DeepFloat: Resource-Efficient Dynamic Management of Vehicular Floating Content

Abstract

Opportunistic communications are expected to playa crucial role in enabling context-aware vehicular services. A widely investigated opportunistic communication paradigm for storing a piece of content probabilistically in a geographica larea is Floating Content (FC). A key issue in the practical deployment of FC is how to tune content replication and caching in a way which achieves a target performance (in terms of the mean fraction of users possessing the content in a given region of space) while minimizing the use of bandwidth and host memory. Fully distributed, distance-based approaches prove highly inefficient, and may not meet the performance target,while centralized, model-based approaches do not perform well in realistic, inhomogeneous settings. In this work, we present a data-driven centralized approach to resource-efficient, QoS-aware dynamic management of FC.We propose a Deep Learning strategy, which employs a Convolutional Neural Network (CNN) to capture the relationships between patterns of users mobility, of content diffusion and replication, and FC performance in terms of resource utilization and of content availability within a given area. Numerical evaluations show the effectiveness of our approach in deriving strategies which efficiently modulate the FC operation in space and effectively adapt to mobility pattern changes over time.

Full text

DeepFloat: Resource-Efficient Dynamic
Management of Vehicular Floating Content
Gaetano Manzo1,2, Sebastian Ot´
alora1, Marco Ajmone Marsan3, Torsten Braun2, Hung Nguyen 4, Gianluca Rizzo1
1University of Applied Science Western Switzerland − {name.surname}@hevs.ch
2University of Bern, Switzerland −braun@inf.unibe.ch
3IMDEA Networks Institute, Spain & Politecnico di Torino, Italy −ajmone@polito.it
4University of Adelaide, Australia −hung.nguyen@adelaide.edu.au
Abstract—Opportunistic communications are expected to play
a crucial role in enabling context-aware vehicular services. A
widely investigated opportunistic communication paradigm for
storing a piece of content probabilistically in a geographical
area is Floating Content (FC). A key issue in the practical
deployment of FC is how to tune content replication and caching
in a way which achieves a target performance (in terms of
the mean fraction of users possessing the content in a given
region of space) while minimizing the use of bandwidth and
host memory. Fully distributed, distance-based approaches prove
highly inefficient, and may not meet the performance target,
while centralized, model-based approaches do not perform well
in realistic, inhomogeneous settings.
In this work, we present a data-driven centralized approach
to resource-efficient, QoS-aware dynamic management of FC.
We propose a Deep Learning strategy, which employs a Con-
volutional Neural Network (CNN) to capture the relationships
between patterns of users mobility, of content diffusion and
replication, and FC performance in terms of resource utilization
and of content availability within a given area. Numerical
evaluations show the effectiveness of our approach in deriving
strategies which efficiently modulate the FC operation in space
and effectively adapt to mobility pattern changes over time.
I. INTRODUCTION
Vehicular communications are considered one of the key
vertical application domains for 5G [1], and a version of the
IEEE 802.11 standards is specifically devoted to vehicular en-
vironments [2]. The two standards serve different purposes. In-
deed, while some vehicular services are better served through
a cellular infrastructure, others are more naturally mapped
onto opportunistic communication schemes, especially those
that are based on vehicle location and require extremely low
latency.
Several opportunistic communication paradigms for the
local dissemination of contextualized information have been
proposed. Relevant examples are Floating Content [3]–[7],
or Hovering Information [8], [9]. They all aim at the local
dissemination of information to end users over a defined
geographic area (called Anchor Zone or AZ) through direct
terminal-to-terminal connectivity. In what follows, we will
refer to this approach as FC. FC aims to minimize the usage
of resources (such as bandwidth and memory) within the AZ,
while achieving a given performance (e.g., minimum message
availability). Typically, FC performance at a given time instant
is defined in terms of success ratio, i.e., the average fraction
of nodes with content at a given location (henceforth denoted
as Zone of Interest or ZOI). FC paradigms find applications
in several scenarios such as urban environments [10] and
university campus contexts [11], for location-aware services.
Performance studies of FC schemes can be categorized
into two classes. On one side, many works focus on content
persistence over time and propose heuristics for guaranteeing
content persistence within a predefined region, which usually
coincides with the AZ itself [3]. Such heuristics are often
tailored to a specific context (e.g., highways or pedestrians in
city squares [12]), and assume that a high likelihood of content
persistence is sufficient to successfully support applications
such as notifications of car accidents or traffic congestion. But
they are hard to generalize for other applications that need a
minimum amount of delivered contents within a given area.
Another class of proposals adopts a model-based approach
to FC dimensioning in terms of AZ. Solutions in [3]–[6],
[10]–[13] are based on the mean field approximation of the
dynamics of the population of users with content [14], and
on the assumptions of stationarity of the mobility patterns and
uniformity of user distribution in space. In these formulations,
the FC dimensioning problem boils down to finding the
minimum radius of a circular AZ area, which guarantees a
given minimum content lifetime, and/or a target success ratio.
However, models based on mean field approximation require a
large enough population of users to yield accurate results, and
this imposes a coarse granularity in the way in which resources
are partitioned and managed. This leads to suboptimal AZ
configurations, often including users who do not contribute
significantly to FC performance, and hence large resource
inefficiencies.
Such approaches do not lend themselves to generalizations
to more realistic settings characterized by heterogeneous user
distributions (and in particular to user clustering, which is very
common during traffic congestion or car accidents) and to
settings in which mobility patterns vary significantly over time.
