Design and Implementation of a Telemedicine System

Technology and Health Care 13 (2005) 199–219 IOS Press

199

Design and implementation of a telemedicine system using Bluetooth protocol and GSM/GPRS network, for real time remote patient monitoring
Yousef Jasemian∗ and Lars Arendt Nielsen
Center for Sensory-Motor Interaction (SMI), Department of Health Science & Technology, Aalborg University, Fredrik Bajers Vej 7, Bld. D-3, DK-9220, Aalborg E., Denmark
Received 1 September 2004 Accepted 12 February 2005 Abstract. This paper introduces the design and implementation of a generic wireless and Real-time Multi-purpose Health Care Telemedicine system applying Bluetooth protocol, Global System for Mobile Communications (GSM) and General Packet Radio Service (GPRS). The paper explores the factors that should be considered when evaluating different technologies for application in telemedicine system. The design and implementation of an embedded wireless communication platform utilising Bluetooth protocol is described, and the implementation problems and limitations are investigated. The system is tested and its telecommunication general aspects are verified. The results showed that the system has (97.9 ± 1.3)% Up-time, 2.5 × 10−5 Bit Error Rate, 1% Dropped Call Rate, 97.4% Call Success Rate, 5 second transmission delay in average, (3.42 ± 0.11) kbps throughput, and the system may have application in electrocardiography. Keywords: Telemedicine, Bluetooth, GSM, GPRS, data security, performance, reliability, safety, healthcare

1. Introduction In a modern healthcare service, where the health authorities optimise the resources most effectively, it is in many cases an option to treat/monitor as many patients as possible at their home. Telemedicine implies telecommunication technology, information technology and biomedical engineering, and it is showing its value in a rapidly increasing number of clinical situations [1,6,20,36]. There have been many studies around the world that have proved the feasibility and usefulness of telemedicine for monitoring e.g. electrocardiography (ECG) [2,17,26]. Fixed communication network has been used in different telemedicine setup for some years and it has shown its values, whereas wireless and cellular technologies within telemedicine have been developed only in the last few years [1,6,7,14,20,25,28,34,36]. There is a need for development of a real-time mobile telemedicine system for long term continuously monitoring patients in a distance.
∗ Address for correspondence: Yousef Jasemian, M. Sc.EE., Soendergade 58, 1.sal, 9000 Aalborg, Denmark. Tel.: +45 23 45 18 29; Fax: +45 98 15 40 08; E-mail: yj@hst.aau.dk.

0928-7329/05/$17.00 © 2005 – IOS Press and the authors. All rights reserved

200

Y. Jasemian and L.A. Nielsen / Design and implementation of a telemedicine system using Bluetooth protocol

One of the sole decisive factors that will make a telemedicine system successful, in urban and rural areas, is the application of modern communication technology for information exchange between a homebound patient and the medical specialists providing care. By integrating telemedicine in home healthcare policy a radical improvement in disease management, diagnoses and treatment can be achieved and a lot of resources and money can be saved. Moreover, patient’s comfort e.g. mobility will be obtained. In spite of communication’s link improvement, and despite all progress in advanced telecommunication technologies, the progress in the development of telemedicine system using advanced telecommunication technologies were not intact with that, and there are still very few functioning commercial wireless mobile monitoring device presented at the market, which are most off-line, and there are still a number of issues to deal with. The present study designed, developed and implemented a real-time wireless telemedicine system using modern communication technologies. The aims of this study are as follows: 1. To design and implement a generic wireless and real-time Multi-purpose Health Care Telemedicine system. 2. To design and implement a generic real-time wireless communication platform for the system realisation utilising Bluetooth protocol, Global System for Mobile Communication network, General Packet Radio Service (GSM/GPRS) and High Speed Circuit Switched Data (HSCSD). 3. To evaluate and validate the designed system by investigating its Telecommunication General Aspects. The concept was formed and defined in May 1999, and the study started in January 2000 and finished with evaluation results in July 2004. 2. Requirements and related works The intended system must fulfil the following requirements. 1. The system must be secure and safe (Specific Absorption Rate (SAR) < 2.0 W/Kg in Europe, according to the international commissions’ recommendations [11–13,35] and SAR < 1.6 W/Kg (USA) [World Wide Web at: http://www.conformity.com/0207specific.pdf, July 2002], high data security and privacy) 2. The system must be user friendly [24] a. The operation must be fast and simple [Design requirement]. b. The start-up must be fast and easy (< 25 s the time required for link establishment) [15] 3. The system must have near – 100% uptime (Up-time > 99.95% [SONOFON, the Danish network provider], and must have = 3.4 kbps minimum transfer rate [Design requirement]) 4. The system must have a high performance (Delay < 7.5 s application-level Round Trip Time (RTT), Dropped Call Rate (DCR) < 1.8%, Call Success Rate (CSR) > 95%, Packet Error Rate (PER) < 10−4 , Packet Lost Rate (PLR) < 10 −4) [18,24]. 5. Medical data recordings must be transferred without information loss or artefact gain (Bit Error Rate (BER) < 10−4 at application level [18], Signal to Noise Ratio > 22 dB. [Design requirement]) Telemedicine systems can transfer medical information via different transmission ways. This leads to variety of the possibilities as regards the applied technologies and data content. Design, development,

Y. Jasemian and L.A. Nielsen / Design and implementation of a telemedicine system using Bluetooth protocol

201

system simulation and several trials have been carried out by different research group, and following are some examples. In 1998, R.H. Stepanian et al. designed a mobile telemedicine system using Interim Standard 54 (IS-54) and GSM cellular telephone standard. They studied the feasibility of transmitting and receiving Photoplethesmography (PPG) data using a simulated GSM and IS-54 cellular channels, the results showed successful PPG transmission over both simulated standards. When a GSM standard was simulated, the signal-to-noise ratio was 12 dB, signal-to-interference ratio was 100 dB, and BER was 1.5 × 10 −3 bps for mobile station moving with 50 km/h speed, but data transmission delay was not announced in the paper [14]. In 1999, S.B. Freedman addressed two important issues in a telemedicine system, namely, timeliness and ease of ECG transmission, using a GSM cellular phone and GSM standard. To study the mentioned issues an ECG transmission system to a mobile phone was developed. The system was an off-line system (save and forward) using the mobile phone’s build in fax software and modem. The results showed that the system could solve those two above mentioned issues. No details about technical issues were addressed [7]. In 1999, K. Shimizu proposed a concept providing a mobile telemedicine system for emergency care in moving vehicle. An experimental system for transmission of color images, an audio signal, threechannel ECGs and blood pressure was developed. The system used a satellite link, and both fixed and cellular communication network for a real-time medical data transmission. Some problems with the used technique were identified, but the theoretical analysis verified the feasibility of the proposed technique. When a cellular network was used, the data transmission delay was 5 seconds in average using a mobile phone. They also experienced 10 −5 BER when the system was used in a usual situation. The system experienced Electromagnetic Interference (EMI). When the system was used in a moving vehicle delays up to several tens of seconds and interruptions in ECGs were observed. Moving in a vehicle when a new connection link was good, connection establishment took 10–30 seconds, but when connection li nk was bad it took several tens of second or even several minutes for reestablishing a connection. The most appropriate data transmission rate was found to be 4.8 kbps, transferring 3 ECG channel, a voice channel, and color image [28]. In 2001, B. Woodward et al. presented the design of a prototype integrated mobile telemedicine system. The system utilised a standard infrared (IrDA), GSM and Public Switched Telephone Network (ISDN). The system was in its first phase of development and the research was carried out producing a working system that allowed the transmission of just one parameter of data (ECG). The system was simulated but not fully developed. For a 900 MHz GSM based model, a BER of 1 × 10 −5 for a mobile station moving with 100 km/h speed was observed [34]. In 2003, E. Kyriacou et al. designed and tested a Multi-purpose HealthCare Telemedicine System with satellite or Plain Old Telephony System (POTS) and GSM mobile communication link support, which combined real-time and store and forward facilities. It seems that the final system afterwards was installed and used in Greece and Cyprus, and the results from the system application were very promising using the GSM system. A 4.4 kbps transmission rate for two ECG channel was found to be appropriate. Testing the performance of the TCP/IP over the GSM network, they found that 431 bytes could be a proper buffer size [16]. In 2004, X. Zhao et al. designed and developed a Telemedicine System for Wireless Home Healthcare Based on Bluetooth and the Internet. The system uses Bluetooth as wireless technology connecting the client side to the server side and vice versa via internet using dial-up modem, Digital Subscriber Line (DSL), or cable modem. The system was tested and showed promising performance for the wireless

