Textbook Mobile Networks and Management Jiankun Hu Ebook All Chapter PDF

Mobile Networks and Management
Jiankun Hu
Visit to download the full and correct content document:
https://textbookfull.com/product/mobile-networks-and-management-jiankun-hu/
More products digital (pdf, epub, mobi) instant
download maybe you interests ...
Data and Energy Integrated Communication Networks Jie

Hu
https://textbookfull.com/product/data-and-energy-integrated-
communication-networks-jie-hu/
5G Green Mobile Communication Networks Xiaohu Ge
https://textbookfull.com/product/5g-green-mobile-communication-
networks-xiaohu-ge/
Principles of Chinese Management 1st Edition Haibo Hu
https://textbookfull.com/product/principles-of-chinese-
management-1st-edition-haibo-hu/
From GSM to LTE-Advanced Pro and 5G: An Introduction to

Mobile Networks and Mobile Broadband 4th Edition Martin
Sauter
https://textbookfull.com/product/from-gsm-to-lte-advanced-pro-
and-5g-an-introduction-to-mobile-networks-and-mobile-
broadband-4th-edition-martin-sauter/
Evolutionary Computing and Mobile Sustainable Networks
Proceedings of ICECMSN 2020 V. Suma
https://textbookfull.com/product/evolutionary-computing-and-
mobile-sustainable-networks-proceedings-of-icecmsn-2020-v-suma/
Mobile Networks and Management 5th International

Conference MONAMI 2013 Cork Ireland September 23 25
2013 Revised Selected Papers 1st Edition Minji Kim
https://textbookfull.com/product/mobile-networks-and-
management-5th-international-conference-monami-2013-cork-ireland-
september-23-25-2013-revised-selected-papers-1st-edition-minji-
kim/
Mobile Networks and Management 8th International

Conference MONAMI 2016 Abu Dhabi United Arab Emirates
October 23 24 2016 Proceedings 1st Edition Ramón Agüero
https://textbookfull.com/product/mobile-networks-and-
management-8th-international-conference-monami-2016-abu-dhabi-
united-arab-emirates-october-23-24-2016-proceedings-1st-edition-
ramon-aguero/
Machine Learning and Cognitive Computing for Mobile

Communications and Wireless Networks 1st Edition
Krishna Kant Singh
https://textbookfull.com/product/machine-learning-and-cognitive-
computing-for-mobile-communications-and-wireless-networks-1st-
edition-krishna-kant-singh/
Technologies and Eco innovation towards Sustainability

II Eco Design Assessment and Management Allen H. Hu
https://textbookfull.com/product/technologies-and-eco-innovation-
towards-sustainability-ii-eco-design-assessment-and-management-
allen-h-hu/
Jiankun Hu · Ibrahim Khalil
Zahir Tari · Sheng Wen (Eds.)
235
Mobile Networks
and Management
9th International Conference, MONAMI 2017
Melbourne, Australia, December 13–15, 2017
Proceedings
123
Lecture Notes of the Institute
for Computer Sciences, Social Informatics
and Telecommunications Engineering 235
Editorial Board
Ozgur Akan
Middle East Technical University, Ankara, Turkey
Paolo Bellavista
University of Bologna, Bologna, Italy
Jiannong Cao
Hong Kong Polytechnic University, Hong Kong, Hong Kong
Geoffrey Coulson
Lancaster University, Lancaster, UK
Falko Dressler
University of Erlangen, Erlangen, Germany
Domenico Ferrari
Università Cattolica Piacenza, Piacenza, Italy
Mario Gerla
UCLA, Los Angeles, USA
Hisashi Kobayashi
Princeton University, Princeton, USA
Sergio Palazzo
University of Catania, Catania, Italy
Sartaj Sahni
University of Florida, Florida, USA
Xuemin Sherman Shen
University of Waterloo, Waterloo, Canada
Mircea Stan
University of Virginia, Charlottesville, USA
Jia Xiaohua
City University of Hong Kong, Kowloon, Hong Kong
Albert Y. Zomaya
University of Sydney, Sydney, Australia
More information about this series at http://www.springer.com/series/8197
Jiankun Hu Ibrahim Khalil
•
Zahir Tari Sheng Wen (Eds.)

•
Mobile Networks
and Management
9th International Conference, MONAMI 2017
Melbourne, Australia, December 13–15, 2017
Proceedings
123
Editors
Jiankun Hu Zahir Tari
SEIT, UNSW Canberra RMIT University
Canberra Melbourne, VIC
Australia Australia
Ibrahim Khalil Sheng Wen
RMIT University Swinburne University of Technology
Melbourne, VIC Hawthorne, Melbourne, VIC
Australia Australia
ISSN 1867-8211 ISSN 1867-822X (electronic)

Lecture Notes of the Institute for Computer Sciences, Social Informatics
and Telecommunications Engineering
ISBN 978-3-319-90774-1 ISBN 978-3-319-90775-8 (eBook)
https://doi.org/10.1007/978-3-319-90775-8
Library of Congress Control Number: 2018941835
© ICST Institute for Computer Sciences, Social Informatics and Telecommunications Engineering 2018
This work is subject to copyright. All rights are reserved by the Publisher, whether the whole or part of the
material is concerned, specifically the rights of translation, reprinting, reuse of illustrations, recitation,
broadcasting, reproduction on microfilms or in any other physical way, and transmission or information
storage and retrieval, electronic adaptation, computer software, or by similar or dissimilar methodology now
known or hereafter developed.
The use of general descriptive names, registered names, trademarks, service marks, etc. in this publication
does not imply, even in the absence of a specific statement, that such names are exempt from the relevant
protective laws and regulations and therefore free for general use.
The publisher, the authors and the editors are safe to assume that the advice and information in this book are
believed to be true and accurate at the date of publication. Neither the publisher nor the authors or the editors
give a warranty, express or implied, with respect to the material contained herein or for any errors or
omissions that may have been made. The publisher remains neutral with regard to jurisdictional claims in
published maps and institutional affiliations.
Printed on acid-free paper
This Springer imprint is published by the registered company Springer International Publishing AG
part of Springer Nature
The registered company address is: Gewerbestrasse 11, 6330 Cham, Switzerland
Preface
We are delighted to introduce the proceedings of the 9th European Alliance for
Innovation (EAI) International Conference on Mobile Networks and Management
(MONAMI 2017). This conference has brought together researchers, developers, and
practitioners from around the world who are leveraging and developing mobile com-
puting, wireless networking, and management.
The technical program of MONAMI 2017 consisted of 30 full papers in the main
conference track. The conference had a special session on Trust, Privacy and Security
in Internet of Things (IoT) and Cloud. Aside from the high-quality technical paper
presentations, the technical program also featured five keynote speeches. These key-
note speakers were Prof. Wanlei Zhou from Deakin University Australia, Prof. Jie Lu
from University of Technology Sydney Australia, Prof. Joe Dong from UNSW Sydney
Australia, Prof. Yongsheng Gao from Griffith University Australia, and Dr. Shui Yu
from Deakin University.
Coordination with the steering chairs, Dr. Imrich Chlamtac from Create-Net, Italy,
and Prof. Jiankun Hu and Conference General Co-Chair Prof. Yang Xiang from
Swinburne University Australia from UNSW Canberra, Australia, was essential for the
success of the conference. We sincerely appreciate their constant support and guidance.
It was also a great pleasure to work with such an excellent Organizing Committee who
worked hard in organizing and supporting the conference. In particular, the Technical
Program Committee, led by our TPC co-chairs, Dr. Zahir Tari and Dr. Ibrahim Khalil
from RMIT Australia, completed the peer-review process for the technical papers and
compiled a high-quality technical program. We are also grateful to other chairs
including the publication chair, Dr. Sheng Wen from Swinburne University of Tech-
nology, the PC members for their support, and all the authors who submitted their
papers to the MONAMI 2017 conference and workshops.
We strongly believe that MONAMI 2017 provided a good forum for all researchers,
developers, and practitioners to discuss all scientific and technological aspects that are
relevant to mobile networks and management. We also expect that future MONAMI
conferences will continue being successful and stimulating, as indicated by the con-
tributions presented in this volume.
April 2018 Jiankun Hu

Ibrahim Khalil
Zahir Tari
Conference Organization
Steering Committee
Imrich Chlamtac (Chair) EAI/CREATE-NET, Italy
Jiankun Hu UNSW Canberra, Australia
Organizing Committee
General Co-chairs
Jiankun Hu UNSW Canberra, Australia
Xiang Yang Deakin University, Melbourne, Australia
Technical Program Committee Chairs and Co-chairs

Zahir Tari RMIT University, Melbourne, Australia
Ibrahim Khalil RMIT University, Melbourne, Australia
Web Chair
Naveen Chilamkurti Latrobe University, Melbourne, Australia
Publicity and Social Media Chair/Co-chair

Chan Yeob Yeun Khalifa University, UAE
Workshops Chair
Sponsorship and Exhibits Chair

Jiankun Hu UNSW, Australia
Publications Chair
Sheng Wen Deakin University, Australia
Panels Chair
Tutorials Chair
VIII Conference Organization
Demos Chair
Posters and PhD Track Chair

Local Chair
Honorary Co-chairs
Albert Zomaya The University of Sydney, Australia
Mohammed Atiquzzaman University of Oklahoma, USA
Hsiao-Hua Chen National Cheng Kung University, Tainan
Ernesto Damiani Khalifa University, UAE
Conference Manager
Alzbeta Mackova EAI (European Alliance for Innovation)
Technical Program Committee

Neda Aboutorab UNSW, Australia
Adnan Al-Anbuky AUT, New Zealand
Abderrahim Benslimane University of Avignon, France
Abbas Bradai University of Poitiers, France
Claudia Canali UNI More, Italy
Chi Chen Chinese Academic of Science, China
Naveen Chilamkurti La Trobe University, Australia
Baldomero Coll-Perales UWICORE, Israel
Antonio De Domenico CEA, France
Mounia Fredj ENSIAS, France
Ying Guo Central South University, China
Song Guo The Hong Kong Polytechnic University, SAR China
Slimane Hammoudi Université d’Angers, France
Xinyi Huang Fujian Normal University, China
Qin Jing Shandong University, China
Ibrahim Khalil RMIT University, Australia
Weifa Liang Australia National University, Australia
Qin Liu Hunan University, China
Constandinos Cyprus
Mavromoustakis
Emad Eldin Mohamed Swinburne University, Australia
Miklos Molnar University of Montpellier 2, France
Conference Organization IX
Paul Pan Unitec, New Zealand

Yi Qian University of Nebraska – Lincoln, USA
Abderrezak Rachedi Université Paris-Est Marne-la-Vallée, France
Javier Rubio-Loyola Mexico Teléfono, Mexico
Henning Sanneck Nokia, Germany
Zhili Sun University of Surrey, UK
Zahir Tari RMIT University, Australia
Abdelkamel Tari University of Bejaia, Algeria
Craig Valli ECU, Australia
Anna Maria Vegni University of Rome, Italy
Sheng Wen Deakin University, Australia
Qianghong Wu Beihang University, China
Chan Yeob Yeun Khalifa University, UAE
Arkady Zaslavsky Data61, Australia
Sherali Zeadally University of Kentucky, USA
RongBo Zhu South-Central University for Nationalities, China
Contents
Offloading of Fog Data Networks with Network Coded Cooperative

