Abstract
LoRa technology is extensively utilized in the Internet of Things world. It allows a transmission of a low volume of data through small wireless devices. The principle of LoRa networks is to transmit data over the air from sensors with low transmission range, for about tens of kilometers. Those sensors are not expected to be powered by electricity, and they are powered by batteries. We understand that visits to hospitals cannot be eliminated and that visits for full examinations were necessary, but technological progress nowadays could reduce the burden on hospitals thanks to remote controls and treatments in homes using those wireless sensors. So, the use of LoRaWAN protocol could greatly make diagnostic of patients more easily by transmitting data between doctors and patients in a real time manner. The aim of this work is to evaluate the performance of a network that contains numerous mobile sensors. Those sensors connect the doctors, nurse, and patient through a reliable and secure wireless network. Here, we want to evaluate various factors of LoRaWAN protocol that have a big effect on power consumption and data transmission delay.. Moreover, our LoRa-based networking implementation, based on software simulations, appears to be an option that allows for a robust, reliable, and lower overall cost IoT deployment and low bandwidth requirements. With LoRa, we can achieve similar or better link quality to IEEE 802.15.4, with higher data rate and lower costs.
1. Introduction
Wireless body area networks (WBAN), which use the human body as a data transmission medium, have attracted the attention of several researchers around the world. If we confirm that the future is associated with the Internet of Things (IoT) and artificial intelligence (AI), the speed of data transmission will be a critical issue to be improved in the network since the speed of data transmission per second has a great impact on the development of services of all kinds. We know that the wireless sensor network (WSN) suffers from the low power of transmission between sensors; this weakness can lead to problems in transmission reliability. Today, developers use the 5G technology that has a high transmission rate to transmit information and signals. For this reason, a marriage between 5G technology and wireless sensor networks has been completed to improve the next generation of the telecommunications world. To use WBAN with this new technology, it is prohibited to extend wireless communication system to support high frequency band of transmission and multiantenna system that should be developed and taking into consideration the power efficiency of each node.
New generations of related items will best be capable of expand if it is miles feasible to cause them to completely power self-sufficient, even if the usage of batteries or cells solves part of this hassle by ensuring autonomy that can be important with relatively low costs. Indeed, for many applications, connected objects are deployed in environments that are difficult to be reached by humans. So, replacing or maintaining batteries becomes complicated or even impossible. In addition, another major concern is linked to the limited lifetime duration of the batteries, which makes the autonomy of the communicating object totally dependent on its energy source. In recent years, many technologies have been developed and introduced to solve the problem of energy consumption such as ZigBee, Bluetooth, and LoRaWAN. With the rise of WSN, low-power wide area networks (LPWAN) are taken into consideration as one of the key solutions for optimizing energy consumption and it has improved its efficiency to enable the new human healthcare and wireless monitoring applications.
The important challenge of many countries is to improve their healthcare system to be more efficient, more coordinated, and to be ready for any pandemic like coronavirus disease 2019 (COVID-19). To be a good competitor in the healthcare system, the use of recent trends of technologies will be a requirement. The use of technology can be a feasible solution to reduce the expenditure of handicap monitoring, people in quarantine, and people with chronic disease. The IoT provides a world of connected devices to a cloud-based applications and services, with different cooperation mechanisms, a proper standardization, progressed sensors with inexpensive, and low-power microprocessors [1]. According to Table 1, LoRaWAN is considered one of the best IoT solutions based on the healthcare monitoring system due to its high communication range and perfect interoperability between IoT sensors.
This work opens a way to compare the main LPWAN technologies as presented in the following table and to assess their performance and choose the best solution for transforming to smart cities that contains many wireless connected devices. Remaining parts of this paper are organized as follows: Section 2 gives associated work to LPWAN technologies and energy consumption improvement. Next, in Section 3, a survey about wireless body area network will be discussed. Section 4 provides an overview of the LoRaWAN protocol. In Section 5, we discuss the theoretical parameters that may affect energy consumption and data transmission using the LoRaWAN protocol. Simulation results proving the efficiency of LoRaWAN parameters on the energy per useful bit will be carried out. Finally, Section 7 concludes the paper.