For these reasons, how to efficiently allocate resources (i.e.,
memory, bandwidth, and infrastructure support for content
download) in an FC scheme with a realistic setting is still
an open issue to date.
arXiv:1906.07098v1 [cs.NI] 11 Jun 2019

In this work, we consider a setup where infrastructure
support to FC (in the form of a collection of data on user mo-
bility, and orchestration of content replication) is ubiquitous.
We propose DeepFloat, a data-based approach for dynamic
management of FC in vehicular scenarios that minimizes the
overall use of bandwidth and of user storage space, as well
as the amount of content seeding (i.e., content downloads)
performed by the infrastructure over the target content lifetime.
Our approach adopts a Convolutional Neural Network (CNN)
architecture, typically used in image processing, in order to
effectively capture the complex relations between elements
of a multidimensional set of system parameters, their impact
on content diffusion and persistence, and the FC success
ratio. Different from existing results, our approach applies to
arbitrary mobility scenarios and spatio-temporal patterns of
vehicle density distributions, by proactively modulating the
content replication and storage strategies over time. We use a
configurable cost function to flexibly account for heterogeneity
in vehicle populations (in terms of resource availability) and
different resource cost models, as well as for the variations
of these parameters over time and space. In particular, and
to the best of our knowledge, our approach is the first to
optimally adapt the balance between opportunistic replications
(i.e., vehicular-to-vehicular communication) and direct content
delivery by the infrastructure to the specific context.
The numerical assessment on a set of realistic scenarios
shows very high levels of accuracy of our system, while
substantially decreasing resource costs, with respect to model-
based strategies. In particular, in contexts of rapidly varying
mobility patterns (in which traditional approaches fail in
delivering the target performance and in ensuring content
persistence), we show its ability to proactively adapt content
replication and availability over time in order to minimize the
amount of infrastructure support to content replication in a
QoS aware manner.
The paper is structured as follows. Section II presents
the system model, followed by the problem formulation in
Section III. Our deep learning algorithm is illustrated in
Section IV and assessed numerically in Section V. Finally,
Section VI concludes the paper.
II. SYSTEM MOD EL
We consider a set of wireless nodes moving on the plane
according to a given road grid. These nodes are modeling
as vehicles and/or pedestrians in an urban area with their
user equipment (UE). We assume that the area is served by
a cellular access network, which collects data about users
mobility (floating user/car data [15]) for optimizing its op-
erations, consistently with the 5G (and beyond) paradigm. We
assume that each UE is equipped with one or more wireless
technologies for direct exchange of information with other
UEs via opportunistic communications, such as WiFi Direct,
Bluetooth, or device-to-device communication.
We assume that the surface of the road grid to be partitioned
into a set of L road links (henceforth simply denoted as links).
The number and the shape of each road link is based on the
tradeoff between computational complexity and accuracy of
our approach described in Section IV.
We assume that two nodes are in contact when they are
in range to exchange information. We assume that at each
time instant, the system maintains, for each road link, a list of
contacts, e.g., by having nodes periodically exchange beacons.
At any time, each node knows its exact position in space (i.e.,
its GPS coordinates).
A. DeepFloat FC operation
We assume that at some point in time a content object (e.g.,
a piece of text, a picture, or a short movie) is generated either
by the infrastructure (e.g., an advertisement from cellular or
WiFi users), or by mobile nodes themselves (e.g., a warning
about a road accident).
We consider a finite time interval, the floating period,
corresponding to the period during which the content should be
made available to users in a target portion of the road grid (the
Zone of Interest or ZOI). The typical duration of the floating
period ranges from half an hour up to a whole day. We assume
that the floating period is partitioned into Tintervals, each
of duration dt,t= 1, .., T . At the beginning of the floating
period, we assume that the infrastructure possesses some
aggregate information about node mobility. This assumption is
realistic as pedestrian and vehicular mobility patterns in urban
scenarios exhibit predictable (and often periodic) behaviors
over time [16].
An FC scheme is identified, for every road link l∈
{1, ..., L}and time interval t∈ {1, ..., T }, by the values of
content infectivity al,t, of recovery rate bl,t, and of seeding
ratio sl,t (with al,t, bl,t , sl,t ∈[0,1]).
At the beginning of each time interval t, the system makes
sure that in each link lthe percentage of nodes possessing
the given content at that time (henceforth denoted as content
availability at that time) is exactly equal to sl,t. At the
beginning of each interval t, if in each link lthe content
availability is less than sl,t, the system transfers the content
object from the infrastructure to a number of users sufficient to
achieve the target value of availability. If the initial availability
is larger than the seeding ratio for that interval, the system
triggers an appropriate number of users to drop the content.
In the FC scheme, at the beginning of the floating period,
parameters al,t, bl,t are made available to each user (e.g.,
broadcast to all users in the area through the cellular in-
frastructure). In each time interval t, when two nodes come
into contact, if the content object is present at only one
of the two nodes, the node with content object transfers it
to the other node with probability al,t, where lis the link
where the sender node resides. Such transfer is subject to
all limitations of the communication capacity between the
two nodes, which is influenced by propagation effects such
as fading and interference, among others. When a content
object is successfully transferred, the receiver (residing on link
l0) keeps it with probability bl0,t, and discards it otherwise.
Moreover, for all links l, whenever a node enters link lwith
content, it keeps it with probability bl,t and drops it otherwise.
2

In this paper, we consider the ideal case in which nodes
do not replicate content when it is not needed (i.e., when
both nodes in contact already possess the content). We also
assume that content exchanges are always unicast (one-to-
one). However, this FC scheme can be easily extended to
include the effects of multicasting and broadcasting.
