Saeid Sanei

Body Sensor Networking, Design and Algorithms


Скачать книгу

though for many different objectives, from monitoring undersea wildlife to volcanic activity [7]. Echoing the investments made in the 1960s and 1970s to develop the hardware for today's Internet, the United States Defense Advanced Research Projects Agency (DARPA) started the Distributed Sensor Network (DSN) programme in 1980 to formally explore the challenges in implementing distributed WSNs. With the birth of DSN and its progression into academia through partnering universities such as Carnegie Mellon University and the MIT Lincoln Laboratory, WSN technology soon found its place in academia and civilian scientific research.

      Governments and universities eventually began using WSNs in applications such as air quality monitoring, forest fire detection, natural disaster prevention, weather stations, and structural monitoring. Then as engineering students made their way into the corporate world of the technology giants of the day, such as IBM and Bell Labs, they began promoting the use of WSNs in heavy industrial applications such as power distribution, wastewater treatment, and specialised factory automation.

      Although BSNs' objective and technology have their own requirements, they owe their birth and early development, particularly with regards to data communication, to the WSN technologies, which enable fruitful use of permitted wireless communication features and frequency range.

      BSNs are also called wireless body area networks (WBANs) as often the transmission is through wireless systems. In their current form, BSNs are wireless networks of wearable devices with recording and some processing capabilities [4, 7–9]. Such devices may be embedded inside the body, implants, surface-mounted on the body in fixed positions, or carried in one way or another [10]. From its start of development, there have been tremendous attempts in reducing the size and cost, and increasing the flexibility, of such devices–particularly those with direct contact with the human body [11, 12]. The development of BSN technology started in 1995 around the idea of using WPAN technologies to implement communications on, near, and around the human body. Later in early 2000, the term ‘BAN’ came to refer to the systems where communication is entirely within, on, and in the immediate proximity of a human body [13, 14]. A WBAN further expands WPAN wireless technologies as gateways to reach longer ranges. Through gateway devices, it is possible to connect wearable devices on the human body to the Internet. This allows medical professionals to access patient data online using the Internet independent of patient location [15].

      BSNs may also be considered a subset of WSN often used in various industrial applications to monitor a large connected system. In many cases, however, each group of sensors, such as those for an EEG, can be wired up to a central recording system, such as the EEG machine, which can then be processed together. For BSNs the sensors often sample the physiological and metabolic variables from human body. Using BSNs for health monitoring, the necessary warning or alarming states for risk prevention can be generated and the diagnostic data for long-term inspection by clinicians can be recorded and archived.

      The main components of the BSN technology are sensors, data processing, data fusion, machine learning, and low- and long-rage communication systems. Groups of researchers in sensor design, microelectronics, integrated circuit fabrication, data processing, machine learning, short- and long- range communications, security, data science, and computer networking, as well as clinicians, have to work together to design an efficient and usable BSN.

      The advances in sensor technology, data analytics for large datasets, distributed systems, new generation of communication systems, mobile technology, and cooperative networks have opened a vast research platform in BSN as an emerging technology and an essential tool for the future development of ubiquitous healthcare monitoring systems [28]. Researchers should (i) enable seamless data transfer through standards such as Bluetooth, ZigBee, or ultrawideband (UWB) Wi-Fi to promote information exchange and the efficiency of migration across networks and uninterrupted connectivity, (ii) the sensors used in an BSN should be of low complexity, small size lightweight, easy to use, reconfigurable, and compatible with the existing tools and software, (iii) the transmission should be secure and reliable, and (iv) the sensors should be convenient to use and ethically approved.

      Looking at the BSN with respect to WSN, WSNs have more general applications. For example, they can be deployed to inaccessible environments, such as forests, sea vessels, swamps, or mountains. In such cases, many redundant or spare nodes may be placed in the environment, making more dense distribution of the sensors to avoid any negative impact of node failures. In BSNs, however, the nodes are located in clinically more informative zones around or even inside the human body. This makes the total number of nodes limited, and generally rarely more than a few dozen. Each node is mounted properly to ensure more robust and accurate results [29]. However, there are cases where the sensors are movable and deployed for short duration recordings. An example of such sensors is endoscopic capsules, also called esophagogastroduodenoscopy (EGD), for monitoring human intestine and internal abdomen tissues.

      Also, in terms of functionality attributes, the nodes in WSNs often record data of the same modality (although, in recent applications, different modalities such as sound and video have been taken into account by WSNs), whereas, in BSNs, various sensors collect different physiological and biological data.

      Some limitations in sensor design – such as their geometrical dimensions, weight, shape, appearance, and size – may be less important for the WSN nodes than those of BSNs. Different sensor types are used in a BSN for recording various data types from the human body [8]. For a WSN there may be large-size sensors which are very resistive to a rough and hostile environment. In BSNs the nodes are supported by more robust electronic circuits which are less sensitive to noise, such as well-tuned differential amplifiers, to enable the recording of very low amplitude signals such as scalp EEG or surface electromyography (EMG). The sensors are often small and delicate enough to be wearable, less intrusive, easily deployable within the human body, and in many cases biocompatible [30].

      There are other considerations and limitations for BSNs, for example in many applications the human body is in motion and the BSN nodes move accordingly. Also, unlike for WSNs, where