202

Y. Jasemian and L.A. Nielsen / Design and implementation of a telemedicine system using Bluetooth protocol

Bluetooth connection. They found 2.2 seconds transmission delay over a Bluetooth link and 3.4 seconds delay trough an internet link during different sessions of ECG transmission with 6.4 kbps data collection rate, but they did not mention the found proper transmission rate for their system. They could not find any data loss during Bluetooth and Internet communication. They found that the major source of data error was arisen from the serial port communication, when working in high noise environments [36]. Even though none of the above mentioned studies have investigated the technology combination (Bluetooth protocol, GSM system, GPRS and HSCSD services, and TCP/IP protocol) introduced in the present study, the applicability, bandwidth, throughput, performance, reliability of a telemedicine system, and the quality of the transmitted vital data, were common issues investigated in all these studies, which are also main issues investigated and evaluated in the present study. 3. System design and technical implementation 3.1. Materials To test and verify the designed telemedicine system including the Bluetooth communication platform, electrocardiography (ECG) was chosen as the monitored medical parameter. A system composed of a patient unit, public mobile communication network, a Protocol Interpreter, and a Central Monitoring Station was designed and implemented. 3.1.1. Patient unit The patient unit composed of an ECG telemetry device, an Ericsson Bluetooth module and a Bluetooth enable Ericsson GSM mobile phone. 3.1.1.1. The telemetry device The ECG telemetry device is an in-market T3300 commercial product from Danica Biomedical A/S (Rødovre, Denmark). The T3300 can measure ECG, Pulse Oximetry (Spo2) and Non-invasive Blood Pressure (NIBP). T3300 operates with 2 × 1.5 V alkaline batteries, and has 72 hours operating time. The T3300 was modified by the software designed and implemented by first author in cooperation with Ericsson A/S Denmark, and the hardware was modified by Danica Biomedical A/S, Denmark. 3.1.1.2. The Bluetooth module and the designed and implemented bluetooth components The Bluetooth module is a type-approved class 2 module (ROK 101 008), which is compliant with Ericsson Bluetooth version 1.0 B. The module consists of three major parts; a base band controller, 1 Mb flash memory, and a radio that operates in the globally available 2.4 GHz free Industrial Scientific Medicine, ISM band. The module has 72 Kbytes RAM and 1 Mbps gross data rate over air interface (Bluetooth data specification). On the top of the Bluetooth stack a Dial-Up Networking Profile, a Point to Point Protocol (PPP) and Transport Control Protocol/Internet Protocol (TCP/IP) client, a Data Handler, a Bluetooth Profile Manager and a Universal Asynchronous Receiver Transmitter (UART) Handler were designed and implemented. 3.1.1.3. The mobile phone The mobile phone is an Ericsson T610, which can be utilised in 900 MHz, 1800 MHz, and 1900 MHz GSM system, and has 95 gram weight. Ericsson T610 supports both GSM Circuit Switched and Packet Switched (GPRS) data services. The T610 has approximately 6 hour’s continuous operating time [29].

Y. Jasemian and L.A. Nielsen / Design and implementation of a telemedicine system using Bluetooth protocol

203

Fig. 1. The block-diagram of the implemented software in the Bluetooth (ROK 101 008) module. On the top of the embedded Ericsson Bluetooth (BT) stack a Universal Asynchronous Receiver Transmitter (UART) Handler, Bluetooth Profile Manager, Data Handler, a Dial-Up (DUN/SPL) Networking Profile, a Point to Point Protocol (PPP) and Transport Control Protocol/Internet Protocol client (TCP/IP) are designed and implemented. The grey areas show the designed and implemented part in the present work, and the rest show the Ericsson Bluetooth standard protocol stack.

3.1.2. The communication network The system utilises a public mobile communication network (SONOFON, Denmark), consisting of GSM, and GPRS network. 3.1.3. The Protocol Interpreter The Protocol Interpreter is a Personal Computer utilising Windows 2000, where the data packets are interpreted and converted to a synchronised data stream, by a developed program, before sending it to the Central Monitoring Station. 3.1.4. The Central Monitoring Station The Central Monitoring Station is a Personal Computer utilising Windows NT platform, using a developed LabView program for monitoring ECG.

204

Y. Jasemian and L.A. Nielsen / Design and implementation of a telemedicine system using Bluetooth protocol

3.2. System design In this progress the technical possibilities, data security and practical aspects were taken into consideration. In order to select the most appropriate technology for the system realisation, properties of the existing wireless technologies such as Bluetooth, Digital Enhanced Cordless Telecommunications (DECT), Wireless Local Area Network (WLAN), Infra Red (IR), GSM, High Speed Circuit Switched Data (HSCSD), and General Packet Radio Service (GPRS) have been investigated. Finally, the wireless Bluetooth technology and the GSM/GPRS network services have been chosen in this setup. 3.2.1. GSM and High Speed Circuit Switched Data (HSCSD) GSM is the most popular digital system in Europe. GSM admits of integration for different telephony and communication services. In this way the system can be upgraded by applying modern technologies and services such as GPRS, Wireless Local Area Network and Bluetooth. GSM adapts the existing network and interplays with transit networks such as Public Switched Telephone Network (PSTN), Integrated Services Digital Network (ISDN), and Public Switched Packet Data Network (PSPDN) [27]. A normal GSM data call is circuit switched, and only one time slot is used for each call. The data speed is therefore limited to 9.6 kbps [18]. This makes the GSM a suitable communication system in the present study, as the minimum required transmission rate is 3.4 kbps. Utilising the transparent bearer services, which only use the functions of the physical layer, GSM apply Forward Error Correction (FEC) in order to improve transmission quality [27]. The High Speed Circuit Switched Data (HSCSD) is designed to enhance the data transmission capabilities in GSM system. It is circuit switched as the basic GSM and based on connection-oriented traffic channel of 9.6 kbps each channel. It enhances the data transmission capabilities by combining several channels to increase the bandwidth. A Mobile Station, in theory, could use all eight time slots within a Time Division Multiple Access (TDMA) frame to achieve an air interface user rate of 115.2 kbps. 3.2.2. GPRS For one GPRS radio channel the GSM can allocate 1–8 time slots within one Time Division Multiplexing frame and the time slots are allocated on demand and not fixed [27]. All time slots can be shared by the active users. This means that each slot is multiplexed to up to 8 active users. The GSM system limits the ability to use all eight time slots. Utilising GPRS data service the applied Ericsson T610 uses up to three time slots for receiving, and one time slot for transmitting. This means the speed for receiving data is up to 39,600 bps, and for sending data up to 13,200 bps. In GPRS, data are sent in packets. GPRS keeps an open connection, staying online to receive and send data at all times when it is needed. GPRS is an IP-based connection, which means that a high transmission capacity is only used when it is necessary. To improve data quality and achieve high reliability, the GPRS uses Automatic Repeat reQuest (ARQ) and Forward Error Correction (FEC) mechanisms for Point-to-Point (PTP) services [27]. For GPRS air-interface, a maximum Temporary Block Flow (TBF) blocking-rate of 1%, for packet data, and BER is a goal for Quality of Service (QoS). In GSM/GPRS the quality measurement (BER) is mapped into one of 8 RXQUAL (0–7) levels, and the RXQUAL0 (BER < 0.2%), which refer to best quality, is the goal [SONOFON, the Danish network provider]. The BER here is measured at the Network Management System level before channel decoding [18]. GPRS provides internet/intranet access, for a PC, PDA or any handheld device connected via Bluetooth wireless technology, infrared or cable.

Y. Jasemian and L.A. Nielsen / Design and implementation of a telemedicine system using Bluetooth protocol

205

Fig. 2. The T3300 internally communicates with the Bluetooth module via a UART interface. The Bluetooth module is instructed to establish a Dial-Up Networking connection towards the Central Monitoring Station using either GPRS or GSM.