The resulting array A={al,t, bl,t , sl,t}of dimensions L×T
completely describes an FC scheme. Each entry in Ais a 3-
tuple identifying the content seeding, replication and caching
strategies over all links during the whole floating period.
In each interval, if enough content replications take place
after the initial seeding, the content persists over time in a
portion of the road grid even when the nodes which received
the initial seeding have moved out of the road grid, or
discarded the content. We say in this case that the content
floats, i.e., it persists probabilistically in the area. This usually
happens for a duration which is determined by the mobility
patterns, and the choice of the FC parameters A, among others.
As a consequence, the primary performance metric for FC
is the mean content availability, i.e., the mean fraction of users
with content over a given time interval. Indeed, as mentioned,
the goal of an FC scheme is to make the content available
in a subset L0of links in the Zone of Interest (ZOI) of the
content, with a given mean availability, which we henceforth
name as success ratio and denote with αt. More specifically,
an FC scheme aims at ensuring that, in every interval t, the
condition
αt=Pl∈L0nc
l,t
Pl∈L0nl,t
≥α0(1)
is satisfied, where nl,t (nc
l,t) is the mean number of users
(resp. mean number of users with content) in the road link
lat interval t, and α0is the target minimum value of success
ratio. The value of α0, the choice of size, shape, and location
of the ZOI depend entirely on application requirements.
III. FORMULATION OF THE OPTIMIZATION PROBLEM
A cost function models the impact of an FC scheme on the
number of employed resources. For each link and time interval,
such function is given by the sum of three components.
The first component accounts for the utilization of storage
resources. For each link and time interval, it is measured by
the product of the mean total number of users with content
nc
l,t and the content size in bits, D.
The second cost component accounts for the mean amount
of UE communication resources used to exchange content and
it is proportional to the mean total number of UEs transmitting
the content at a given time instant. Specifically, let Ube the
set of communication technologies available (e.g., Bluetooth,
WIFi direct, D2D, LiFI), such component is given by
DX
l,t X
u∈U
θl,t,uγl,t,u
where γl,t,u is the mean total number of UEs transmitting
the content at time interval tand link lwith the u-th com-
munication technology. The coefficient θl,t,u can be used for
adjusting the cost of opportunistic content exchanges across
technologies, road links and time intervals, in order to take
into account the availability of resources in that technology
(e.g., due to utilization, to interference levels), and of the
way it varies over time. In general, γl,t,u depends on content
size, available bandwidth, and distance between transmitter
and receiver.
The third cost component takes into account the number
of content transfers, implemented directly through the infras-
tructure (e.g., through the cellular network), for seeding the
content in those links in which the availability at the beginning
of a given time interval is inferior to the minimum content
availability in the given road link. Hence, it is given by
DX
l,t
[sl,t −vl,t]+
(where ∀x, [x]+stands for max(x, 0)), and it is based on the
assumption that content suppression “comes for free”. vl,t is
the availability at the beginning of interval t, before content
seeding. It is given by
vl,t =(0t= 1
nc
l,t−1
nl,t−1t > 1(2)
An optimal FC scheme Ais hence a solution to the following
problem:
Problem 1 (FC resource optimization):
min
AX
l,t
dtDnc
l,t +βPu∈U θl,t,uγl,t,u 
PT
t=1 dt
+δD [sl,t −vl,t ]+
(3)
∀t, αt≥α0(4)
∀t, l 0≤al,t ≤1,0≤bl,t ≤1,0≤sl,t ≤1(5)
The overall cost function is given by storage, opportunistic
content exchange, and content seeding components. Those
components are weighted by the ratio between the interval du-
ration and the duration of the whole floating period, PT
t=1 dt.
Coefficients β, δ ≥0modulate the relative weight of the cost
components (memory, bandwidth, and seeding) on the overall
cost of an FC scheme. θl,t,u, which modulates technology and
interference costs, together with β, δ are the instruments by
which network operators can orchestrate the Floating Content
service. Among all the possible FC schemes A, a special role
is played by the one in which ∀l, t, al,t =bl,t =sl,t = 1.
With such choice, the content is replicated at every opportunity
over the whole road grid, and it is never dropped. This scheme,
which we call all-on, allocates all user resources in the system
to the FC communication scheme and guarantees that at the
beginning of each interval, each user possesses the content.
This implies that for each time interval, the all-on scheme
achieves the highest possible value of success ratio in the
considered scenario. Therefore, if the all-on scheme is not a
feasible solution of Problem 1, the problem is infeasible. When
this is the case, Problem 1 has to be reformulated by providing
3

more resources to FC, e.g., by considering a larger road grid,
by including a larger population of users, by changing α0or
communication technologies.
In general, Problem 1cannot be solved efficiently. Indeed, in
general ∀l, t, nc
l,t and γl,t,u depend on Aas well as on a large
set of parameters, in ways which are hard to accurately capture
analytically without strong and unrealistic assumptions, such
as stationarity of mobility patterns and uniformity of user
distributions. In the next section, we present our approach to
this intractable problem, based on Deep Learning.
IV. A DEEP LEARNING ALGORITHM FOR EFFICIENT FC
DIMENSIONING
In this section, we illustrate our DeepFloat approach to
finding an efficient solution of Problem 1. It is based on
Convolutional Neural Networks (CNN), a specialized type
of neural networks that has proven particularly effective in
modeling data with grid-like topology [17]. The key idea
of our approach is to exploit this CNN property in order
to accurately and efficiently model the correlation between
the spatial features of the road grid, and the way in which
the content spreads among links and persists over time. Our
approach is divided into three phases:
•Offline Set Up. The cellular infrastructure collects and
records UE mobility traces over time across the whole road
grid, on a regular basis. The training set is generated, by
associating the communication features to the measured
mobility features. Finally, the CNN is trained.