D2D Communications . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1
Ben Quinton and Neda Aboutorab
Persistent vs Service IDs in Android: Session Fingerprinting from Apps . . . . 14

Efthimios Alepis and Constantinos Patsakis
Towards Developing Network Forensic Mechanism for Botnet Activities

in the IoT Based on Machine Learning Techniques . . . . . . . . . . . . . . . . . . . 30
Nickolaos Koroniotis, Nour Moustafa, Elena Sitnikova, and Jill Slay
Performance Comparison of Distributed Pattern Matching Algorithms

on Hadoop MapReduce Framework. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 45
C. P. Sona and Jaison Paul Mulerikkal
Robust Fingerprint Matching Based on Convolutional Neural Networks. . . . . 56

Yanming Zhu, Xuefei Yin, and Jiankun Hu
A Personalized Multi-keyword Ranked Search Method Over

Encrypted Cloud Data . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 66
Xue Tian, Peisong Shen, Tengfei Yang, Chi Chen, and Jiankun Hu
Application of Fuzzy Comprehensive Evaluation Method for Reservoir

Well Logging Interpretation While Drilling . . . . . . . . . . . . . . . . . . . . . . . . 79
Zhaohua Zhou, Shi Shi, Shunan Ma, and Jing Fu
Factor Effects for Routing in a Delay-Tolerant Wireless Sensor Network

for Lake Environment Monitoring . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 87
Rizza T. Loquias, Nestor Michael C. Tiglao,
Jhoanna Rhodette I. Pedrasa, and Joel Joseph S. Marciano
Estimating Public Opinion in Social Media Content

Using Aspect-Based Opinion Mining. . . . . . . . . . . . . . . . . . . . . . . . . . . . . 101
Yen Hong Tran and Quang Nhat Tran
An Approach for Host-Based Intrusion Detection System Design

Using Convolutional Neural Network . . . . . . . . . . . . . . . . . . . . . . . . . . . . 116
Nam Nhat Tran, Ruhul Sarker, and Jiankun Hu
A Robust Contactless Fingerprint Enhancement Algorithm . . . . . . . . . . . . . . 127

Xuefei Yin, Yanming Zhu, and Jiankun Hu
XII Contents
Designing Anomaly Detection System for Cloud Servers by Frequency

Domain Features of System Call Identifiers and Machine Learning . . . . . . . . 137
Waqas Haider, Jiankun Hu, and Nour Moustafa
A Variant of BLS Signature Scheme with Tight Security Reduction . . . . . . . 150

Tiong-Sik Ng, Syh-Yuan Tan, and Ji-Jian Chin
Quantum Authentication Scheme Based on Fingerprint-Encoded

Graph States. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 164
Fei Li, Ying Guo, and Jiankun Hu
Cooperative Information Security/Cybersecurity Curriculum Development . . . 178

Abdelaziz Bouras, Houssem Gasmi, and Fadi Ghemri
An Energy Saving Mechanism Based on Vacation Queuing Theory

in Data Center Networks . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 188
Emna Baccour, Ala Gouissem, Sebti Foufou, Ridha Hamila, Zahir Tari,
and Albert Y. Zomaya
Homomorphic Evaluation of Database Queries . . . . . . . . . . . . . . . . . . . . . . 203

Hamid Usefi and Sudharaka Palamakumbura
A Cache-Aware Congestion Control for Reliable Transport

in Wireless Sensor Networks . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 217
Melchizedek I. Alipio and Nestor Michael C. Tiglao
A New Lightweight Mutual Authentication Protocol to Secure

Real Time Tracking of Radioactive Sources . . . . . . . . . . . . . . . . . . . . . . . . 231
Mouza Ahmed Bani Shemaili, Chan Yeob Yeun,
Mohamed Jamal Zemerly, Khalid Mubarak, Hyun Ku Yeun,
Yoon Seok Chang, Basim Zafar, Mohammed Simsim,
Yasir Salih, and Gaemyoung Lee
Fog Computing as a Critical Link Between a Central Cloud

and IoT in Support of Fast Discovery of New Hydrocarbon Reservoirs . . . . . 247
Andrzej M. Goscinski, Zahir Tari, Izzatdin A. Aziz,
and Eidah J. Alzahrani
Performance Assessment of Cloud Migrations from Network

and Application Point of View . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 262
Lukas Iffländer, Christopher Metter, Florian Wamser, Phuoc Tran-Gia,
and Samuel Kounev
A Cloud Service Enhanced Method Supporting

Context-Aware Applications . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 277
Zifan Liu, Qing Cai, Song Wang, Xiaolong Xu,
Wanchun Dou, and Shui Yu
Contents XIII
Application of 3D Delaunay Triangulation in Fingerprint

Authentication System . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 291
Wencheng Yang, Guanglou Zheng, Ahmed Ibrahim,
Junaid Chaudhry, Song Wang, Jiankun Hu, and Craig Valli
The Public Verifiability of Public Key Encryption with Keyword Search . . . . 299
Binrui Zhu, Jiameng Sun, Jing Qin, and Jixin Ma
Malicious Bitcoin Transaction Tracing Using Incidence

Relation Clustering . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 313
Baokun Zheng, Liehuang Zhu, Meng Shen, Xiaojiang Du, Jing Yang,
Feng Gao, Yandong Li, Chuan Zhang, Sheng Liu, and Shu Yin
Cryptanalysis of Salsa and ChaCha: Revisited . . . . . . . . . . . . . . . . . . . . . . 324

Kakumani K. C. Deepthi and Kunwar Singh
CloudShare: Towards a Cost-Efficient and Privacy-Preserving

Alliance Cloud Using Permissioned Blockchains. . . . . . . . . . . . . . . . . . . . . 339
Yandong Li, Liehuang Zhu, Meng Shen, Feng Gao, Baokun Zheng,
Xiaojiang Du, Sheng Liu, and Shu Yin
Probability Risk Identification Based Intrusion Detection System

for SCADA Systems . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 353
Thomas Marsden, Nour Moustafa, Elena Sitnikova, and Gideon Creech
Anonymizing k-NN Classification on MapReduce . . . . . . . . . . . . . . . . . . . . 364

Sibghat Ullah Bazai, Julian Jang-Jaccard, and Ruili Wang
A Cancellable Ranking Based Hashing Method for Fingerprint

Template Protection. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 378
Zhe Jin, Jung Yeon Hwang, Soohyung Kim, Sangrae Cho,
Yen-Lung Lai, and Andrew Beng Jin Teoh
Author Index . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 391

Offloading of Fog Data Networks
with Network Coded Cooperative D2D
Communications
Ben Quinton(B) and Neda Aboutorab
University of New South Wales, Campbell, ACT 2612, Australia

quintonbj@gmail.com, n.aboutorab@unsw.edu.au
Abstract. Future fog data networks are expected to be assisted by users

cooperation and coding schemes. Given the finite I/O access bandwidth
of the drives in the data servers and the explosive increase in the end
users’ demand for download of the content from the servers, in this paper,
we consider the implementation of instantly decodable network coding
(IDNC) in full-duplex device-to-device (D2D) enabled cooperative dis-
tributed data networks. In particular, this paper is concerned with opti-
mizing D2D communications with efficiently coded transmissions such
that we offload traffic from the expensive backhaul of network servers.
Previous works implementing IDNC have not focused on a cooperative
architecture, therefore a new theoretical-graph model is proposed and the
optimal problem formulation is presented. However, as the optimal solu-
tion suffers from the intractability of being NP-hard, it is not suitable for
real-time communications. The complexity of the problem is addressed
by presenting a greedy heuristic algorithm used over the proposed graph
model. The paper shows that by implementing IDNC in a full-duplex
cooperative D2D network model significant reduction in the number of
downloads required from the servers can be achieved, which will result
in saving valuable servers’ resources.
Keywords: Instantly decodable network coding · IoT · Full-duplex

Backhaul offloading · Cooperative D2D communications
Fog storage networks
1 Introduction
With the modern advancements of wireless communications, wireless networks
have seen an explosion in data traffic over the past decade [6]. This rapid demand
for more data is largely attributed to video and multimedia streaming, where it is
expected that three-fourths of data traffic will be consumed by video [6]. To com-
pound this further, it is expected that the next generation of wireless networks
will encapsulate the new paradigm of the internet of things (IoT). This concept
c ICST Institute for Computer Sciences, Social Informatics and Telecommunications Engineering 2018
J. Hu et al. (Eds.): MONAMI 2017, LNICST 235, pp. 1–13, 2018.
https://doi.org/10.1007/978-3-319-90775-8_1
2 B. Quinton and N. Aboutorab
moves to further integrate more and more devices into communication networks,
where it is foreseen that the IoT will add a further 50 billion heterogeneous wire-
less devices by 2020 [6]. Consequently, this growing demand puts further pressure
on data networks, where the offloading of the servers becomes an increasingly
important problem.
This ever-growing demand for real-time data, where users expect to maintain
their quality-of-experience (QoE) has led to much research to address the data
networks backhaul problem. Multiple areas of research have shown promising
methods to deal with this problem, one such option is to distribute the data
closer to the users with improved redundancy [2,7,8,11]. The idea of distributing
resources to the edge of a network is known as “fog” networking [5]. Motivated
by very high temporal correlation among the “popular” content demanded by
end-users, it is expected that the proactive (i.e. without users requests) diffusing
of such popular content from its storage and transmission clouds behind the
backhaul, and caching it in a “fog” of low-cost storage units close to the end-
users to serve the requests to download this content could largely improve the
network performance and service quality. Using this approach not only the users’
requests can be immediately and efficiently addressed, but also the access to the
backhaul could be significantly offloaded [10,12].
In addition to distributing the data, with the rapid increase in the number
of wireless devices, there are more and more devices in each others proximity.
Such “geographically close” wireless devices form an autonomous local network
over which the users can communicate and exchange files without contacting
the backhaul servers. Such scenario may occur for instance when co-workers are
using their tablets to share and update files stored in the cloud (e.g. Dropbox),
or when users, in the subway or a mall, are interested in watching the same
popular video. Under such scenario, the benefits of communicating over a local
network can be utilized not only to reduce the users’ download time but also
offload the backhaul of the data network (i.e., minimizing the download from it).
Furthermore, network coding (NC), initially introduced in [1], can help in
offloading of the backhaul servers in the considered distributed cooperative data
network scenario by maximizing the number of served users in one transmission,
thus maximizing the backhaul offloading. Although NC was originally imple-
mented at the network layer, more attractive application was found at the data
link layer where there is coded combinations of files to improve throughput.
Multiple areas of study have focused on various types of network coding, where
this paper will focus on opportunistic network coding (ONC) [14], in particular
instantly decodable network coding (IDNC) [13]. This technique has recently
gained much attention due to its instant decodability (as the name suggests)
by using a simple XOR operation that results in reducing the computational
complexity of the decoding at the end users. It also provides a significant benefit
to real-time communications, where studies in [3,7,13] show through a heuristic
algorithm that utilizing IDNC results in shown significant performance improve-
ments over uncoded transmissions in both centralized point-to-multipoint (PMP)
and decentralized network settings.
Offloading of Fog Data Networks 3
Although much work has focused on implementing IDNC in various network

