In traditional airplane monitoring system (AMS), data sensed from strain, vibration, ultrasound of structures or temperature, and humidity in cabin environment are transmitted to central data repository via wires. However, drawbacks still exist in wired AMS such as expensive installation and maintenance, and complicated wired connections. In recent years, accumulating interest has been drawn to performing AMS via airborne wireless sensor network (AWSN) system with the advantages of flexibility, low cost, and easy deployment. In this review, we present an overview of AMS and AWSN and demonstrate the requirements of AWSN for AMS particularly. Furthermore, existing wireless hardware prototypes and network communication schemes of AWSN are investigated according to these requirements. This paper will improve the understanding of how the AWSN design under AMS acquires sensor data accurately and carries out network communication efficiently, providing insights into prognostics and health management (PHM) for AMS in future.
In a typical commercial/military aircraft, the AMS includes safety-critical system (e.g., engine control system, flight control system) and nonsafety critical systems (e.g., structural and engine health monitoring system, cabin environmental control system, and inflight entertainment system) [
Till now, the progress achieved in embedded sensors technologies and wireless data transmission has extended the monitoring capability of aeronautical structures, spacecraft and ground testing equipment, and cabin environment [
An attractive use of AWSN in AMS is sensing. In aircraft, due to high angles of attack in takeoff/landing, sudden pilot manoeuvres, turbulence, wind gusts, and normal shock waves on the wing at transonic speed, boundary layer separation on the wings the occurs, resulting in the phenomenon of parasitic drag and stall. For this reason, deployment of wireless airflow control actuators on strategic locations especially wing for providing real-time airflow information and decision metrics to the local system sheds light on the construction of efficient closed-loop airflow control operation [
This idea of operating AWSN for AMS was initially introduced by researchers in the early 21 century and several researchers remarked the potential benefits of this technology over traditional AMS systems. Nevertheless, issues should be still underscored when these systems kept in a long-term operation. As a matter of fact, the first AWSN prototypes employed low-processing MCU coupled with low-resolution analog-to-digital converter (ADC) and low sampling rate, but the technique makes sophisticated sensing and conditioning elements available. Recently, a number of scholars have been dedicated to investigating key issues (e.g., powerful hardware prototype, network protocol, time synchronization, and passive sensing) in this disciplinary field, indicating that AWSN becomes increasingly practical in AMS [
Several well-known research institutes have invested adequate funds for AWSN based AMS. For example, the Wireless Interconnectivity and Control of Active Systems (WICAS) project funded by the Engineering and Physical Sciences Research Council (EPSRC) applies AWSN to aircraft wing active flow control [
Rapid advances on composite materials and piezoelectric sensors have presented new opportunities to AMS, essential to make more comprehensive analysis for damage, impact, and crack monitoring [
This paper is organized as follows: Section
In this section, the context of AWSN is discussed, from both a general and a specific perspective. Firstly, a brief description of AMS is presented to illustrate the role of AWSN in this novel concept and framework. Then, an AWSN diagram is introduced to show the structure of general AWSN in which the architecture of airborne wireless sensor nodes, communication networking, and node deployment are demonstrated. Finally, we summarize the characteristics required for AWSN to accommodate new requirements of AMS.
In this review, the AMS mainly includes airplane structural health monitoring (SHM) and airplane cabin environmental monitoring. The AMS collects data from various sensors deployed on airplane structures and installed inside airplane cabin, implementing structural health monitoring and cabin environmental monitoring, respectively. With the rapid development of new materials and advanced technology in airplane, the modem structures are becoming complicated increasingly. The airplane SHM, including Lamb wave damage detection technology, optical fiber based global condition perception, multisensor fusion detection technology, and structural health assessment, provides approaches for assessing the health condition and ensures the safety of the complicated structures. Especially, structural response is obtained through strain, vibration, ultrasound, and piezoelectric sensor online structural health monitoring on airplane structures. The response is used to evaluate structural health status and assess the residual life of the airplane structure, further to develop PHM. Overall, the SHM is aim to save money spent on maintenance or replacement and ensure the structure to operate efficiently during its whole intended life.