2. Related Work
In recent years, many applications are achieved to widely lay out the evaluation of the next generation of mobile communication and wireless sensor network and to improve the healthcare monitoring system. The authors in [5, 6] integrated a 5G system in WSN application that improves energy consumption and provides a secure healthcare system. They are interested in the new generation of mobile communication in the WBAN, where they use a connection between small smart devices via 5G technology to establish connection between numerous intelligent sensors at the same time. However, authors in [7] have used new three-factor rapid authentication technologies to hide user recognition in wireless sensors networks based on 5G technology. The use of the 5G technology and wireless sensor network contributes to the progress of the electronic healthcare system. Each user should be identified by a biometric, password, and a smart card with a time bounded identification to register with the server. The concept of hospital of future was discussed in [8] where authors proposed a reconfigurable hybrid radio communication system based on radioconnectivity and wireless optics. This proposed solution uses an optical transmission, enhance the redundancy, minimize the throughput, and minimize the spectrum congestion. Enemy warfare attract some authors in [9], where they implemented an automatic healthcare monitoring system to track location of soldiers in real time; they use Arduino board kit and IoT technology with a Wi-Fi module to control problems that can face soldiers as gas leakage and bombs by using sensors to detect any sudden health problems.
Due to the continuous monitoring of patient‘s health, sensor nodes will consume a lot of energy during sensing and transmitting data. Without an estimated technique for data transmission and data gathering, the performance of the WBAN can be degraded. Several works like [10–13] address the energy consumption problems in WSN using a fuzzy algorithm to schedule the data transmission under MAC layer. Many works have been carried out on the energy efficiency in WSN and especially in WBAN as investigated in the review paper [14]. That work conducts a survey on LPWAN in WBAN to improve the healthcare monitoring of patients while satisfying quality of system requirement in terms of energy efficiency, latency, and data transmission rate. They cited the advantage and disadvantage of different communication techniques of such as LoRa, LPWAN, NB-IoT, and SigFox. Recently, many researchers focus on the relation between patients and doctors to assure a reliable connection between them especially in a harsh environment such as rural area, while at the same time, they focus on reducing the energy consumption of IoT devices [15].
Energy efficiency in LoRa-based networks, especially when using the LoRaWAN protocol, is a well-studied topic. The authors in [16] studied the effect of LoRaWAN parameters together with stated transmission, scatter factor, encoding speed, payload length, and data range. This optimization study presented a trade-off among LoRaWAN communication range, propagation factor, and power transmission. [16] was very interesting for the choice and configuration of LoRa parameters. Several researchers tried to enhance the quality of service (QoS) to find the better solution that guarantee a high data transmission delay and a secure link that does not affect the energy consumption. Table 2 presents recent works in the literature that uses LoRaWAN technology where each one used a different method compared with others; they used different types of LoRaWAN sensor transceiver chip that support devices of class A, B, and C. In almost of those researches, energy saving in LoRa networks has been improved; most of them deal with LoRaWAN parameters and the trade-off between energy consumption, location, reliability, and distance. Since minimizing information is a realistic energy-saving approach, we will focus on this work on the data transmission period and its impact on energy consumption.
3. Wireless Body Area Network: WBAN
WBAN uses the electric field of the human body to transmit data wirelessly, for example, from the music player to the headphones, from the electronic key of the car in the pocket to the unlocking system of the doors, or from the cardiac sensor to the control unit worn around the belt. This amazing technology, pioneered in the United States by IBM and MIT for the wearable computer perspective, is the subject of intense research around the world. In WBAN, sensors are used to monitor, to collect, and to transmit medical signals and other information about the human body like electroencephalography (EEG), electrocardiography (ECG), and temperature, directly to a node known as “sink node” as shown in Figure 1. Collected data will be then be transmitted via the LoRaWAN gateway that is installed in the hospital, and then, it will be sent to the doctors and nurse to be analyzed.