models to address the above problem, the studies have focused on centralized
PMP and distributed architectures. Furthermore, there has currently been no
work to consider implementing IDNC in a cooperative setting where there is a
focus on reducing the number of downloads required from the network servers.
A network coded cooperative D2D-enabled architecture is considered in this
paper, as it provides an attractive solution to offload the servers in a distributed
data networks. Therefore, in this paper we aim to address the following ques-
tion: How should we encode files amongst users in a cooperative D2D-enabled
transmission, such that the remaining requests from the users (if any) can be
delivered (using IDNC) with a minimum number of transmissions from the net-
work servers? Although much research has focused on implementing IDNC in
a PMP setting and even a distributed architecture with multiple servers, these
approaches cannot be directly applied to a full-duplex cooperative network with
the current graph modelling technique, therefore there is a need to develop a
new model.
To address the question above, we first need to model the problem with a
new graphical representation, namely the IDNC graph with induced subgraphs.
The new graph representation is developed due to limitations of the conventional
graphical representation when we wish to implement full-duplex communications
into the system model. With the new graph modelling of the system, the optimal
solution is formulated and shown to be NP-hard and not applicable for real-time
applications [9]. The paper then proposes an online greedy heuristic algorithm
that employs a maximum weighted vertex search over the new graph model.
Simulation results show that the proposed algorithm when employed over the
new graph model in a full-duplex D2D-enabled environment significantly outper-
forms the conventional uncooperative IDNC approach in reducing the downloads
required from the servers of distributed data networks.
In this paper, we first present the system model and mathematical notation
in Sect. 2. In Sect. 3, we formulate the problem, where a motivating example is
first presented then followed by a mathematical optimal solution to the problem
utilizing the new graph model. As the solution is found to be NP-hard, we then
present the proposed greedy heuristic scheme in Sect. 4, followed by simulation
results and discussion in Sect. 5. Lastly, this paper is concluded in Sect. 6.
2 System Model
A distributed wireless data network model is considered in this paper and

is illustrated in Fig. 1. In this model, there is a set of Nu users defined as
U = {u1 , . . . , uNu }. In the system model, the assumption is made that all users
are capable of full-duplex communications. The users will request to receive one
file in the current time epoch, from a library of files defined as F = f1 , . . . , fNf
with Nf files that are collectively stored at the servers. The servers are defined
in the set S = {s1 , . . . , sNs } with Ns servers. All servers are assumed to have
full coverage, where the users in the coverage area are denoted by U(si ) and
Fig. 1. An illustration depicting our system model for a distributed storage network,
showing two servers, six users, and two proximity based wireless networks, where users
can conduct cooperative D2D communications.
must satisfy U(si ) ∩ U = U. The model shows a distributed setting where the
users are in coverage of multiple servers. Also in the model, multiple proximity
networks (possibly Wi-Fi or LAN) are shown.
The proximity regions are defined
as the proximity set P = p1 , . . . , pNp with Np proximity-enabled D2D com-
munication networks. The proximity networks contain a subset of the users in U,
defined as U(pi ), that is the users in the coverage area of the proximity-enabled
network pi . It is assumed that there is no overlap of the users in each proximity
set, that is, the users in each proximity network that are “geographically close”
can communicate locally but not outside this network.
In our model, we assume the users have received some files in the initial
transmission phase1 . That is, a user ui has partially downloaded some of the files
from a transmitted frame which constitutes the users Has set Hui . Furthermore,
the remaining files wanted by user ui in the frame form the user’s Wants set,
denoted as Wui . Similarly, the servers will store a subset of the files in F, however
the union of all files at the servers should contain the complete set of F (with
possible repetition). Here, a server’s Has set is defined as Hsi . It is assumed
that the servers will maintain a global knowledge of the system state during
the initial transmissions, that is the users will respond with positive/negative
feedback depending if they receive their files successfully or not. At completion
of this phase the system will move into the recovery transmission phase.
IDNC can now be utilized to exploit users’ side data to optimize the trans-
missions in the current time epoch. In the recovery transmission, we assume an
erasure free channel, where the different users and servers will operate on orthog-
onal channels. It is also assumed that the servers have an unlimited capacity,
1
This first phase of the transmission is known as the initial transmission phase. During
the initial transmission the servers will attempt to serve all files to the users in the
network. However, some users will have received only a portion of the files requested
due to channel erasure.
Fig. 2. Conventional IDNC approach from [2], where there is a total of three downloads
required for the optimal solution.
Fig. 3. A D2D-enabled approach showing the potential for offloading servers, where
only one download is required.
such that after all cooperative D2D communications all requests remaining will
be served by the server in the current time epoch. Here, the main goal is to opti-
mize the selection of the files for network coding for the users and the servers,
where priority is given to the cooperative D2D communications, such that we
reduce the amount of downloads required from the servers.
3 Problem Formulation
3.1 Motivating Example
If we consider the system model depicted in Fig. 1, it can be shown by example

that by finding the optimal solution, that is, to solve our question defined earlier,
there are numerous allocations of coded transmissions that can lead to different
results. For way of a motivating example, we present two different allocations:
in the first solution, we assume a conventional IDNC approach without the

cooperative D2D enabled transmissions. In the second solution, we show the
scenario for a network with cooperative D2D enabled communications.
Solution 1. In this solution (depicted in Fig. 2), we employ the method used in
[2], this solution utilizes a conventional IDNC approach without D2D-enabled
cooperation. One possible optimal solution using this method is:
– s1 transmits f1 to u1 and f2 ⊕ f3 to u2 and u3 .

– s2 transmits f1 ⊕ f4 to u4 and u5 .
This scenario results in consuming three downloads from the servers.
Solution 2. In this solution (depicted in Fig. 3), we show a scenario where a

cooperative D2D setting is incorporated. One possible optimal solution is:
– s1 transmits f1 to u1 .
– u1 transmits f2 ⊕ f3 to u2 and u3 .
– u4 transmits f1 to u5 and u5 transmits f4 to u1 .
In the second solution, it can be seen that we only need one download from one of
the servers, while no download is required from the other server. This approach
shows that even in a small network setting, there is a download reduction of two
thirds of the previous solution, freeing up valuable servers’ resources.
Although much work has focused on implementing IDNC in various network
models, the graph-theoretical modellings used in these cases are limited in a
cooperative full-duplex environment. In previous approaches, the graph models
incorporated assume that there is a clear differentiation between the sender and
the receiver. In our setting, we remove this restriction and allow users the ability
of full-duplex communications, therefore the existing IDNC graph models are
not appropriate and there is a need for a new model.
3.2 Graph-Based Solution
To be able to formulate the optimal solution to the above problem, stated in

Sect. 3, we will propose a new IDNC graph that represents coding opportunities.
The IDNC graph when formulated, will represent all the possible files that can be
XORed together to create a network coded transmission that can be decoded by
the targeted end users. To form the model, we first define the graphs of interest
in our system model as follows: Graph G = G1 , . . . , GNp with the subgraph Gi
representing each discrete D2D network, as well as the graph Ψ = {Ψ1 , . . . , ΨNs }
that is representing all servers, where the subgraph Ψi represents each individual
server.
To construct each of the graphs previously mentioned, we proceed as follows:
Generate Vertex Set. Vertices are generated from a server and user perspec-
tive under the two conditions:
– Generate a vertex set for every server si in S that is represented

in the
subgraph Ψi , generating the vertices vijk , ∀si ∈ S and fk ∈ Hsi ∩ Wuj . The
vertices of the subgraph are defined as Ψi(ijk) .
– Generate a vertex set for every user ui ∈ pn in U that is represented
inthe
subgraph Gn , generating the vertices vijk , ∀ui ∈ U and fk ∈ Hui ∩ Wuj on
the conditions ui = uj and both ui , uj ∈ pn . The vertices of the subgraph are
defined as Gn(ijk) .
Generate Coding Opportunity Edges. In each individual subgraph in G

and Ψ , we connect two vertices vijk and vlmn with an edge if they satisfy one of
the following two conditions:
– fk = fn , uj = um and ui = ul if in G (or si = sl if in Ψ ), meaning the two
requested files are the same, and these files are requested by two different
users.
– fn ∈ Huj and fk ∈ Hum , representing a potential coding opportunity, so that
when fn and fk are XORed both users can successfully decode and retrieve
their requested file.
In the formulation so far we have incorporated graphs that represent coding

opportunities from a user/server viewpoint. To further create a global awareness,
we have to incorporate induced subgraphs (subgraphs of G and H) that will
represent the transmission conflicts (subgraph K) and a subgraph to ensure only
one transmission per user is permitted in the current time epoch (subgraph L).
In Fig. 4, we depict the implementation of IDNC with the prescribed the-
oretical graph model for the example shown in Fig. 1. In our case, we show
independently, firstly in subgraphs Gi , the IDNC subgraphs for each individual
proximity network. While in subgraphs Ψi , we show the potential coded trans-
missions in maximal cliques2 for each server si . In the graph model shown in
Fig. 4, it is clear that there is no interconnection between the graphs of G and Ψ
(no edges connecting vertices). Therefore, we introduce the induced subgraphs
approach to allow us to represent particular conditions that need to be accounted
for in our network setting. We will introduce the graph K, a set of subgraphs
that ensures conflict free transmissions. Additionally, we introduce the graph L
that contains a set of subgraphs that ensure users in a proximity network will
not transmit more than once in the current time epoch.
Generate Induced Subgraphs. The two induced subgraphs as described are

generated as follows:
2
A clique is a sub-set of the graph, where every distinct pair of vertices in the induced
subgraph are pairwise adjacent. A maximal clique is one that cannot be a subset of
a larger clique [4].
Fig. 4. A visualization of the IDNC induced subgraph methodology proposed for sys-
tem model shown in Fig. 1. The figure shows coding opportunities represented by edges,
transmission conflicts represented in subgraphs Kj and limitation of one transmission
per user represented in subgraph L3 (that is, u3 may only transmit to either u2 or u1
in the current time epoch).
– First, we define the set of induced subgraphs K = {K1 , . . . , KD } as a subset

of both graphs G and Ψ , where the subgraphs may contain the null-set of
either G or Ψ but not both. To generate the subgraph Kj , each vertex vijk in
both G and Ψ will form a member of the subgraph Kj for every vertex that
has the same user uj and file fk .
– Similarly, we define the set of induced subgraphs L = {L1 , . . . , LE } as a
subset of both graphs G and Ψ . A subgraph Li is formed for any two vertices
vijk and vlmn , where ui = ul but uj = um or fk = fn .
3.3 The Proposed Optimal Problem Formulation

