A Complete Interoperable eHealth System Based on ZigBee Pro Standard

The eLife research group at University of Deusto is trying to socialize new technologies through projects focused on health. Telemedicine and eHealth systems are being used widely to improve health services or to develop value added ones. The main aim of this project is to design an eHealth system that increases life quality. It is focused on people who need daily support because of age, disability or disease. A wireless sensor network has been developed and it is integrated in a Set-Top-Box that allows controlling the network from TV. Besides, some devices have been designed so the article describes them as well as the software development.


Introduction
Biomedical Engineering and Information and Communication Technologies are growing together at high speed.The European Commission has been promoting them for two decades through its Framework Programmes (FP), which are focused on multiple topics.One of most important areas where both participate is telemedicine, better known nowadays as telehealth.In the FP6 (2002FP6 ( -2006) ) health research projects were priorities in many cases (Iakovidis, Le Dour, & Karp, 2007).More specifically, personal eHealth systems and rehabilitation technologies can be found among the Information Society Technologies priorities, neurosciences and medical imaging in Life Sciences, Genomics, and Biotechnology for Health Programme, and biosensors as part of the Nanotechnologies Programme.The FP7 is organized into five different blocks encompassing both collaborative and individual group research (in the latter case, without need for internationalization).It also supports researcher mobility and career development.The Collaboration block, comprising two thirds of the total budget, is also divided into different topics.One of them is called Health and there are also topics related to nanotechnologies and ICTs, as in FP6.
Telemedicine (Krol, 1997) defined as the use of computer and communication technologies to provide healthcare services between distant locations, started by the sharing of information between medical centers.The first applications were focused on audio and video links between these locations in order to resolve problems all together if necessary, to establish remote consultations and to access common databases.Nowadays, with the rapid growth of technologies, a wide range of health services have been developed with the support of sizeable research programs.Common applications are based on image and signal processing, specialized medical systems or telecommunication standards, among others.There is also great support for wearable systems (Lymberis & Dittmar, 2007;Chiu, 2006) and applications in order to complement eHealth and monitoring systems (McAdams, 2011).
Nevertheless, these eHealth advances have not been introduced into the market nor adopted by it.There are a lot of technological, legal or financial barriers (Paul, Pearlson, & McDaniel, 2009;Gibbons, 2008) involved in this process, sometimes too difficult to resolve.In technological terms, we can encounter problems not only in the user needs (if technology can solve the particular problem) but also in the user capabilities (if end-user can manage new technologies and applications).It is important because the users' capabilities vary, and rural people are, on average, the least technologically proficient but the neediest of such services (Sudhahar, y otros, 2010).Technological barriers are cited as a significant cause of the low utilization rates of eHealth services, but a reduction of these barriers does not guarantee wider utilization since other barriers are as important as the technological ones.
There is also a disagrement over costs (Wells, 2002;Poon & Zhang, 2008): whether eHealth systems cause a cost reduction or not (Nitzkin, 1996), and who such reduction would be for.We should realize that the population is aging (Mauksch, 1987), thus increasing the utilization rates of health services (Moullec & Ray, 2009).The elderly is a collective that generates higher costs than the rest of the population.This idea, together with the ever increasing population, shows that health care costs are also increasing.The development of new prevention-oriented services can be translated into medium to long term investments (Iluyemi & Briggs, 2009;Hui, Yanting, & Xue, 2010).However, the inclusion of technology in the health sector implies a need for training (for both specialists and patients) due to possible deficiencies in technological knowledge.This failing results in the development of adaptive user interfaces (Vasilyeva, Pechenizkiy, & Puuronen, 2005).Health services have to ensure not only accurate and up-to-date, but also easy to understand, information.These interfaces should take into account both content and navigation adaption.The contents should be adapted to the patients' situation, illness and treatment.Users could choose how to see the content, but it is the system that offers possibilities based on the specific patient.
Another important feature of eHealth services is the information shared through electronic medical records (Detmer, Bloomrosen, Raymond, & Tang, 2008;De, 2005).As you can see, many factors contribute to the economic effects of eHealth systems, and thousands of hours could be spent discussing them.All these changes should be well organized with the aim of avoiding waste and inefficiency, so a better management must be achieved (Sheng, Hu, Wei, & Ma, 1999).A well organized service reduces the number of treatment days (Akematsu & Tsuji, 2009).If there are affordable systems for patients, the diagnostics and some tests could be performed at a patient's home without the need for new appointments using telematic services.
Apart from multimedia (Cabral & Kim, 1996) and telematic services, wireless sensor networks (Yan, Huo, Xu, & Gidlund, 2010) have a place in the end users' home.These networks provide the link between the user and the service itself.If mass adoption of eHealth systems is desired, not only is user adaption to the application needed, but also the standardization of systems with the aim of different services and devices operating together, in other words, promoting interoperability.Bear in mind that often standardization, and therefore interoperability, is contrary to the interests of manufacturers, but it encourages competitiveness.And this competitiveness means that prices fall.Also important is the collaboration of government agencies such as the European Telecommunications Standards Institute (ETSI) (Hine, Petersen, Pluke, & Sund, 2008), as well as partnerships and alliances between companies (Piniewski, Muskens, Estevez, Carroll, & Cnossen, 2010).On one hand, the ETSI proposes the standardization of both application framework and architecture.On the other hand, Continua Alliance has designed some guidelines for the standardization of health devices.There are also social features which are truly important, such as security.Health information is very sensitive and technology should be capable of maintaining its privacy.So, both technical and social factors must be analyzed for a strong standardization (Chadwick, 2007).
Although there are a lot of factors involved in the development of telehealth services, the main aim of this project is to try to promote standard-based systems in order to achieve interoperable applications for eHealth.This aim could increase the number of interoperable devices and thus competitiveness.Therefore, the prices of end users' systems will be reduced.In accordance with this aim, several specific goals can be highlighted:  Design of a standard-based system to permit interoperability between devices and to reduce the final costs.