WBANs differ from typical wireless sensor networks on a large scale, and they are characterized by mobility in the network which follows the movements of the human body and the quality of the links which varies according to the posture of the wearer. Also, the transmitting power of the sensors is kept low in order to improve their autonomy and to reduce the exposure to the electromagnetic waves of the carriers. Therefore, considering the absorption, reflection, and interference effects of the body, it is difficult to maintain a direct connection between the “sink” and the other nodes. LoRaWAN can support wireless devices connected in the same network to a distance of 30 km, it means that it can be a good solution for transforming to smart cities and a good IoT solution to improve the healthcare domain specially in recent years where we had faced many chronic diseases. In the current context of health crisis and in the face of COVID-19, European health authorities have put in place new recommendations imposing temperature and CO2 thresholds to be respected. They have installed new IoT sensors to measure the indoor air quality in hospitals, for example, it is considered that above 1000 ppm in the indoor environment, the health of occupants is at risk [27]. The objective of those IoT solutions is to improve the health of occupants and nursing staff in hospitals and to consider the recommendations of the health authorities. As an example, a total of 87 of wireless radio sensors were installed in the two Madrid hospitals on LoRaWAN wireless technology in a private mode. 79 of transmitters, specifically designed for indoor applications, have been installed in order to check whether the sanitary conditions (CO2, VOC, ambient temperature, and humidity) are respected. As a study case, we have five companies operating within the subject of the Internet of Things. [28] [29] [30] [31] have come together to design a system intended for healthcare personnel and aimed at equipping hospitals with emergency call buttons connected with LoRaWAN long-range, low-power radiotechnology. This system flags patients requiring immediate attention and allows medical personnel to locate and identify these calls in real time. The call button device allows patients to report an emergency at any time. This waterproof, compact, and easy-to-clean “sensor” incorporates a Bluetooth detector that can immediately identify and locate the equipment or person concerned. It can also measure ambient temperature and detect physical contact through a built-in speaker. Nurses then have real-time access to essential information about patients’ emergency calls and the location of their beds.
4. LoRaWAN Overview
4.1. LoRaWAN Protocol
IoT is a network where objects are related to the Internet. Its implementation requires the use of low-power wide area network (LPWAN) in order to support connected objects. These are low-power, long-range communication networks. It is precisely to install these networks that the LoRaWAN protocol was developed. This protocol reduces the consumption of connected devices. LoRaWAN is a media access control type protocol allowing low-data rate communication and low energy consumption. This protocol is widely used in the context of smart cities, industrial monitoring, or even agriculture. LoRaWAN has becomes an interesting method access in smart sensing application using IoT. This new technology belongs to the LPWAN networks. LoRaWAN has improved a greater efficiency regarding other access method to the medium like CSMA and ALOHA [32]. It has enhanced the collision ratio, but it is not considered a good optimizer for energy consumption for many nodes. Due to its low cost, high data rate, energy efficiency, and improved packet delivery ratio, LoRa attracts a lot of researchers and industrial in different fields ranging from environment monitoring to application in smart cities and smart homes. LoRaWAN support three types of end devices, namely, class A, B, and C where each device is characterized by a specific data frame transmission [33, 34], while end devices can always send uplinks, the device’s class determines when it can receive downlinks. Class A has the lowest energy consumption where the data frame transmission support only in the uplink. However, a device in group B allows a compromise among power consumptions and the need for two-way communication. These devices open reception windows at a scheduled interval by periodic messages sent by the server. Devices from group C have the highest power consumption but it provides bidirectional communications that are not programmed where the devices have an everlasting listening window.
Figure 2 gives the architecture of LoRa topology. The network topology supported here is called a star-of-star because a network server is connected to a multitude of gateways that are themselves connected to a multitude of end devices. In the network, the end devices are not connected to the gateways; they have only used them as relays to reach the server that manages the network, which is itself connected to one or more application servers. The packets sent by the end devices are retransmitted by the gateways after adding only information concerning the quality of the signal received.
