diff --git a/paper/main.bib b/paper/main.bib
index 0f6719f38be948676a4e66ee5af328228805a313..88801b6bfcc2c73333e5580cfa557702c2019c17 100644
--- a/paper/main.bib
+++ b/paper/main.bib
@@ -317,4 +317,24 @@ month={Sept},}
  isbn = {0470041900},
  publisher = {Wiley-Interscience},
  address = {New York, NY, USA},
-} 
+}
+
+@incollection{bib:A,
+  title={Inertial Sensing, GPS and Odometry},
+  author={Dudek, Gregory and Jenkin, Michael},
+  booktitle={Springer Handbook of Robotics},
+  pages={737--752},
+  year={2016},
+  publisher={Springer}
+}
+
+@inproceedings{bib:B,
+  title={An introduction to event-triggered and self-triggered control},
+  author={Heemels, WPMH and Johansson, Karl Henrik and Tabuada, Paulo},
+  booktitle={Decision and Control (CDC), 2012 IEEE 51st Annual Conference on},
+  pages={3270--3285},
+  year={2012},
+  organization={IEEE}
+}
+
+
diff --git a/paper/sections/02-related-work.tex b/paper/sections/02-related-work.tex
index 83ddfc5ed7ac4d7e84c188b80644f9867569a124..60d3dd15748d1b6bd69d2c9194c7a3533ba7661b 100644
--- a/paper/sections/02-related-work.tex
+++ b/paper/sections/02-related-work.tex
@@ -56,6 +56,17 @@ model. We argue that this is needed, since the GPS receiver can behave
 differently in response to the same input, depending on its internal
 state.
 
+To end this brief review, we would like to notice that this work
+combines, in a purpose-specific manner, problems and ideas from
+different domains. To quote just a couple, there is a vast literature
+on correcting drift in accelerometer-based odometry, see
+e.g.~\cite{bib:A}, and the idea of taking action -- in our case,
+accessing the GPS -- only when performance is deemed unsatisfactory,
+is shared by the more general domain of event-triggered
+sampling~\cite{bib:B}. However, the addressed problem has specific
+characteristics, detailed in the following, that make our solution
+effective despite -- and in some sense thanks to -- its simplicity.
+
 % \begin{itemize}
 % \item many works discuss the problem of power consumption in GPS
 %   devices from different points of view. those are:
diff --git a/paper/sections/04-fusion.tex b/paper/sections/04-fusion.tex
index 34312df5df48fe41b1b1ddbe6b5a3f6be4346764..9186ff650a604f106fd1a56f0cc91cd92eff81b9 100644
--- a/paper/sections/04-fusion.tex
+++ b/paper/sections/04-fusion.tex
@@ -46,12 +46,13 @@ estimates and compensates the IMU biases.
 
 We want to estimate the following quantities: the tridimensional
 position, the tridimensional velocity and the orientation (attitude)
-using its quaternion representation. We denote with $p(k)$ the
-position at time $k$ and measure it in meters with respect to a
-reference point, i.e., $p(k) \in \mathbb{R}^3 [m]$. We use $v(k)$ to
-indicate the velocity at time $k$ measuring it in meters per seconds,
-i.e., $v(k) \in \mathbb{R}^3 [m/s]$. Finally, we denote with $q(k)$
-the attitude at time $k$, using its (adimensional) quaternion
+using its quaternion representation. We define the state
+$x(k) = [p(k), v(k), q(k)]'$, where we denote with $p(k)$ the position
+at time $k$ and measure it in meters with respect to a reference
+point, i.e., $p(k) \in \mathbb{R}^3 [m]$. We use $v(k)$ to indicate
+the velocity at time $k$ measuring it in meters per seconds, i.e.,
+$v(k) \in \mathbb{R}^3 [m/s]$. Finally, we denote with $q(k)$ the
+attitude at time $k$, using its (adimensional) quaternion
 representation, i.e., $q(k)\in \mathbb{R}^4 [-]$.
 
 The IMU provides the following measurements at time $k$: the