorizes sensor nodes based on data significance, enabling network balancing
and responsive emergency measures. A selection algorithm for cooperative nodes in multi-
hop communication considers residual energy, congestion control, and signal-to-noise ratio
to optimize QoS. This approach demonstrates significant improvements in QoS, surpassing
existing state-of-the-art methods.

4. Experimental Validation
We discuss the study design and implementation specifications, followed by a perfor-
mance evaluation of the three baselines to optimize WBAN communication. The experi-
ments are designed to demonstrate the effectiveness of data-driven approaches by incor-
porating machine learning, statistical methods, and coding-based techniques to address
key challenges in WBANs, such as reliability (measured in terms of predictive accuracy of
significant events) and quality of service (in terms of data loss and communication delay).

4.1. Study Design

4.1.1. Simulation Platform
We developed a customized simulation platform using the Python discrete event
simulation library called simulation in Python (SimPy) [58]. The simulator features an
event generator (EG) module that associates physiological data collected using a prototype
device (see Section 4.1.2) or sourced from the MIMIC clinical dataset (details discussed
hereafter in Section 4.1.3), with location coordinates (sampled from a uniform random
distribution) within a predefined simulation area representing the human body. Each event
generated by the EG is a physiological data point and is tagged with a unique event
identifier and location coordinates given by ⟨ID, data, x, y⟩. In our simulation setup, we
deploy n sensors throughout the designated area, each with functionalities such as sensing,
sending, and receiving event data in its vicinity, governed by a sensing range. These
energy-constrained sensors can intercommunicate and transmit data to a local processing
unit known as the base station (BS). Each biosensor node is implemented as a generator
function within the SimPy simulation framework, simulating its real-world operations and
energy consumption.
To evaluate three approaches for redundancy mitigation, the WBAN is simulated
under various conditions in separate runs. These approaches are representative of statistical,
machine learning or coding-based models designed to reduce data redundancy. The static
sensors communicate through selective flooding, except when they are in the neighborhood
of the base station (BS). In these cases, they directly forward the aggregated information
received from their peers to the BS. For each baseline scenario, the BS assesses the following
two key metrics: the predictive accuracy of anomalous clinical events and the energy
savings achieved through the elimination or compression of redundant data. This dual
evaluation provides a holistic understanding of each algorithm’s effectiveness in optimizing
the predictive accuracy and energy efficiency of the WBAN.

4.1.2. Real Data Collection Using a Wearable Biosensor Prototype Device

We developed a wearable prototype device to collect real data comprising blood oxy-
gen level and pulse rate. It employs photoplethysmography (PPG), which is a non-invasive
optical technique used to detect blood volume changes in the microvascular bed of tis-
sue. It is used to measure various cardiovascular parameters, including heart rate and
blood oxygen saturation (SpO2 ). Pulse oximetry leverages the differential light absorp-
tion properties of oxygenated hemoglobin (HbO2 ) and deoxygenated hemoglobin (Hb)
at specific wavelengths. A fingertip probe transmits red (around 660 nm) and infrared
(around 940 nm) light, chosen for their contrasting absorption rates by HbO2 and Hb.
A photodetector measures the absorption at each wavelength as the light passes through
the fingertip vasculature. The red-to-infrared light absorption ratio is used to estimate SpO2
through an empirically derived algorithm based on prior calibrations with blood samples
Sensors 2024, 24, 6531 11 of 19

of known oxygen saturation levels, providing continuous, non-invasive monitoring of

SpO2 . The oxygen saturation percentage is the ratio of HbO2 in the arteries to the total
Hb, i.e.,