Today, there are a wide number of technologies that are used under the MAC layer in order to access the medium, like LoRaWAN, SigFox, and CSMA/CA. Also, there are several numbers of researches that use LoRaWAN technology in such application like data monitoring and energy efficiency. Compared to other wireless technology, LoRaWAN is the most useful technology in the wireless sensor network due to its flexibility, low energy consumption, and its low cost. Technological progress, nowadays, has led to use LoRa technology and the 5G network together to support smart cities that can be adapted to different environmental conditions and especially in healthcare applications. Using WBAN, healthcare monitoring of patient status can be done easily using LoRa technology even in the rural area. In this work, a study of different wireless technology supported in the IoT are discussed and compared to prove the weakness and advantage of each one of them using simulation analysis.
4.2. LoRaWAN End Devices
Sensors are used to monitor and pick up any information in different environments to prevent any issues before having major problems. Researchers focus of data transmission between nodes to enhance cost and power consumption while improving workplace productivity. Data transmission depends on itself by the protocol of communication used to handle communication between nodes. RA ecent standard used in IoT application is the well-known LoRaWAN standard that approved its efficiency in long range and high performance in network operations. IoT sensors harvest information and send them using LoRaWAN gateways; then, those gateways transfer data through a cloudy system that process data and store them in a database to be retrieved later by the user. Here, the user received data by end device’s that are considered a radio bridge console. Technology is the key secret of success of most of organizations whatever its service: healthcare, manufacturing, aerospace, and telecommunication. So, improving their productivity depends on the system they used to handle data and treat it.
LoRa devices are an electronic system belonging to the IoT world having low consumption, small size, low power, and low cost. They have a LoRa radio to reach the gateways. The gateways are not addressed specifically: all those present in the coverage area receive the messages and process them. Any LoRa device is composed of a customized electronic board composed of a microcontroller, energy module, LoRaWAN radio module, wireless module, sensing module, and interface USB and converter as presented in Figure 3. This chip has a small size and can be integrated in any space. The microcontroller is considered a central data processing unit which is connected to almost all of the components of the sensor device. A serial peripheral interface SPI is used to connect between the microcontroller and the LoRaWAN radio module along with digital input output lines I/O. An 868 MHz monopole antenna is connected to the wireless module for radio communication. A universal serial bus (USB) interface is used for microcontroller firmware upload.
4.3. LoRaWAN Packet Structure
To calculate the LoRa packet time transmission on air, we need to start by calculating the size of the payload. As mentioned in Figure 4, an implicit or explicit LoRaWAN packet consists of three main parts. The packet begins with a preamble which is used for synchronization between end devices and the gateway; after, we found the header which is considered optional and is used to indicate the scale of the payload and some other data related to LoRa parameters, its value is fixed to and then the payload who comes with an adjustable value of CR and with a cyclic redundancy check (CRC) who comes on the end of the packet frame.
5. Discussion and Performance Metrics
LoRaWAN consumes extremely low energy. It used a simplified version of the additive links on-line Hawaii area (ALOHA) protocol for the MAC layer and cyclic transmission in each subband. Parameters used in the optimization of energy during the transmission-reception cycle are configured based of node status (transmission, reception, standby, and sleep), the information transmission strategy, and the power transmission. The power consumption of nodes (end equipment) of the LoRaWAN network is chosen according to the use of the four main modes (transmit, receive, standby, and sleep) and the time spent in each mode. The authors in [14] modelled the energy consumption from these four modes, in order to compare the consumption of a long-range linear wireless sensor network designed by ZigBee or LoRaWAN. Energy consumption depends mainly on several factors: the quantity of data transferred (in number of messages and/or size of messages), as well as the transmission power required to transmit this data, and the spectrum spreading factor (SF).
