Progress

Scott Haseley
1 parent 9aa9847d
Showing 5 changed files with 100 additions and 4 deletions
badwords
design.tex
introduction.tex
paper.tex
references.bib
+AppInsight
+AsyncTask
+AsyncTasks
+cgroup
+Challen
+challen
+edu
+Haseley
+HotMobile
+ondemand
+pdfinfo
+prioritization
+QoE
+shaseley
+\section{design}
+A QoE-centric OS must continuously prioritize resources based on their 
+effect on QoE. Doing so requires the OS to both quantify QoE and understand
+which tasks contribute to it. Given the multi-threaded nature of modern apps,
+the complexity of the systems they run on, and the operating system's vantage
+point, this is a challenging task. Unlike approaches like
+AppInsight~\cite{appinsight-osdi2012} that requires app instrumentation, or
+QoE Doctor~\cite{qoe-doctor-imc14} which involves human intervention, using QoE
+to do continuous resource prioritization for all apps on all devices requires
+the ability to measure QoE on unmodified apps and without developer
+participation.
+
+% Task prioritization
+As principle, if a user is waiting for something, priority should
+be given to any task that contributes to that end.
+
+%Network prioritization...
+
+%I/O prioritization...
+
+%New approaches...
+\section{introduction}
+
+Current operating systems are already proficient at managing certain system
+resources, such as the CPU, memory, and disk. But on interactive mobile devices,
+users care more about resources such as time, battery life, and money, that are
+unmanaged or poorly managed by today's smartphone platforms.
+When apps slow down our typing, they waste our time. When apps download
+unnecessary resources, they waste battery life and potentially money, if using
+a metered data plan. It is the degree two which mobile devices effectively
+manage these human-facing resources that determines a user's
+\textit{quality of experience} (QoE), and it is QoE which should drive not just
+policy, but decisions on mobile devices.
+
+\begin{sloppypar}
+While modern operating systems such as Android make decisions based on policies
+meant to improve QoE, it is unclear that these static policies always result
+in the right decisions. For example, Android uses two mechanisms to control the
+underlying task scheduling in Linux: thread priority and control groups (cgroup).
+Background threads - any thread running in another app, or low priority threads
+running in the foreground app - are placed in a cgroup that receives a very
+limited share of the processor, \( \approx \) 5\% on Nexus 5 running stock
+Android 5.1.1. Furthermore, the scheduler considers priority for tasks withing
+the same cgroup. However, thread priorities can be set by app developers both 
+directly through the API, and indirectly by creating an \texttt{AsyncTask}.
+While this policy was created based on the assumed effect on QoE, it is still
+possible to negatively affect QoE by assigning the wrong priority to a thread
+or by inappropriately using \texttt{AsyncTasks}.
+\end{sloppypar}
+
+The \texttt{ondemand} CPU frequency governor that is enabled in Android
+Linux is another example of where static policy meant to improve QoE may not
+result in correct decisions being made. This governor changes the CPU frequency
+depending on the current CPU load with the goal of improving performance.
+However since running at higher frequencies is less efficient, this increased
+performance comes at the cost. If QoE would not be improved by increasing the
+frequency, such as for time-insensitive background tasks or when the performance
+increase would not be perceivable, negative QoE may manifest itself through a
+quicker draining battery.
+
+Static policies can lead to resource allocation decisions that are less than
+optimal in terms of QoE. To remedy this, we propose designing QoE-centric
+mobile operating systems that can accurately quantify QoE and use it as feedback
+to drive resource allocation decisions.
 \input{./include/start.tex}
  
 \def\theconference{HotMobile'16}
-\def\thetitle{Poster: Quantifying Quality of Experience Using the Screen as a Sensor}
-\def\theauthors{Scott Haseley$^\star$, Geoffrey Challen$^\star$}
+\def\thetitle{Poster: QoE-centric Mobile Operating System Design}
+\def\theauthors{Scott Haseley, Geoffrey Challen}
 \def\theemails{\{shaseley,challen\}@buffalo.edu}
  
 \ifdefined\isdraft
@@ -32,7 +32,7 @@
  
 \author{
 \alignauthor\theauthors\\[0.05in]
- \affaddr{$\star$ University at Buffalo}\\
+ \affaddr{University at Buffalo}\\
 \email{\theemails}
 }
  
@@ -48,7 +48,9 @@
  
 \maketitle
  
-\input{abstract.tex}
+%\input{abstract.tex}
+\input{introduction.tex}
+\input{design.tex}
  
 {\scriptsize
 \renewcommand{\baselinestretch}{0.8}
@@ -581,3 +581,19 @@ pages = {1-14},
 year={2008},
 publisher={Elsevier}
 }
+
+@inproceedings{qoe-doctor-imc14,
+ title = {QoE Doctor: Diagnosing Mobile App QoE with Automated
+UI Control and Cross-layer Analysis},
+ author = {Qi Alfred Chen and Haokun Luo and Sanae Rosen and Z. Morley Mao and
+Karthik Iyer and Jie Hui and Kranthi Sontineni and Kevin Lau},
+ booktitle = proc # {ACM Internet Measurement Conference},
+ year = {2014}
+}
+
+@inproceedings{appinsight-osdi2012,
+ title = {AppInsight: Mobile App Performance Monitoring in the Wild},
+ author = {Lenin Ravindranath and Jitendra Padhye and Sharad Agarwal and Ratul Mahajan and Ian Obermiller and Shahin Shayandeh},
+ booktitle = proc # {OSDI},
+ year = {2012}
+}