•Bootstrap. It starts when an application (which may or
may not reside within a UE) requests to the system, via
the cellular infrastructure, the activation of the FC scheme,
with a given ZOI, target success ratio, and floating period.
The system, given the forecasted resource availability and
mobility patterns, uses the trained CNN to compute in real
time a strategy A∗that achieves the target success ratio α0.
•Deployment. The system provides to all UEs present in
the considered road grid at the beginning of the floating
period (as well as to all those UEs which enter the road
grid during the floating period) the coefficients a∗
l,t and b∗
l,t,
which identify the replication and caching strategy derived
by the CNN. At the beginning of each time interval, in each
link, the content is seeded according to s∗
l,t. If necessary
(i.e., if, during the floating period, mobility patterns in the
road grid deviate significantly from the forecasted ones),
new forecasts are elaborated, and a new strategy A∗∗ is
adopted for the rest of the floating period.
In what follows, we describe in detail the three phases, and
the algorithms involved.
A. Offline Set Up
1) Data collection: As we have mentioned, the system
collects, on a regular basis, data about UEs mobility in the
grid. The system divides content lifetime into intervals to
provide a more efficient strategy that follows UEs mobility.
For each time interval, the system records the trajectories of
each UE in the grid.
Starting from these trajectories, for each interval i, and
every link l, the system computes a set of aggregate metrics
relative to node mobility and to wireless communications.
In what follows, these metrics are the average node speed,
the average number of nodes, and average contact duration.
In addition, it computes the mean number of nodes that, in
a given time instant, are in contact (i.e., able to exchange
beacons) with a node. Notice that nodes of different road
links can be in contact. These parameters have been chosen as
they are typically used as input in the main existing analytical
models of FC [4], [5]. Of course, however, different choices
are possible. In general, the choice of the set of parameters,
as well as the aggregation level in space (i.e., size and shape
of links) and time (i.e., duration of an interval d) affects the
degree of accuracy with which our model accounts for the
spatio-temporal features of the system, and specifically, for
the mobility patterns. The mobility parameters values of the
whole map, for each interval i, are stored in a mobility feature
array mi.
2) Label generation: In the next step, called label genera-
tion, a randomization procedure is applied. Specifically,
•for each mi, a set of Krandom FC schemes Ak,k=
1, ..., N are generated, each with its seeding strategy.
•For each iand each pair (mi,Ak), a simulation is per-
formed based on the random strategy Akand the user
trajectories during interval i. For each communication tech-
nology, these simulations assume a given model for the
channel capacity between nodes over time. The parameters
measured, for each link, are the mean number of nodes
with content at a given time instant nc
l,i, and the mean
number of nodes that are transmitting at a given time
instant for each communication technology, γl,i,u. These
parameters constitute the communication feature vector,ci.
Given a choice of ZOI, these parameters are the basis for
the computation of the success ratio, as well as for the
estimation of the resource utilization and of the associated
costs. However, note that the knowledge of the ZOI or of the
target success ratio is not required to produce the training
set. Hence this phase can be performed entirely offline.
For each k and i, with Pi= (mi,ci)we denote the
Link Features Vector associated with Ak(Table I). Finally,
each vector Piis normalized, in order to avoid numerical
issues in the subsequent phases of the process. The output
of the offline phase is thus a set of pairs (Pi,Ak), which
we denote with Π. The set Πis therefore enriched over time
with new elements derived by simulations. In addition to those
elements, the system includes in Πalso those elements derived
by directly measuring FC performance (specifically, those in
the communication feature vector) during the operation of the
FC schemes. Note that, while the presence of such measured
elements helps in achieving a high level of accuracy in real
scenarios, the contribution to Πof those elements derived
through simulations plays a key role in enabling the system to
appropriately configure an FC scheme in those scenarios which
occur with a relatively low probability (e.g., road accidents,
4

TABLE I: The set of link features for time interval iand link l.
Name Description
nc
l,i average number of nodes with content
nl,i average number of nodes
λl,i average number of nodes in contact
τl,i average contact duration [s]
νl,i average speed [m/s]
γl,i,u average number of nodes transmitting
in the u−th technology
sport events, or disasters).
B. Bootstrap and Deployment
The bootstrap phase starts when a request for FC service is
made. A request is characterized by an indication of the region
of the road grid which constitutes the ZOI, by the target value
of success ratio α0, and by the floating period.
In what follows we assume the floating period to be
composed by Tintervals, and let mtbe the mobility feature
array for the t−th interval, t= 1, ..., T . For all intervals
t, we assume the arrays mtto be perfectly known at the
time of the service request, possibly by forecasts based on
the mobility traces which the system collects on an ongoing
basis (we will later describe how our approach accounts for
forecast errors). 1. Given M=m1, ..., mTthe forecasted
mobility features for the floating period T, the definition of
the ZOI (i.e., the set L0), and the target success ratio, the
CNN computes the replication, caching, and seeding (i.e.,
the vehicles downloading the content from the infrastructure)
strategies A∗=A∗
1, ..., A∗
Tfor the whole floating period,
which achieve the target success ratio while trying to minimize
the resource cost as defined in (3). For these strategies, the
content infectivity (i.e., replication probability), as well as the
recovery rate (i.e., the caching probability) for each link and
time interval within the floating period are communicated to
all nodes in the scenario. In those cases in which, during
the operation of an FC scheme, better forecasts of mobility
become available, the system computes a new strategy A∗∗
for the remaining portion of the floating period, based on such
forecasts, and it injects the new values of content infectivity
and recovery rates to all the nodes.
