A novel energy-efficient adaptive superframe structure for OWC-based real-time bio-sensor networks

Biosensors are becoming more common in the healthcare industry, but battery life is a significant barrier to broader adoption. Another concern is that these sensors heavily rely on radio frequency technology, which is dangerous to the human body and the environment. This research presents an optical wireless communication (OWC) based battery-efficient adaptive superframe for real-time biosensors. To begin, OWC uses visible light for downlink and infrared for uplink, making it human-safe. Second, our retrofit technique makes sensor transmission delays and energy usage more efficient. Third, as we look into wearable sensors, we see that the one-of-a-kind continuous superframe structure saves energy and allows for emergency data handling via the deployment of an emergency beacon. Fourthly, we compare the simulation results to the IEEE 802.15.7 standard superframe timeline to evaluate the practicality of our suggested technique. Analytical expressions are formulated to predict and analyze endto-end delays, average energy consumption, and savings. The results show a significant improvement in energy savings and delays without incurring any additional overheads.


Introduction
A Biosensor network (BSN) is a communication paradigm that connects implantable and wearable healthcare devices.These nanoscale, low-power devices can be attached to or embedded inside the human body.Implantable BSN devices can include a pacemaker, cardiac defibrillator, or retinal sensor, while wearables include Electroencephalogram (EEG), Electrocardiogram (ECG), SpO2 pulse oximeters, and others.The data gathered by these sensing devices is processed and transmitted to the gateway, communicating with the medical server or doctor.The technology enables medical practitioners to monitor, diagnose, and treat patients remotely.For instance, any abnormalities in the body organs of a sportsman or soldier can be discovered and tracked virtually before any life-threatening incident [1].
Additionally, the coronavirus (SARS-CoV), which causes the severe acute respiratory syndrome, has recently attacked the world.A range of procedures has been undertaken with the assistance of BSNs to detect its early symptoms [2].An artificial intelligence (AI) framework with body sensors to forecast the severity and course of illness caused by this fatal virus is proposed in [3].The suggested design allows the use of cellphone sensor data, which doctors and radiologists can access from any location and at any time via their mobile phones.
BSN can utilize numerous RF-based technologies, including Zigbee, 6LowPAN, WiFi, Bluetooth, etc., to conduct these operations.IEEE 802.15.6 and IEEE 802.15.4 standards are the basis of numerous Wireless sensor networks (WSN) protocols' implementation.These standards specify the physical layer (PHY) and medium access control (MAC) requirements for sensor networks.IEEE 802.15.4 is meant for limited-power, lowcost wireless personal area networks (WPANs), whereas IEEE 802.15.6 is designed only for wireless body area networks (WBANs).Despite the widespread use of these well-established technologies in research, considerable limitations remain.
Moreover, while the PHY layer for these technologies is mature, the MAC layer still confronts challenges.In addition, none of these technologies are safe for individuals or the environment.In addition, the world is currently experiencing a spectrum scarcity as the number of Internet-connected gadgets increases every day [4].
According to [5], a piece of battery-powered biomedical implant equipment should have a minimum life expectancy.Additionally, the MAC must be secure and compatible with various devices and data kinds.The BSN-MAC layer is a software layer that enables the wireless connection of sensors and actuators to the human body.BSN-MAC must meet various criteria, including low power consumption, energy efficiency, and long battery life.
As an alternative to the prior discussed technologies, we suggest OWC technology, which is more secure when utilized near the human body and offers higher data speeds [6,7].OWC can merge lighting and data communication by using LEDs and LASERS that are intensity changed at a rate significantly faster than the endurance of the human eye.Further, the recent technological advancements in OWC have enabled the development of nanoscale medical implants.The human body's stimulus is recorded and transformed into an electrical signal for further processing.OWC provides medical healthcare with unparalleled bandwidth, better safety, and compact designs [8].
The IEEE 802.15.7 standard [9] for short-range OWC, published in 2018, claims to transmit enough data to support audio and video multimedia services.It specifies six application-specific PHY layers and a MAC layer that can manage visible and other light wavelengths.
The standard maintains a timing structure known as a superframe for MAC functionality.This structure comprises an active and inactive segment delimited by the beacon frame.The contention-free time (CAP) and the contentionaccess period are further divisions of the active part (CFP).The CFP provides tiny slots for guaranteed data transmissions (as described in Sect.3).This work intends to offer a modified superframe structure for energy efficiency and emergency data handling.
The novel contributions of the work are as follows: This article covers current BSN technologies in detail to provide a technical understanding of innovative developments (Sect.2).
The work describes the conventional architecture of an OWC-based BSN.(Sect.3).
The research describes an innovative continuous superframe construction for enhanced energy efficiency without adding overheads (Sect.4).
The suggested superframe is retrofitted to the standard and can handle any emergency using its sleep mode (Sect.5).
The results indicate a substantial improvement in energy consumption and delays (Sect.6).
The following sections constitute the remainder of the article: Sect. 2 examines the relevant state-of-the-art emergency handling and energy-efficient Guaranteed time slot (GTS) allocation techniques.Section 3 discusses the design, limitations, and difficulties of BSN implementation.Sect.IV presents the proposed superframe structure.The simulation settings and performance metrics are presented in Sect.V, followed by an explanation and interpretation of the results in Sect.6. Section 7 finishes the paper with potential directions for further research.