Use of a single technology for healthcare devices and home automation.


Design a gateway device in order to control the system from a PC user application.

Methods
It is not new that each wireless and wired technology has its own advantages and disadvantages.However, mobility is one of most important features in eHealth systems or other kinds of patient monitoring systems and it can only be provided by wireless technologies.Therefore, wireless technologies have gained much ground in the eHealth market.Another important feature is low power consumption, so the technology selected should be wireless and with reduced power consumption (Chien, Xu, & Molloy, 2010).Bluetooth (in its low-energy version), ZigBee and ANT are technologies with these features.Each one has its advantages and disadvantages, in terms of number of nodes, topologies supported, etc.But it is difficult to point out one of them as the best or the favorite.There are a lot of reviews comparing them (Gomez & Paradells, 2010), but it is quite clear that each one has its own market.
Continua Alliance, mentioned above, has selected Bluetooth-LE and ZigBee for its second version of guidelines (Piniewski, Muskens, Estevez, Carroll, & Cnossen, 2010).Bluetooth stands as the technology for mobile devices (e.g., a watch with an ECG monitor connected to the mobile phone), while ZigBee focuses toward low-power sensors (bed pressure sensors, fall detectors, etc).
The problem facing this choice is the need for platforms that support different technologies, thus adding complexity to the system.It is not a complexity in development, but for interoperability.It is difficult to achieve interoperability using a single technology, so adding different technologies makes it more difficult.In addition, it involves the development of gateways (Hwang, In, Park, & Eom, 2003;Song, Zhou, Zhang, & Song, 2008).These gateways are common elements used to interconnect different systems or technologies in order to provide a more complete service.Thus, indoor sensor networks can be connected to telematic systems to send or record information (Farshchi, Pesterev, Nuyujukian, Mody, Judy, & Bi-Fi, 2007), and also to control or manage the network from remote sites.However, the inclusion of different technologies in the indoor system involves the development of gateways within the sensor network itself.For this reason, the project develops a platform based on a single technology.
The technology selected for this project was ZigBee, which provides the best performance for a single technology based system and it is based on IEEE802.15.4 standard (Bandyopadhyay, 2010).ZigBee adds network and routing services to IEEE802.15.4.ZigBee was created by the ZigBee Alliance, a non-profit association whose main aim is to develop standards suiting the real evolving needs of manufacturers and developers.The Alliance designed some application profiles with industrial experts to meet the market needs.Specifically, the Home Automation Profile offers a global standard for smart homes, enabling the control of lights, environment or security among others.There is also a Health Care Profile, which enables reliable monitoring of non-critical healthcare services, targeted as independent life, health or wellness.Nowadays the focus is the end user, that is the customer, and they are demanding standard systems that can operate together, in order not to fill the house with numerous independent systems.

