-
Notifications
You must be signed in to change notification settings - Fork 0
/
Copy pathreport.tex
113 lines (99 loc) · 4.75 KB
/
report.tex
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
%% ----------------------------------------------------------------
%% GDP.tex
%% ----------------------------------------------------------------
\documentclass{.style/ecsgdp} % Use the GDP Report Style
\graphicspath{{./figures/}} % Location of your graphics files
\usepackage{array}
\usepackage{listings}
\usepackage{pdflscape}
\usepackage{pgfgantt}
\usepackage{pgfplots}
\usepackage{enumitem} % For making numbered indent lists
\usepackage{color}
\usepackage{graphicx,subfigure}
\usepackage{multirow}
\usepackage{todonotes}
\usepackage{tikz}
\usetikzlibrary{shapes,automata,positioning}
\usepackage{gensymb} % For extra symbols such as Ohm
\usepackage{dirtree}
\lstset{language=C}
\hypersetup{colorlinks=true} % Set to false for black/white printing
%\input{Definitions} % Include your abbreviations
\newcommand{\namedsection}[2]{\section[#1]{#1 \hfill {\small \color{gray} #2}}} % Use this to create sections with authorship: \namedsection{Section Name}{Your name}
\newcommand{\namedsubsection}[2]{\subsection[#1]{#1 \hfill {\small \color{gray} #2}}} % Use this to create subsections with authorship: \namedsubsection{Section Name}{Your name}
\newcommand{\note}[1]{\todo[inline]{#1}}
% Fix for todo notes rendering incorrectly
\setlength{\marginparwidth}{2cm}
%% ----------------------------------------------------------------
\begin{document}
\frontmatter
\title{Ultra-Low-Power Exercise Monitoring Applications for Sub-Threshold Micro-Controllers}
\authors {\\ \texorpdfstring
{\href{mailto:[email protected]}{Daniel Playle}}
{Daniel J. A. Playle}
\\
\texorpdfstring
{\href{mailto:[email protected]}{Emily Shepherd}}
{Emily Shepherd}
\\
\texorpdfstring
{\href{mailto:[email protected]}{Mohit Gupta}}
{Mohit Gupta}
\\
\texorpdfstring
{\href{mailto:[email protected]}{Toby Finch}}
{Toby Finch}
\\
\texorpdfstring
{\href{mailto:[email protected]}{Calin Pasat}}
{Calin Pasat}
}
\addresses {\groupname\\\deptname\\\univname}
\date {\today}
\subject {}
\keywords {}
\supervisor{Dr. Geoff Merrett \\ Dr. Alex Weddell}
\examiner{Dr. Klaus-Peter Zauner}
\degrees {Master of Engineering}
\maketitle
\begin{abstract}
This paper details the process by which the team developed an exercise detection algorithm capable of running on a sub threshold Cortex M0+. The report explains the process by which a user study was carried out to obtain movement data of multiple people performing the exercises, then discusses the methods by which this was processed. It provides a review of the exercise detection systems in use today and compares various Machine Learning algorithms, settling on a Multilayer Perceptron with added case-specific heuristics. The paper explains the choice of emulating the proposed device on a Cortex M0, running on an mBed platform and compares this to an FPGA before describing the process of designing, building and working with the hardware. The work to develop the software for a constrained system is then analysed, focusing on the optimisations to the mBed library to reduce required size and the removal of floating point from the algorithm. Finally, the results of this work are reviewed, proving that the working algorithm requires 3456 bytes and can run on a $861 \mu W$ system. The report finishes with an overview of the project management and concludes that the project is a successful proof of concept.
\end{abstract}
\tableofcontents
\listoffigures
\listoftables
\lstlistoflistings
\acknowledgements{Our group would like to take the time to thank the people that helped us with this project. Firstly, our supervisors, Dr Geoff Merrett and Dr Alex Weddell, for their continued and weekly support. Thanks must also be given to ARM for providing such an interesting and thought provoking topic of investigation. Particularly, we would like to thank James Myers and Rohan Gaddh for acting as a point of contact, and for providing access to the resources we required.}
%\dedicatory{To \dots}
\mainmatter
%% ----------------------------------------------------------------
% Introduction chapter
\input{intro/intro}
% User study chapter
\input{study/study}
% Software chapter
\input{research/research}
% Embedded
\input{embedded/embedded}
% Software chapter
\input{software/software}
% Results chapter
\input{results/results}
% Project management chapter
\input{management/management}
% Conclusion chapter
\input{conclusion/conclusion}
\appendix
\input{appendices/brief}
\input{appendices/email}
\input{appendices/source}
\input{appendices/risks}
\input{appendices/gantt}
\input{appendices/data_gathering}
%
\backmatter
\bibliographystyle{.style/IEEEtran}
\bibliography{ECS}
\end{document}
%% ----------------------------------------------------------------