Background and related work
The designing of MAC protocols is a vast area of study, and significant progress has already been made utilizing RF; meanwhile, BSN views OWC as a rapidly growing field.This section summarizes the published literature on BSN-MAC performance enhancement strategies.
Based on IEEE 802.15.4 specifications, [10] seeks to enable energy-efficient GTS allocation for internet of things (IoT) devices.The suggested technique adjusts the duty cycle dynamically based on the remaining energy to guarantee that high-energy devices are given priority while low-energy devices are given power while they are waiting.Simulations show that using the suggested protocol instead of the IEEE 802.15.4 standard can extend the life of IoT nodes and improve the amount of data they can supply by up to 94 percent and 79 percent, respectively.The suggested protocol, however, makes no mention of the delays that are added for the low-priority devices.
In [11], the wake-up radio (WuR) technique manages emergency data transmission in WBAN-based ultralowpower healthcare applications.By delivering a WuR message, wearable sensors can relay their data on demand.For analyzing the protocol's performance, the author proposes a modified superframe structure with analytical modeling based on the M/G/1/2 queuing theory.The author has simulated the proposed superframe structure for throughput, delay, and energy consumption, but with the caveat that all results were plotted using only one value of beacon order, that is 4. [12] describes an energy-aware multi-group composite MAC protocol for biomedical sensors to extend battery life and conserve energy.To improve overall network lifetime, the proposed solution combines time-division multiple access (TDMA) with carrier sense multiple access (CSMA/CA).A sleep-wake-up mechanism with a scheduling algorithm for non-critical packets extends battery life even further.In terms of throughput, responsiveness, and hop-to-hop reliability, the suggested MAC outperforms the former state-of-the-art, but at the expense of some overhead.The sensor node must send a preamble along with every payload.
A dynamic distribution-based GTS allocation mechanism is presented in [13], in which allocation is entirely reliant on real-time network traffic.The Knapsack technique is used to allocate GTS, and fairness is also maintained under severe latency constraints, however, energy efficiency is completely disregarded.
A redesigned superframe structure with squeezed CFP slots is offered in [14], doubling the GTS slots from seven to fourteen.The author uses the adaptive duty cycle to accommodate changing network traffic requirements.
Furthermore, instead of employing the 'first come, first serve' strategy, the 'shortest job first' technique reduces the network delay.However, the sensors cannot in any way request the transmission of emergency data.
[15] presents another traffic prioritization-based energyefficient MAC for WBANs.Unconstrained, delayconstrained, reliability-constrained, and critical data are considered the four types of sensor node data.The priority of data is determined by altering the traditional superframe structure.Further, guard bands are also used in the proposed strategy to prevent channel interference, resulting in reduced delays.However, throughout the simulations, the values of SO, BO, and the duty cycle are kept fixed at 2, 3, and fifty percent respectively.
Based on traffic analytics, [16] presents an adaptable MAC for WBANs.This TDMA-based protocol reduces synchronization overhead and increases energy efficiency by dynamically adjusting the network's QoS needs.Further, by utilizing adaptive packet size and flexible data delivery, the system extends the device's sleep time to over ninety-nine percent.
The document [17] presents a hybrid technique for reliable and timely communication in cyber-physical medical systems.The proposed approach dynamically allocates GTS services to nodes based on their respective data types.The suggested protocol includes a time-triggered methodology, a priority queue framework, and an adaptive way to allocate mini-slots.Consequently, the critical data is guaranteed to be communicated reliably and in real-time, while less important information is exchanged at other times.
A unique energy-efficient MAC algorithm for sensorbased WBAN is introduced in [18].A central coordinator connects and manages all the wearable sensors.Unless the coordinator initiates a communication, all nodes remain in sleep mode, saving energy and reducing collisions.
Numerous solutions for efficient GTS allocation have been presented to improve the quality of Service (QoS), reduce latency, and maximize bandwidth consumption.However, most of the work is focused on changing the CAP and CFP periods without regard for the idle period.Our proposed scheme is distinguished from existing approaches in two ways.First, we've made the inactive period adaptable and opportunistic in response to traffic priority, making the system more reliable in any emergency.Secondly, we used a continuous superframe structure to reduce energy consumption significantly.