3.2.3. Bluetooth The protocol components in Bluetooth (ROK 101 008) are realised by using the embedded Ericsson Bluetooth stack. On the top of the Bluetooth stack a Dial-Up Networking Profile, a Point to Point Protocol (PPP) and Transport Control Protocol/Internet Protocol client (TCP/IP), Data Handler, Bluetooth Profile Manager and a Universal Asynchronous Receiver Transmitter (UART) Handler are designed and implemented by the first author in cooperation with Ericsson Denmark. Figure 2 shows the blockdiagram of the implemented software in the Bluetooth module. PPP is a client/server-based packet transport system used on dial-up links, which is used to carry packets from the higher IP layer across Bluetooth’s RFCOMM serial port emulation layer using Serial Port Profile (SPP) and is responsible to convert the TCP/IP packet into data-stream and vice versa. PPP handles error detection allows negotiation of IP addressing at connection time, permits authentication, flow control and header and data compression [3,30]. The Bluetooth base band uses Forward Error Correction (FEC) code of 1/3 providing reliable in-sequence delivery of byte stream. This encoding replicates the data three times [3]. The grey area in Fig. 2 illustrates the developed part by the first author in cooperation with Ericsson Denmark. The Bluetooth module is connected to T3300 telemetry device by a Universal Asynchronous Receiver Transmitter interface (UART interface) in order to enable the T3300 to communicate with the mobile phone wirelessly. This is illustrated in Fig. 3. 3.2.3.1. UART Handler The UART Handler is composed of an interface protocol and a Medical Communication Profile (MedCom Profile). UART Handler controls the send and receipt of data on the UART interface. It is responsible for detecting message frames and dispatching to both the Bluetooth Profile Manager and the Data Handler. All messages sent towards the T3300 telemetry device go through the UART Handler which makes the required message framing. The Medical Communication Profile (MedCom Profile) configures the communication set-up for each specific medical application utilising a specific profile. 3.2.3.2. Bluetooth Profile Manager The Bluetooth Profile Manager controls the base band and link manager of the Bluetooth protocol, initiating inquiry, service discovery and handling security related functionality. The Bluetooth Profile Manager makes use of the Dial-Up Networking Profile (DUN/SPL) component for all DUN specific handling. Additionally, it also communicates with the PPP layer to trigger establishment and release of PPP sessions and forwarding data between the RFCOMM (Protocol for RS-232 serial cable emulation) of the Bluetooth stack and the PPP layer.

206

Y. Jasemian and L.A. Nielsen / Design and implementation of a telemedicine system using Bluetooth protocol

Fig. 3. The left side shows the patient has the telemetry device and the mobile phone on him, and the right side shows a block diagram of the implemented system. The T3300 implies a NEC 78P4038 microprocessor (for the logic control), and a Field Programmable Gate Array (FPGA) is used as I/O extender. The T3300 can measure ECG, Pulse Oximetry (Spo2) and Non-invasive Blood Pressure (NIBP), but only the ECG is measured in the present work.

3.2.3.3. TCP/IP client Internet Protocol (IP) client is responsible for addressing, routing, segmentation, and reassembling of IP-packets. The Transport Control Protocol (TCP) is a reliable connection-oriented protocol that allows a byte stream originating on one machine to be delivered without error on any other machine in the internet. The TCP client is responsible for handling of the data flow control. 3.2.3.4. Data Handler The Data Handler is responsible for the data handling and data delivery to the TCP/IP stack. 3.2.3.5. Dial-Up Networking Profile The Dial-Up Networking Profile is a generic profile that provides an easier interface towards the Bluetooth stack. When T3300 telemetry device sends information to the Bluetooth module the Bluetooth module uses Dial-Up Networking Profile to establish a Dial-Up connection towards the Protocol Interpreter and Central Monitoring Station using either GPRS or GSM depending on the capabilities of the used mobile terminal. 3.2.4. Patient safety The radio wave exposure guidelines use a unit of measurement known as the Specific Absorption Rate (SAR). To test the SAR value standardized methods are utilised, with the phone transmitting at his

Y. Jasemian and L.A. Nielsen / Design and implementation of a telemedicine system using Bluetooth protocol

207

highest certified power level in all used frequency bands [12,23]. The most commonplace handset is 2 W and in real terms this usually only emits around 1200–1400 mW at the aerial, when used a long distance from base station. A research result from Ericsson, regarding health and safety in mobile telephony, showed that Ericsson’s GSM terminals comply with International Commission of No-Ionising Radiation Protection [11,23], World Health Organization [35], and International Radiation Protection Association [5,13,22]. Both the Ericsson Bluetooth module and the Ericsson Bluetooth enabled GSM mobile phone have maximum 0.89 W/Kg SAR value. This is much less than the recommended SAR value by these mentioned International organization and commission, which is 2.0 W/Kg. Thus, it can be contemplated that the mobile phone is safe to use. The T3300 telemetry device is designed in accordance with the following standards: I-ETS 300 220, IEC 529, EN60601-1 and EN 60601-1-2 (Device specification from Danica Biomedical A/S), which specifies requirements intended to reduce risks of fire, electric shock or injury for the operator and layman who may come into contact with the equipment. Both the T3300 telemetry device and the phone are operating by battery power and no hazard appended to the patient, regarding patient’s safety. 3.2.5. Data security GSM and GPRS offer Access Control, Authentication, Data Encryption, and User Anonymity [27]. Bluetooth technology offers also a sophisticated mechanism for authentication and encryption [3].

4. Problems, limitations and improvement 4.1. Problems and limitations The interplay between the Data Handler and TCP/IP client in the Bluetooth module was gradually adjusted to achieve best flow control, by checking the available buffer capacity in the TCP/IP client before sending any data to that. Moreover, a TCP/IP performance test was done, by testing appropriateness of different data buffer test [16], in order to find the best packet size. The technical tests showed that Bluetooth module could buffer only 2500 bytes; it could also send 800 byte/s TCP/IP-package to the server continuously. With 7.5 kbps data rate there was 0.5% packet lost having the system online in 4 weeks, 24 hours a day. On account of the GPRS network had a number of dropouts, and the TCP/IP’s have non-deterministic activity, there has been succession system dropout. Changing the flow-control logic towards the T3300 telemetry device, and making a large TCP/IP packages did not enhanced the situation. The mobile phone could not transmit data in peak situations, consequently data-transmission were stopped. With more intelligent flow-control between the Data Handler, in the Bluetooth module, and the TCP/IPstack the result was run out of TCP/IP buffer and in that way the data transmission was deadlocked. Flow-control optimization appeared to be the way for enhancing the transmission capacity. The investigation showed that the GPRS-network, Bluetooth module and the phone terminal are the bottleneck, because the standard mobile phone can get 9.6 kbps uplink, and also the TCP/IP protocol in the Bluetooth module is not a “streaming” protocol, that is, timing requirement could not always fulfill the desired transmission rate. It has been shown that when the GPRS transmission’s condition is good (depending on the communication traffic on the network), there has been capacity surplus in the network, and the transmission was well performed. This condition makes the reliability of the designed system

208

Y. Jasemian and L.A. Nielsen / Design and implementation of a telemedicine system using Bluetooth protocol

traffic-time dependent. Moreover, the buffer capacity is limited in both the T3300 telemetry device and the Bluetooth module; this makes the data flow-control unreliable. Specifying a telemedicine dedicated GPRS transmission’s service or fixing an order of priority for the telemedicine services, in order to enhance the capacity of the transmission channel, could improve the system reliability. Additionally, improving the RAM capacity in both the Bluetooth module and T3300 telemetry device can enhance the buffering and flow-control mechanism, which again improve the reliability. Neither the network provider commercially offered the dedicated GPRS network capacity nor did manufacturer offer the Bluetooth module with more RAM capacity. The battery life-time, which is 6 hours for continuous operation for the Ericsson T610 mobile phone, sets also a limitation regarding system’s operation-time. Thus, to be able to test the hypotheses of the present study, design modification was a necessity. 4.2. Modification and improvement Considering the study time limitation, in order to override the mentioned limitations, reduce the complexity of the system, and investigate the applicability of that instead, a serial cable replaced the Bluetooth connection between the T3300 and the mobile phone, a traditional GSM data service was utilized instead of GPRS, and a tiny modification to flow control software was performed. The Protocol interpreter PC was replaced by a Modem Server with tiny modification to the software. The Modem Server is a Personal Computer using a Sony Ericsson GM29 machine to machine modem (asynchronous -non transparent up to 9.6 kbps, class B terminal, and GPRS class 8, up to 19.2 kbps) utilising Windows 2000 platform. The Modem Server receives the medical data over the air interface via GSM system and sends that in bit streams trough a serial cable to the Central Monitoring Station. The Central Monitoring Station was not changed. An extra battery pack was arranged and employed for the T610 mobile phone, in order to increase the T610’s operating time to up to 30 hours. These design modifications made the system more reliable, in order to illustrate the applicability of the system in remote monitoring patients from distance. 5. Performance analysis 5.1. System function As shown in Fig. 3, four disposable electrodes trough four electrode-leads carry the ECG signals from the patient to the T3300 ECG telemetry device. Analogue ECG signals are amplified, filtered, and sampled to an accuracy of 10 bits at 100 Hz before being assembled into 340 bits frames for transmission. The data frames are sent every 80 ms. Thus; the transmission rate is 4.25 kbps. T3300 performs error correction using Cyclic Redundancy Check. The T3300 telemetry device is connected to a T610 Sony Ericsson mobile phone through RS232 serial cable with 9.6 kbps rate. Data rate from the T3300 telemetry device to T610 is approximately 4.25 kbps; in this way there is almost less than 50% load on the serial cable connection. The Field Programmable Gate Array (FPGA) in the T3300 is used as I/O extender. The T3300 telemetry device collects the medical data and invokes the T610 mobile phone to establish a traditional GSM data service connection to the public mobile communication network automatically. When the connection is established, the GSM phone acts as a mobile modem to the T3300 telemetry device for the transmission of the data over the GSM system to the Modem Server.