HbO2
SpO2 = × 100%
Hb + HbO2
Oximetry devices utilize red and infrared LEDs in conjunction with photosensors to
gauge the light intensity passing through fingertips. This intensity reduction is attributed
to blood flow in veins, arteries, and tissues. Venous blood flow and tissue generate a
stable DC signal, while arterial blood flow results in a less stable AC signal. The ratio of
ACred /DCred
red-to-infrared light absorption producing AC and DC components is R = AC ,
ired /DCired
where ACred , DCred , ACired , and DCired are the AC and DC voltages from the absorption of
red and infrared light, respectively. Lastly, given the voltage ratio (R) resulting from the
absorption of red and infrared light (ACred ), SpO2 is SpO2 = 10 − 25R.
The portable healthcare system utilizes an IC MAX30100 oximetry sensor to monitor
health indicators. This integrated module features red and infrared LEDs, alongside signal
conditioning components, all housed within a single package. By capturing voltage,
the sensor module calculates SpO2 levels and detects pulses. A schematic representation of
the sensor module is shown in Figure 4a. The MAX30100 sensor module (manufactured by
Maxim Integrated Products, Inc., a subsidiary of Analog Devices) positioned on thin body
tissues like a fingertip, employs alternating red and infrared light from its second LED.
This light is emitted onto the fingertip, where some is absorbed and reflected. The reflected
light is detected by the photodiode on the IC MAX30100, generating a voltage that is
transmitted to the integrated signal conditioner within the IC MAX30100. The resulting
signal from the MAX30100 sensor module is conveyed to the XIAO ESP32-S3 using the I2C
communication protocol. The system implemented for this study comprises the following
open-source electronic components:
1. ESP32-S3 microcontroller: A cost-efficient tiny development board with dimensions of
21 mm × 17.5 mm, featuring Wi-Fi and Bluetooth capabilities and designed around
the ESP32-S3 microcontroller. With 11 general-purpose input/output pins, support
for interrupt/PWM/I2C/one-wire (except A0), and a single analog input (3.3 V
max), it provides versatile functionality. The microcontroller is compatible with
MicroPython, Arduino, and ESP-IDF, boasting a clock speed of up to 240 MHz, 8 MB
PS-RAM, and 8 Mbytes of Flash storage. This device can be powered via a micro
USB connection, accepting a 5 V input. Leveraging its Wi-Fi capabilities, the Seed
Studio XIAO ESP32S3 facilitates the creation of applications for wireless data transfer
to cloud infrastructure.
2. MAX30100 pulse oximeter module: With dimensions of 19 mm × 14.5 mm × 3 mm,
this module is capable of embedding pulse oximetry and heart rate monitoring
applications. Comprising two LEDs and a low-noise analog signal processing unit,
it calculates heart rate and SpO2 measurements using raw sensor data. Operating
within a range of 1.8 V to 3.3 V, the sensor module includes an inbuilt voltage regulator.
Equipped with a photo sensor, red LED (660 nm), and IR LED (880 nm) with radiation
power of up to 9.8 mW and 6.5 mW, respectively, the sensor can achieve a high sample
rate of up to 1000 Hz. Its programmable nature allows for precise control of a high
data output. The module interfaces using the I2C protocol with a four-pin connection
(VCC, GND, SDA, and SCL) to the MCU pins.
3. PKCELL LIPO Battery: A 350 mAh 3.7V lithium-ion polymer battery with dimension
of 36 mm × 19.6 mm × 5.2 mm), a standard discharge rate of 0.2C, and a maximum
continuous discharging current of 525 mA.
4. Raspberry Pi 4 Model B: This component features a Broadcom BCM2711, Quad-core
Cortex-A72 (ARM v8) 64-bit SoC @ 1.8 GHz, 8 GB LPDDR4-3200 SDRAM, 2.4 GHz
and 5.0 GHz 802.11ac [59] wireless, Bluetooth 5.0, gigabit Ethernet, operating on 5 V
DC via a USB-C connector and consuming less than 500 mA.
Sensors 2024, 24, 6531 12 of 19

(a) (b) (c)

Figure 4. Prototype device for the collection of physiological data (namely, blood oxygen levels and
pulse rates). (a) Block diagram and IC MAX30100 placement. (b,c) Circuit connection between the
XIAO ESP32-S3 development board and the sensors.

This wearable sensor system targets the continuous monitoring of blood oxygen levels
and pulse rates. The data are processed by the XIAO ESP32-S3 MCU from Seeed Studio,
which features IoT communication technology. The XIAO ESP32-S3 MCU is a compact
unit embedded with Wi-Fi and Bluetooth, along with integrated power management and
battery protection features. A sensor attached to the MCU via I2C pins sends the collected
data over short-range Wi-Fi to a gateway device hosted by a Raspberry Pi Single-Board
Computer (SBC), where they are further processed. The IoT gateway receives data from
the MCU every five seconds and transfers them to the cloud (see Figure 4b).
Overall, the prototype is capable of various functions, including monitoring of patient
data using sensors directly attached to the patient’s body. The sensors and MCU are
integrated into a wearable device that patients can easily wear. This prototype can wirelessly
transmit data to the cloud without the need for extra wires. An access point receives data
from the wearable device via Wi-Fi. A schematic diagram of the circuit connection between
the XIAO ESP32-S3 development board and the sensors is shown in Figure 5.

Figure 5. Components of the prototype. (a) Overview. (b,c) Gateway device connection with cloud
and power supply to MCU.

4.1.3. MIMIC
The MIMIC-IV Database [14] is a repository of physiological health features and indica-
tors obtained from intensive care patients. This repository contains various measurements
of patients under critical care, namely electrocardiogram readings, blood oxygen level,
Sensors 2024, 24, 6531 13 of 19

respiratory rate, etc. The data are collected using medical equipment directly from the
bedside of the patients. Patient identification related to the data is removed when placing
the records in the database. We consider the following four physiological features for our
analysis: heart rate (HR), pulse, breathing rate, and blood oxygen level (SpO2 ).