A. BSN Architecture
A typical BSN is equipped with several sensors engaged to or implanted in the body to monitor various biological parameters such as temperature, respiration rate, blood glucose, and calorie burn.Figure 1 illustrates the suggestive BSN communication architecture comprising intra-BSN, inter-BSN, and beyond-BSN.
Tier, one represents the application of OWC-based BSN biosensors to the human body.Communication between nodes of the same BSN is termed intra-BSN communication.These sensing devices are generally connected in a star topology, and the data obtained is aggregated by an OWCenabled lamp acting as a gateway or coordinator.A classifier embedded in the coordinator then classifies the exit data as regular traffic (RT) or contingency traffic (CT).The information is then delivered via ethernet cable to the doctor's PDA or laptop for additional processing.This is accomplished using the Inter-BSN.The data can then be relayed to the ambulance, another medical center, or hospital database through the internet.This is an example of communication that occurs outside of the BSN.

B. Applications
BSN has applications in both the medical and nonmedical domains.Wearable, implantable, and remote sensing sensors are the three types utilized in the medical industry.BSN finds numerous applications in the non-medical field due to its low cost and limited computational and energy requirements.Multiple applications exist in entertainment, recreation, sports, defense, and the military [19,20].

C. Design issues and Challenges
Tier 1 is critical in the creation of medical and consumer MAC protocols.It is accountable for the overall performance parameters of a typical BSN, including energy consumption, robustness, dependability, and scalability.As a result, we may achieve high throughput, high energy efficiency, and low latency by using efficient MAC layer protocols at the tier-one level.
BSN protocols must be capable of processing different forms of data traffic to maintain their priority.Additionally, the MAC protocol must be flexible enough to accommodate changing resource allocation requirements.The subsequent major challenge is energy efficiency, as sensors have limited battery life.While the battery of wearable sensors is simple to change, this cannot be done frequently with implantable sensors due to the possibility of requiring costly invasive surgery [21,22].
The hardware design of wearable sensors is particularly critical, as they may overheat due to antenna radiations, causing irritation and impairment to the body's heat-sensitive tissues.Finally, body movement and environmental changes might cause sensor readings to fluctuate abruptly, which is another concern [22].

Proposed superframe structure
This section addresses the proposed scheme with IEEE 802.15.7 standard superframe structure adjustments to allow contingency traffic.This novel superframe decreases latency, optimizes network utilization, and requires less energy.The fundamental purpose of shifting periods is to deliver critical data with the least bandwidth and power practical.Figure 2 represents the standard superframe structure starting with the beacon, followed by CAP, CFP, and an optional sleep period.Figure 3 illustrates the proposed BSN-MAC superframe design with a 50% duty cycle.The traditional structure utilized the inactive period of the sensors as a standby mode, whereas the suggested system does so on an as-needed basis.The design includes two additional fields: time to report contingency (TTRC) and contingency beacon (CB).These two sections have been taken out from the active period.And an optional adaptive CFP part for GTS (CFP G1 ) allocation to devices in an emergency has been taken out from the inactive portion.Following that, there are the standard CAP and CFP parts.
A ZIGBEE-based sensor network can also make use of the suggested superframe structure.A similar approach can be used to examine the work, with the exception that RFbased technology will be used instead of visible light communication/optical wireless communication (VLC/OWC) in this instance.The frequency bands used will differ, and as a result, so will the lengths of the superframe slots.Rest certain that every component will stay the same.