C. CNN Model Architecture
In this section we describe the architecture of the CNN
adopted in our approach.
In the FC dimensioning problem —definition of the repli-
cation, caching and seeding strategies— the information about
proximity or relative position between links is not part of the
link features. In general, the correlations between features of
different road links are not due to proximity only. Rather, they
are also the result of wireless propagation effects and, most
importantly, of spatio-temporal patterns in node mobility (i.e.,
typical patterns of vehicular traffic).
1The mechanism by which these arrays are forecasted is out of the scope
of the present paper
Conv2D
Activation
MaxPooling
Conv2D
Activation
Flatten
Dense
Activation
Step 1 Step 3
Step 2
Conv2D
Activation
Dropout
Dense
Activation
Step 4
Conv2D
Activation
MaxPooling
Conv2D
Activation
Flatten
Dense
Activation
Step 1 Step 3
Step 2 Dropout
Dense
Activation
Step 4
Fig. 1: Outline of the architecture of our Convolutional Neural
Network. It takes as input the mobility features M, as well the set
of links composing the ZOI, L0, and the target success ratio α0. Its
output are the replication, caching, and seeding strategies A∗.
The Convolutional Neural Network (CNN) architecture en-
ables capturing both intra-link and inter-link relations between
features. The features extracted by our CNN are associated
with the correlation between road links. As a result, the
characteristics extracted from each layer of the CNN do not
consist only in local correlations, but also in long-spatio-
temporal relations between links, which have a strong impact
on FC performance.
The detailed structure of the CNN architecture consists of
four steps (Fig. 1). The goal of step 1 is to learn features
corresponding to basic patterns in the input (e.g., the strongest
local correlations, such as the spatio-temporal correlations
between adjacent road links and between the features of the
same road link.) [17]. Step 1 consists in a convolution layer,
which filters the input data (layer Conv2D in Fig. 1). The
size of the convolution kernel determines the extent of the
locality of the correlations captured. The output is fed into
an activation layer, followed by a discretization process (max
pooling layer). The purpose of these two layers is to filter
out information which will not be relevant for subsequent
parameter optimization.
In step 2, to improve model accuracy and to avoid over-
fitting, the same operations as step 1 is repeated. Note that,
in step 2 the max-polling layer is unnecessary given the data
structure. The number of these layers directly defines the depth
of the CNN and the level of complexity of the correlations able
to capture. Hence, their number depends on the characteris-
tics of the modeled system, but also on a tradeoff between
computational complexity and accuracy of the output. The
choice of kernel size in the convolutional layers (Conv2D)
should also be done according to such tradeoff. Indeed, a small
kernel forces the use of several convolution steps in order
to capture long-range correlations (i.e., correlation between
road links). A large kernel instead decreases the amount of
Conv2D layers required in step 2, but it also prevents the CNN
from accurately modeling complex correlations. The purpose
of step 3 is to reshape the output of the previous step, in
order to facilitate the subsequent computations. Its output is a
set of candidate strategies for orchestrating DeepFloat system.
Finally, step 4 performs an optimization over these candidate
strategies, by selecting out iteratively those who do not satisfy
the performance constraint on success ratio, and by giving as
output the one which minimizes the cost function in (3).
5

D. Complexity
A crucial aspect of the performance in our approach is
the computational load required by the three phases which
compose it. Given the potentially large amount of data to
be collected and pre-processed, the offline set up is the most
computationally intensive phase, being directly related to the
size of the set Π, as well as to the number of links and features
considered. However, given that all the elaborations in the
offline setup phase are executed offline, they do not impact
the delay with which a request for Floating Content is served.
The computations required by the bootstrap phase are
performed in real time at the moment a request is made. There-
fore, those effects heavily impact the overall FC performance,
particularly for those applications for which a quick and
effective deployment of the FC scheme is often necessary such
as car accident warning or medical emergency notification. In
order to enable a fast deployment in these scenarios, several
ways exist to decrease the computational load at the bootstrap
time within our approach. A possibility, as discussed, is to
adopt a low number of blocks at step 2 in the CNN.
V. NUMERICAL EVAL UATIO N
In this section, we numerically assess the performance of
DeepFloat, in terms of model accuracy and resource efficiency,
and we characterize the spatio-temporal strategies emerging
from it.
A. System Setup
We assumed nodes using a single wireless communication
technology, with a path loss attenuation of 3(typical of urban
scenarios), a SINR value equal to 5dB (which models setups
with high interference, common in crowded city centers, e.g.,
at the 2.4GHz frequency), content size D= 8 MB (to emulate
the exchange of multimedia content such as pictures and short
videos), bandwidth B= 1 MHz (as in Bluetooth) and the
channel capacity resulting from Shannon’s formula. Unless
otherwise specified, the FC parameters and those of the CNN
have been chosen conservatively in order to fit a worst case
scenario, in which applications require a fast bootstrap phase,
a relatively short content floating period (1h), and a high level
of target success ratio (0.9). Both coefficients βand δin the
cost function (3) are set to 1 (to have replication, caching, and
seeding equally contributing). As we show in this section, a
CNN architecture with a single block at step 2 and a Conv2D
kernel of size 3-by-3, together with the seeding approximation
described in Section IV-C have proven sufficient to achieve
a good level of accuracy while minimizing computational
load. In order to avoid over-fitting, a standard 10-fold cross-
validation is performed.