4.2. Implementation Specifications

4.2.1. Data Labeling
We employed k-means clustering with k = 2 to identify two clusters, where the
majority class is classified as not interesting (0), while the minority class is marked as
interesting (1). Furthermore, a generative algorithm called SMOTE [60] was used to augment
the dataset size to 1000 data points, ensuring that 70% of the data points belong to the
0 class, while the remaining 30% are in the 1 class.

4.2.2. WBAN Deployment

The simulation setup is designed to ensure consistency in the evaluation of the
three methods. Several physical parameters are kept constant for comparative analy-
sis. Specifically, a total of 15 nodes are utilized, each with a sensing radius of 0.5 m.
These nodes are randomly deployed within an area of approximately 2.5 × 2 square meters
(see Figure 6a). The simulation duration is set to 200 min, providing a substantial period
for data collection and analysis. For the statistical and machine learning-based approaches,
a buffer size of 25 is implemented. This buffer size is chosen to balance between computa-
tional efficiency and accuracy. Additionally, the statistical method operates with a period
of 10 min, allowing for periodic data aggregation and analysis. This setup ensures that the
evaluation scores obtained from the three baseline approaches can be directly compared
under identical conditions.

(a) Deployment (b) Precision

(c) Recall (d) F1 score

Figure 6. WBAN deployment and the predictive accuracy of the isolation forest-based redundancy
mitigation approach [48] (MIMIC dataset).
Sensors 2024, 24, 6531 14 of 19

4.3. Comparative Analysis of Baseline Approaches

Finally, we evaluate the machine learning (ML) [48], statistical [49], and coding-
based [53] approaches implemented using the customized simulator modeling a WBAN
with biosensors exchanging biomedical datasets taken from the MIMIC dataset and real
physiological data collected from six subjects through a hardware prototype. The proce-
dures collected heart rate and blood oxygen data using non-invasive wearable sensing
devices according to ethical standards and causing no risk to individuals. We discuss
the performance of the stated approaches in terms of predictive accuracy and achieved
compression averaged across 10 runs.
Figures 6 and 7 show the precision, recall, and F1 score for the Isolation Forest (IF)-
based ML approach and the ANOVA-based statistical counterpart (called ATP) on the
MIMIC dataset, as well as the real collected data. The overall performance improves
with the number of data points. The IF approach shows better predictive accuracy (F1
score∼0.80) with the MIMIC than with real data (F1 score∼0.55), whereas the ANOVA
approach shows the opposite trend, (i.e., F1 score∼0.3 on MIMIC and F1 score∼0.45 on the
collected data). Notably, the IF approach exhibits an improvement in precision for both
datasets. We attribute its superior performance to its use of recursive partitioning, which
effectively differentiates between redundant and non-redundant data. Furthermore, IF
leverages an ensemble of trees during the training process, making it less susceptible to false
positives. This ensemble method contributes to the high precision (see Figure 6b). Overall,
the robustness of the IF-based approach to varying data distribution assumptions enhances
its performance and reliability. Finally, we measure the compression factor, calculated as
number of compressed bits
1 − number of uncompressed bits , and the average number of bits transmitted by the coding-
based approach (called GROWN) [53]. This approach takes advantage of the correlation
between consecutive sensed values, which reduces the frequency of data to be transmitted
during the lossy acquisition phase. The compression code also reduces the number of bits
to be transmitted by the sensors. The GROWN approach is compared against the average
number of bits transmitted without compression, which is based on a 14-bit ADC with
higher resolution.

(a) Precision (b) Recall (c) F1 score

Figure 7. Predictive accuracy of the statistical redundancy mitigation approach [49] (collected data).

The analysis (on 500 data points) of the QoS of the IF and ANOVA approaches shows
similar trends as the earlier analysis on predictive accuracy. The data loss rate is measured in
terms of the proportion of unique events not received by the BS, while latency is measured
in terms of the time elapsed between the occurrence of the event and its receipt at the BS.
For ANOVA, we analyze the effects of two significance levels (SLs), i.e., 0.05 and 0.10, on
the overall performance. Figure 8 shows the loss for ANOVA and latency to be significantly
higher than for IF. Since a higher SL makes it easier to reject the null hypothesis that the
two points are similar, ANOVA with an SL of 0.10 exhibits marginally improved results
relative to those achieved with an SL of 0.05.
Two consecutive data points sensed by a given node are considered the same if
the mean squared error (MSE) between them is within a predefined threshold range.
As illustrated in Figures 9 and 10, an increase in the MSE threshold results in greater
Sensors 2024, 24, 6531 15 of 19

compression, captured in terms of an increased compression factor and a reduction in the

number of transmitted bits for both MIMIC and the real collected data.