Y. Jasemian and L.A. Nielsen / Design and implementation of a telemedicine system using Bluetooth protocol

209

The Modem Server receives the data, multiplex it onto 115,200 bps rate, and convert it to pre-defined format, which is readable for the Central Monitoring Station. The data then is sent to the Central Monitoring Station via serial cable. Central Monitoring Station interoperates and converts the received data to graphical ECG. 5.2. Methods To test the present system, electrocardiography was picked out as an application area. To simulate a patient a MTK phantom 320 ECG-simulator with 256 Hz sampling frequency, from Danica Biomedical A/S (Rødovre, Denmark), was used. Test and verification of the system were done by examining the general communication aspects of that. The patient unit was placed in different landscape (In urban, on the hilly area, in the forest and in to the country), and in this relation, reliability, performance, and data quality of the system were investigated. To monitor and examine these aspects, the system log-file and the statistical information from the Network Management System at the GSM radio level (SONOFON, the Danish Network provider) were utilised. From the collected information by the log-file following performance and reliability key-indicators were calculated. These key-indicators are standard indicators defined and recommended by European Telecommunication Standard and are frequently used key-indicators [4,18,32]. – – – – – Throughput (Transmission rate, [bits/sec].) Packet Error Rate (PER, [No. of erroneous packets/sec]) Packet Lost Rate (PLR, [No. of lost packets/sec]) Down-time interval [Minutes] Down-time frequency [Reconnection Rate]

5.2.1. Exp. 1, Interferences (DECT, GSM, Microwave oven) GSM mobile phones operate at 890–960 (GSM 900) and 1710–1880 (GSM 1800) MHz Ultra High Frequency (UHF) range and have approximately 2 w transmission power. Digital Enhanced Cordless Telephone (DECT) operates at 1880–1900 MHz UHF range and has a maximum transmission power of 250 mW. A microwave oven operates at 2.4 GHz frequency band, and a modern microwave oven normally consumes about 1000 Watts. The Food and Drug Administration’s (FDA) microwave oven standard is an emission standard (as opposed to an exposure standard) that allows leakage (measured at five centimetres from the oven surface) of 1 mW/cm2 at the time of manufacture and a maximum level of 5 mW/cm 2 during the lifetime of the oven [World Wide Web at: http://www.fda.gov/cdrh/consumer/microwave.html#4, Updated March 8, 2000] this is far below the level known to harm people, but 5 mW/cm 2 is enough power density for introducing interference on the GSM mobile phone’s antenna, the electrode cables, and the Bluetooth module’s antenna. Electromagnetic power emitted from the above mentioned devices could have an influence on the reliability and performance of the present telemedicine device, as the electrode cables and the serial port cable between the T3300 and the mobile phone act as an antenna too. Since the risk of a patient, with remote monitoring telemetry device, being very close to the mentioned devices, when he/she is monitored at home, is very high, the investigation of these interferences became evident. Interference between Wireless Local Area Network (WLAN) and Bluetooth, which both operate in the 2.4 GHz unlicensed ISM frequency band, has been studied before [9], and the achieved results indicate

210

Y. Jasemian and L.A. Nielsen / Design and implementation of a telemedicine system using Bluetooth protocol

that the performance of the Bluetooth and WLAN are mutually affected by each other, and the error caused by the interference are more often too much to correct. The ECG electrodes were connected to a MTK phantom 320 ECG simulator in order to simulate a heart patient. The patient unit (T3300 and the mobile phone) was exposed to interferences from another GSM phone, a Digital Enhanced Cordless Telephone (DECT), and the adjacency of a microwave oven at 15 cm test position (closest distance a patient could be to a microwave oven at home), in order to verify the patient-unit’s degree of security against these mentioned interferences. The experiment was repeated 5 times for following 4 scenario: Baseline (without interferences), and with GSM, DECT and a 1000 W microwave oven (Daewoo, model 4K-63D7, that has been estimated to have 5–8 mW leakage at 5 cm position from the surface) as interference sources. These experiments lasted 1 hour each. The variations in Up-time, Packet Error Rate, Packet Lost Rate and Throughput in this different scenario were studied. 5.2.2. Reliability (exp.2) and performance (exp.3) Based on the distance and the pathway of the radio wave propagation, the received signal will experience different type of fading, which represents the average signal power attenuation due to motion over large areas in relation to the base station. In this case the receiver is often mentioned as being shadowed by prominent ground, such as hills, forest, billboards, clumps of buildings, etc. To investigate the reliability and performance of the system in different landscape, the ECG electrodes were connected to a MTK phantom 320 ECG simulator. The system was set-up and put in to service for 5 weeks 24 hours a day continuously, and the data were transferred to the modem server via a live GSM system. The experiments were performed when the telemetry transmitter was in different landscapes, such as urban, in the forest, on the top of the hill, and in to the country. The frequency and the duration of the system down-time, system up-time, in addition to reconnection rate, were the key-indicators used for the reliability evaluation. Packet Error Rate, Packet Lost Rate, Transmission Delay and Throughput were the performance key-indicators used for the verification of the system performance. 5.2.3. Quality (exp.4) Data quality is evaluated using Bit Error Rate and Signal to Noise Ratio, as quality key-indicators, at the receiver side. To measure an experimental end-to-end Bit Error Rate a number of well known data streams were sent in a loop from a portable computer, using a GSM machine to machine modem, simulating the telemetry transmitter. The received data on the server side were compared with the sent data, at bit level. To measure the signal to noise ratio a developed Wavelet algorithm was employed. Thirty minutes ECG were transmitted by the above mentioned set-up and saved as reference data, then the same 30 minutes ECG were polluted by noise (motion noise + 50 Hz noise) and transmitted by the same set-up. The received data at the server side were de-noised, and the separated noise by the mentioned algorithm was compared with the mentioned reference data. This experiment was repeated 10 times. 5.2.4. Data analysis One-way analysis of variance (ANOVA) with repeated measures, utilising SPSS, was used to determine if there were significant differences between and within each test group. P < 0.05 was considered as significant difference. The results are presented as (mean ± SD). 5.3. Results 5.3.1. Exp.1, Interferences (DECT, GSM, Micro oven area test) Regarding the interference test, the experiments showed that the Up-time was (97.47 ± 0.93)%. The Up-time was poor, when the system was interfered by adjacency of microwave oven, testing at 15 cm

Y. Jasemian and L.A. Nielsen / Design and implementation of a telemedicine system using Bluetooth protocol

211

Fig. 4. Comparing the system up-time in percent, when the patient unit (T3300 and the mobile phone) was interfered by microwave oven, a Digital Enhanced Cordless Telephone (DECT), and GSM phone. The Up-time was poor, when the system was at the adjacency of microwave oven, and a significant difference was observed (P < 0.05). The Baseline is measured, when no interference was occurred.

Fig. 5. Comparing the system throughput (Kbps), when the patient unit (T3300 and the mobile phone) was interfered by microwave oven, a Digital Enhanced Cordless Telephone (DECT), and GSM phone. The Throughput was poor, when the system was at the adjacency of microwave oven, and a significant difference was observed (P < 0.05). The Baseline is measured, when no interference was occurred.

distance, comparing to the other situations and a significant difference was observed (P < 0.05). The results are depicted in Fig. 4. The Throughput was (3.38 ± 0.04) kbps, which is very close to the target (3.4 kbps). The throughput was less, when the system was exposed to interference by a microwave oven (3.20 ± 0.12) kbps, comparing to the other situations, and significant difference was observed (P < 0.05). Figure 5 illustrates this. The Packet Error Rate during runtime was (6.46 ± 6.64) × 10 −5 , and when the system was adjacent to a microwave oven, more erroneous data packets were delivered (2 ± 0) × 10 −4 compared to the other situation and a significant difference was observed (P < 0.05). Figure 6 presents that. The Packet Lost Rate was (3.59 ± 5.21) × 10 −5, but no significant difference was observed (P > 0.05) between these different situations. Table 1 summarizes the target, the achieved results and the critical comments for each reliability and performance key-indicator. 5.3.2. Reliability (Exp.2), comparison between a “good” and a “bad” radio connection The experiments showed that the system up-time was (97.90 ± 1.3)% and it was less when the patient unit was in the forest (−79.0 dB signal strength with 98% probability) in day 4 and 5 in week 4 and 5, comparing to other landscape but no significant difference was observed (P > 0.05), neither between weeks nor within week days.