The airplane cabin full of a mixture of outside and recirculated air is a semienclosed structure. Generally, the airplane cabin is in a low humidity, low pressure dynamic condition. Additionally, various concentrations of ozone (O3), carbon monoxide (CO), carbon dioxide (CO2), and other chemicals are generated and spread over the airplane cabin. Different locations (e.g., on the ground, in ascent, at cruise, or in decent) of the airplane determine the level of contaminants invaded from outside sources. If the level of contaminants is not real-time monitored and adjusted timely, it is harmful and dangerous for passengers and crews.
To further illustrate the framework of AWSN, we will concentrate on wireless communication from the perspective of AWSN deployed in AMS. As shown in Figure
The airborne wireless network schematic diagram.
The basic block of any wireless sensor network is the airborne wireless sensor board. The appropriate selection of board is favorable for the performance of wireless monitoring. As shown in Figure
The schematic diagram of airborne wireless sensor board.
The computation core is the primary difference between a airborne wireless sensor board and its wire-based counterpart. The presence of a microcontroller unit (MCU) allows for onboard data processing, data storing, and preparing for communication. To fulfill these tasks, the measured data and executable program (such as damage detection routines) are embedded in random access memory (RAM) and read only memory (ROM), respectively. The size (in bits) of internal data bus for microcontroller is classified as 8-, 16-, or 32-bits, determining processing speed and power consumption. Many different memory sizes and employed algorithms are commercially available, which are tailored in conformity with the particular monitoring activity to be performed.
The sensing section is dedicated for converting analog output into a digital representation that can be handled by digital electronics. Some typically used sensors for AMS application include strain gauges, temperature sensors, accelerometer and piezoelectric sensors. Many sensing sections integrate more than one type of sensing elements, while others incorporate one sensor concentrating on one kind of physical quantity for accuracy and power-saving reasons. Usually, the section includes amplifier, linear, compensator, and filter. The sensing resolution relies on the ADC effective number of bits and measurement range in Volts, coupled with sensitivity of sensors. For most AMS applications, ADC resolution of 16 bits or higher is preferred for detecting signals. For example, generally, low sampling rates (e.g., less than 500 HZ) are adequate for aircraft structural health monitoring. However, wireless sensors are increasingly investigated for applying in acoustic and ultrasonic NDE; therefore, there has been a growing desire for higher sampling rates in excess of 500 Hz.
The presence of radio frequency (RF) communication allows each board to interact with other nodes and to forward sensing data. For this reason, more stress on effective communication needs to be laid for the sake of AMS’s reliability and high-performance transmission. This is particularly true as high data sampling rate, high fidelity sensing, high transmission rate, and large transmission range are often involved in AMS. RF communication is real challenge on aircraft structures made of composite or steel components.
The last subsystem would be the actuation interface in which the core element is the digital-to-analog converter (DAC). It allows converting digital data generated by MCU into a continuous analog voltage output for exciting active sensors (e.g., piezoelectric elements) interplayed with the physical structures. Actuators and active sensors installed on physical system can both be handled by an actuation interface.
In Table
Comparisons of different Wireless Avionics Intra-Communications.