B. Manhattan Scenario
In the first set of experiments, we evaluated the performance
of our proposed approach in a scenario with synthetic mobility,
in order to perform, in a controlled scenario, a first-order
characterization of the strategies emerging from the DeepFloat
approach, and of their effectiveness and efficiency.
150 m road link ZOI
Fig. 2: Manhattan road grid composed of 5-by-4square blocks of
side 150 m, for a total of 31 road links and 12 road intersections.
We considered the road grid in Fig. 2. Nodes enter and exit
the grid from road links at the border, with an arrival rate
equal to 1.5nodes/s per road link, equal for all road links. At
every crossroad, a new direction is chosen at random between
straight, right and left. The floating period is composed by a
single interval of 1h duration. The training set was built as
follows. The road link features were measured over 1h, using
a sampling interval of 1s. In order to collect communication
features, we considered 103different random strategies. The
resulting simple randomization, in each scenario, contained
about 4·106pairs (P,A). Only for the model accuracy study,
we assume that mobility is predicted perfectly.
1) Model Accuracy: In order to evaluate the accuracy of
DeepFloat, Fig. 3 shows the values of F-score [18] versus the
training set size, for three different combinations of transmis-
sion radius and speed: r= 100 m and speed ν= 30 km/h;
r= 100 m and νuniformly distributed in (20,30) km/h —
emulating vehicles queuing; r= 500 m and ν= 30 km/h.
In each scenario, we also computed the performance of other
algorithms for multi-labeled classification problems, i.e., K-
Nearest Neighbor (KNN), Decision Tree (DT), and Random
Forest (RF). Note that for each algorithm, all scenarios are
tested using over 1.5·105registers of the test set, obtaining a
high sample population for the confidence interval.
Results in Fig. 3 show that DeepFloat substantially out-
performs the other algorithms in terms of accuracy. In par-
ticular, our approach achieves high accuracy even with small
training sets. This confirms that for the problem of efficient
dimensioning of an FC scheme, the specific ability of CNNs
to capture the inter-link correlations is key for satisfactory
accuracy performance. This is also suggested by the fact that,
by increasing the transmission radius (and hence the amount
of inter-link content exchanges), the relative accuracy of our
approach improves.
A key aspect of any learning approach to FC dimensioning
is that even approaches with high accuracy may produce
configurations that are unfeasible, i.e., which do not achieve
the target performance. Table II shows, for all considered
scenarios and algorithms, the probability that the output of
the learning processes does not satisfy constraint (4). For our
approach, the rejection probability is around 3%, an order of
magnitude lower than for the other considered approaches.
6

103104105106
Training set size
0.7
0.75
0.8
0.85
0.9
F-score
CNN
KNN
DT
RF
r=100m, v=30km/h
r=100m, v=[20,30]km/h
r=500m, v=30km/h
Fig. 3: F-score versus size of the training sample, for the DeepFloat
approach (DF), as well as for K-Nearest Neighbor (KNN), Decision
Tree (DT), and Random Forest (RF), for the three scenarios consid-
ered. All curves are with a 98% confidence interval of 7%
TABLE II: Rejection probability (i.e., probability of not achieving
the target success ratio) of the four approaches considered. Values are
with a 98% confidence interval within 6% of interval size. Training
set size: 1.5·105.
Method r= 100 mr= 100 mr= 500m
ν= 30 km/h ν= [20,30] km/h ν= 30 km/h
DF 0.029 0.031 0.032
KNN 0.274 0.287 0.284
RF 0.324 0.320 0.333
DT 0.347 0.345 0.349
Indeed, CNNs tend to decrease the false positive predictions
rather than the false negative ones. This is due to the fact
that, for any input, in the search for the strategy, which
minimizes resource costs, our approach always considers the
all-on configuration as a fallback option.
2) Resource Optimization: A crucial performance parame-
ter of our CNN approach is the percentage of resources saved
with respect to the all-on configuration. Such a configuration
is considered as a reference, as by hypothesis it is the only
one that is always feasible. Fig. 4 shows the percentage of
resources saved with respect to the trivial case of the all-on
configuration. They have been computed assuming that those
configurations that do not achieve the target success ratio have
the same cost of the all-on configuration. In Figure 4, we
present both the ideal case, where no predicted configuration
has been rejected, and the case, which takes into account the
rejected outputs. In the ideal case, by using the Decision Tree
technique we save up to 39% of the resources with respect
to the all-on configuration. However, not all configurations
resulting from DT satisfy the success ratio constraint. These
results suggest that our CNN architecture is able to shape
the content replication and storage strategies much more
efficiently than classical, non-deep ML techniques. Indeed, our
CNN is able to take advantage of the very high number of
parameters used for training, achieving high accuracy without
requiring a large training set or any geographical information
(such as a road map).
KNN DF RF DT
0
10
20
30
40
50
% resources saved over all-on configuration
all outputs
outputs not rejected
Fig. 4: Percentage of resources saved with respect to the all-
on strategy, for the four approaches considered. r= 100mand
ν= 30km/h. Values are with the confidence interval of 98% with
interval size of 3%.
RESIDENTIAL
CITY CENTER
Fig. 5: Luxembourg City map, with the borders of the City center
and Residential areas.
C. Luxembourg City Scenario
In order to perform a more realistic assessment of the
performance of our approach, we considered a second sce-
nario, shown in Fig. 5. It consists of two districts (denoted as
Residential and City center). In both districts, the street grid
and the measurement-based mobility traces for the rush traffic
hour (i.e., between 7 AM and 8 AM) were derived from [19].