212

Y. Jasemian and L.A. Nielsen / Design and implementation of a telemedicine system using Bluetooth protocol Table 1 Interference test, the telemetry device at 15 cm distance to GSM, DECT and microwave oven

Key-indicators Target Result (mean ± SD) Critical comments Up-time (% of the run-time) > 99.50% (97.47 ± 0.93)% Poor, when microwave oven interfere (P < 0.05) Throughput (Kbps) > 3.4 (3.38 ± 0.04) Kbps Poor, when microwave oven interfere (P < 0.05) Packet Error Rate < 10−4 (6.46 ± 6.64) × 10−5 Poor, when microwave oven interfere (P < 0.05) Packet Lost Rate < 10−4 (3.59 ± 5.21) × 10−5 Good Summarises the target, the achieved results and the critical comments for each reliability and performance key indicators, when the Patient Unit (The T3300 telemetry device and the mobile phone) were exposed to interference, caused by the electromagnetic emission from another GSM phone, a Digital Enhanced Cordless Telephone and in the vicinity of a microwave oven. Table 2 Reliability test Key-indicators Target Result (mean ± SD) Critical comments Up-time (%) > 99.50% (97.90 ± 1.3)% Not critical Down-time duration < 3 minutes (2.88 ± 2.44) minutes Poor, but not critical Reconnection rate < 5 times (8.32 ± 11.74) times Poor, but not critical Summaries the target, the achieved results and the critical comments for the reliability key indicator, when the system was set-up and put in to service for 5 weeks 24 hours a day continuously. In row three the Reconnection Rate refers to the number of times (in a week) the T3300 telemetry device or the phone were turned off to change the batteries or the connection was terminated because no data was received or the modem hung up. Table 3 The measured connection establishment time for Bluetooth, GSM, GPRS, and HSCSD Bluetooth GSM HSCSD GPRS 2.5–5 s (mean = 3 s) 5–25 s (mean = 10 s) 5–25 s (mean = 10 s) 3–8 s (mean = 5s) Summarises the measured connection establishment time for Bluetooth, GSM, GPRS, and HSCSD.

The duration of the down-time was (2.88 ± 2.44) minutes in average. The down-time duration was higher in day 2 and 3 in week 2, but no significant difference was observed (P > 0.05), neither between weeks nor within week days. The Reconnection Rate was (8.32 ± 11.74) in a week and it was higher when the patient unit was placed in the forest in day 4 and 5 in week 4 and 5, but no significant difference was observed (P > 0.05), neither between weeks nor within week days. Table 2 summarizes the target, the achieved results and the critical comments for the reliability key-indicators. In row three (Table 2) the Reconnection Rate refers to the number of times (in a week) the T3300 telemetry device or the phone were turned off to change the batteries or the connection was terminated because of no data was received or the modem hung up. The mobile phone’s battery-pack should be changed five times in a week, in order to recharge it. Table 3 summaries the measured connection establishment time for Bluetooth, GSM, GPRS, and HSCSD. 5.3.3. Performance (Exp.3), comparison between a “good” and a “bad” radio connection Regarding the performance test, the experiments showed that the throughput was (3.42 ± 0.11) kbps in average, which is very close to the target. There were fewer throughputs in week 5 day 4 and week 4 day 5 in average (3.23 ± 0.14) kbps, when the patient unit was placed in the forest, and significant differences were observed (P < 0.05) between weeks but not within week. Figure 7 demonstrates that. In average the Packet Error Rate was (8 × 10 −5 ± 8 × 10−5 ) during the run-time, which is close to the target. The Packet Error Rate was higher in week 4 day 2 (1 × 10 −4), and week 5 day 1 (1 × 10 −4),

Y. Jasemian and L.A. Nielsen / Design and implementation of a telemedicine system using Bluetooth protocol

213

Fig. 6. Comparing the system Packet Error Rate (PER), when the patient unit (T3300 and the mobile phone) was interfered by microwave oven, a Digital Enhanced Cordless Telephone (DECT), and GSM phone. The Packet Error Rate was higher, when the system was at the adjacency of microwave oven, and a significant difference was observed (P < 0.05).The Baseline is measured, when no interference occurred.

Fig. 7. Comparing the variation in averaged Throughput (Kbps), between 5 weeks, 7 days a week. There were fewer throughputs in week 5 day 4 and week 4 day 5 in average, when the patient unit was in the forest, and significant differences were observed (P < 0.05) between weeks but not within weeks.

when the patient unit was placed in the forest, but no significant difference was found (P > 0.05), neither between weeks nor between those mentioned landscapes. The Packet Lost Rate was (1 × 10 −5± 6.7 × 10−7 ). The Packet Lost Rate was higher in week 3 day 2 (7 × 10 −5), but no significant difference was observed (P > 0.05), neither between weeks nor between those mentioned landscape. The measured Transmission Delay was approximately 5 second in average. Table 4 summarizes the target, the achieved results and the critical comments for the Performance key-indicators. The information driven from GSM Network Management System (NMS) is more representative than those gathered by tests, as the statistics NMS represent the overall performance of the network [17]. In addition to measured performance keys the information from NMS showed the following results for 23 weeks test. Dropped Call Rate (DCR), which measures the percentage of connection lost, was reported to DCR = 1% in average, which is less than the target (< 1.8%). Call Success Rate (CSR), which measures the success rate of the signalling process in relation to origination, was reported to CSR = 97.4% in average, which is higher than the target (> 95%).

214

Y. Jasemian and L.A. Nielsen / Design and implementation of a telemedicine system using Bluetooth protocol Table 4 Performance test (In urban, on the hill, in the forest and in to the country) Key-indicators Target Result (mean ± SD) Critical comments Throughput (Kbps) > 3.4 Kbps (3.42 ± 0.11) Kbps Poor, in the forest (P < 0.05) Packet error rate < 10−4 (8 ± 8) × 10−5 Good Packet lost rate < 10−4 1 × 10−5 ± 6 × 10−7 Good Transmission delay < 7.5 second 5 second Good Dropped Call Rate (DCR) < 1.8% 1% Good Call Success Rate (CSR) > 95% 97.4% Good Summarises the target, the achieved results and the critical comments for the Performance key indicators, when the system was set-up and put in to service for 5 weeks 24 hours a day continuously. Table 5 Quality test Key-indicators Target Result (in average) Critical comments Bit error rate < 10−5 2.5 × 10−5 Good Signal to Noise Ratio (dB) > 22 dB 28 dB Good Summarises the target, the achieved results and the critical comments for the Quality key indicators, using a laboratory set-up.

5.3.4. Data quality (exp.4) Regarding the quality test, the experiments showed that the measured Bit Error Rate was 2.5 × 10 −5 , which is a bit higher than the target, but it is acceptable. The measured Signal to Noise Ratio was 28 dB, which is higher than the target. Table 5 summarizes the target, the achieved results and the critical comments for the Quality key-indicators, using the laboratory setup. 5.4. Discussion 5.4.1. The first design and implementation For the first time, we designed and implemented a real-time wireless telemedicine system for long term continuously patient monitoring purpose, utilising the Bluetooth for short-range and GPRS/GSM for long-range together. The first prototype showed that the system was capable of transferring reliably ECG data. Closer examination of the system revealed that when the GPRS transmission’s condition was good (depending on the network’s traffic), there was capacity surplus in the network, the transmission was well performed, and the reliability and performance were satisfactory. The experiments testified that Bluetooth module had a buffer limitation, but was capable of sending 800 bytes TCP/IP-package to the server continuously. Problems crop up when data packet delays vary because of network congestion, asymmetric network capacities, dropped packets, or asymmetric server/client processing capacities. The degree of network congestion is inferred by the calculation of changes in Round Trip Time (RTT): that is the amount of delay attributed the network. This is measured by computing how long it takes a packet to go from sender to receiver and back to the client. The present system uses different protocols (Bluetooth, GSM/GPRS, and TCP/IP). Interaction between different layers within these protocols includes tune and negotiation of a number of adjustable parameters (QoS. packet retransmission timeout, packet size, etc.). In Bluetooth, for instance, the Link Manager (LM), the Logical Link Control and the Adaptation protocol (L2CA) are involved in peer to peer negotiations to configure Qos for a link. The QoS in Bluetooth protocol is configured by parameters such as: service type, token rate, token bucket size, peak bandwidth, latency and delay variation [3]. There is interplay