Standards | Standard | Max. | Frequency | Free-space range | Spectrum |
---|---|---|---|---|---|
WiFi | IEEE 802.11 | 54 Mbps | 2.4 GHz | 150 m | Unlicensed |
Zigbee | IEEE 802.15.4 | 250 Kbps | 868/915 MHz/2.4 GHz | 300 m | Licensed |
Bluetooth | IEEE 802.15.1 | 24 Mbps | 2.4/5 GHz | 150 m | Unlicensed |
RFID | ISO/IEC24791 | 640 Kbps | 125 kHz/13.56 MHz | 10 cm–10 m | Unlicensed |
LoRaWAN | IEEE 802.15.4 | 50 Kbps | 433/780/868/915 MHz | 14 Km | Unlicensed |
SigFox | IEEE 802.15.4 | 100 Kbps | 868/902 MHz | 17 Km | Unlicensed |
NB-IOT | IEEE 802.15.4 | 200 Kbps | 900 MHz | 22 Km | Licensed |
WirelessHART | IEEE 802.15.4 | 250 Kbps | 2.4 GHz | 150 m | Unlicensed |
ISA100.11a | IEEE 802.15.4 | 250 Kbps | 2.4 GHz | 150 m | Unlicensed |
WiMAX | IEEE 802.16 | 300 Mbps | 11 GHz | 100 m | Unlicensed |
60 GHz | 60 GHz | 3000 Mbps | 5–7 GHz | 10 m | Unlicensed |
UWB | UWB | 200 Mbps | 3.1–10.6 GHz | 10 m | Unlicensed |
WAIC | C-Band | 250 Kbps–200 Mbps | 2.2–3.4 GHz, | 100 m | Unlicensed |
General aircraft body is consisted of left and right wing, cockpit or cabin, engine, vertical tail, left and right horizontal stabilizer, landing gear, front, middle, and rear sections which are installed in subsystems of aircraft. Due to the dispersing deployment characteristic of the subsystems, cluster-star network topology is more suitable for AWSN in AMS [
The AWSN has some requirements particularly for the environment of AMS. Generally, the architecture of ordinary WSN is constituted with different layers including application, transporting, routing, medium access control (MAC), and physical layer. However, this traditional layer architecture cannot satisfy the requirements for AWSN applied in AMS. The requirements cover accuracy, real-time, reliability, time synchronization, throughput, longevity and safety and security in terms of their constraints, challenges, and design goals for realizing AWSN, as shown in Figure
The characteristics for AWSNs.
The accuracy of acquired data is an important data quality aspect. In the AWSN, data accuracy is directly related to the accuracy of the airplane prognostics and health management, such as fault detection, fault isolation, fault prognosis, and prognosis of the remaining life. Additionally, the accuracy of acquired data affects the safety of airplane, the economic profits, and flight efficiency. In AWSN, the context is also associated with time synchronization, node number, hop number, and sampling rate.
WSNs have been applied in smart plants, industrial environment monitoring, and automation factories for low latency wireless communication, which sets an example for the AWSN. Real-time performance is an important issue for the AWSN. Communications between wireless nodes require latency to improve productivity. However, different applications may have different real-time requirements. Table
Real-time requirements in different AMS.
AMS | Latency |
---|---|
Automation system | <3 ms |
Environmental monitoring system | <1 s |
Flight motion control system | <60 ms |
Wing control | <100 ms |
The overview of all aeronautical spectrum bands and their services.
Band | Frequency | Service |
---|---|---|
HF | 3–30 MHz | A/G communication system, A/A communication system |
VHF | 117.95–137 MHz | A/G communication system, satellite-based aeronautical communication system |
L-Band | 960–1215 MHz | A/G communication system, Extended Squitter (ES) signals, and AWSN |
X-Band | 10 GHz | Satellite-based aeronautical communication system |
Ku-Band | 14.4–14.3 GHz | |
Ka-Band | 26–40 GHz | |
S-Band | 2.2–3.4 GHz | AWSN, WAIC |
C-Band | 4.2–4.5 GHz | WAIC, Radio Altimeters |
C-Band | 5.091–5.15 GHz | Airport surface applications, WAIC. |
Each airborne wireless sensor node has its own local clock, which is not initially synchronized with other nodes. Two jitters, namely, temporal jitter and spatial jitter, occur inside node and between different nodes, respectively, due to variation in oscillator crystals. Time synchronization errors between different devices mean that obtaining the proper mode shapes without deviating from reality or theoretical calculations of structure is impossible [
For AWSN application, it is required to improve the network throughput for collecting large amount of acquired data in AMS. Theoretically, the baseline throughput between two nodes with single-radio at IEEE 802.15.4 band is 250 Kbps. Furthermore, Osterlind and Dunkels proved that the maximum data throughput at IEEE 802.15.4 band is 225 Kbps [
Many studies have pointed that WSNs are still constrained by energy limits because most of airborne wireless sensor nodes are usually battery-powered. In this case, it is impractical to replace and recharge a large number of batteries. Furthermore, in AWSN, batteries not only provide energy for wireless communication but also for mechanical systems, so the energy limitation is a critical challenge for prolonging network life. Typical WSN deployments employ battery-powered nodes. These onbattery nodes are usually deployed randomly and widely. It is impossible for them to be recharged once deployed. We should try our best to provide staple networking functionality within limited energy budgets and reduce their energy consumption by using innovative routing algorithm. In this section, we will concentrate on two aspects to prolong the operation of AWSN and discuss the longevity.