(a) Precision (b) Recall

Figure 8. Predictive accuracy of the statistical redundancy mitigation approach [49].

(a) MIMIC (b) Real collected

Figure 9. Redundancy minimization through compression achieved using GROWN [53].

(a) MIMIC (b) Real collected

Figure 10. Reduction in the number of transmitted bits achieved using GROWN [53].

5. Discussions
Future WBANs, especially those incorporating lightweight, energy-constrained biosen-
sor nodes, must carefully balance reliability with energy and storage efficiency. As high-
lighted in this survey, energy efficiency is paramount due to the limited battery life of
biosensor nodes [61]. Leveraging the inherent spatial and temporal patterns in biomedical
data streams can enhance both data transmission reliability and robustness against failures
while managing energy consumption. To address these challenges, strategies like adaptive
correlation-based optimization can dynamically adjust communication parameters based on
network conditions and energy levels. Additionally, energy-efficient encoding, decoding,
Sensors 2024, 24, 6531 16 of 19

and data aggregation methods will be critical in minimizing storage overhead. Optimal
node placement, utilizing overlapping radio ranges to provide redundancy, can improve
network reliability and ensure continuity in cases of node failure [62]. In WBANs, effec-
tively managing spatial and temporal correlations minimizes data loss risks and ensures
uninterrupted monitoring, which is especially vital in critical healthcare settings.
The emergence of software-defined wireless body area networks (SD-WBANs) represents
another significant advancement in the field, offering a flexible and programmable network
architecture where a centralized controller can dynamically regulate network topology
and traffic flow [63]. This is particularly useful in scenarios where the WBAN exhibits
heterogeneity in terms of power sources and configuration [64]. Understanding spatial and
temporal trends in data streams becomes particularly important in SD-WBANs, as they can
substantially enhance fault tolerance in time-critical applications. By introducing multiple
controllers within the SD-WBAN framework, the network can achieve higher levels of
reliability and resilience [61]. These additional controllers can act as backup nodes, taking
over control and data transmission responsibilities in the event of a failure, thereby ensuring
uninterrupted service and reducing the risk of data loss. The centralized nature of SD-
WBANs allows for sophisticated management strategies, such as real-time monitoring
and adaptive reconfiguration based on data correlations, optimizing resource utilization,
and maintaining high-quality service levels. Integrating machine learning algorithms to
preemptively address potential failures can also enhance their effectiveness in SD-WBANs.
Another security consideration is the possibility of compromising nodes to push incor-
rect health information. In future WBAN applications, it will be crucial to adopt consensus
or data fusion techniques to minimize adverse health outcomes from misreporting. Com-
promised nodes can potentially provide false data that may lead to incorrect diagnoses
or treatments, posing significant risks to patient safety. Consensus, as well as data fusion
techniques, involve multiple nodes reaching an agreement on the reported data, ensuring
that a single compromised node cannot influence the overall system [27]. These techniques
can help validate the accuracy of the data by comparing inputs from multiple biosensors
and identifying possible anomalies or inconsistencies. By requiring a majority agreement
before data are considered valid, consensus mechanisms can effectively mitigate the impact
of compromised nodes. Incorporating these security measures into the design of future
WBANs will be essential for maintaining the integrity and reliability of health monitoring
systems. By leveraging spatial and temporal trends to improve network performance and
enhance security, researchers can develop robust WBANs capable of providing accurate
and reliable health data, even in the presence of potential threats.

6. Conclusions
This survey examined strategies to address inherent spatial and temporal trends in
biomedical data, focusing on time-critical optimization goals such as robustness, optimized
placement, routing approaches, sampling rate calibration, machine and statistical learning,
and coding-based methodologies in wireless body area networks (WBANs). Our analysis,
utilizing a combination of the MIMIC clinical database and physiological data collected
using a prototype biosensing device built by our team, reveals that machine learning
(ML) approaches, particularly the isolation forest algorithm, significantly outperform
statistical methods in leveraging these correlations. Our results indicate that larger sample
sizes substantially enhance the predictive accuracy in identifying anomalous readings in
healthcare data. These insights highlight the importance of advanced ML techniques and
adequately sized datasets in improving energy efficiency, security, reliability, and robustness
in WBANs. We conclude with a discussion on the future of WBAN research to enhance the
quality of service and accuracy for real-world healthcare solutions.

Author Contributions: P.M. and S.R. curated the articles for review and conducted the computational
experiments. S.S. designed the prototype device and carried out data collection. P.G. and N.G. led the
manuscript writing. All authors have read and agreed to the published version of the manuscript.
Sensors 2024, 24, 6531 17 of 19

Funding: This work was partially supported by the National Science Foundation (CBET-1802588), the
Commonwealth Health Research Board (CHRB-2360623), and the VCU Quest and Breakthrough Grant.
Conflicts of Interest: The authors declare no conflict of interest.