These areas were selected because of their difference in vehicle
density distribution. In the city center, during the rush hour,
the vehicle density distribution is high and relatively uniform,
while in the residential district it is lower and more clustered.
In both scenarios, we considered as ZOI the closest link to
the center of the district. In both districts, we collected data
from 169 road links, using a sampling rate of 1sto accurately
capture mobility dynamics. The transmission radius is taken
to be 100m. Note that, Bluetooth Low Energy transmission
range is greater than 100m, whereas IEEE 802.11p (WAVE)
reaches 1000m[2].
We assumed a floating period composed by a single interval
of 1h, (adequate for a traffic warning e.g., about a temporarily
restricted area). By choosing a single interval, we focus on
the ability of the considered approaches to capture spatial
correlations in the input data.
In Table III, we list the test accuracy values obtained by
using 1.5·105registers for testing, and 106for training.
7

TABLE III: F-score for the Manhattan and Luxembourg scenarios,
with a training set size of 106, and test set size of 1.5·103.
Scenario DF KNN RF DT
Manhattan
r= 100m0.892 0.824 0.810 0.738
ν= 30km/h
r= 100m0.893 0.834 0.816 0.740
ν= [20,30]km/h
r= 500m0.894 0.819 0.802 0.736
ν= 30km/h
Lux.
city center 0.897 0.802 0.800 0.726
residential 0.896 0.798 0.801 0.722
TABLE IV: Rejection probability of the considered ML approaches.
Values are with a 98% confidence interval within 6% of interval size.
Luxembourg scenario.
Method City Center Residential
DF 0.018 0.021
KNN 0.224 0.229
RF 0.307 0.314
DT 0.332 0.341
While non-deep learning techniques have a worse accuracy
in the Luxembourg scenario than in the Manhattan one, the
contrary holds for DeepFloat, which is able to account for
spatial inhomogeneities in the mobility features.
The main existing analytical result which addresses the
issue of resource-optimal FC dimensioning in a road grid
was presented in [4]. It assumes that content is replicated and
stored within a circular area (called anchor zone - AZ) within
which content infectivity and recovery rate are both set to 1.
For the two districts of Luxembourg City, we have evaluated
the resource efficiency of our DeepFloat approach, as well
as of the KNN, Decision Tree, Random Forest approaches,
and Full-Infrastructure —a base station providing the content
object to the vehicles within the ZOI— in terms of percentage
of resources saved with respect to the approach from [4] (see
Figure 6). As in the Manhattan scenario, we have assumed
that the cost of rejected outputs —when the strategy does not
satisfy the performance target— is the same as the one of
the all-on configuration. As we can see, all ML approaches
outperform the analytical result. This is most likely due to the
fact that the analytical approach is based on an assumption
of uniformity of the main mobility features within the AZ.
Indeed, the gains over the analytical result are larger in the
less-uniformly distributed residential scenario than in the city
center.
Table IV provides the output rejection probability for the
Luxembourg scenarios. Similarly to the Manhattan scenario,
the lowest values are provided by the DeepFloat approach in
the city center (1.8%).
Fig. 8 shows the box-plots, with interquartile range (IQR)
and population median (in red), of the success ratio obtained
by simulating the strategies from the DeepFloat approach.
Outliers exit the Qx±1.5IQR, with x∈ {1,3}quartile
Q. As shown, our strategies allow the system to achieve in
both intervals the minimum value of success ratio (0.9).
As we have observed, strategies emerging from our Deep-
Float approach try to minimize, for each link, the difference
KNN CNN RF DT FI
-10
0
10
20
30
40
% resources saved over circular AZ
city center
residential
outputs not rejected
Fig. 6: %of resource saved with respect to the Circular Anchor Zone
approach [4], for our DeepFloat approach (DF), as well as for the K-
Nearest Neighbor (KNN), Decision Tree (DT), Random Forest (RF)
approaches, and Full-Infrastructure solution (FI) in the rush hour time
interval. Confidence interval of 98% with interval size of 3%.
0 500 1000 1500 2000 2500 3000 3500
time[s]
0
20
40
60
80
100
number of vehicle
t1 t2
#vehicles
mean #vehicles
Fig. 7: Vehicle density over time in the Luxembourg city center
scenario, for the first time interval t1= (7 AM - 7:30 AM), and
for the second, t2= (7:30 AM - 8 AM).
between the amount of seeding required at the beginning of
each interval, and the mean availability at the previous interval.
It produces strategies that tune the distribution of content in
an interval taking into account the seeding requirements at
the following interval. With the given choice of coefficients
(β, δ = 1), in the whole set of simulations no seeding support
has been necessary at the beginning of the second interval, in
any link. This suggests the effectiveness of our approach in
accurately controlling content availability in space and time
as a function of resource cost, and in particular, in decreasing
infrastructure support whenever its cost becomes higher than
the cost of using hosts resources.
Another evidence of the dynamic adaptation of the temporal
strategy to traffic conditions is shown in Figure 9 and 10.
To handle the increase in vehicle density while meeting the
performance target at t2, the system effectively adapts both
replication and caching areas. Figures 9(b) and 10(b) show
8

0.4 0.5 0.6 0.7 0.8 0.9 1
success ratio
t1
t2
interval
Fig. 8: Distribution of the measured success ratio derived with the
strategies from the DeepFloat approach, in time interval t1 = (7 AM
- 7:30 AM) and t2 = (7:30 AM - 8 AM), for the Luxembourg city
center scenario.
(a) (b)
Fig. 9: Luxembourg city center replication strategy at t1 = (7 AM -
7:30 AM) (a) and t2 = (7:30 AM - 8 AM) (b).