The first approach to ensure longevity of AWSN is efficient energy management and conservation. Energy efficient strategies have been implemented in different layers. For instance, in the physical layer, unnecessary actions can be reduced, and the physical parameters can be optimized to achieve strong power-saving performance. In the application layer, several useful methods such as event-driven techniques, application-driven techniques, and efficient data/messaging can be used to decrease energy consumption. Additionally, as for networking and communication, the protocols of MAC routing can be designed to reduce the energy consumption. For instance, sleeping and working mechanism is adopted to achieve good energy conservation.
Energy harvesting and transferring power wirelessly are other approaches to improve the longevity of AWSN, which redefine the traditional design of battery-operated AWSN. The WSNs coupling with ambient energy harvesting can prolong the system’s lifetime or possibly enable perpetual operation. Meanwhile, the self-powered nodes have longer life-time for routing and path selections for data transmission than on-battery nodes. Many researchers have studied energy harvesting by exploring environmental energies such as solar, vibration, wind, and microwave [
Some sources of energy harvesting for AWSN.
Authors | Harvesting energy |
---|---|
Samson et al. [ | Thermoelectric |
Lu et al. [ | Vibration |
Siu et al. [ | Solar |
Azevedo and Santos [ | Wind |
Zhao et al. [ | Microwave |
Safety and security are an important consideration in safety-critical avionics applications. In both wired network and AWSN, they should be considered in the OSI protocol stack including physical, MAC, routing, transportation, and application layer. The AWSN is vulnerable to malicious attacks in all layers, and security vulnerabilities with these layers are separately protected at each layer.
Jamming [
Characteristics and countermeasures of attacks in various OSI layers.
Attacks | Characteristics | OSI layer affected | Countermeasures |
---|---|---|---|
Malware attack | Trojan horse, worms, key-loggers, and viruses | Application | Firewalls and antiviruses |
SQL injection | Acquiring unauthorized access to websites | Application | Firewalls and antiviruses |
TCP flood | Sending massive ping requests | Transport | Reducing packets response |
UDP flood | Sending massive UDP packets | Transport | Reducing packets response |
IP hijacking | Legal users IP address impersonation | Routing | Firewalls |
Smurf attack | Sending massive ICMP requests | Routing | Reducing packets response |
Mac spoofing | MAC addresses falsification | MAC | ARP packets |
MITM attack | Communicating nodes impersonation | MAC | Virtual private networks (VPNs) |
Jamming attack | Interrupting legal data transmission | Physical | Spread spectrum techniques |
Denial of service | Sending abundant packets | Physical | Temper-proof packaging |
In Section
Summary of airborne wireless hardware prototypes.
Wu et al. [ | Becker et al. [ | Gao et al. [ | Pook et al. [ | Liu et al. [ | Lu et al. [ | |
---|---|---|---|---|---|---|
Computing specifications | ||||||
Processor | Atmel AVR ATMega 128L | Texas Instruments | Texas Instruments | Atmel AT32UC3A3256 | Altera Cyclone II | Texas Instruments |
Clock speed | 8 MHz | 8 MHz | 30 MHz | 66 MHz | 50 MHz | 32 MHz |
Bus size | 8-bit | 16-bit | 16-bit | 32-bit | 16-bit | 16-bit |
Program memory | 128 KB | 32 KB | 64 KB | 256 KB | 64 KB | 128 KB |
Data memory | 128 KB | 64 MB | 64 KB | 128 KB | 8 KB | |
| ||||||
Data acquisition specifications | ||||||
Sensing type | Strain gauge | Strain gauge | Piezoelectric | CO, CO2, pressure, | Piezoelectric | MEMS |
A/D channels | 8 | 8 | 4 | 8 | 8 | 8 |
A/D resolution | 12-bit | 12-bit | 16-bit | 10-bit | 12-bit | |
| ||||||
Wireless specifications | ||||||
Radio | ChipCon CC1000 | ChipCon CC2420 | Atmel | ChipCon CC2420 | ChipCon CC2420 | iM22A |
Frequency band | 433 MHZ | 2.4 GHz | 2.4 GHz | 2.4 GHz | 2.4 GHz | 2.4 GHz |
Outdoor range | 70 m | 75 m | 100 m | 75 m | 75 m | 100 m |
Data rate | 76.8 Kbps | 250 Kbps | 1 Mbps | 250 Kbps | 250 Kbps | 250 Kbps |
Comparison of the characteristics features for airborne wireless prototypes.