References
1. Pandey, I.; Dutta, H.; Banerjee, J.S. WBAN: A smart approach to next-generation e-healthcare system. In Proceedings of the
2019 3rd International Conference on Computing Methodologies and Communication (ICCMC), Erode, India, 27–29 March 2019;
IEEE: Piscataway, NJ, USA, 2019; pp. 344–349.
2. Bhatt, V.; Chakraborty, S. Real-time healthcare monitoring using smart systems: A step towards healthcare service orchestration
Smart systems for futuristic healthcare. In Proceedings of the 2021 International Conference on Artificial Intelligence and Smart
Systems (ICAIS), Coimbatore, India, 25–27 March 2021; IEEE: Piscataway, NJ, USA, 2021; pp. 772–777.
3. Jagadeeswari, V.; Subramaniyaswamy, V.; Logesh, R.; Vijayakumar, V. A study on medical Internet of Things and Big Data in
personalized healthcare system. Health Inf. Sci. Syst. 2018, 6, 14. [CrossRef] [PubMed]
4. Alfian, G.; Syafrudin, M.; Ijaz, M.; Syaekhoni, M.; Fitriyani, N.; Rhee, J. A personalized healthcare monitoring system for diabetic
patients by utilizing BLE-based sensors and real-time data processing. Sensors 2018, 18, 2183. [CrossRef] [PubMed]
5. Philip, V.; Suman, V.; Menon, V.; Dhanya, K. A review on latest internet of things based healthcare applications. Int. J. Comput.
Sci. Inf. Secur. 2017, 15, 248.
6. Roy, S.; Das, S.K. Bio-Inspired Design of Biosensor Networks. Encycl. Sens. Biosens. 2023, 4, 86–102.
7. Kim, J.; Campbell, A.S.; de Ávila, B.; Wang, J. Wearable biosensors for healthcare monitoring. Nat. Biotechnol. 2019, 37, 389–406.
[CrossRef]
8. Meng, X. Virtual Reality Solutions for Sports Training Data Management with 5G UWB Communication Technology and Wearable
Devices. Comput.-Aided Des. Appl. 2024, 21, 92–108. [CrossRef]
9. Egbogah, E.; Fapojuwo, A. Achieving energy efficient transmission in wireless body area networks for the physiological
monitoring of military soldiers. In Proceedings of the MILCOM 2013–2013 IEEE Military Communications Conference, San
Diego, CA, USA, 18–20 November 2013; IEEE: Piscataway, NJ, USA, 2013; pp. 1371–1376.
10. Patel, M.; Wang, J. Applications, challenges, and prospective in emerging body area networking technologies. IEEE Wirel.
Commun. 2010, 17, 80–88. [CrossRef]
11. Kaur, N.; Verma, S.; Jhanjhi, N.; Singh, S.; Ghoniem, R.; Ray, S. Enhanced QoS-aware routing protocol for delay-sensitive data in
Wireless Body Area Networks. IEEE Access 2023, 11, 106000–106012. [CrossRef]
12. Mehmood, G.; Khan, M.; Abbas, S.; Faisal, M.; Rahman, H. An energy-efficient and cooperative fault-tolerant communication
approach for wireless body area network. IEEE Access 2020, 8, 69134–69147. [CrossRef]
13. Sahar, G.; Bakar, K.; Zuhra, F.; Rahim, S.; Bibi, T.; Madni, S. Data redundancy reduction for energy-efficiency in wireless sensor
networks: A comprehensive review. IEEE Access 2021, 9, 157859–157888. [CrossRef]
14. Moody, B.; Hao, S.; Gow, B.; Pollard, T.; Zong, W.; Mark, R. MIMIC-IV Waveform Database. PhysioNet 2022, 101, e215–e220.
[CrossRef]
15. Movassaghi, S.; Abolhasan, M.; Lipman, J.; Smith, D.; Jamalipour, A. Wireless body area networks: A survey. IEEE Commun.
Surv. Tutor. 2014, 16, 1658–1686. [CrossRef]
16. Choudhary, A.; Nizamuddin, M.; Zadoo, M. Body node coordinator placement algorithm for WBAN using multi-objective swarm
optimization. IEEE Sens. J. 2021, 22, 2858–2867. [CrossRef]
17. ul Huque, M.; Munasinghe, K.; Jamalipour, A. Body node coordinator placement algorithms for wireless body area networks.
IEEE Internet Things J. 2014, 2, 94–102. [CrossRef]
18. Xing, J.; Zhu, Y. A survey on body area network. In Proceedings of the 2009 5th International Conference on Wireless
Communications, Networking and Mobile Computing, Beijing, China, 24–26 September 2009; IEEE: Piscataway, NJ, USA, 2009;
pp. 1–4.
19. Yazdandoost, K.Y. Channel model for body area network (BAN). In Proceedings of the 2008 30th annual International Conference
of the IEEE Engineering in Medicine and Biology Society, Vancouver, BC, Canada, 20–25 August 2008.
20. ElSalamouny, M.; Halawa, H.; Daoud, R.; Amer, H. Performance of fault-tolerant WBAN for healthcare. In Proceedings of the
2017 International Conference on Engineering & MIS (ICEMIS), Monastir, Tunisia, 8–10 May 2017; IEEE: Piscataway, NJ, USA,
2017; pp. 1–5.
21. Sarkar, S. Internet of Things—Robustness and reliability. In Internet of Things; Elsevier: Amsterdam, The Netherlands, 2016;
pp. 201–218.
22. Sinche, S.; Polo, O.; Raposo, D.; Femandes, M.; Boavida, F.; Rodrigues, A.; Pereira, V.; Silva, J.S. Assessing redundancy models for
IoT reliability. In Proceedings of the 2018 IEEE 19th International Symposium on “A World of Wireless, Mobile and Multimedia
Networks” (WoWMoM), Chania, Greece, 12–15 June 2018; IEEE: Piscataway, NJ, USA, 2018; pp. 14–15.
23. Rullo, A.; Serra, E.; Lobo, J. Redundancy as a measure of fault-tolerance for the Internet of Things: A review. In Policy-Based
Autonomic Data Governance; Springer: Cham, Switzerland, 2019; pp. 202–226.
24. Samouni, N.; Jilbab, A.; Hammouch, A. A Critical Review on Recent Research on Reliable Communication for Wireless Body
Area Network. Int. J. Sens. Wirel. Commun. Control 2019, 9, 432–442. [CrossRef]
Sensors 2024, 24, 6531 18 of 19