(a) (b)
Fig. 10: Luxembourg city center storage strategy t1 = (7 AM - 7:30
AM) (a) and t2 = (7:30 AM - 8 AM) (b).
how the system activates only those resources near the ZOI
that are necessary to achieve the performance target. On
the other hand, the system lowers replication and caching
probabilities where possible, to avoid unnecessary resource
usage (see the north-west areas in Figures 9 and 10).
D. Time considerations
In the Luxembourg scenario, characterized by 167 links,
and on a I7 desktop PC with 16 GB of RAM, the first two
phases of our approach (training set buildup and CNN training)
together took 3s for a training set of 103batches, and about 3
min for 106batches. Hence, at least in the considered scenario
(i.e., a whole city center), and with the given choices of links
and parameters, the computational load of the offline phase
does not require a large number of resources. As for the
bootstrap phase, in all our experiments, the computation of
the FC strategy by the CNN always took less than one second,
suggesting that, despite the large road grid and the large size
of the training set considered, our approach is fit for real-time
operation while keeping an excellent level of accuracy.
VI. CONCLUSIONS
In this work, we have outlined a Deep Learning approach
for efficient FC dimensioning, which exploits a Convolu-
tional Neural Network to modulate over time the parame-
ters governing an FC service. We have shown on realistic,
measurement-based scenarios that our approach efficiently
shapes the communication area and optimizes replication,
caching, and seeding strategies while achieving the desired
QoS targets over time. Among the possible extensions to our
approach, we intend to introduce temporal components in the
learning models, and to consider those to scenarios in which
mobility patterns are influenced by the spreading of content
(e.g., traffic jams notification).
VII. ACKNOWLEDGMENT
This research was supported/partially supported by the
Swiss National Science Foundation (SNSF, project CON-
TACT, no. 164205), by Hasler MOBNET, and by COST
RECODIS.
REFERENCES
[1] 5GPP, “White paper: 5G empowering vertical industries,” Tech. Rep.,
February 2016.
[2] D. Jiang and L. Delgrossi, “IEEE 802.11p: Towards an international
standard for wireless access in vehicular environments,” in IEEE VTC
Spring, May 2008.
[3] E. Hyytia, J. Virtamo, P. Lassila, J. Kangasharju, and J. Ott, “When
does content float? Characterizing availability of anchored information
in opportunistic content sharing,” IEEE INFOCOM, April 2011.
[4] G. Manzo, R. Soua, A. Di Maio, T. Engel, M. R. Palattella, and G. Rizzo,
“Coordination mechanisms for floating content in realistic vehicular
scenarios,” in IEEE MobiWorld, April 2017.
[5] G. Manzo, M. A. Marsan, and G. Rizzo, “Performance modeling of
vehicular floating content in urban settings,” in ITC 29, September 2017.
[6] J. M. Duarte, E. Kalogeiton, R. Soua, G. Manzo, M. R. Palattella, A. D.
Maio, T. Braun, T. Engel, L. A. Villas, and G. A. Rizzo, “A multi-
pronged approach to adaptive and context aware content dissemination
in VANETs,” Mobile Networks and Applications, January 2017.
[7] G. Manzo, M. A. Marsan, and G. A. Rizzo, “Analytical models of
floating content in a vehicular urban environment,” Ad Hoc Networks,
vol. 88, pp. 65–80, 2019.
[8] A. A. V. Castro, G. D. M. Serugendo, and D. Konstantas, “Hovering
information–self-organizing information that finds its own storage,” in
Autonomic Communication. Springer, 2009, pp. 111–145.
[9] T. Nikolovski and R. W. Pazzi, “A Lightweight and Efficient Approach
(LEA) for Hovering Information Protocols,” in DIVANet. New York,
NY, USA: ACM, 2017, pp. 31–38.
[10] J. Kangasharju, J. Ott, and O. Karkulahti, “Floating content: Information
availability in urban environments,” in IEEE PERCOM, March 2010.
[11] S. Ali, G. Rizzo, V. Mancuso, and M. A. Marsan, “Persistence and avail-
ability of floating content in a campus environment,” IEEE INFOCOM,
March 2015.
[12] M. S. Desta, E. Hyytia, J. Ott, and J. Kangasharju, “Characterizing
content sharing properties for mobile users in open city squares,” in
10th WONS, March 2013.
[13] S. Ali, G. Rizzo, M. A. Marsan, and V. Mancuso, “Impact of mobility on
the performance of context-aware applications using floating content,” in
International Conference on Context-Aware Systems and Applications.
Springer, 2013, pp. 198–208.
[14] J. L. Boudec, D. McDonald, and J. Mundinger, “A generic mean field
convergence result for systems of interacting objects,” in QEST, Sep.
2007, pp. 3–18.
[15] A. Dhamaniya and S. Chandra, “Speed prediction models for urban ar-
terials under mixed traffic conditions,” Procedia - Social and Behavioral
Sciences, vol. 104, January 2013.
[16] L. Ghouti, T. R. Sheltami, and K. S. Alutaibi, “Mobility prediction in
mobile ad hoc networks using extreme learning machines,” Procedia
Computer Science, vol. 19, March 2013.
[17] I. Goodfellow, Y. Bengio, and A. Courville, Deep learning. MIT press
Cambridge, March 2016.
[18] Y. Sasaki, “The truth of the F-measure,” Teach Tutor Mater, January
2007.
[19] L. Codeca, R. Frank, and T. Engel, “Luxembourg SUMO Traffic (LuST)
Scenario: 24 hours of mobility for vehicular networking research,” in
VNC, December 2015.
9