In order to formulate the optimal solution we need to select the combination of
disjoint maximal cliques from G such that when these vertices are removed, thus
removing the union of the associated subgraph from Ψ , we reduce the remaining
maximal cliques of Ψ . That is, we wish to minimize the number of maximal
cliques in Ψ , which is equivalent to minimizing the number of downloads from
the servers. Therefore, we can either find an expression to minimize the maximal
cliques of Ψ , or equivalently we can minimize the number of maximal independent
sets3 of the complementary graph of Ψ , which we refer to as Ψ . The minimum
number of maximal cliques in a graph can be found by finding the chromatic
number of a complementary graph [4]. Therefore, the optimal solution can be
expressed in mathematically in (1)
3
An independent set is a set of vertices in a graph, no two of which are adjacent.
A maximal independent set is an independent set that is not a subset of any other
independent set [4].
⎡ ⎛⎛ ⎞ ⎞⎤
C
D E
min X ⎣Ψ \ ⎝⎝Ii Kj ⎠ Ii Lk ⎠⎦ (1)
I1 ,... ,IC
i=1 j=1 k=1
subject to Ii ⊆ G, Ii ∩ Ψ = ∅
∃ul ∈ U where F(Ii ) ⊆ Hul

where is the disjoint set union operator and the first constraint ensures that
the independent set Ii is selected only from the vertices that belong to the
graph G. This is to ensure the selected coded file combinations that the users
serve reduces the chromatic number of the servers graph (optimal solution),
as the chromatic number of the remaining graph Ψ is equal to the number of
downloads required from the servers. The second constraint shown in (1) ensures
that for all files selected in the independent set Ii , denoted by F(Ii ), there exists
a user that posses the files and can XOR them. If this is not satisfied then the
coded transmission cannot be sent and the conditions in (1) are not met.
Solving for the optimal solution that has been presented requires that we
determine the chromatic number of a graph (equivalent to finding all maximal
cliques). It is well known that determining the chromatic number of a graph is
proven to be NP-hard [9]. This is further compounded by the fact that we not
only need to calculate the chromatic number of one graph, but we need to find
the selection of independent sets such that we minimize the chromatic number of
the remaining subgraph. Hence, the optimal solution is not applicable for online
and real-time communications. Therefore, we will propose a heuristic scheme in
the following section to solve sub-optimally.
4 The Proposed Greedy Heuristic Algorithm

In this section, we propose a greedy heuristic approach that can be solved in
real-time and efficiently reduce the number of downloads from the servers. The
fall back of a greedy heuristic scheme is that it does not in fact guarantee a
global optimum, although we hope that this scheme will on average, give a good
approximation to it.
An attractive feature of the graph-based formulation proposed in Sect. 3.2
is that we can directly apply a maximal weighted vertex search under a greedy
policy on the graph model. With the graph already established from the problem
formulation, we can carry out the maximum clique listing, using a maximum
weighted vertex search as follows:
Firstly, we associate a weight to each vertex in the graph G, each vertex’s
weight is proportional to δijk which is the degree4 of vijk . The weight is calculated
in (2),

wijk = δi j k (2)
vi j k ∈N (vijk )
4
The degree of a vertex (δ) in a graph is equal to the number of incident edges to
that vertex [4].
Algorithm 1. Algorithm for Maximum Weight Vertex Search

Require: :
Initialisation :
– Construct Graphs G, Ψ , K and L

– Gs ← {G}
– Γ ←∅
1: repeat
2: ∀vijk ∈ Gs : Compute wijk using (2)
∗
3: vijk ⇐= argmaxvijk wijk , ∀vijk ∈ Gs
∗
4: add vijk to Γ
∗
5: Gs ← Gs ∪ vijk
6: until Gs = ∅
7: return the clique listing Γ
where N (vijk ) is the set of adjacent vertices to vijk . Therefore, each vertex
in graph G will have a large weighting if it has a large number of adjacent
vertices, which themselves have a large number of adjacent vertices. The search
will then select the vertex with the largest weighting, or between those with the
same largest weight with equal probability. The algorithm then removes all non-
adjacent vertices to vijk , and then checks if the vertex vijk belongs to a subgraph
Kj or Li and will remove all other vertices that are a member of either subgraph.
Secondly, the algorithm will then update all weights in G before selecting the
next (if any) adjacent vertex in graph G that forms a clique with all previously
selected vertices. The algorithm then continues to iterate these steps until no
more vertices can be added to the clique. Finally, once a maximal clique listing is
found and removed, we iterate the whole procedure until no more vertices are left
in the graph G. The steps of algorithm described is summarized in Algorithm 1.
At this stage, the algorithm has removed all possible D2D cooperations avail-
able in hopes to minimize the amount of downloads from the servers. Therefore,
we now need to serve the remaining vertices in graph Ψ that were not served
locally from D2D cooperation. We now conduct the exact same procedure on
the remaining vertices in graph Ψ , where each maximal clique represents one
download from a server and continue until all vertices are removed from the
graph. Once all vertices have been removed from the graph Ψ the system will
have reached absorption, that is, all users will have received the file in their
Wants sets.
5 Simulation Results
In this section, we present our simulation results for the proposed algorithm
in a cooperative D2D setting in comparison with a uncooperative decentralized
conflict free IDNC approach that was incorporated in [2]. In both cases, the aim
of the approaches is to reduce the number downloads from the servers.
20
Uncooperative IDNC
18 Cooperative IDNC
Avg. Number of Server Downloads

16
14
12
10
0
10 20 30 40 50 60 70 80 90 100
Number of Users
Fig. 5. The average number of downloads required from the servers as a function of
the number of users.
In the simulations, each user is interested in receiving one file and has two
files already received and stored in the Has set (when fixed), where the recovery
downloads are to be completed in one time epoch. We assume each users’ Has
and Wants sets to be determined probabilistically, with uniform distribution
over all files in the library. In all simulations, there are two servers available
with total coverage of all users in the network, while in the cooperative model
we consider a dual network where the users are split evenly between the two
proximity networks p1 and p2 (similar to Fig. 1).
Firstly in Fig. 5, we show the average number of downloads required from the
servers for a fixed number of files in the transmission frame of Nf = 20, as a func-
tion of the number of users Nu . The result shows that for the algorithm imple-
mented for a cooperative D2D-enabled setting, as the number of users increase
the average number of server downloads tends to monotonically decrease. Intu-
itively, this result is expected as more users in the network will result in a greater
likelihood that the users can serve themselves independently from the servers, as
the collective Has set of the users in the network will cover the files in the frame
F. Additionally, it can be seen that in comparison to a conventional uncoopera-
tive conflict-free IDNC approach, as the network size increases there is significant
improvement, where we see an improvement of approximately 550% with only
20 devices in the network setting. Furthermore, approximately no downloads
from the servers are required as the number of users approach 60 in this network
setting, that is, 30 users in each D2D-enabled network.
Now if we consider fixing the number of users to 20, while varying the amount
of files per transmission frame, we can see the results in Fig. 6 for cooperative
versus uncooperative IDNC transmission schemes. In both cases, it can be seen
as the number of files increase, both schemes show a similar increase on the
20
Uncooperative IDNC
18 Cooperative IDNC
Avg. Number of Server Downloads

16
14
12
10
0
10 20 30 40 50 60 70 80 90 100
Number of Files
Fig. 6. The average number of downloads required from the servers as a function of
the number of files.
number of downloads required from the servers. Although the two schemes tend
to converge if we consider an asymptotic limit, the cooperative scheme still shows
reasonable improvement of approximately 50% for up to 100 files. Again, this
result is expected as increasing the number of files in a frame reduces the poten-
tial to leverage a coded transmission. Additionally, as the number of files increase
the likelihood of a users ability to diffuse the wanted packets is diminished. Nev-
ertheless, the cooperative approach still shows significant ability to reduce the
number of downloads required from the network servers.
6 Conclusion
In this paper, we investigated the problem of offloading the expensive backhaul of

data network servers through a network coded cooperative D2D network model.
The problem was formulated using the IDNC induced subgraph model, where
the optimal solution requires finding maximal cliques of multiple graphs. In the
problem formulation it is found that an optimal solution is intractable and not
solvable in real-time, therefore a greedy heuristic algorithm is employed using
a maximum weighted vertex search approach. The paper utilizes the proposed
subgraph model again in the heuristic approach, where the simulation results
showed a significant improvement over the conventional method that incorpo-
rates IDNC in a distributed fashion without D2D enabled cooperation.
References
1. Ahlswede, R., Cai, N., Li, S.-Y.R., Yeung, R.W.: Network information flow. IEEE
Trans. Inf. Theory 46(4), 1204–1216 (2000)
2. Al-Habob, A.A., Sorour, S., Aboutorab, N., Sadeghi, P.: Conflict free network
coding for distributed storage networks. In: 2015 IEEE International Conference
on Communications (ICC), pp. 5517–5522. IEEE (2015)
3. Baran, P.: On distributed communications networks. IEEE Trans. Commun. Syst.
12(1), 1–9 (1964)
4. Bondy, J.A., Murty, U.S.R.: Graph Theory with Applications, vol. 290. Macmillan,
London (1976)
5. Bonomi, F., Milito, R., Zhu, J., Addepalli, S.: Fog computing and its role in the
internet of things. In: Proceedings of 1st Edition of the MCC Workshop on Mobile
Cloud Computing, pp. 13–16. ACM (2012)
6. Cisco: Cisco visual networking index: global mobile data traffic forecast update.
Technical report, February 2016
7. Dimakis, A.G., Godfrey, P.B., Wu, Y., Wainwright, M.J., Ramchandran, K.: Net-
work coding for distributed storage systems. IEEE Trans. Inf. Theory 56(9), 4539–
4551 (2010)
8. Dimakis, A.G., Ramachandran, K., Wu, Y., Suh, C.: A survey on network codes
for distributed storage. Proc. IEEE 99(3), 476–489 (2011)
9. Edwards, C.S., Elphick, C.H.: Lower bounds for the clique and the chromatic
numbers of a graph. Discret. Appl. Math. 5(1), 51–64 (1983)
10. Golrezaei, N., Molisch, A., Dimakis, A.G., Caire, G.: Femtocaching and device-
to-device collaboration: a new architecture for wireless video distribution. IEEE
Commun. Mag. 51(4), 142–149 (2013)
11. Papailiopoulos, D.S., Luo, J., Dimakis, A.G., Huang, C., Li, J.: Simple regenerating
codes: network coding for cloud storage. In: 2012 Proceedings of IEEE INFOCOM,
pp. 2801–2805. IEEE (2012)
12. Shanmugam, K., Golrezaei, N., Dimakis, A.G., Molisch, A., Caire, G.: Fem-
toCaching: wireless content delivery through distributed caching helpers. IEEE
Trans. Inf. Theory 59(12), 8402–8413 (2013)
13. Sorour, S., Valaee, S.: On minimizing broadcast completion delay for instantly
decodable network coding. In: 2010 IEEE International Conference on Communi-
cations (ICC), pp. 1–5. IEEE (2010)
14. Sorour, S., Valaee, S.: An adaptive network coded retransmission scheme for single-
hop wireless multicast broadcast services. IEEE/ACM Trans. Netw. (TON) 19(3),
869–878 (2011)
Persistent vs Service IDs in Android:
Session Fingerprinting from Apps
Efthimios Alepis(B) and Constantinos Patsakis
Department of Informatics, University of Piraeus,

