User Tools

Site Tools


index

Presentation

In a society irreversibly marked by the everyday use of technology, new ways to automatically share this data and automate systems and decisions are constantly emerging. Machine-to-Machine (M2M) solutions are becoming increasingly popular, for the great scalability they provide, and because they generally make people's lives easier. A quick M2M example is GPS systems. There is no human interaction, and yet by gathering and processing data from satellites, it is able to compute the user's current position. The ease of use of these systems grant them the massive adoption they hold in today's society.

This dissertation, executed by Ricardo Morgado, is going to study the use of a smartphone as an enabler of M2M services, in an health-care scenario, aiming to allow the user to effortlessly make blood pressure or weight measurements and store them in the Cloud. This will allow the user, for example, to conveniently search his medical measurements history.

First, the context of project in which this dissertation is inserted is explained, followed by the motivation for studying the subject at hand. Finally, the objectives this dissertation aims to achieve are stated, before the approach definition.

Context of Project

This dissertation is integrated in a joint project between the Instituto de Telecomunicações (IT Porto) and Portugal Telecomunicações Inovação (PT Inovação). It involves the creation of a Mobile M2M Gateway (following ETSI's M2M standard) as an Android application. This gateway will aggregate sensor data (either internal or external) and manage the communication with the M2M network domain, as a proxy for the sensors. The Gateway was designed to be multi-protocol, and originally implemented with MQTT (Message Queue Telemetry Transport. CoAP (Constrained Application Protocol support is currently in development, and AMQP remains a possibility for later addition.

Back when the project started in May 2013, it focused in the architectural design and implementation of the M2M Gateway, with MQTT. The aim was to develop it as modularly as possible, to ease the addition of protocols and sensors. At the beginning of this dissertation (September 2013), version 0 of the Gateway had been released with the MQTT protocol working (sending data to PT’s Broker), as well as the Gateways base architecture, which had the ability to gather the internal sensors data, store it and then send in defined intervals or when the buffer got full. It also featured the modularity which was desired, having each significant module running in its own thread, to improve parallel processing and thus overall performance.

Motivation

Nowadays, the ascension of technology is a given fact, as evident by the well-known Moore's Law, which has been fairly accurate since the paper was published back in 1965. As today's trends focus on the ubiquity of smartphones, there is a massification of applications that aim to ease people's lives. Despite this, there are not many applications focused on M2M systems, let alone mobile M2M scenarios, which inspired this dissertation's theme.

M2M communications, have an important role in interconnecting sensors and services, providing the means to allow data to be seamlessly stored and then available to applications authorized to access it. As an example, cities are starting to explore the possibility of using M2M Smart Grids, that allow meter measurements to be automatically uploaded to the service provider, either it is water, electricity or gas. There are also other scenarios in smart homes, where M2M communication can be used to autonomously optimize and manage energy consumption of any applicable appliance that runs on electricity.

These and many other scenarios are becoming a reality every day, but the lack of mobile solutions for the M2M architecture limits its possibilities. Evaluating an health-care scenario, it should be possible for someone who suddenly felt bad to make a blood pressure measurement, with a friend's or pharmacy's sphygmomanometer, and upload for his doctor's assessment, storing it in the service for future reference in his medical history.

These needs provided the proper motivation for the study of this subject, since it will only be possible if the M2M modules are mobile. Smartphones are then an excellent choice, since they are permanently around the user, possess the necessary processing, battery, and connectivity power needed for a Gateway, while retaining the mobility necessary to fulfill a new wider range of scenarios.

Having a smartphone working as an M2M Gateway has its advantages, but since users usually do their best to save battery power, the smartphone in this dissertation's scenario is being considered constrained, so that the application working as the M2M Gateway uses the absolute minimum resources possible, to fulfill this requirement.

Objectives

This dissertation's first objective is the conclusion of the mobile M2M Gateway, with the ability to communicate in MQTT and CoAP with PT's Broker, using ETSI's M2M communication standard. The Gateway will also feature the versatile modular architecture, which aims to provide the flexibility necessary for adding new sensors or protocols, without the need to re-write the entire application.

Thereafter, in sequence of the M2M Gateway development, an Application that is works as proof-of-concept in the mobile health-care scenario, is also going to be developed.

Given the constrained nature of smartphones, there is a special interest in saving its scarce resources, and so one of the objectives is also to evaluate the performance of the M2M Gateway with each protocol, specifically in terms of battery drain, and network usage on both cellular and Wi-Fi. This could be accomplished by sending the same amount of information though each protocol for the same duration, monitor the behavior of the smartphone and the network. The goal of this study is to make a real scenario comparison between the protocols, that may end up with the suggestion of a smart algorithm to control the communication channels, to both improve battery life and reduce network traffic.

Approach

To tackle the M2M mobility issue explained above, this dissertation's approach relies on a versatile mobile Gateway, that will have to accommodate the following functional and non-functional requirements.

Functional requirements:

  • Support disconnection and save any unsent data for next connection.
  • Support a Web Server that allows the reception of requests and subscriptions from the Server.
  • Support health-care external sensors:
    • Sphygmomanometer.
    • Scale.
  • Support API for local access.

Non-functional requirements:

  • Support a modular design.
  • Multi protocol support:
    • HTTPS.
    • CoAP.
    • MQTT.
    • AMQP (Low Priority).
  • ETSI M2M standard compliant.
  • Support a modular design.

These requirements will then be demonstrated by the proof-of-concept application, in a health-care scenario that fits the ETSI standards and health-care requirements.

index.txt · Last modified: 2014/03/25 17:12 by ricardomorgado