Y. Jasemian and L.A. Nielsen / Design and implementation of a telemedicine system using Bluetooth protocol

215

between high throughput (TP) and high QoS. To increase the probability of errorless packet transmission, the TCP/IP packet size should be treated of in relation to the TCP/IP transmission buffer size and timing. To increase the QoS, the TCP/IP packet size was decreased, and by this the transmission traffic was reduced, as the transmission of a long packet generates more traffic. On the other hand, the error correction mechanism in GPRS and RFCOMM reduces the bandwidth, as we deal with a distributed data line. Thus, TCP/IP might be triggered to retransmit a packet, which is stuck in the buffer waiting for retransmission; this will further increase the traffic unnecessarily. Thus, it has been demonstrated that Bluetooth technology and GPRS data service, principally, are useful technologies for the present system. But, the limitation of the buffer capacity in both T3300 telemetry device and the Bluetooth module, become evident if one considers the transmission timing, network congestion and data flow control. Nevertheless, the commercial dedicated telemedicine channel within GPRS data service was also not offered by the network providers at the time of the study. 5.4.2. Reducing the complexity and testing the applicability In order to reduce the complexity of the system and investigate the applicability of that, the Bluetooth wireless connection between the T3300 telemetry device and the Bluetooth enabled mobile phone, was replaced by a serial cable, and the GPRS data service was replaced by a traditional GSM data service. These replacements had advantages and disadvantages. 5.4.3. Wireless technology, contra serial cable connection Using serial cable instead of Bluetooth wireless technology made the communication between the T3300 telemetry device and the mobile phone less sophisticated, saving up to 3 second connection time, performing bit stream data transmission. On the contrary, data security, the interferences protection and the mobility were reduced. The cable acts as an antenna collecting possible noise and electromagnetic interferences, and therefore more data error was observed when the patient unit was used in noisier environment. This is, of course, improved by a 50 cm long shielded serial cable. Using Bluetooth avoids these interferences. Moreover, the Bluetooth system has high safety arrangement. The possibility of recharging the mobile phone, during remote monitoring, when applying Bluetooth technology, was also cancelled. Applying Bluetooth one could recharge the phone without having any galvanic connection to the telemetry device, while this is not possible in the case of using serial cable. The last mentioned obstacle limits the operation time of the mobile phone, which is very important issue for a real-time and long-term remote monitoring system. 5.4.4. GPRS data packet switched, contra HSCSD and the traditional GSM data service The employment of a traditional GSM data service instead of GPRS, made the offered service expensive. For a traditional GSM data service and High Speed Circuit Switched Data (HSCSD), the user gets charged for the entire connection time, while for the GPRS the user pay just for the data to send. In the other hand, by using a traditional GSM data service, the performance and the reliability was significantly enhanced, as for the GPRS Packet Switched data service one GSM time slot is multiplexed to up to 8 GPRS active users. This will clearly reduces the GPRS channel capacity when there is more than one active GPRS user on the network, which is mostly the case in real life. The HSCSD is designed to enhance the data transmission capabilities in GSM system. It is circuit switched as the basic GSM and based on connection-oriented traffic channel of 9.6 kbps each. It enhances the data transmission capabilities by combining several channels to increase the bandwidth. A Mobile

216

Y. Jasemian and L.A. Nielsen / Design and implementation of a telemedicine system using Bluetooth protocol

Station, in theory, could use all eight time slots within a Time Division Multiple Access (TDMA) frame to achieve an air interface user rate of 115.2 kbps. Although it appears attractive and promising, the HSCSD exhibits some major disadvantages. Allocating several circuit switched channel is reflected in the service costs. Moreover, for n channels connection-oriented HSCSD requires n times signalling during handover, connection setup and release, and each channel is treated separately. Then the probability of blocking or service degradation increase during handover, as in this case a Base Station Controller (BSC) has to check resources for n channel, instead of one channel. As the implemented patient unit in the present study is intended to be fully mobile and the patients are moving freely in the town and neighbourhood, the number of handover is increasing and the probability of blocking or service degradation increase during handover, proportional to that. 5.4.5. Improvement suggestions From the preceding discussion it is clear that increasing the buffer capacity in both the ECG device and the Bluetooth module yields better reliability and performance. Nevertheless, a commercial dedicated telemedicine channel within GPRS Packet Switched data service, offered by the network providers, improve the performance and the reliably. An alternative improvement method would be, the selection of the most suitable GPRS coding scheme, depending on the quality radio condition, in order to gain a link adaptation, which results in the maximum possible performance in each link [10]. Furthermore, allocation of GPRS user on non hopping Broadcast Control Channel (BCCH) would results in a higher user throughput [10]. Moreover, application of High Speed Circuit Switched Data (HSCSD) for GSM or Enhanced Data Rates for GSM (EDGE), which uses a deeper modulation scheme to increase the basic data rates over GSM, are yet other alternatives to GPRS [8,18], which will be studied in the future. 5.4.6. Reliability and performance of the modified system From the achieved results concerning reliability and performance, application of a traditional GSM data service shows great promise. The reached system Up-time is close to the target and is not critical, which is in agreement with the reliability requirement defined by SONOFON [SONOFON, the Danish Network Provider] see Table 2. In agreement with previous work [14,33], we found the same resultant Bit Error Rate utilising GSM 900 standard. But In opposite to the work results of [14], but not very far from that, we found a higher signal to noise ratio. The resultant throughput in average is very close to the target (design requirement). The maximum data transfer rate over GSM system is never as high as the theoretical data transfer rate. Because, a practical data rate, depends on the data-traffic time, the weather, area and landscape, where the system is used. It has been experienced that in the forest the reliability and the performance were significantly affected. The existing factors in urban areas such as building and other man-made obstacles influence the signal strength. In rural areas, substantial path losses come about as a result of shadowing, scattering and absorption by trees and other vegetation, particularly at higher frequencies. Most reported measurements conclude that the extent of signal attenuation depends on the season of the year, the propagation distance within the vegetation and the frequency of the transmitted signal [21]. This part of the work needs further investigation. The results from the interference test support the notion that when the patient unit is very close (15 cm or less) to a microwave oven the performance and the reliability were affected, and frequent line cuts and break up were experienced. To best of our knowledge we could not found any other studies show the affect of microwave oven’s interference on a GSM or DECT terminal, but it is evident that a common microwave oven, with some

Y. Jasemian and L.A. Nielsen / Design and implementation of a telemedicine system using Bluetooth protocol

217

1000 W, induces electromagnetic fields to the adjacent environment. As the electrode cables and the serial cable connected between the T3300 and the mobile phone act as antenna, the interference from the microwave oven may occur. This part of the work, also, needs further investigation.

6. Conclusion We investigated the design of a generic wireless and real-time Telemedicine system for patient monitoring purpose. It has been shown that the development of such wireless real-time monitoring system, utilising Bluetooth technology, GSM and GPRS is entirely possible. A generic real-time wireless communication platform, utilizing Bluetooth protocol, GSM system and GPRS service, for the system realisation was designed, implemented and tested. Although HSCSD appears attractive and promising, but it exhibited some disadvantages, such as higher service costs comparing to a traditional GSM data service. Moreover, HSCSD requires more signalling during handover, connection setup and release, thus the probability of blocking or service degradation increase during handover, like this the QoS will be reduced. From the preceding discussion it is clear that the Bluetooth, GSM/GPRS solution is in the vein of such telemedicine system, but while promising, further works need to be carried out before reaching an appropriate reliability and performance. Universal Mobile Telecommunications System (UMTS) is a third generation of mobile network proposed by Europe to the International Telecommunication Union (ITU) for the International Mobile Telecommunications (IMT-2000), which is supported by Japan. The number 2000 in IMT-2000 indicates the start of the system and also indicates the used frequency spectrum (around 2000 MHz) [27]. This third generation mobile system operates internationally and includes different environments such as indoor and outdoor use, vehicles, satellites and pedestrians. Furthermore, UMTS provides different services such as real-time, non real-time, circuit and packet switched transmission with many different data rates (minimum 144 kbps and maximum 2 Mbps). Three G. supports many of the already existing networks such as GSM, ATM, IP and ISDN- based networks, it offers more bandwidth and with that a higher throughput and QoS can be achieved. We believe that UMTS is the future mobile communication network applied in a modern telemedicine system. The Telecommunication General Aspects of the system was investigated. Comparing the characteristics of the designed system with the system requirements, we conclude that the designed and implemented system fulfils the requirements. We have evaluated and validated the designed system by a number of well defined testes and experiments. The results showed that in the forest and when the patient unit was exposed by interference from a microwave oven the reliability and the performance of the system were affected, but are not critical. We conclude that the hypothesis of the present study has been confirmed by our experiments and results.