80, Karaoli & Dimitriou, 18534 Piraeus, Greece
talepis@unipi.gr
Abstract. Android has conquered the mobile market, reaching a mar-

ket share above 85%. The post Lollipop versions have introduced radical
changes in the platform, significantly improving the provided security
and privacy of the users. Nonetheless, the platform offers several fea-
tures that can be exploited to fingerprint users. Of specific interest are
the fingerprinting capabilities which do not request any dangerous per-
mission from the user, therefore they can be silently shipped with any
application without the user being able to trace them, let alone blocking
them. Having Android AOSP as our baseline we discuss various such
methods and their applicability.
1 Introduction
Mobile devices, especially smartphones have become an indispensable part of our
daily lives, as a big part of our communications and daily activities is processed
and monitored by them. One of the main reasons for their wide adoption is that
they have a plethora of embedded sensors that allow them to understand their
context and adapt accordingly. For instance, through luminosity and proximity
sensors as well as accelerometers, mobile phones may adapt the UI to fit better to
user expectations. Moreover, thanks to GPS, mobile devices are location aware
enabling them to render content according to the spacial restrictions significantly
improving the user recommendations.
Data mining and data profiling can be used in order to collect valuable infor-
mation about a particular user or group of users, in order to generate a profile
[12], which can be further used by companies to gain profit. As stated in [14],
this kind of information, namely user profiling, is valuable also for advertisers
who want to target ads to their users and in return, advertisers may pay more
money to their hosting applications’ developers. Building user profiles requires,
as the authors state, sensitive privileges in terms of permissions, such as Internet
access, location, or even retrieving installed applications in a user’s device [14].
To this end, we may infer that collecting and successfully fusing user data from
more than one service can create even better and more complete user profiles,
which will consequently translate in higher monetization. Looking back in 2009,
it was quite clear that:
c ICST Institute for Computer Sciences, Social Informatics and Telecommunications Engineering 2018
J. Hu et al. (Eds.): MONAMI 2017, LNICST 235, pp. 14–29, 2018.
https://doi.org/10.1007/978-3-319-90775-8_2
Session Fingerprinting from Apps 15
“Once an individual has been assigned a unique index number, it is possible

to accurately retrieve data across numerous databases and build a picture
of that individual’s life that was not authorised in the original valid consent
for data collection” [19].
The above has been realised by tech giants. For instance quoting a statement
from Google’s current privacy policy [10]:
“We may combine personal information from one service with information,
including personal information, from other Google services - for example
to make it easier to share things with people you know. Depending on your
account settings, your activity on other sites and apps may be associated
with your personal information in order to improve Google’s services and
the ads delivered by Google”
In order to “enable” data fusion from different sources and services, one could
argue that unique identifiers should be either implicitly or explicitly present. In
particular, a value describing a quantity of some valuable variable might be
useless if it is not accompanied by a unique identifier that would allow us to
track its source. Contrariwise, identifiers coming from different services that
are matched, may act as a “bridge” between these services to combine their
corresponding datasets and integrate them.
During the last decade relevant surveys have revealed that the majority of
both iOS and Android apps were transmitting the phone’s unique ID and the
user’s location to advertisers. These findings are confirmed by “The Haystack
Project” [11] which revealed that nearly 70% of all Android apps leak personal
data to third-party services such as analytics services and ad networks [20].
All this wealth of information apart from the benign usage for the user benefit
has been a constant target by companies who wish to monetize it, mainly through
targeted advertisement. The recent advances in big data and data mining have
enabled the extraction of information from theoretically diverse data, leading to
the revealing of a lot of sensitive data through data fusion. To this end, many
fingerprinting techniques have been introduced in order to link data flows to
specific individuals. Apparently, since Android in currently the prevailing mobile
platform, most companies are targeting it with their apps, under the freemium
model, harvesting user data to monetize them. There is even a common saying
in the privacy community suggesting that “If you’re not paying for the product,
you are the product”. To this end, If users are not paying for an app, they are
usually selling their profiles (with or without their knowledge/consent) to an ad
network, which will use their unique identifiers to track and target them.
In view of the above and targeting at improving the OSes privacy, the new
coming Android O, makes a number of privacy-related changes to the platform,
many of which are related to how unique identifiers are handled by the system [9],
and in particular aiming to help provide user control over the use of identifiers
[2]. One of the most important improvements concern “limiting the use of device-
scoped identifiers that are not resettable”.
16 E. Alepis and C. Patsakis
Clearly, during app environment of Android hardware identifiers can greatly

facilitate companies’ attempt to deanonymise users. Therefore, Google has been
gradually introducing specific measures to restrict them. In fact, Google decided
to introduce further restrictions in one more identifier; not hardware based,
namely Android ID. While this attempt might seem noble, in this work we show
that these restrictions do not actually serve the purpose, while apps may be
deprived of many persistent identifiers, ephemeral IDs can actually serve their
purposes in the attempt to deanonymize their users and fuse their data.
1.1 Main Contributions

The main contribution of this work is to study user fingerprinting from mobile
apps, which correspondingly and as already discussed can lead to user profiling.
In this regard we assume that apps and software services which profile users, want
to correlate the information that each one of them has collected to fine tune their
profiles. Conceptually, in order to provide proof for our claims, we explore all
the available communication channels that mobile apps could utilize in Android
AOSP in order to identify that specific profiles are installed in the same device.
Furthermore, we suggest that the existing underlying mechanism is able to func-
tion without using unique hardware identifiers, nor dangerous permissions which
could alert the user, or demand further user interaction. While many of the com-
munication mechanisms are apparent, e.g. inter-process communication, there is
a wide misconception that the upcoming changes in Android O will eradicate
many such issues. Therefore, initially we analyse each possible communication
channel and ways it can be used to transfer the needed information. Moreover,
by providing statistical evidence we discuss when these changes are expected to
be noticed by the average user. Finally, despite the touted changes in Android
ID, we detail new methods that can provide permanent cross-app IDs that can
be collected even if an app is uninstalled.
1.2 Organisation of This Work

The rest of this work is organized as follows. In the next section we present
the related work. Section 3 provides the problem setting and our basic assump-
tions. Then, Sect. 4 presents all the available communication channels. In Sect. 5
illustrates possible temporary and ephemeral identifiers that can be used to link
users between applications using Android AOSP as our reference point. Finally,
the article concludes with some discussion about our research findings and pro-
viding statistics regarding the adoption timeline of the expected anonymization
mechanisms of Android O.
2 Related Work
Unique identifiers have been used for a long time and facilitate many tasks in
modern database systems as they allow us to perform record linkage between
Session Fingerprinting from Apps 17
different entities and extract the necessary information and thus knowledge from
the corresponding database tables. The most typical example of a unique iden-
tifier is the Social Security Number, which allows us to distinguish two people
from each other. However, in the digital era, unique identifiers can be considered
hardware identifiers like the MAC address of the network card, or a set of prop-
erties such as browser fingerprints which consist among others of the browser
version, OS, fonts, and browser plugins.
In the Android ecosystem there is a plethora of unique identifiers which have
so far been extensively exploited by advertisement companies to track users and
their interests as ad libraries have become more and more greedy and rogue
[3,18] while apps may deliberately leak information to the ads [4,21] harness-
ing arbitrary amounts of users’ sensitive information directly or indirectly [5].
A key role in this procedure is the use of unique identifiers [17]. Acknowledg-
ing this situation, Google initially introduced some recommendation guidelines
for the proper use of unique identifiers in Android [1]. Then, Google gradually
started requesting more permissions from the apps to allow them access to these
identifiers. For instance, a typical unique identifier for mobile phones are IMEI
and IMSI, however, after Marshmallow, the user has to grant the dangerous
READ PHONE STATE permission to an app to access them. While many users may
ignore app permissions [8], for many others it works as an obstacle, forcing many
companies to comply with the rule.
Despite the ads, apps may collaborate in order to perform malicious acts
which independently would not be allowed to perform. Orthacker et al. [15]
study this problem from the aspect of permissions. In this regard, the malicious
apps which are installed in the victim’s device may result in “possessing” and
correspondingly using dangerous permissions that other normal apps do not. The
concept is that the user would not allow camera and microphone permission to a
single app. However, since the permissions are requested by two apps which are
seemingly independent, the permissions are “spread” so the user grants them,
yet an adversary controls both of them getting access to the desired resource.
Contrary to Orthacker et al. we do not aim to resources, but access to information
that the user would not share to one specific app to prevent his profiling.
In Nougat, the current stable version of Android, Google prohibited unpriv-
ileged access to even more hardware identifiers, such as the MAC address of
the WiFi card, by restricting access to /proc. While the latter measure creates
many issues with applications targeting towards security and privacy services
as Google has not provided any permission so far to access this information,
undoubtedly, it leaves little space to adversaries to exploit.
3 Temporary and Ephemeral Identifiers