25. Abdullah, W.A.N.W.; Yaakob, N.; Elobaid, M.E.; Warip, M.N.M.; Yah, S.A. Energy-efficient remote healthcare monitoring using
IoT: A review of trends and challenges. In Proceedings of the International Conference on Internet of Things and Cloud
Computing, Dalian, China, 22–23 October 2016; pp. 1–8.
26. Wu, G.; Ren, J.; Xia, F.; Xu, Z. An adaptive fault-tolerant communication scheme for body sensor networks. Sensors 2010,
10, 9590–9608. [CrossRef]
27. Salayma, M.; Al-Dubai, A.; Romdhani, I.; Nasser, Y. Wireless body area network (WBAN) a survey on reliability, fault tolerance,
and technologies coexistence. ACM Comput. Surv. (CSUR) 2017, 50, 1–38. [CrossRef]
28. Cavallari, R.; Martelli, F.; Rosini, R.; Buratti, C.; Verdone, R. A survey on wireless body area networks: Technologies and design
challenges. IEEE Commun. Surv. Tutor. 2014, 16, 1635–1657. [CrossRef]
29. Kaleem, M.; Devarajan, G. Energy-efficient classification strategy for detecting interference and malicious sensor nodes in wireless
body area Networks. Cyber Secur. Appl. 2024, 2, 100048. [CrossRef]
30. Bedi, P.; Das, S.; Goyal, S.; Rajawat, A.; Kumar, M. Energy-Efficient and Congestion-Thermal Aware Routing Protocol for WBAN.
Wirel. Pers. Commun. 2024, 137, 2167–2197. [CrossRef]
31. Cai, G.; Fang, Y.; Wen, J.; Han, G.; Yang, X. QoS-aware buffer-aided relaying implant WBAN for healthcare IoT: Opportunities
and challenges. IEEE Netw. 2019, 33, 96–103. [CrossRef]
32. Ali, A.; Khan, F. Energy-efficient cluster-based security mechanism for intra-WBAN and inter-WBAN communications for
healthcare applications. EURASIP J. Wirel. Commun. Netw. 2013, 2013, 216. [CrossRef]
33. Rakhee; Srinivas, M. Energy efficiency in load balancing of nodes using soft computing approach in WBAN. In Harmony Search
and Nature Inspired Optimization Algorithms: Theory and Applications, ICHSA 2018; Springer: Berlin/Heidelberg, Germany, 2018;
pp. 423–430.
34. Samal, T.; Kabat, M. A prioritized traffic scheduling with load balancing in wireless body area networks. J. King Saud Univ.-Comput.
Inf. Sci. 2022, 34, 5448–5455. [CrossRef]
35. Shankar, V.; Natarajan, A.; Gupta, S.; Schwiebert, L. Energy-efficient protocols for wireless communication in biosensor networks.
In Proceedings of the 12th IEEE International Symposium on Personal, Indoor and Mobile Radio Communications (PIMRC 2001),
San Diego, CA, USA, 30 September–3 October 2001; Proceedings (Cat. No. 01TH8598); IEEE: Piscataway, NJ, USA, 2001; Volume 1.
36. Vyas, A.; Pal, S. Exploiting redundancy and mobility in energy-efficient communication for WBANs. In Proceedings of the 21st
International Conference on Distributed Computing and Networking, Kolkata, India, 4–7 January 2020; pp. 1–8.
37. Masroor, K.; Jeoti, V.; Drieberg, M. Improving the energy efficiency of a wireless body area network using a redundant coordinator
for healthcare applications. In Proceedings of the 2018 International Conference on Intelligent and Advanced System (ICIAS),
Kuala Lumpur, Malaysia, 13–14 August 2018; IEEE: Piscataway, NJ, USA, 2018; pp. 1–5.