Through this paper, we desire to improve the performance of LoRaWAN in terms of energy consumption. We used NS-3 as a simulator tool for the wireless sensor network [35]. In order to optimize the energy consumption, LoRaWAN protocol allows a dynamic set of SF parameter to reduce the consumption of the end devices, as well as to free up radio bandwidth and therefore to limit collisions. So, in the first step, we are going to study in more detail the impact of these different parameters in the network performance. The energy consumption of a LoRa system depends on several parameters such as (i)Quantity of data to be sent (payload)(ii)Spreading factor(iii)Collisions during transmission (and therefore retransmission)(iv)Request for acknowledgment of the frames sent(v)Duty-cycle(vi)Transmit energy of the transceiver(vii)Power consumed in standby between 2 transmissions
6. Network Model
6.1. Algorithm Specification
The network model is a star topology that contains 100 nodes. Each one can be connected to the personal area network (PAN) coordinator. The good advantage of this type of topology is that it is highly reliable. This network contains a combination of static and mobile nodes, so the failure of one or some nodes cannot cause a significant problem in the network operations since many other nodes are available to maintain stability in the network. In this network, there are some nodes considered end devices and gateway router devices that are used to forward each received router packet to the network server and execute transmission request coming from the server. In the developed algorithm, we suppose that the network contains mobile nodes and static nodes.
| ||||||||||||||||||||||||||||||
6.2. Energy Model Assumption
To compute the energy consumed in the network, we should suppose some assumptions; first, we considered that the energy was consumed while the radio module transmit data is constant. Whereas the energy consumed while the processing data is considered a variable value and it depends on the function of the sensor node, it is known that sensor node goes into a sleep period while they are neither sensing nor sending data and the energy consumed during this mode is statistically considered zero. So, to compute the energy consumed of the rest of operation mode, we assume that all its elements are active during a specified period and inactive during the rest of the cycle. The initial energy of all nodes is the same for all devices which is equal to 1 J. Lets define the hole energy consumed as follows: where , , and are, respectively, the initial energy of the node, the energy consumed by the node, and the energy consumed during the sleep phase. We said that the energy of inactive period is nearly about zero so the total energy consumed can be calculated as presented in
with where , , , , , and are the energy consumed during the initialization of the system, the energy consumed while sensing, the energy consumed in data processing, the energy consumed in the radio, the energy used to send data, and the energy consumed in the sensing for the channel, respectively.
To study the autonomy of a sensor node, it is prohibited to model the power consumption of each unit of the node as follows:
Clearly, the energy of each unit depends of the frequency of the processor and the power consumed by the microprocessor . The time needed by the processor to process data is equal to where the latter can be calculated as follows: where is the number of instructions executed by the processor, and here, we assume that we have one instruction per clock period. Also, we can compute the time needed to transmit data as presented in
Where presents the number of sent bits and presents the transmission time of an information bit, respectively. Finally, we can define as the operating time of the microprocessor. It depends on the operating time of all the sensor units. is expressed as follows:
The following formula is used to calculate the LoRa Packet duration: where calculates the symbol duration in millisecond, PL calculates the payload size indicated in bytes, calculates the preamble duration, indicates the payload and header duration, and payloadSymbNb is the number of symbols in payload period. As an example, Table 3 calculates the spreading factor (SF) with time on air mentioned for a bandwidth equal to 250 kHz.
To evaluate the effect of LoRaWAN parameter on the energy consumption of each node, we can calculate the energy per useful bit denoted , which is used in order to evaluate the energy consumption of the sensor node. The energy is calculated in the following equation:
So, we have where PL is the size of the payload and is the total power. Using equation (12), the expression of energy per useful bit is calculated as shown in
7. Simulation Results
7.1. Test Performance and Node Placement
To evaluate the performance of LoRaWAN end devices which include 100 end devices, the topology described in Figure 5 is used. We particularly evaluate the power consumption and ToA of LoRa to optimize the best configuration parameters that may affect the network performance. Simulation parameters are presented in Table 4.
Various operating scenario modes of a sensor node are presented in Figure 6. Each scenario allows us to define different functional modes, which are managed by the processing unit. In order to measure and transmit information, sensors begin by awakening the system from its hibernation state. Then, the sensor makes periodic measurements to verify the displacement of the structure. After that, it should run out the necessary processing of the measured data.
This process mainly used for the most part depends on the frequency of the microcontroller. Finally, the LoRaWAN module is awakened in order to transmit the information and receive an acknowledgment of reception to verify the correct transmission of data. To optimize the consumption, a microcontroller is put into a standby state at the end of the operating cycle.