3.1 Problem Setting
While the aforementioned issues have led to the introduction of many changes
to Android, improving the security and privacy of the OS, in terms of user
fingerprinting, from the side of apps, we argue that little has been achieved.
Another random document with
no related content on Scribd:
method is therefore not applicable in such cases, but is useful in
water examination.
Electrolytic Reactions.—Solutions of lead are easily
electrolyzed, and give a precipitate of lead at the cathode;
simultaneously the peroxide is produced at the anode, and the
reaction is acid. In nitric acid solutions Riche pointed out that the
whole of the lead is carried to the anode, and this is the reaction
made use of in the determination of lead present in the urine (see p.
172).
The presence of copper in an electrolyte regulates the
precipitation of lead oxide, copper alone being deposited at the
cathode, and at the same time the presence of a small quantity of
copper promotes the destruction of organic materials.
REFERENCES.
[1] Pliny: lxxxiii., 11, N.c.v.
[2] Stockhusen: De Litharg. Fumo, etc. Goslar, 1656.
[3] Tronchin: De Colica Pictonum. 1758.
[4] John Hunter: Observations of Diseases of the Army in Jamaica.
London, 1788.
[5] Meillère, G.: Le Saturnisme. Paris, 1903.
[6] Bisserie: Bull. Soc. Pharmacol. May, 1900.
[7] Houston: Local Government Board Annual Report, 1901-02,
supplement, vol. ii.
CHAPTER II
ÆTIOLOGY
Lead poisoning of industrial origin rarely occurs in the acute form.
Practically all cases coming under the notice of either appointed surgeons,
certifying surgeons, or even in the wards of general hospitals, are of the
subacute or chronic type. There is no reason to suppose that lead
compounds are used more frequently by the workers in lead industries as
abortifacients than by other persons.
The compounds of lead which are responsible for poisoning in industrial
processes are for the most part the hydrated carbonate, or white lead, and
the oxides of lead, whilst a comparatively small number of cases owe their
origin to compounds, such as chromates and chlorides.
The poisonous nature of any lead compound from an industrial point of
view is proportional to (1) the size of the ultimate particles of the substance
manufactured, and therefore the ease with which such particles are capable
of dissemination in the air; and (2) the solubility of the particles in the normal
fluids of the body, such as the saliva, pharyngeal and tracheal and bronchial
mucus, etc., and the fluids of the stomach and intestine. An instance of the
variation in size of the particles of lead compounds used industrially is the
difference between ground lead silicate (fritted lead) used in the potteries,
and the size of the particles of ordinary white or “raw” lead. By micrometric
measurements one of us [K. W. G.[1]] found the average size of the particles
of fritt to be ten times that of the white lead particles. Further, direct
experiment made with equal masses of the two compounds in such a manner
that the rate of settling of the dust arising could be directly compared in a
beam of parallel light showed presence of dust in the white lead chamber
fifteen minutes after the fritt chamber was entirely clear. It is found as a matter
of practice that where dust is especially created, and where it is difficult to
remove such dust by exhaust fans, the greatest incidence of lead poisoning
occurs. The association of dusty processes and incidence of lead poisoning is
discussed in relation to the various trades in Chapters XV. to XVII. Fume and
vapour given off from the molten metal or compounds, such as chlorides
(tinning), are only a special case of dust.
The channels through which lead or its compounds may gain entrance to
the animal body are theoretically three in number:
1. Respiratory tract.
2. Gastro-intestinal.
3. Cutaneous.
For many years most authorities have held that industrial poisoning by
means of compounds of lead takes place directly through the alimentary
canal, and that the poison is conveyed to the mouth mainly by unwashed
hands, by food contaminated with lead dust, and by lead dust suspended in
the air becoming deposited upon the mucous membrane of the mouth and
pharynx, and then swallowed. As evidence that lead dust is swallowed, the
classical symptom of colic in lead poisoning has been adduced, on the
supposition, in the absence of any experimental proof, that the lead
swallowed acted as an irritant on the gastro-intestinal canal, thus causing
colic, and, on absorption from the canal, setting up other general symptoms.
Much of the early treatment of lead poisoning is based upon this assumption,
and the administration of sulphuric acid lemonade and the exhibition of
sulphate of magnesia and other similar compounds as treatment is further
evidence of the view that the poisoning was considered primarily intestinal.
One of the chief objections to this view, apart from the experimental
evidence, is that in those trades where metallic lead is handled, particularly
lead rolling, very few hygienic precautions have ever been taken in regard to
washing before meals, smoking, etc. Although in these trades the hands
become coated with a lead compound (oleate), and the workers frequently
eat their food with unwashed hands, thus affording every opportunity for the
ingestion of lead, the incidence of poisoning is by no means as high or so
pronounced in these occupations as in those giving rise to lead dust, such as
the white lead industry, where special precautions are taken, and where the
incidence of poisoning is always related to the dust breathed.
Respiratory Tract.—In a report on the incidence of lead poisoning in the
manufacture of paints and colours, one of us [T. M. L.[2]] in 1902 laid stress on
the marked incidence of poisoning in the specially dusty lead processes.
Following on that report special attention was given to the removal of dust by
means of exhaust ventilation. With the introduction of precautionary
measures, the incidence of poisoning underwent a marked decrease, this
decrease being most definite in those industries where efficient exhaust
ventilation could be maintained (see p. 47). Experience shows that cases of
poisoning in any given trade or manufacturing process are always referable to
the operations which cause the greatest amount of dust, and where,
therefore, the opportunity of inhaling lead dust is greatest.
The investigations of Duckering[3], referred to on p. 203, show the amount
of dust present in the air in certain dangerous processes. His results clinch
the deductions made from general observation, that dusty processes are
those especially related to incidence of industrial poisoning. Ætiologically,
therefore, the relationship of dust-contaminated air and poisoning is
undeniable, and in not a few instances on record persons residing at a
distance from a lead factory have developed poisoning, although not
employed in any occupation involving contact with lead, aerial infection
through dust remaining the only explanation. The actual channel through
which the lead dust suspended in the air gains entrance to the body is,
therefore, of especial importance; one of two channels is open—gastro-
intestinal and respiratory.
The investigations of one of us (K. W. G.) on the experimental production of
lead poisoning in animals has shown conclusively that the dust inhaled was
far more dangerous, and produced symptoms far earlier than did the direct
ingestion of a very much larger quantity of the same compound by way of the
mouth and gastro-intestinal canal. There is no doubt whatever that the chief
agent in causing lead poisoning is dust or fume suspended in the air. That a
certain amount finds its way into the stomach direct is not denied, but from
experimental evidence we consider the lung rather than the stomach to be
the chief channel through which absorption takes place (see p. 81).
The following table gives a specific instance of the incidence of lead
poisoning in a white lead factory, and demonstrates clearly the ætiological
importance of dust. The increase in reported cases, as well as in symptoms
of lead absorption not sufficiently severe to prevent the individual from
following his usual occupation, was associated with the rebuilding of a portion
of the factory in which the packing of dry white lead had been carried on for a
large number of years. The alterations necessitated the removal of several
floors, all of which were thoroughly impregnated with lead dust. Before the
alterations were undertaken it was recognized that considerable danger
would arise; stringent precautions were therefore taken, and the hands
engaged in the alterations kept under special observation. Notwithstanding
this there was an increase in the number of reported cases, which were all
mild cases of colic; all recovered, and were able to return to their work in a
short time.
Table I.—Lead Poisoning in a White Lead Factory.
The figures refer to the weekly examination of the whole of the men. For example, if a man
was returned as suffering from anæmia on three occasions, he appears as three cases in
Column 7.
Year Total Total Cases Cases Cases of Cases of Cases of Blue Line
Number Cases of in Dusty in Other Suspen- Anæmia Tremor
of Poisoning Processes Processes sion
Exami-
nations
(1) (2) (3) (4) (5) (6) (7) (8) (9)
1905 5,464 9 8 1 20 78[B] 249[B] 311[B]
1906 [A] 5,096 18 16 2 9 256 215 532
1907 4,303 4 3 1 6 62 81 38
1908 3,965 4 3 1 5 40 25 11
[A] Structural alterations in progress, including cutting up “lead floor,” saturated with
white lead dust.
[B] These numbers for the half-year only, the inspection being taken over in June,
1905.
Meillère[4] goes to considerable trouble to show that absorption of lead dust