38. Khan, R.; Memon, S. Sensors Transposing Technique for Minimizing the Path Loss Effect and Enhancement of Battery Lifetime in
Wireless Body Area Sensor Networks. In Proceedings of the 3rd International Conference on Networking, Information Systems
& Security, Marrakech, Morocco, 31 March–2 April 2020; pp. 1–8.
39. Peng, Y.; Wang, X.; Guo, L.; Wang, Y.; Deng, Q. An efficient network coding-based fault-tolerant mechanism in WBAN for smart
healthcare monitoring systems. Appl. Sci. 2017, 7, 817. [CrossRef]
40. Yadav, D.; Tripathi, A. Load balancing and position-based adaptive clustering scheme for effective data communication in WBAN
healthcare monitoring systems. In Proceedings of the 2017 11th International Conference on Intelligent Systems and Control
(ISCO), Coimbatore, India, 5–6 January 2017; IEEE: Piscataway, NJ, USA, 2017; pp. 302–305.
41. Shukla, S.; Sachan, V.; Sinha, A.; Pandey, S.; Rao, G.; Shah, M.; Choudhary, A.; Tamrakar, B. WHOOPH: Whale optimization-based
optimal placement of hub node within a WBAN. Sci. Rep. 2024, 14, 3422. [CrossRef] [PubMed]
42. Patra, S.; Samal, T.; Kabat, M.; Mishra, M.; Barik, R. Optimizing relay node placement and routing in WBANs using free search
krill herd and harmony search algorithm. Multimed. Tools Appl. 2024, 1–39. [CrossRef]
43. Singla, R.; Kaur, N.; Koundal, D.; Lashari, S.; Bhatia, S.; Rahmani, M.K.I. Optimized energy efficient secure routing protocol for
wireless body area network. IEEE Access 2021, 9, 116745–116759. [CrossRef]
44. Selem, E.; Fatehy, M.; Abd El-Kader, S. mobTHE (mobile temperature heterogeneity energy) aware routing protocol for WBAN
IoT health application. IEEE Access 2021, 9, 18692–18705. [CrossRef]
45. Shawqi J., A.; Kadhum Idrees, A. Adaptive rate energy-saving data collecting technique for health monitoring in wireless body
sensor networks. Int. J. Commun. Syst. 2020, 33, e4589. [CrossRef]
46. Mehrani, M.; Attarzadeh, I.; Hosseinzadeh, M. Sampling rate prediction of biosensors in wireless body area networks using
deep-learning methods. Simul. Model. Pract. Theory 2020, 105, 102101. [CrossRef]
47. Kadhum Idrees, A.; Alhussein, D.; Harb, H. Energy-efficient multisensor adaptive sampling and aggregation for patient
monitoring in edge computing-based IoHT networks. J. Ambient. Intell. Smart Environ. 2023, 15, 419.
48. Jain, S.; Jain, P.; Upadhyay, P.; Moualeu, J.; Srivastava, A. An Energy Efficient Health Monitoring Approach with Wireless Body
Area Networks. ACM Trans. Comput. Healthc. (HEALTH) 2022, 3, 1–22. [CrossRef]
49. Harb, H.; Makhoul, A.; Couturier, R.; Medlej, M. ATP: An aggregation and transmission protocol for conserving energy in
periodic sensor networks. In Proceedings of the 2015 IEEE 24th International Conference on Enabling Technologies: Infrastructure
for Collaborative Enterprises, Larnaca, Cyprus, 15–17 June 2015; IEEE: Piscataway, NJ, USA, 2015; pp. 134–139.
50. Mohammadi, R.; Shirmohammadi, Z. Optimizing energy harvesting in wireless body area networks: A deep reinforcement
learning approach to dynamic sampling. Alex. Eng. J. 2024, 109, 157–175.
Sensors 2024, 24, 6531 19 of 19