Acknowledgements Thanks to the Danish operator SONOFON for their network support, and to Ericsson Denmark for the financial support to the study. Special thanks to Professor Preben Mogensen from Center for Personkommunikation, Aalborg University, Denmark, for his technical review of and scientific advices on the present paper.

218

Y. Jasemian and L.A. Nielsen / Design and implementation of a telemedicine system using Bluetooth protocol

References
[1] [2] [3] [4] [5] [6] [7] [8] [9] [10] [11] [12] [13] [14] [15] [16] Z. Agha, R.M. Schapira and A.H. Maker, Cost effectiveness of telemedicine for the delivery of outpatient pulmonary care to a rural population, Telemedicine Journal and e-Health 8 (2002), 281–291. J. Bai, Y. Zhang, D. Shen, L. Wen, C. Ding, Z. Cui, F. Tian, B.Yu, B. Dai and J. Zhang, A portable ECG and Blood Pressure Telemonitoring System, IEEE Engineering in Medicine and Biology 18(4) (1999), 63–70. J. Bray and C.F. Sturman, BLUETOOTH 1.1, in: Connect Without Cables, J. Bray, C.F. Sturman and B. Goodwin, eds, New jersey, 2002, pp. 12, 33, 52–53 , 112, 224–230, 320–328. E.B. Carne, Telecommunications primer, in: Data, Voice and Video communications, B. Goodwin, ed., Prentice-Hall PTR, New Jersey, 1999, pp. 403–414. CENELEC ENV 50166-2, Human exposure to electromagnetic High-frequency Fields: (10 kHz–300 GHz), European Prestandard, European Committee for Electro technical Standardization (January, 1996). K.H. Dansky, L. Palmer, D. Shea and K.H. Bowles, Cost analysis of telehomecare, Telemedicine Journal and e-Health 7 (2001), 225–232. S.B. Freedman, Direct Transmission of Electrocardiograms to a Mobile Phone for Management of a Patient with Acute Myocardial Infarction, Journal of Telemedicine and Telecare 5 (1999), 67–69. L. Goleniewski, in: Telecommunications Essentials, P. Education, ed., Addition-Wesley, Boston, 2002, pp. 444–449. N. Golmie, R.E. Van Dyck, A. Soltanian, A. Tonnerre and O. Rebala, Interference evaluation of Bluetooth and IEEE 802.11b systems, kluwer academic publishers, Wireless Network 9 (2003), 201–211. P.J.A. Guti´ rrez, J. Wigard, P.N. Andersen, H.C. Damgaard and P. Mogensen, Performance of Link Adaptation in GPRS e Networks, IEEE 2 (2000), 492–499. ICNIRP, Health issues related to the use of hand-held radiotelephones and base transmitters, International Commission on Non-Ionizing Radiation Protection, Health Physics 70-4 (1996), 587–593. IEEE Standard for Safety levels with respect to human exposure to radio frequency electromagnetic fields (3 kHz to 300 GHz), ANSI/IEEE C95.1-1992, NJ 08854-1331, New York, 1991–1999. IRPA, Guidelines on limits of exposure to radio frequency electromagnetic fields in the frequency range from 100 kHz to 300 GHz, International Radiation Protection Association, Health Physics, IRPA 54 (1988), 115–123. R.H. Istepanian, B. Woodward, E. Gorilas and P.A. Balos, Design of mobile telemedicine systems using GSM and IS-54 cellular telephone standards, Journal of Telemedicine and Telecare 4(Spplement 1) (1998), 80–82. J. Kammann, Proposal for a Mobile Service Control Protocol, Institute for Communications and Navigation German Aerospace Center (DLR), 2001, PDCS Anaheim (USA), (http://www.kn-s.dlr.de/kammann/papers/iasted/iasted.pdf.). E. Kyriacou, S. Pavlopoulos, A. Berler, M. Neophytou, A. Bourka, A. Georgoulas, A. Anagnostaki, D. Karayiannis, C. Schizas, C. Pattichis, A. Andreou and D. Koutsouris, Multi-purpose HealthCare Telemedicine System with mobile communication link support, BioMedical Engineering OnLine (2003), http://www.biomedical-engineeringonline.com/content/2/I/7, 1–12. F. Magrabi, N.H. Lovell and B.G. Celler, A Web-based approach for electrocardiogram monitoring in the home, International Journal of Medical Informatics 54 (1999), 145–153. J. Melero, J. Wigard, T. Halonen and J. Romero, Basics of GSM Radio Communication and Spectral Efficiency, in: GSM, GPRS and EDGE performance, Evolution Towards 3G/UMTS, T. Halonen, J. Romero and J. Melero, eds, John Wiley & Sons LTD, West Sussex, 2002, pp. 145–190. P.E. Mogensen and J. Wigard, On Antenna – and Frequency diversity in GSM related systems (GSM-900, DCS-1800, and PCS1900), IEEE (1996), 1272–1276. O.I. Orlov, D.V. Drozdov, C.R. Doarn and R.C. Merrell, Wireless ECG monitoring by telephone, Telemedicine Journal and e-Health 7 (2001), 33–38. J.D. Parsons, Propagation over irregular terrain, in: The Mobile radio propagation channel, J.D. Parsons, ed., John Wiley & Sons LTD, West Sussex, 2000, pp. 52–53. D. Poljak, in: Human Exposure to Electromagnetic Fields, D. Poljak, ed., WIT press, Southampton, 2004, pp. 11–40. L. Ramqvist, Health and Safety in Mobile Telephony, EN/LZT 123 4060, Ericsson Radio Systems AB, 164 80 Stockholm, 1997, pp. 1–14. J. Romero, J. Martinez, S. Nikkarinen and M. Moisio, GPRS and EGPRS Performance, in: GSM, GPRS and EDGE performance, Evolution Towards 3G/UMTS, T. Halonen, J. Romero and J. Melero, eds, John Wiley & Sons LTD, West Sussex, 2002, p. 291. R. Satava, P.B. Angood, B. Harnett, C. Macedonia and R. Merrell, The physiologic cipher at altitude: telemedicine and real-time monitoring of climbers on Mount Everest, Telemedicine Journal and e-Health 6 (2000), 303–313. S. Scalvini, E. Zanelli, D. Domenighini, G. Massarelli, P. Zampini, A. Giordano and F. Glisenti, Telecardiology community (a new approach to take care of cardiac patients), Cardiologia (1999), 921–924. J.H. Schiller, in: Mobile Communications, Jochen H. Schiller, ed., Addition-Wesley, London, 2000, pp. 83–92, p. 86, pp. 105–113, pp. 119–125.

[17] [18] [19] [20] [21] [22] [23] [24] [25] [26] [27]

Y. Jasemian and L.A. Nielsen / Design and implementation of a telemedicine system using Bluetooth protocol [28] [29] [30] [31] [32] [33] [34] [35] [36]

219

K. Shimizu, Telemedicine by Mobile Communication, IEEE Engineering in Medicine and Biology (1999), 32–44. Sony Ericsson, White Paper T610/T612, EN/LZT 108 6196 Sony Ericsson Mobile Communication AB, Sony Ericsson, ed., Lund, 2002. S.W. Strode, S. Gustke and A. Allen, Technical and clinical progress in telemedicine, JAMA Journal of the American Medical Association 281-12 (1999), 1066–1068. A.S. Tanenbaum, in: Computer Networks third edition, A.S. Tanenbaum, ed., Prentice Hall, Inc., New Jersey, 1996, pp. 231–234. Telecommunication standardization sector of ITU series P: Telephone Transmission Quality (Methods for objective and subjective assessment of quality), Telecommunication standardization sector of ITU, (1996), 800. S.B. Uldal, S. Manankova and V. Kozlov, Choosing a PC-based ECG System for a Mobile Telemedicine unit, The Health News magazine (www.Telemedtoday.com), 1999, pp. 30–31. B. Woodward, R.S.H. Istepanian and C.I. Richards, Design of a Telemedicine System Using a Mobile Telephone, IEEE (2001), 13–15. World Health Organization WHO, Environmental Health Criteria 137, Electromagnetic Fields (300 Hz to 300 GHz), Geneva, 1993. X. Zhao, D. Fei, C.R. Doarn, B. Harnett and R. Merrell, A telemedicine system for wireless home healthcare based on bluetooth and the internet, Telemedicine Journal and e-Health 10 supplement 2 (2004), 110–116.