Study (-) | Throughput | Max. sample rate (Hz) | Synchronization | Target application (-) |
---|---|---|---|---|
Wu et al. [ | 250 | 32 | - | Strain monitoring for aircraft carbon fiber reinforced plastic wing box |
Wu et al. [ | 250 | 32 | - | |
Liu et al. [ | 250 | - | - | Impact monitoring for aircraft wing box |
Delebarre et al. [ | - | - | - | Impact monitoring for aircraft wing box |
Gao et al. [ | 909 | 32 | 50 | Strain monitoring for aircraft wing box and an UAV composite wing |
Gao et al. [ | 2000 | 10000 | - | Aircraft aluminum plate |
Loo et al. [ | 250 | - | - | Aircraft corrosion monitoring |
Pook et al. [ | 250 | - | - | |
Becker et al. [ | - | 40000 | - | Laboratory environment |
Zhao et al. [ | 250 | 15000 | - | Ultrasound monitoring for aircraft wing inspection |
Demo et al. [ | 250 | 1000 | - | Aircraft corrosion monitoring |
Hall et al. [ | 250 | 50000 | - | Aircraft corrosion monitoring |
Arms et al. [ | 250 | 512 | 39000 | Strain and vibration monitoring Bell Model 412 helicopter |
Blanckenstein et al. [ | 250 | - | - | RSSI and BER for Airbus A330-300 |
Lu et al. [ | - | 5300 | - | Lab environment |
Some of the proposed wireless sensor nodes: (a) Samson et al. [
In 2007, Zhao et al. [
From 2007 to 2009, Wu et al. [
In 2008 and 2009, Arms et al. [
Generally, the WSNs for AMS mostly relied on active RF transmission. Some researchers described specific wireless passive devices under harsh conditions in aerospace vehicles. In 2011, Elmazria and Aubert [
In 2012, as for online aircraft impact damage monitoring, Delebarre et al. [
In 2016, starting from previous work proposed by Yuan et al. [
Due to minor damage detection ability of Lamb wave produced by piezoelectric sensors for AMS, piezoelectric sensors showed great potential promises for online aircraft structural health monitoring. For this reason, Gao et al. [
In 2016, Lu et al. [
In 2009, Becker et al. [
In 2009, Loo et al. [
In 2010, Demo et al. [
In 2012, Hall et al. [
In 2014, Blanckenstein et al. [
Efficient medium access control (MAC) scheme and routing protocol are highly required in AWSN. Many transmission protocols have been proposed for WSN. Demirkol et al. [
The feasibility and reliability of collision-free protocol on the basis of IEEE 802.15.4 standard in wireless links for AWSN are fully considered [
Although a number of improved CSMA protocols have been proposed for WSN in AMS, they still suffer from many issues due to persistent collisions. The contention-free protocol such as TDMA algorithm provides better performance on latency, jitter, and spatial reuse than improved CSMA protocols. Meanwhile, to achieve high accuracy or high resolution of damage localization in aircraft structure, high-precision data synchronized acquisition in AMS should be underlined. To cope with these issues of star-cluster WSNs, Zhou and Jing [
In Table
Comparison of network communication schemes for AWSN.
