MOBILE REMOTE MONITORING SYSTEM

Abstract - Low bandwidth has long been a reason for the unsuitability of wireless internet in telemedicine. However with the advent of extended third generation wireless as an economically accessible high speed network, more opportunities are being created in this area of telemedicine. This paper explores the opportunity created by the latest wireless broadband technology for remote monitoring of patients in the home.

I.                  INTRODUCTION

One of the first records of wireless communication in telemedicine was the use of a GSM (second Generation) cellular network to transmit an electrocardiogram (ECG) from a patient in 2001. Since then there have been a number of developments related to improving the speed and capacity of networks. The main wireless technologies have been GPRS (2.5G), satellite and wireless LAN . The typical speed of 2G system is 9.6 Kbps. However the evolution of mobile telecommunication systems from second generation (2G) to 2.5G (iDEN 64 Kbps, GPRS 171 Kbps, EDGE 384 Kbps) and 3G (W-CDMA, CDMA2000, TD-CDMA) systems have provided faster data transfer rates but these rates are still not fast enough for many telehealth applications. Satellite communication is another wireless communication medium, which despite having high operation costs, has the advantage of world wide coverage and a variety of data transfer speeds from 10 Kbps to a high speed data rate of 100 Kbps. The latest technology in use in Australia is 3.5G, also called BeyondG and NextG. It is based on a High Speed Packet Access (HSPA) network which consists of HSDPA (High Speed Downlink Packet Access) for downlink and HSUPA (High Speed Uplink Packet Access) for uplink, uplink being slower than downlink. 4G, e.g. Orthogonal Frequency and Code Division Multiplexing (OFCDM), is the future generation of wireless network that boasts potential speeds of up to 100Mbps.

Our research focuses on use of a 3.5G based system for home monitoring and control of drug consumption in patients such as those on narcotic replacement medication. The research work has three parts:

1. Define the requirements for monitoring and control and identify the limitations of the extended third-generation mobile network in this application.

2. Design and implement hardware and software for a Mobile Remote Monitoring System (MRMS) capable of monitoring patients through audio and video and controlling remote assessment and medication dispensing devices through command signals based on extended 3G.

3. Test the MRMS firstly in the laboratory and then in a controlled external environment.

II. MOBILE REMOTE MONITORING SYSTEM (MRMS)

The MRMS consists of a dedicated server and at least one Client (Patient) side computer linked via cable or wireless broadband internet (extended G). A prominent feature of the system is its portability, so there is no location restriction. The system is designed to replicate the dosing conditions of a patient being dosed under the supervision of a clinician.

                                                                    An overview of the remote monitoring system

The program accommodates both types of Wireless modem providing the user with flexibility of choosing the appropriate modem based on network availability and cost.

1) The Client side comprises a laptop, a wireless modem, two webcams and a headset with a microphone. The Client has the following functions running on parallel Threads

a. Video communication: This function captures images from the webcams as well as receives images from the Server and plays them

b. Audio communication: This function captures mono audio at 11025 HZ at 8 bits per sample for 743 ms. Both the video and audio data are sent to the Server by User Datagram Protocol (UDP).

2) The Server side comprises a computer, USB webcam and a headset with a microphone. It is connected to the internet by broadband internet with a static IP. The Server has control of the Client (e.g. which webcam to use). The Server has the following functions.

a. Video communication: Unlike the Client, there is only one webcam at the Server.

b. Audio Communication: Audio has the same specification as for the Client. Both the video and audio send data to Client by TCP or UDP

c. Control Signal: This function allows the user to select between the alternative webcams at the remote end.

B. System Methodology

The MRMS comprises two parties – the Staff (Server side) and the Patient (Client side). The Client sends images and audio to the Server which has a fixed IP using UDP. UDP, which is also termed Unreliable Data Protocol since it doesn’t guarantee the delivery of packets, is, however, suitable for sending video images since loss of few packets is not as important as timely delivery of the latest images. UDP is faster than TCP as UDP does not wait for connection to establish and has less overhead. Congestion control is an important issue which has to be implemented at the application layer when sending data by UDP.  The webcams capture images at five different resolutions - 160x120, 176x144, 320x240, 352x288, and 640x480. Joint-Picture-Experts-Group (JPEG) compression was applied to the images.

III. EXPERIMENTAL

Experiments were conducted using this system in the lab with normal people. The Server was based on a Pentium IV 2.8GHz processor running Windows XP. The Client was based on a Dell Latitude Laptop with a Pentium M 1.5 GHz processor running Windows XP. Our university’s broadband internet was used at the Server end and the two wireless modems at the Client end. We had two webcams for general viewing, and headphones with microphones at the Server and Client ends. The Wireless internet had 7.2 Mbps downlink and 1.9 Mbps uplink speeds on a UMTS network incorporating HSPA running at 850 MHz. The questions thet arise are- quality of image required, frame rate, UDP perform compared to TCP, frame loss rate, bandwidth of the network, congestion.

IV. RESULTS AND DISCUSSION

The optimum range of frame rate for 640 x 480 was around 6 to 10 fps whereas for 352 x 480 the optimum rate was 10 to 20 fps in our experiment. 10 fps gave a satisfactory video output.

Two compression ratios 35% and 65% for image were used by the program. The average size for 352x288 using JPEG compression was around 6 KB at 65% compression & was 8 to 11 KB at 35% compression. UDP as expected could transmit frames faster than TCP which would wait for acknowledgement before sending the next frame. The upload frame rate was found to be around

9.00 fps with 65% compression by TCP, In HSPA, download is faster than upload. With UDP the download rate was around 9.90 at 65% compression There was no frame loss when the image size was 6.26 KB Congestion control is an important issue when using UDP  unlike TCP. However with our image size and frequency and the upload speed of 1.9 Mbps supported by

Telstra’s NextG, this was not a major problem.

V. CONCLUSION

Our results show that a usable mobile remote monitoring system with two way audio and video

communication capability can be achieved using 3.5G wireless technology. It can also easily be used for wired broadband internet like ADSL. The next stage of our research is implementing a congestion control mechanism. The system should also support remote assessment techniques. Along with that further work includes compressing JPEG images into M-JPEG instead of

 ending one frame per packet.


Reference :

Title : 3.5G Based Mobile Remote Monitoring System

Authors : Aman Bajracharya, Timothy J. Gale, Clive R. Stack and Paul Turner

30th Annual International IEEE EMBS Conference

Vancouver, British Columbia, Canada, August 20-24, 2008