A. How an emergency is declared (priority-based data segmentation)
We considered two kinds of traffic in our suggested scheme: regular and contingency.Segmentation is achieved via data received from a particular sensor.For instance, the average standard heart rate (at rest) range is around 60 to 100 beats per minute (bpm).Thus, anytime the pulse rate surpasses this range, the data is critical, and contingency is declared.
Table 1 highlights the allowable readings from several sensors considered acceptable under regular traffic.By combining the data from one or more sensors, we may picture any irregularity in the human body.For instance, accelerometer and heart rate data may reveal a lot about Take the case of a patient in a hospital bed as our example for now.His vital signs are fine while the doctor or the nursing staff are present, but as soon as they left, the patient's heart rate begins to rise and eventually goes above the allowed limit.Now, the coordinator inside the lamp receives the signal from the sensor on the patient's body and generates a contingency signal.And this is how the emergency is declared.

B. How energy consumption is reduced (continuous superframe structure)
A continuous-superframe design is suggested with n beacons to adapt to the algorithm for energy efficiency.Further, the inactive component of the superframe is made adaptable to accommodate emergency needs.
As mentioned in the previous sections, two data types can be generated from the sensors.As an analogy, the sensors and their readings can also be categorized into periodic and priority.For instance, blood glucose is a regular reading, so the sensor may not need to ask for emergency GTS slots.On the other hand, a cardiac sensor may be classified as a priority sensor.
Assume that s1 and s2 are two sensors that produce periodic data, and the time interval after which their data is required can span several superframes.If we make them skip those superframes, we can save the energy they will consumewhile receiving those beacons that are useless to them.This can be performed if we use a continuous superframe structure, as shown in Fig. 4. It can be seen that s1 and s2 can be allotted a GTS slot every 3rd and 2nd superframe, respectively.Hence both the sensors can go to sleep mode during that duration and save energy.This is how energy efficiency is achieved.

Performance evaluation
We have considered ten wearable sensors with six that may generate emergency data for simulation purposes.These sensors can communicate with the central coordinator (gateway) via single-hop.The data communication within the CAP and CFP G2 are with ack, whereas we have not considered ack for the transmissions in CFP G1 as the packet size is minimal, and we cannot compromise on delay.Table 2 lists the various simulation parameters and their range of values.We have focused mainly on the delay and energy consumption of the nodes while comparing our structure with the standard.
The relationship between Superframe order (SO) and Supeframe duration (SD) is expressed as follows: (1) SD = aBaseSuperframeduration × 2 SO , 0 ≤ SO ≤ 14 Similarly, the value of Beacon order (BO) and Beacon interval (BI) are associated in the following way: Data transmission is available in CAP and CFP via the CSMA/CA mechanism and GTS allocation.All communications commence at the start of the backoff slot since the superframe runs in slotted, beacon-enabled mode.Whenever a node intends to send a data packet, it first enters a random backoff phase by choosing a number between (2 macminBE − 1) and then executes a Clear chan- nel assessment (CCA) that lasts one backoff slot.The data packet is sent if the medium is deemed free; the node returns to the backoff phase.The initial value of the Backoff exponent (BE) is assigned to macMinBE (3), the default value specified by the standard.Consequently, the backoff time durationT &( is computed as: Further, in accordance with the standard, we have opted for an essential line speed of 48Mbps and an optical clock rate of 120MHz.The Run Length Code (RLL) is 8B10B, while the Forward Error Correction Code (FEC) is taken as RS (64,32).
The data packet size ( T PHYheader , T MACheader ) is cal- culated by adding the PHY and MAC layer headers ( T PHYheader , T MACheader ) to the payload (T packetsize ).4) and ( 5), the overall transmission time for a data packet can be expressed as: A. Delay analysis The average transmission delay according to the standard and the proposed superframe for a packet to be transmitted in CFP can be calculated as: (refer to Figs. 2 and 3) When two or more nodes attempt to report an emergency, the CSMA/CA technique is utilized to transmit the request message.In addition, the emergency beacon and data transmission do not affect regular data transfer.This is because