Study | Platform | Target application | Reliability | Energy | Real-time | Type of network | Framework | Data flow |
---|---|---|---|---|---|---|---|---|
Krichen et al. [ | Matlab | Vibration monitoring | Medium | Medium | | Homogeneous | CSMA-CA and | Continuous |
Barcelo et al. [ | Experiment | Aircraft condition monitoring | Medium | Medium | | Homogeneous | CDA and CSMA-CA | Event based |
Ma et al. [ | Experiment | Aircraft cabin environment monitoring | Medium | No | | Homogeneous | CSMA-CA | Event based |
Guanglin and Hongyu [ | NS2 | Aircraft vibration monitoring | Medium | No | | Homogeneous | Zigbee/CSMA-CA | Event based |
Notay and Safdar [ | OPNET | Aircraft condition monitoring | Medium | No | | Homogeneous | Zigbee/CSMA-CA | Event based |
Ma et al. [ | OPNET | Air flow control for aircraft | High | High | | Homogeneous | TDMA | Continuous |
Zhou and Jing [ | OPNET | Aircraft health monitoring | High | High | | Homogeneous | TDMA | Continuous |
Ma et al. [ | OPNET | Vibration monitoring of | High | High | | Homogeneous | TDMA | Continuous |
Arms et al. [ | Experiment | Bell Model 412 helicopter | High | Medium | | Homogeneous | CSMA, TDMA | Continuous |
Blanckenstein et al. [ | OPNET | Airbus A330-300 | High | High | | Homogeneous | TDMA | Continuous |
In recent years, more practical WSNs including system design and network protocol have been successfully proposed for AMS, evaluating safety of aircraft flight and engine control, the quality of aircraft cabin environmental condition, and the flexibility of aircraft structural health. Nevertheless, there is room for improvement of WSNs for AMS. More work is needed to allow WSNs to fulfill the requirements for large-scale AMS. One important challenge researchers are now facing is turning the sensor node from ordinary data acquired device into online data processing intelligent “brain,” making the WSN more powerful and efficient [
We should also identify how the achievement of high-accuracy time synchronization for large-scale AMS (especially in aircraft structural monitoring) is attained. The accuracy of time synchronization is closely correlated with network protocol and transmission bandwidth, sampling rate, and other factors. A large amount of data transmission results in network congestion, lower synchronization quality, and higher power consumption, whereas smaller data amount might reduce the accuracy of data analysis and increase network delay. It should be highlighted that tradeoff algorithm among the amount of data transmission, synchronization and network quality, and power consumption needs to be explored. More energy facilitates maximization of operating time for WSN. However, the battery life of WSN is too short for performing long-term AMS. Energy harvester [
Nonsafety critical systems (e.g., wireless smoke and fire detection system, cabin emergency wireless-controlled lighting system) have been certified by the Federal Aviation Administration (FAA) which is typically operated in an unlicensed spectrum [
The AWSN is a promising technology which plays an increasingly key role in the AMS applications. However, few AWSN surveys consider its application background in AMS, so this is motivation of this review. The AWSN creates a new set of challenges in terms of accuracy, real-time, reliability, time synchronization, throughput, longevity, and safety and security. In this paper, a brief survey of AWSN and AMS including a range of areas from the general to the specific. The survey focuses on the prototype architecture and network communication schemes of the developed AWSN in AMS. It is outlined in this review that the node architecture and network communication schemes proposed in the early 21st century are being developed and updated continuously, providing amounts of solutions for AMS. The first AWSN prototypes employed low-processing MCU coupled with low-resolution ADCs, but technology makes sophisticated sensing and conditioning elements available. Recent modifications and improvements made on board’s MCU, RF transceiver, signal conditioning unit, and time synchronization show how WSN becomes mature gradually for applied in AMS. We further demonstrate that wireless piezoelectric platform based on Lamb wave algorithm shows great potential promises for online AMS. Besides, most traditional network communication schemes cannot satisfy all requirements of real-time AWSN in AMS. Intelligent algorithm (e.g., game theory) or machine learning might act as a new approach to resolve issues intrinsic to these protocols. These findings not only provide theoretical evidence and appropriate solutions to WSN design according to AMS requirements but also novel insights into airplane prognostics and health management (PHM).
The authors declare that there are no conflicts of interest regarding the publications of this paper.
This work was supported in part by Nanjing University of Science & Technology under Research Start-Up grant (no. AE89991/032).