by the lung is hypothetical; that it may take place, but that it is not a channel
of absorption of practical importance. He cites a number of opinions and
experiments by various observers on the absorption of lead through the
mucous membrane of the mouth, alimentary canal, conjunctiva, etc., and he
regards the absorption of lead as one peculiarly confined, in the majority of
instances, to the intestinal canal.
The usual view is that, in the passage of the respired dust-laden air through
the nose, the larger particles of dust are deposited first of all upon the
mucous membrane in the interior chambers of the nose; further, a second
deposit takes place on the posterior wall of the pharynx and in the throat,
where the eddies produced by the current of air inhaled through the nostrils
allow the finer particles to become more easily deposited. Finally, should a
small trace gain access to the larynx, it is said to be deposited there upon the
mucous membrane, to be subsequently ejected, and only a very small
proportion of the total may ever find its way into the lung.
In all arduous labour, directly the respiration rate rises through extra calls
made upon the muscles of the body, an increase in the depth of respiration
takes place; yet even under these circumstances Meillère and others incline
to the view that the dust is deposited on the mucous surfaces of the mouth
and swallowed. Experimental evidence is entirely opposed to these
suppositions. In the first place, unless particles of dust readily find their way
into the lung, it is difficult to understand how the lung itself becomes the site
of so much deposit of carbon, and of flinty material in stonegrinder’s
pneumokoniosis. The staining of the lung by means of carbon particles,
particularly in dwellers in cities, is too well known to warrant more than a
passing reference. Moreover, experimental work has shown that fine powders
suspended in the air easily reach the lung. Armit[5] has shown that the nickel
in nickel carbonyl poisoning gains direct access to the lung, and becomes
deposited there, the metallic particles being readily demonstrated in the lung
tissue itself. Further, the experiments (see p. 84) demonstrate that white lead
dust and other forms of lead dust definitely gain access to the lung, and thus
inhaled produce all the symptoms of lead poisoning in animals subjected to
the inhalation. White lead, litharge, or red lead, are not easily suspended in
water, and long-continued mixing is necessary to make a suspension. Great
difficulty is found in “laying” lead dust by water, as the following experiment
demonstrates: Five wash-bottles are arranged in series; in the first ground dry
white lead is placed, and the other three bottles are filled with water, and a
tube laid under the surface of the water in such a way that the air from the
first bottle must pass the whole of the water seals in each subsequent bottle.
In the last bottle is dilute nitric acid saturated with sulphuretted hydrogen. If
the series is now attached to an aspirating jar, and air drawn slowly over at
the rate of ordinary respiration, the white lead powder in the first bottle being
at the same time shaken so that the air is fully charged with finely powdered
dust, lead is quickly detected in the air passing through the last bottle of the
series, by the darkening of the solution. In this way the presence of lead dust
has been demonstrated after passing through four 2-inch water seals and 8
feet of ¹⁄₄-inch wet rubber tubing. Such an experiment negatives the theory
that all, or even a large quantity, of a finely divided powder becomes
deposited on the upper portion of the respiratory tract.
Particles of lead present in the air in industrial processes are exceedingly
minute, and even in ground white lead the average size of the particle is
under 1 μ. Finally, Tanquerel[6] and Stanski[7] succeeded in producing lead
poisoning experimentally by blowing lead dust through a tube inserted in a
tracheotomy opening. There remains, therefore, no room for doubt that the
lung is the pre-eminent portal for lead absorption, particularly in industrial
processes; from which it follows, as has been extensively shown in actual
practice, that the diminution of dust in workshops and factories by means of
exhaust ventilation is invariably followed by a diminution in the number of
cases of plumbism.
Gastro-Intestinal.—We have dealt with absorption by way of the lung,
and have insisted that such inhalation of dust is of greater importance in
giving rise to industrial lead poisoning than gastro-intestinal absorption.
Gastro-intestinal absorption can take place, and is by no means negligible, in
ordinary industrial conditions. One of the most interesting and important
confirmatory evidences of the absorption of lead by the gastro-intestinal canal
is to be found in the large outbreaks of poisoning in which water-supplies
have been contaminated, either at their source or locally. We have already
seen that electrolysis may play an important part in the solution of lead in
water, and also learnt from Gautier[8] that the carbon dioxide content of water
is not necessarily the sole predisposing element in the solution of lead. In this
connection an important case is described by Thresh[9], where water by no
means soft, but holding some 30 degrees of hardness, produced lead
poisoning in an isolated family. The water in question was distinctly acid to
litmus-paper, and contained a very high percentage of nitrates; the compound
or salt of lead present was therefore one easily absorbed from the alimentary
canal (see p. 86).
In all instances of water-borne lead poisoning the amount of lead present in
the water was small; but as such lead would not be removed by boiling, the
amount of water consumed per person from the contaminated source was
probably large. As the signs of poisoning did not appear until a considerable
time had elapsed, a much larger quantity of lead was probably absorbed than
would appear from the simple statement that the water contained ¹⁄₁₀ grain
per gallon.
A number of cases have been reported from use of diachylon as an
abortifacient, and the symptoms in these cases are invariably those which
occur in other severe forms of poisoning such as are met with in industrial
processes. In nearly every case colic was the first symptom, followed later by
paresis of various types—amaurosis, albuminuria, albuminuric retinitis,
melancholia, encephalopathy—and not a few of the persons succumbed. In
most of the reported cases abortion was produced, but in some, particularly in
one[10], three dozen pills containing diachylon were taken in a month,
producing acute lead poisoning, colic, and paresis, but not abortion.
In fifteen recorded cases of the use of diachylon, fourteen showed a lead
line, in many cases distinct and broad. This point has considerable interest,
as such a line cannot have been produced by oral contact. The drug in the
form of pills would be rapidly swallowed, and little opportunity afforded for
particles to remain in the mouth. Its presence, therefore, suggests excretion
from circulating blood of lead which has been absorbed in the intestine. The
blue line will be referred to again later (see p. 122).
Practically all cases of water poisoning and of swallowing of lead
compounds have developed colic. Further, colic is cited in all the early
recorded cases, even in the very earliest cases referred to in the historical
note, of lead poisoning; and as poisoning in those cases had invariably taken
place by swallowing the drug, it may be presumed from this association has
arisen the belief that lead must be swallowed to produce gastro-intestinal
symptoms. No attention has been paid to the fact that a few cases of definite
cutaneous absorption of lead from the use of hair lotions have been followed
by colic. Gastro-intestinal symptoms, therefore, can be produced without the
direct ingestion of the drug, and colic is a symptom of generalized blood-
infection rather than a localized irritative action on the intestinal mucosa. This
question, again, is more related to pathology than ætiology, and is dealt with
in that section. But mention may be made here of the fact that a number of
observers, more lately Meillère, have laid it down as an axiom that
experimental production of lead poisoning in animals gives no criterion or
evidence of lead poisoning produced in man industrially. Very grave exception
must be taken at once to such a statement. In the majority of experiments
quoted by Meillère the quantity of lead given for experimental purposes has
been large—much larger, indeed, than is necessary to produce small and
characteristic effects—and instead of chronic poisoning an acute lead
poisoning has generally been set up; and even where chronic poisoning has
supervened, the condition has as a rule been masked by the severer initial
symptoms. On the other hand, the evidence to be derived from comparison of
the various observations from animal experiments brings out with remarkable
unanimity the similarity of the symptoms to those produced in man, and, as
will be seen later in the section devoted to Pathology, experiments by one of
us (K. W. G.) have so far confirmed this surmise; in fact, a description of a
case of encephalopathy coming on after lead poisoning of a chronic nature,
described by Mott, agrees in practically every particular with the train of
symptoms as observed in these experimental animals. Certain slight
differences as to the muscles first affected are observed, but it is practically
always the homologous muscle (the physiological action of which more nearly
resembles the human muscle) which is the one to be affected in the animal,
not the anatomical homologue. Thus, for instance, in the cat the spinal
muscles, and particularly the quadriceps extensor, is the muscle which is first
affected through the medium of the anterior crural nerve. This extensor
muscle is one which only performs a slight amount of work in extending the
knee-joint, the amount of work being, however, disproportionate to the size of
the muscle. The extensors of the fore-feet ultimately do become weakened,
but it is the hind-limb upon which the stress first falls.
Attention has been given to the solubility of lead salts in gastric juices, the
majority of such experiments having been performed with artificial gastric
juice. The method at present in use, prescribed by the amended rules of
August, 1900, for earthenware and china factories, is based on some, if
slight, consideration of the physiology of digestion. The method described by
Rule II. states that the estimation of the quantity of lead present in the lead
fritt shall be performed as follows:
A weighed quantity of dry material is to be continuously shaken for one
hour at room temperature with one thousand times its weight of an aqueous
solution of hydrochloric acid, containing 0·25 per cent. of HCl. This solution is
thereafter to be allowed to stand for one hour, and to be passed through a
filter. The lead salt contained in a portion of the clear filter is then to be
precipitated as lead sulphide, and weighed as lead sulphate.
This method has been adopted on the supposition that the solubility of a
lead salt in the gastric juices is the chief source of the lead poisoning in the
Potteries, and that the hydrochloric acid content of the solution determines,
for practical purposes, the quantity of lead dissolved out of a given sample.
The temperature, however, at which this estimation is made—namely, room
temperature—is one considerably lower than that of the body, and the
quantity of lead taken up into solution at this temperature is less than that
which occurs at the ordinary temperature of the body—37° C. Practically
twice as much lead is dissolved out of fritt at 37° C. for an hour as is rendered
soluble at the ordinary temperature of the room—about 15° C. Thomason[11],
who made some experiments in this direction, gives a figure of 2·35 lead
oxide dissolved at 15° C. and 4·54 at 37° C. In another estimation—a matter,
too, of some considerable importance—it was found that acetic acid dissolved
1·97 per cent, at 15° C., and 3·27 at 37° C. In lactic acid the figure was 2·28
at 15° C., and 3·53 at 37° C. It is therefore a low estimation of the solubility of
any substance by the gastric juices if the substance is operated on at a
temperature below that of the body.
The question of the solubility of a lead salt in the gastric contents is
important in view of the small quantities of dust swallowed; and in addition to
hydrochloric acid, other substances are also present in the gastric juice,
which is by no means a simple aqueous solution of the mineral acid. Further,
the gastric juice, except in cases of pathological type, is not acid in periods of
gastric rest, unless such acidity may be represented by the presence of
fermentative acids—acetic, lactic, and butyric.
The activity of the gastric juice on lead is directly caused by the quantity of
organic acids present in addition to the hydrochloric acid, and by the
presence of foodstuffs—(1) in the undigested and (2) in the semidigested
condition. In considering the absorption of lead products from the gastro-
intestinal canal, the normal digestive processes should not be lost sight of—
that is, the sequence of events which occur during digestion of food. On
swallowing food, no definite acidity is present in the stomach for fifteen to
twenty minutes, and even after that time the hydrochloric acid is only
commencing to be secreted. As digestion proceeds, and the whole mass
becomes partially dissolved, such portions as are in a soluble condition are
passed through the pyloric opening at intervals, and the whole contents of the
stomach do not pass straight through the pyloric opening as through an
ordinary straight drain-pipe. As each mass of food passes onwards through
the pylorus, it comes into contact in the duodenum with pancreatic juice, and
with the bile, these alkaline fluids rapidly change the reaction, and allow the
other ferments, trypsin, etc., to become active. As the mass proceeds
onwards through the intestine, the succus entericus also exerts its function.
Finally the fluid contents of the intestine are passed onwards through the ileo-
cæcal valve. During the passage from the pylorus to the ileo-cæcal valve, the
reaction of the intestinal contents undergoes variations, from an alkaline in
the duodenum or upper parts of the jejunum, to acid at the ileo-cæcal valve.
Practically no absorption takes place from the stomach itself; a small quantity
of water and such highly volatile fluids as alcohol may be absorbed, but the
main absorption is not commenced until the food has left the stomach; in fact,
the stomach contains no mechanism for food absorption. The work of
absorption of the products of digestion is carried on actively through the small
intestine until finally the materials have reached the large intestine through
the ileo-cæcal valve; water is then mainly absorbed, and albuminous fluids
and substances in solution to some extent, but the amount of absorption
which takes place is infinitesimal as compared with that of the small intestine.
These points in the physiology of digestion require to be taken into account
when discussing the absorption of lead salts in the gastro-intestinal canal.
When human gastric juice is obtained direct from the stomach in man, and
lead is submitted to its action, definite quantities of lead pass into solution;
and, curiously enough, in the normal gastric juice lead sulphate is as soluble
as both white lead and litharge. The following two tables give the results of
the estimation of the direct action of human gastric juice upon lead. The
particular point is that the juice was obtained by the stomach tube from
persons who had been given a simple test meal preceded by a twelve hours’
fast; the juice was therefore in a normal condition. The tests gave the
following results in the normal stomach:
Lead sulphate 0·080 per cent.
White lead 0·048 „
Litharge 0·040 „
In the second digestion, in which the analysis of the contents showed the
patient to be suffering from the condition known as “hyperhydrochloridia,” the
results were—
Lead sulphate 0·046 per cent.
White lead 0·042 „
Litharge 0·340 „
A very large number of experiments have also been performed for the
purpose of determining the solubility of raw lead glaze, and white lead, in
artificial digestions, the digestions having been made up in such a way that
they resembled as far as possible in every particular the ordinary stomach
contents. The type of digestion used was as follows:
Dry breadcrumbs 140 grammes.
Hydrochloric acid 5 c.c.
Lactic acid 0·1 c.c.
Acetic acid 0·1 c.c.
Pepsin 1·2 grammes.
Milk 1,200 c.c.
Digestions were performed with this mixture, and in every case the digest
was divided into two portions; each portion was retained at body temperature,
with agitation for a couple of hours, and at the end of that time one portion
was submitted to analysis. The second portion was neutralized, sodium
carbonate and pancreatic ferment added, and digestion carried on for another
two and a half hours at body temperature. At the end of this time the
pancreatic digest was examined.
Thirty-five digestions were performed. When 1 gramme of white lead was
used—that is, 0·01 per cent., containing 0·75 per cent. of lead oxide—the
quantity of lead found as lead oxide in the acid digest varied from 2 to 3 per
cent., whilst the amount found in the pancreatic digest varied from 4 to 6·5
per cent. of the added salt. On increasing the amount to 12 grammes—that
is, 1 per cent.—the quantity returned in the digest only increased from 1·5 to
2 per cent. In other words, in the addition of larger quantities of material the
ratio of solubility did not rise in proportion to the quantity added. Where a
direct pancreatic digestion was performed without the preliminary digest of
the gastric contents, the amount of lead present in the digest was only about
0·2 per cent. of the quantity added; indeed, it was very much smaller than the
amount dissolved out after preliminary acid digestion—that is, if the normal
sequence of digestion is followed, the solubility progresses after the gastric
digest has been neutralized and pancreatic ferment has been added,
whereas very slow action indeed occurs as the result of action of the
pancreatic digest alone. Some experiments described by Thomason[12],
although carried out without special regard to the physiological question of
the progressive nature of digestion, distinctly confirm the point raised. Thus,
in a digest of gastric juice, milk, and bread, 5·0 per cent. of lead was
dissolved, whereas when pancreatic juice alone was used only 0·4 per cent.
was found to be dissolved, a remarkable confirmation of the point under
discussion.
The difficulty of estimating lead present in these gastric digestions is a very
real one, as, owing to the precipitation of lead by various fluids of an
albuminoid nature, it is difficult to determine the amount of lead present in a
given quantity of digest; moreover, in making such a digest, much of the
material may become entangled among the clot of the milk in a purely
mechanical fashion, and, in attempting to separate the fluid from the other
portion of the digest, filtration no doubt removes any lead which has been
rendered soluble first of all, and reprecipitated as an albuminate. An
albuminate of lead may be formed with great ease in the following way: A 5
per cent. solution of albumin in normal saline is taken, 0·02 per cent. of
hydrochloric acid is added, and 10 per cent. solution of lead chloride added
as long as a precipitate is formed. The precipitate is then filtered off, and
washed in a dialyser with acidulated water until no further trace of lead is
found in the washings. A portion of this substance taken up in distilled water
forms a solution of an opalescent nature, which readily passes through the
filter and gives the reaction of protein with Millon’s reagent, and the lead
reaction by means of caustic potash and sulphuretted hydrogen, but very
large quantities of mineral acid are required to produce any colour with
hydrogen sulphide. Lead which gains access to the stomach, either dissolved
in water or swallowed as fine dust, becomes in all probability converted first
into a soluble substance, chloride, acetate, or lactate, which compound is
then precipitated either by the mucin present in the stomach, or by the protein
constituents of the food, or by the partially digested food (peptonate of lead
may be formed in the same way as the albuminate described above). In this
form, or as an albuminate or other organic compound, it passes the pylorus,
and becomes reprecipitated and redigested through the action of the
pancreatic juice. A consideration of the action of artificial gastric juices and
the properly combined experiments of gastric and pancreatic digestions
suggest that the form in which lead becomes absorbed is not a chloride, but
an organic compound first formed and gradually decomposed during the
normal process of digestion, and absorbed in this manner from the intestine
along with the ordinary constituents of food. Dixon Mann[13] has shown that
about two-thirds of the lead administered by the mouth is discharged in the
fæces, and that the remaining one-third is also slowly but only partially
eliminated. This point is of very considerable importance in relation to
industrial poisoning of presumably gastro-intestinal origin, and consideration
of the experiments quoted suggests that the digestion of albuminate or
peptonate may to some extent be the basis which determines the excretion of
so much of the lead via the fæces. This alteration of solubility has no doubt a
bearing on the immunity exhibited by many animals when fed with lead, and
probably explains the fact that many of the experimental animals fed with lead
over long periods exhibited no symptoms of poisoning (see p. 85), whereas
control animals, given a far smaller quantity of lead by other means and
through the lung, rapidly developed symptoms of poisoning. A diversity of
opinion exists as to the effect of pepsin upon the solubility of lead. Oliver[14]
considers that the pepsin has a retarding influence on the solubility of lead in
the gastric juice, and Thomason’s experiments also support this view,
although it is difficult to see why the action of pepsin alone should be of such
extreme importance. There is also the complicating fact that other added
substances in the food may mask any direct pepsin factor that may be
present. Albumose and peptone rather than pepsin are to be regarded as the
more important substance physiologically in their reaction with lead, and it is
interesting to note that Schicksal[15] found that by exposing lead in the form of
white lead in a 1 per mille solution of hydrochloric acid in the presence of
peptone produced a greater solvent effect on white lead than did the diluted
acid alone, and the same effect was also seen on metallic lead.
Table II.—Schicksal’s Table.
Amount
dissolved
returned
as
Metallic
Solution. Substance. Time. Lead.
(a) 1·0 per cent. peptone 100 White lead, 10 3 days at 37°
- 0·1471 grm.
0·1 per cent. HCl c.c. grms. C.
(b) 1·0 per cent. peptone 100 Metallic lead, 4