51. Li, S.; Yang, H.; Xu, F.; Hu, H.; Hu, F. Energy-Efficient Relay Transmission for WBAN: Energy Consumption Minimizing Design
With Hybrid Supervised/Reinforcement Learning. IEEE Internet Things J. 2024, 11, 17770–17779.
52. Oliveira, C.; Ghamri-Doudane, Y.; Brito, C.; Lohier, S. Optimal network coding-based in-network data storage and data retrieval
for IoT/WSNs. In Proceedings of the 2015 IEEE 14th International Symposium on Network Computing and Applications,
Cambridge, MA, USA, 28–30 September 2015; IEEE: Piscataway, NJ, USA, 2015; pp. 208–215.
53. Passos, C.; Pedroso, C.; Batista, A.; Nogueira, M.; Santos, A. GROWN: Local Data Compression in Real-Time to Support Energy
Efficiency in WBAN. In Proceedings of the 2020 IEEE Latin-American Conference on Communications (LATINCOM), Santo
Domingo, Dominican Republic, 18–20 November 2020; IEEE: Piscataway, NJ, USA, 2020; pp. 1–6.
54. Razavi, R.; Fleury, M.; Ghanbari, M. Rateless coding on a wearable wireless network for augmented reality and biosensors.
In Proceedings of the 2008 IEEE 19th International Symposium on Personal, Indoor and Mobile Radio Communications, Cannes,
France, 15–18 September 2008; IEEE: Piscataway, NJ, USA, 2008; pp. 1–4.
55. Wang, K.; Song, M.; Bian, G.; Shao, B.; Huang, K. A Lightweight Identity-Based Network Coding Scheme for Internet of Medical
Things. Electronics 2024, 13, 1316. [CrossRef]
56. Boussaid, A.; Alaoui, N.; Aghzout, O.; Chakkour, Y.; Alejos, A.; Falcone, F. Highly efficient error correcting codes for ubiquitous
healthcare in Wireless Body Area Networks. In Proceedings of the 2016 International Conference on Electrical and Information
Technologies (ICEIT), Tangiers, Morocco, 4–7 May 2016; IEEE: Piscataway, NJ, USA, 2016; pp. 225–230.
57. Gupta, A.; Chaurasiya, V. QoS supported redundancy balanced data transmission scheme for wireless body area network-
based H-IoT. Wirel. Netw. 2023, 29, 3793–3808.
58. Matloff, N. Introduction to Discrete-Event Simulation and the Simpy Language; Department of Computer Science, University of
California at Davis: Davis, CA, USA, 2008; Volume 2, pp. 1–33.
59. Ong, E.; Kneckt, J.; Alanen, O.; Chang, Z.; Huovinen, T.; Nihtilä, T. IEEE 802.11 ac: Enhancements for very high throughput
WLANs. In Proceedings of the 2011 IEEE 22nd International Symposium on Personal, Indoor, and Mobile Radio Communications,
Toronto, ON, Canada, 11–14 September 2011; pp. 849–853.
60. Arslan, M.; Guzel, M.; Demirci, M.; Ozdemir, S. SMOTE and Gaussian noise-based sensor data augmentation. In Proceedings of
the 2019 4th International Conference on Computer Science and Engineering (UBMK), Samsun, Turkey, 11–15 September 2019;
IEEE: Piscataway, NJ, USA, 2019; pp. 1–5.
61. Rak, J.; Pickavet, M.; Trivedi, K.; Lopez, J.; Koster, A.; Sterbenz, J.; Çetinkaya, E.; Gomes, T.; Gunkel, M.; Walkowiak, K.; et al.
Future research directions in the design of reliable communication systems. Telecommun. Syst. 2015, 60, 423–450. [CrossRef]
62. Sharma, V.; Kumar, S.; Bhushan, S. An overview of data redundancy reduction schemes in WSNs. In Proceedings of the 2017
3rd International Conference on Advances in Computing, Communication & Automation (ICACCA) (Fall), Dehradun, India,
15–16 September 2017; IEEE: Piscataway, NJ, USA, 2017; pp. 1–8.
63. Hasan, K.; Wu, X.; Biswas, K.; Ahmed, K. A novel framework for software-defined wireless body area network. In Proceedings
of the 2018 8th International Conference on Intelligent Systems, Modelling and Simulation (ISMS), Kuala Lumpur, Malaysia,
8–10 May 2018; IEEE: Piscataway, NJ, USA, 2018; pp. 114–119.
64. Cicioğlu, M.; Çalhan, A. HUBsFLOW: A novel interface protocol for SDN-enabled WBANs. Comput. Netw. 2019, 160, 105–117.

Disclaimer/Publisher’s Note: The statements, opinions and data contained in all publications are solely those of the individual
author(s) and contributor(s) and not of MDPI and/or the editor(s). MDPI and/or the editor(s) disclaim responsibility for any injury to
people or property resulting from any ideas, methods, instructions or products referred to in the content.