ZigBee Basics and Wireless Devices
ZigBee is a new technology that provides networking features over IEEE802.15.4 wireless standard.IEEE805.15.4 is a standard defined by the IEEE for low-rate wireless PANs.It defines the link and physical layer (low-power spread spectrum at 2.4GHz with a basic rate of 250kbps) The ZigBee Alliance developed the rest of the layers with the collaboration of companies working in electronics.The Alliance has developed several releases of the protocol.The latest one is ZigBee 2007, more commonly known as ZigBee Pro.
Figure 1.ZigBee stack As it can be seen in Figure 1, there are some layers defined by the Alliance to support all features needed in a WPAN.The last change was the inclusion of a new part in the Application Framework (AF) layer: the application objects and the ZigBee Cluster Library.The AF layer provides some features for specific applications and devices, thus it can achieve interoperability and ubiquity of devices from multiple manufacturers, so that devices with same functionalities should behave similarly.Although this is a huge step into standardization and interoperability, the addition of this new feature brings about greater complexity in application development.
Good-quality research work has been done into this, because it is a really arduous task to understand how the

Gateway Development
Although the network can operate completely independently, the need to control it through an embedded system or a personal computer arises.Thereby, user applications based on any platform could be designed to manage the network.The proposed design is a gateway controller between the network coordinator and the computer.The communication channel is through a UART port of the microcontroller.The computer, meanwhile, establishes a virtual serial port, established through a USB port.
The protocol has been designed with AT-like commands.There are commands from general to the most basic kind.There are some commands used to discover information on the devices: how many endpoints they have their input and output clusters, etc.There are also specific commands according to the clusters developed.There are specific commands to add a device to a group, to send on/off commands to the lights, to read a specific attribute of a device, etc.The command shown in Table 1 is used to obtain information on a device.The command requests the simple descriptor of a specific endpoint of the device.The descriptor contains plenty of information, such as clusters (both input and output), the profile associated to the endpoint, the device indetifier, and so on.It is a general command, not specific of a profile.The command shown in Table 2, however, is a specific cluster command.It corresponds to the On/Off cluster.The command is used to send a unicast frame to switch on, switch off or toggle a light of a device.
During the progress of this project, several devices have been designed in a parallel way.One example of them is the dimmable light device.This device is included in the Home Automation profile (according to ZigBee device profiles).The dimmable light device permits the user to modify the level of brightness in an incandescent light bulb.There are many ways of controlling this level, but we have chosen a digital control circuit.
The level of brightness is controlled by four general purpose output pins of the ZigBee module.The circuit also has a 555 timer and a 4-bit counter.The 555 timer works as the counter clock.The output pins of the microcontroller control the initial value of the counter.Thus, when a zero crossing of the AC signal is detected, it resets the counter.The counter starts from the initial value (provided by the microcontroller pins) to zero, allowing the current to pass through the bulb.Once the counter has reached zero, the borrow signal of the counter stops the current flow into the bulb.The process, allowing the power signal to pass through the bulb at a time interval during the half-cycle, controls the level of brightness.This current flow is controlled by an optotriac.
As the bulb has to be powered by the AC signal, this signal is also used for the rest of the circuit.We need to convert this AC signal into a DC signal, and lower its voltage.Therefore, in the complete schematic shown in Figure 6, a supply signal adaptation stage appears.This consists of a 9 V transformer and a rectifier bridge.The transistors in the center of the figure are used for the zero crossing detector system.The top of the image is another adaptation stage.We need to reduce the voltage from 9 V to the 3.3 V needed by the microcontroller.

Autonomo
The following scenario.Also  Scenario In this first sce The switch im and the light h cluster, and th

Satisfaction Test
We have also carried out several questionnaires with different people in terms of age and technological knowledge in order to analyze whether we are working in the right way.Although the questionnaire was simple, it was sufficient to test the system.The responses of the questionnaire had to be between the values of 1 and 5 (1 for "disagree", 5 for "strongly agree").
Among all the questions asked, the three most important are highlighted below.The users were asked if: • The software application was easy to use.

•
Creating the WSN was simple.

•
The system is useful.We interviewed some 25 people and the results are presented in Table 6.We have come to the conclusion that people with higher technological knowledge are more critical than others with less expertise.Nevertheless, people believe that the project is extremely useful.According to the general mean value, the system is being developed in the right way but a great effort should be made to make it easier to use.