7.2. Effect of SF and CR on the Energy Consumption
LoRa modulation has several configuration parameters such as carrier frequency (CF), scatter factor (SF), bandwidth (BW), and encoding rate (CR). Those parameters present different levels of power consumption and different transmission ranges where each parameter is specified as follows: (i)Carrier frequency: present the central frequency used during transmission. CF is 863-870 MHz in Europe(ii)Spreading factor: define the number of bits per symbol varying between 7 and 12. The higher the SF value, the greater the receiver’s ability to receive data with low signal-to-noise ratio value. Thus, the larger the value of SF, the longer the transmission time(iii)Bandwidth: presents the range of frequencies in the transmission band of LoRaWAN and its value ranges from 125 to 500 kHz(iv)Coding rate: it is used to improve the efficiency of LoRaWAN protocol by adding a cyclic error coding (CRC) to the encoding process of data transmission. The value of CR is equal to with vary from 1 to 4. The increase in the value of CR will increase the reliability of LoRa, but it increases energy consumption because of the large value of CR that will increase time on air
In this part, we will evaluate the effect of LoRa parameters on the power consumption by varying the value of SF, CR, and BW. An optimization of these parameters makes it possible to regulate the power consumption of the sensor node and the data transmission delay. A node presents different modes of transmission which are presented in Table 5 and will be synthesized in the simulation.
As calculated in equation (15) that represents the relation between the consumed energy and the spreading factor (SF), we can conclude that this parameter may affect the energy as shown in Figure 7. This figure presents the evaluation of power consumption per bit as a function of payload for different values of SF. We noted that the consumed energy decreases with the increase in the number of useful bits. As stated previously, the larger value of SF, the longer it takes to transmit one packet and consume more energy to transmit data. When the value of the spreading factor becomes high, the power consumption is high also; here, we have the highest value of SF is 11.
Figure 8 presents the energy consumed per useful bit as a function of payload and under different values of coding rate (CR) (while maintaining SF fixed). When the coding rate decreases (means increasing the number of coding bits), the transmission time of a packet and the power consumption increased too. These results show that an optimization of LoRaWAN parameters such as SF, CR, and payload size is a key element in reducing energy consumption for such a communicating sensor. When the value of CR is low, the energy consumption is high. So, the lowest value of CR is 4/5 in this experiment.
Figure 9 shows the variation of the ToA as function of the payload while varying SF. Here, we can clearly see that if SF is increased the ToA parameter will increase, too. The variation of the transmission time on air depends on the value of SF. For example, suppose that we take for kHz, here, the value of the transmission time will be equal to 60 ms. When we increased the value of SF for example to 12, the value of the transmission time will be increased also to 872 ms. This increase will affect the energy consumption that will be approximately higher; it means that each node will consume more energy while sending data. Here, we can say that a good compromise to increase the ToA is to change between the subbands of the reserved channel.
The impact of the CR on the transmission parameter time on air is shown in Figure 10. It is noted that the increase in the number of coding bits leads to an increase in packet transmission time, which also leads to a higher consumption of the radio module.
8. Conclusions
IoT sensor devices make the work within hospital or homes more sophisticated. The use of recent trends in electronic devices and protocol of communication may improve the data transmission without loss and within a brief time. But some devices need a good choice of batteries and a good management of information to be powered during a long period of time. One aspect of this research work consists of minimizing the energy consumption in order to improve the ease of use and to enhance the lifespan of the batteries, which could ultimately lead to a network. In our future work, we are ready to study the impact of other parameters on the behavior of energy consumption between communicating nodes through LoRaWAN technology, to mention but not limited to, we can cite the packet delivery ratio (PDR), throughput, and jitter.
Data Availability
The processed data are available upon request from corresponding author.
Conflicts of Interest
The authors declare that they have no conflicts of interest.
Acknowledgments
The authors extend their appreciation to the Deanship of Scientific Research at King Khalid University for funding this work through a general research project under grant number (R.G.P.1/194/41).