References: [1] [2] [3] [4] [5] [6] [7] [8] [9] [10] [11] [12] [13] [14] [15] [16] Z. Agha, R.M. Schapira and A.H. Maker, Cost effectiveness of telemedicine for the delivery of outpatient pulmonary care to a rural population, Telemedicine Journal and e-Health 8 (2002), 281–291. J. Bai, Y. Zhang, D. Shen, L. Wen, C. Ding, Z. Cui, F. Tian, B.Yu, B. Dai and J. Zhang, A portable ECG and Blood Pressure Telemonitoring System, IEEE Engineering in Medicine and Biology 18(4) (1999), 63–70. J. Bray and C.F. Sturman, BLUETOOTH 1.1, in: Connect Without Cables, J. Bray, C.F. Sturman and B. Goodwin, eds, New jersey, 2002, pp. 12, 33, 52–53 , 112, 224–230, 320–328. E.B. Carne, Telecommunications primer, in: Data, Voice and Video communications, B. Goodwin, ed., Prentice-Hall PTR, New Jersey, 1999, pp. 403–414. CENELEC ENV 50166-2, Human exposure to electromagnetic High-frequency Fields: (10 kHz–300 GHz), European Prestandard, European Committee for Electro technical Standardization (January, 1996). K.H. Dansky, L. Palmer, D. Shea and K.H. Bowles, Cost analysis of telehomecare, Telemedicine Journal and e-Health 7 (2001), 225–232. S.B. Freedman, Direct Transmission of Electrocardiograms to a Mobile Phone for Management of a Patient with Acute Myocardial Infarction, Journal of Telemedicine and Telecare 5 (1999), 67–69. L. Goleniewski, in: Telecommunications Essentials, P. Education, ed., Addition-Wesley, Boston, 2002, pp. 444–449. N. Golmie, R.E. Van Dyck, A. Soltanian, A. Tonnerre and O. Rebala, Interference evaluation of Bluetooth and IEEE 802.11b systems, kluwer academic publishers, Wireless Network 9 (2003), 201–211. P.J.A. Guti´ rrez, J. Wigard, P.N. Andersen, H.C. Damgaard and P. Mogensen, Performance of Link Adaptation in GPRS e Networks, IEEE 2 (2000), 492–499. ICNIRP, Health issues related to the use of hand-held radiotelephones and base transmitters, International Commission on Non-Ionizing Radiation Protection, Health Physics 70-4 (1996), 587–593. IEEE Standard for Safety levels with respect to human exposure to radio frequency electromagnetic fields (3 kHz to 300 GHz), ANSI/IEEE C95.1-1992, NJ 08854-1331, New York, 1991–1999. IRPA, Guidelines on limits of exposure to radio frequency electromagnetic fields in the frequency range from 100 kHz to 300 GHz, International Radiation Protection Association, Health Physics, IRPA 54 (1988), 115–123. R.H. Istepanian, B. Woodward, E. Gorilas and P.A. Balos, Design of mobile telemedicine systems using GSM and IS-54 cellular telephone standards, Journal of Telemedicine and Telecare 4(Spplement 1) (1998), 80–82. J. Kammann, Proposal for a Mobile Service Control Protocol, Institute for Communications and Navigation German Aerospace Center (DLR), 2001, PDCS Anaheim (USA), (http://www.kn-s.dlr.de/kammann/papers/iasted/iasted.pdf.). E. Kyriacou, S. Pavlopoulos, A. Berler, M. Neophytou, A. Bourka, A. Georgoulas, A. Anagnostaki, D. Karayiannis, C. Schizas, C. Pattichis, A. Andreou and D. Koutsouris, Multi-purpose HealthCare Telemedicine System with mobile communication link support, BioMedical Engineering OnLine (2003), http://www.biomedical-engineeringonline.com/content/2/I/7, 1–12. F. Magrabi, N.H. Lovell and B.G. Celler, A Web-based approach for electrocardiogram monitoring in the home, International Journal of Medical Informatics 54 (1999), 145–153. J. Melero, J. Wigard, T. Halonen and J. Romero, Basics of GSM Radio Communication and Spectral Efficiency, in: GSM, GPRS and EDGE performance, Evolution Towards 3G/UMTS, T. Halonen, J. Romero and J. Melero, eds, John Wiley & Sons LTD, West Sussex, 2002, pp. 145–190. P.E. Mogensen and J. Wigard, On Antenna – and Frequency diversity in GSM related systems (GSM-900, DCS-1800, and PCS1900), IEEE (1996), 1272–1276. O.I. Orlov, D.V. Drozdov, C.R. Doarn and R.C. Merrell, Wireless ECG monitoring by telephone, Telemedicine Journal and e-Health 7 (2001), 33–38. J.D. Parsons, Propagation over irregular terrain, in: The Mobile radio propagation channel, J.D. Parsons, ed., John Wiley & Sons LTD, West Sussex, 2000, pp. 52–53. D. Poljak, in: Human Exposure to Electromagnetic Fields, D. Poljak, ed., WIT press, Southampton, 2004, pp. 11–40. L. Ramqvist, Health and Safety in Mobile Telephony, EN/LZT 123 4060, Ericsson Radio Systems AB, 164 80 Stockholm, 1997, pp. 1–14. J. Romero, J. Martinez, S. Nikkarinen and M. Moisio, GPRS and EGPRS Performance, in: GSM, GPRS and EDGE performance, Evolution Towards 3G/UMTS, T. Halonen, J. Romero and J. Melero, eds, John Wiley & Sons LTD, West Sussex, 2002, p. 291. R. Satava, P.B. Angood, B. Harnett, C. Macedonia and R. Merrell, The physiologic cipher at altitude: telemedicine and real-time monitoring of climbers on Mount Everest, Telemedicine Journal and e-Health 6 (2000), 303–313. S. Scalvini, E. Zanelli, D. Domenighini, G. Massarelli, P. Zampini, A. Giordano and F. Glisenti, Telecardiology community (a new approach to take care of cardiac patients), Cardiologia (1999), 921–924. J.H. Schiller, in: Mobile Communications, Jochen H. Schiller, ed., Addition-Wesley, London, 2000, pp. 83–92, p. 86, pp. 105–113, pp. 119–125. [17] [18] [19] [20] [21] [22] [23] [24] [25] [26] [27] Y. Jasemian and L.A. Nielsen / Design and implementation of a telemedicine system using Bluetooth protocol [28] [29] [30] [31] [32] [33] [34] [35] [36] 219 K. Shimizu, Telemedicine by Mobile Communication, IEEE Engineering in Medicine and Biology (1999), 32–44. Sony Ericsson, White Paper T610/T612, EN/LZT 108 6196 Sony Ericsson Mobile Communication AB, Sony Ericsson, ed., Lund, 2002. S.W. Strode, S. Gustke and A. Allen, Technical and clinical progress in telemedicine, JAMA Journal of the American Medical Association 281-12 (1999), 1066–1068. A.S. Tanenbaum, in: Computer Networks third edition, A.S. Tanenbaum, ed., Prentice Hall, Inc., New Jersey, 1996, pp. 231–234. Telecommunication standardization sector of ITU series P: Telephone Transmission Quality (Methods for objective and subjective assessment of quality), Telecommunication standardization sector of ITU, (1996), 800. S.B. Uldal, S. Manankova and V. Kozlov, Choosing a PC-based ECG System for a Mobile Telemedicine unit, The Health News magazine (www.Telemedtoday.com), 1999, pp. 30–31. B. Woodward, R.S.H. Istepanian and C.I. Richards, Design of a Telemedicine System Using a Mobile Telephone, IEEE (2001), 13–15. World Health Organization WHO, Environmental Health Criteria 137, Electromagnetic Fields (300 Hz to 300 GHz), Geneva, 1993. X. Zhao, D. Fei, C.R. Doarn, B. Harnett and R. Merrell, A telemedicine system for wireless home healthcare based on bluetooth and the internet, Telemedicine Journal and e-Health 10 supplement 2 (2004), 110–116.