Discussion
As shown throughout the article, there is a growing need for health services.Life expectancy is increasing as well as the number of illnesses, which is why telehealth systems are highly demanded by users, professionals and governments.Our goal as developers is to design systems that help people without involving major changes in their lifestyle or without a long training being needed to use it.To avoid these two aspects, adaptive systems are needed.This means having systems that are self-configuring in order to require as little user intervention as possible, thus enhancing its ease of use, one of the goals of this project.
Apart from this feature, another important goal was to develop a system based on a standard, thereby promoting interoperability between devices by different manufacturers.Firstly, a preliminary state-of-the-art analysis of wireless technologies was carried out until ZigBee was selected.ZigBee is a technology that is taking off slowly, step by step.Its entry into the market is hard because it is open standard.There are other protocols that are owned by manufacturers of consumer electronics causing their prompt take-off in the market.Even so, ZigBee has the skills and features necessary for a wide deployment of sensors in a network regardless of the application type (home automation, security, healthcare, safety…).
As presented in the results section, specific and general application features have been developed.These features are: binding, in order to decentralize network management and allow independent operation of devices, and the ability to create groups of devices to manage them as a group rather than individually.
Despite enabling independent operation, the system can be controlled by a user application.The idea is to have an intuitive and easy-to-use application that manages the network.For this reason, a gateway has been also developed allowing information exchange between the network coordinator and any embedded device.A user interface can be designed to run on any platform or operative system and be able to connect with the network through a serial communication port.This provides many possibilities as the user application can use different technologies.In this case, the embedded device could act as a bridge between the sensor network based on ZigBee and the other networks.The gateway developed is oriented to the described scenarios but can easily be expanded or customized to fit in any design.
There are similar types of systems on the market like the one presented in this article.One example is the AlertMe system, which is a complete product for home automation and control.It has sensors that detect if a window or a door has been opened, devices to turn on electric appliances, motion sensors, and so on.It is a scalable system throughout its range of products, but it offers no interoperability with other devices.It is a closed product, based on the first ZigBee release.So if you need a particular device and it is not developed by the manufacturer, you cannot integrate a device from another manufacturer in the AlertMe system.Another company with similar ZigBee compliant products is Simplehomenet, a company experienced in different technologies such as X10, Insteon or ZigBee.It markets gateways of these technologies, but they have some products that still have not been adapted to ZigBee.However, products that are already adapted meet the latest ZigBee release.
There are not a great number of products on the market.There are no complete or customized systems that can interoperate with other devices, or systems that anyone can install at home regardless of their technological knowledge.New technologies are increasingly appearing, not only more, but better ones.This prevents the development of systems with previous technologies and they become unfinished and unsupported systems.We suggest that more technologies are not needed and we have to exploit existing ones.In addition, this fast leaping from one technology to another does not provide sufficient time for companies to recover their investment, causing a lack of interoperable systems that requires more effort.And after all, this is what is needed to create value-added services.
The expansion of profiles used in the devices is proposed as a future line of work.Other future work could be the design of new devices oriented to healthcare, or the adaptation of designed ones to the ZigBee Pro standard.And finally, as regards software development, we propose the design of an intuitive and easy-to-use user application.The application could be adaptable, depending on the collective to which it is oriented.

Conclusions
As a final reflection, below is a review of the aims proposed at the beginning of this article that have been fulfilled.
As regards the implementation of ZigBee in application devices, several scenarios have been presented as a result of the developments carried out.These scenarios present, in a clear and simple way, everything developed in the devices to attain their use in real environments.These developments are the base or the mainstay for future development of specific applications.It is the beginning of a long road towards a complete system that provides value-added services based on a unique technology.
Likewise, the development of a gateway making it possible to connect the system to different systems has been successfully implemented.The gateway can be extended depending on the level of specificity needed, thus enabling easier use in more specific future applications.
The scenarios presented above are based on some tests performed with several devices of both home automation and health profiles.Since it is a work in progress, we have started studying ZigBee standard by developing general devices (such as switches and dimmers).It is a simple but effective way of studying the standard in order to establish knowledge.
Health sensors and devices are being designed in a parallel way to be integrated in the system as soon as possible.So, the work presented is the basis of the complete system and the progress of a long future work including integration of new health devices as well as an user application to control the whole network.
The main aim is to produce a complete system for future deployment in patients' homes.
Figure 3. Zig Figure Figure with ad frames Figure12ributes, becaus and in this cas light has been report, when t

Figure
Figure 16.Co

Table 1 .
Get simple descriptor command

Table 6 .
Test results