B. Energy consumption calculation
We have defined the average energy consumption of a body sensor as the amount of energy dissipated by it during the transmit, receive, and sleep modes.Hence, the average energy consumption according to the standard and the proposed superframe can be computed as:  where, P r P t and P r denotes the power consumed while reception, transmission, and sleep mode, respectively.Calculating the amount of energy saved when a sensor skips a single beacon while following a continuous superframe structure is straightforward.Therefore, the sensor's energy usage is proportionate to the number of beacons it bypasses each time interval.

Results and discussions
We have used the MATLAB simulation tool to evaluate and compare our suggested scheme with the standard superframe.We used three data sets to examine our suggested strategy: end-to-end delay, average consumption, and energy savings.We plotted them against the various values of BO, data payload, and skipped beacons.
Figures 5 and 6 compare the delay and energy usage of the standard and proposed superframes with varied BO levels.Both parameters reflect a substantial level of improvement.Even though we have transmitted an additional beacon for emergency packet transfer, the delay, and energy consumption are low.This is due to the placement of our segment immediately after the standard beacon.
According to Eqs. ( 1) and ( 2), we can see that the Beacon interval is directly dependent on the value of the Beacon order and similarly the total superframe duration depends upon the order of the superframe which is related to the beacon order by the formula mentioned in the Eq.(2).
Hence, as the value of BO rises, SO rises as well, elevating BI and SD in the process.Due to the formula's usage of a log 2, however, the changes are minor until BO reaches a value of 7 or 8.But after that, the scale shifts as BO increases, beyond 8, and now the values of SD and BI are increasing steadily, Higher SD values result in larger TTRC values, which causes long delays.We can observe that delays and energy usage are directly related in Eqs.(7)(8)(9)(10).So, our direct high increase in energy consumption is caused by abruptly growing delays.
Figures 7 and 8 compare standard and proposed superframes' delay and energy consumption with varying data packet sizes.In both instances, an increase in payload size results in a rise in both delay and energy consumption.The reason for this trend is that as the size of the data packet increases, so does the amount of time and energy required to transmit it.Now we shall address the saving on energy consumption when a continuous superframe structure is utilized.The amount of energy saved if regular sensors begin skipping beacons (based on their data interval) is depicted in Figs. 9 and 10.The trend appears to be precisely related to the number of beacons missed.

Fig. 3 Fig. 4
Fig. 3 Proposed continuous superframe structure with m beacons The packet transmission time can be calculated as:where, T IFS denotes the Interframe spacing, T ack is the acknowledgment transmission duration and T SIFS represents (3) D Prop = [T beacon + T ΠRC + T BO + T CB + T data + T IFS + (8)] (4) T data = T packet size + T PHY header + T MAC header (5) T transmission = [T data + T IFS + T ack + T SIFS ]

Fig. 5
End-to-end delay comparison for the standard and the proposed scheme with varying values of BO the short IFS.Using equations (

( 6 )
T transmission = T packetsize + T PHY header + T MAC header +T IFS + T ack + T SIFS (7) D Std = T beacon + T CAP + T BO + T CFP + T IP + T beacon +T CAP + T data + T IFS +T ack + T SIFS (8) D Prop = T beacon + T TTRC + T BO + T CB + T data +T IFS + T ack + T SIFS we derived the emergency reporting and transmission durations from the superframe's inactive period.

EFig. 6 Fig. 7
Fig. 6 Average energy comparison for the standard and the proposed scheme with varying values of BO

Fig. 8
Fig. 8 Average energy comparison for the standard and the proposed scheme with varying values of data packet size

Table 1
Acceptable readings from wearable sensors under regular traffic

Table 2
Simulation parameter settings