- „ 0·0330 „
0·1 per cent. HCl c.c. grms.
(c) 0·1 per cent. HCl, 100 White lead, 10
0·0983 „
c.c. grms.
(d) 0·1 per cent. HCl, 100 Metallic lead, 4
0·0194 „
c.c. grms.
(e) Metallic lead, 4
0·3 per cent. Na2CO3 None
grms.
(f) 0·3 per cent. Na2CO3 White lead „
(g) 0·3 per cent. Na2CO3
- White lead „
0·5 per cent. NaCl
(h) 0·3 per cent. Na2CO3

- Metallic lead „
0·5 per cent. NaCl
The experiments referred to on p. 18 undoubtedly agree with those of

Schicksal. In addition to the presence of peptones, the effect of carbonic acid
must be also considered, as increase in solubility in gastric and pancreatic
digestions was produced when carbonic acid gas was bubbled through the
digest during the period of action. The whole question of solubility of many
materials in the fluids of the stomach and intestinal canal requires entire
revision, not only as regards lead, but as regards a number of other metals,
including arsenic.
The Mechanism of Lead Absorption.
—The final method of absorption of lead particles or lead solution into the
animal body remains to be considered. Experimental phagocytosis of lead
particles—as, indeed, of any minute particles of substance—suspended in an
isotonic solution, may be observed directly under the microscope. Lead
particles show no exception to the rule, and white blood-corpuscles in a
hanging-drop preparation, made by suspending them in an isotonic salt
solution and serum, may be watched englobing particles of lead, and by
appropriate means the ingested lead may be afterwards demonstrated. In
such an experiment, much of the lead absorbed by the individual corpuscles
rapidly loses its property of giving a black precipitate with sulphuretted
hydrogen, and has apparently become converted into an organic compound,
peptonate or albuminate.
In the section devoted to the Chemistry of Lead, it has been noted that the
colloidal solutions of lead are not precipitated by sulphuretted hydrogen, and
that albuminates and peptonates of lead are presumably of colloidal form.
There seems evidence, therefore, that the direct absorption of lead takes
place by means of the phagocytes of the body, and that in them it becomes
converted into a colloidal form, in which it is probably eliminated through the
kidney and intestine, mainly the latter.
Further evidence of the englobement of lead particles by amœbic cells may
be gained if sections of the intestines of experimental animals are examined;
in the lymphoid glands particles of lead may be seen situated in the interior of
the walls, and even in the cells. It does not by any means follow that these
particles of lead sulphide present in the cells have been formed in situ; more
probably the lead has been converted into a sulphide in the intestinal lumen
itself, and subsequently taken up by the amœbic cells situated in its
periphery.
Another solution is possible—namely, that the particles seen in the
intestinal wall are particles of lead in process of excretion into the intestine
itself, and that the pigmentation of the vessel walls and cells is caused by the
staining of the particles of lead passing from the blood into the lumen of the
tube, which have been converted into a sulphide during their passage.
The localization of the staining in the large intestine, especially in the region
of the appendix in animals (cats), tends to support this theory. The large
bowel near the ileo-cæcal valve, the appendix, and even the glands in the
immediate neighbourhood, are found to be discoloured, and to contain lead in
larger quantities than any other portion of the intestine. In extreme cases the
whole of the large intestine may be stained a greyish-blue. The bloodvessels
in the mesentery in this region are also engorged. When, however, a salt of
lead, such as lead carbonate or lead oxide, gains access to the stomach, it
may be easily converted into chloride by the free hydrochloric acid present in
the stomach; and, in addition, should there be any chronic acid-dyspepsia
(hyperchlorhydria), particularly of the fermentative type, in which free lactic
acid and other organic acids are to be found within the viscus, small
quantities of lead swallowed as dust undergo solution and conversion into
chloride or lactate. The pouring out of acid gastric juice from the stomach
glands does not take place immediately after the first bolus of food is
swallowed, and it may be twenty minutes or half an hour before the gastric
contents have an acid reaction. During this time any lead salts previously
swallowed may become incorporated with the bolus of food and escape
absorption.
Lead in solution or suspension in the stomach which becomes mixed up
with the food, and at the same time subjected to the action of various
albuminous constituents of the food in addition to acids, causes an
albuminate or peptonate of lead to be easily formed, and as such can never
be absorbed from the stomach direct; practically no absorption takes place in
the stomach, and the presence of food containing albuminate precipitates any
lead in solution as an organic insoluble salt. The bolus of food impregnated
with small quantities of lead passes onwards to the intestine, where further
digestion takes place. As the mass passes through the intestine the action
gradually results in the reappearance of acidity, but at the same time a certain
quantity of sulphuretted hydrogen is produced, some of it from the
degradation of the sulphur-containing moiety of the protein molecule by
ordinary hydrolytic process and intestinal ferments, quite apart from any
bacterial action. A portion of the lead present in the chyme may be set free
again for absorption. The bile is said to assist in the solution of lead in vitro.
In experiments made by one of us, which are quoted later, it has been
shown that an isolated loop of intestine allows the absorption of a soluble
lead salt (chloride) when there is no food present in the loop. As the food
mass proceeds through the length of the intestine more and more sulphur is
set free, and an opportunity arises for the fixation of the lead as a sulphide,
but even as a sulphide it is slightly soluble. Probably, however, most of the
lead becomes absorbed long before it reaches the stage at which free
sulphur or sulphuretted hydrogen exists for the formation of sulphide. It is
highly probable that lead, in common with a number of other heavy metals,
including arsenic, is absorbed gradually in the upper part of the intestine, and
re-excreted in the lower. Such an hypothesis is undoubtedly strongly
supported by the remarkable staining of the large intestine and the ileo-cæcal
valve.
The exact mechanism of the absorption of lead from its compound with
albumin or peptone as a lead peptonate or albuminate is very difficult to state
at present; lead albuminate is undoubtedly insoluble in water or normal saline
and in albumin. The process of absorption, then, of the metal lead from the
gastro-intestinal canal is very closely related to the absorption of other heavy
metals, and the fact that animals after very large doses of lead salts
administered via the mouth show hæmorrhages in the intestinal wall, in
addition to hæmorrhages in other parts of the body, with occasional distinct
ulceration, suggests a localized coagulative action on the vessels in the wall
of the intestine as the probable origin of the ulceration. A consideration of this
problem of lead absorption from the intestine—probably only the minutest
quantity of lead, if any, is absorbed from the stomach direct—is one of
considerable importance in the prevention of such lead poisoning as is
attributable to swallowing lead. No work in a lead factory should be
commenced in the morning without partaking of food, because if food be
present the opportunities for absorption of lead are greatly diminished, and of
all foods the one to be recommended as the most efficient is milk, or cocoa
made with milk.
The absorption of dust through the lung is probably an exceedingly
complicated reaction, and Armit’s experiments with nickel carbonyl probably
give the clue. He found that in nickel carbonyl poisoning the volatile product
was split up on the surface of the lung cells, the metallic portion passing
onwards into the lung itself, to be eventually absorbed by the serum.
From the pathological and histological investigations described on p. 81,
and from the fact that particles of lead are very readily taken up by white
blood-corpuscles, we can conclude that absorption of the finer lead particles
gaining access to the lung takes place through the medium of these
phagocyte cells, as such cells are well known to exist within the alveoli of the
lung. The stored-up carbon particles found in the lungs in dwellers in cities
show that such transference of particles from the alveoli to the inner portions
of the lung trabeculæ is a constant phenomenon, and it is therefore easily
understood how rapidly any fine particles not of themselves irritant may be
easily taken up by the tissues. Once having gained access to the interior of
the cells, the particles subjected to the action of the serum of the blood in the
ordinary process of bathing the tissues by the exuding lymph—nay, more,
actual particles of lead—may thus be actually transferred bodily into the finer
blood-spaces, and so be carried forward to the general circulation. Such
particles as remain fixed in the lung will undergo gradual absorption, and the
constant presence of carbonic acid in the circulating blood brought to the lung
undoubtedly largely contributes to their solution, and there is no need to
presuppose the necessity of some recondite interaction of organic acid for the
solution of the inhaled lead in the lungs.
In the absorption of the substance from the intestine, it may go direct into
the blood-stream in a similar fashion through the lacteals along the lymph
channels, and so into the thoracic duct, and finally into the general circulation.
On the other hand, a certain amount, probably not an inconsiderable portion,
is taken up by the portal circulation and transferred direct to the liver itself.
Chemical analysis of the liver supports this view, as does also the
considerable amount of stress thrown upon the liver when poisoning has
taken place from the intestinal canal on administration of massive doses of a
highly soluble lead compound. According to Steinberg[16], excretion of lead
takes place partly from the liver by the bile. This is probable, but there is no
experimental evidence at the present time to support the view. If such an
excretion does take place, the form in which the lead is excreted is probably
one in which it is no longer soluble by digestive action. On the other hand, it
may be in so soluble a form as to become reabsorbed from the intestine, thus
setting up a constant cycle. But such a theory is one that would require a
considerable amount of experimental evidence to support it before it could be
relied on.
There is no doubt that, however absorbed, lead remains stored up in the
body in minute quantities in many places, and the close analogy to arsenic is
met with in the curious elimination of the metal by the fæces. Cloetta[17],
quoted by Dixon Mann, discovered that, although dogs were unable to take a
larger dose of arsenic than 0·0035 gramme per day without exhibiting toxic
results, they could nevertheless take arsenic in much larger doses if it were
given in the solid form, and he was able to increase the dose to as much as 2
grammes per diem without showing any toxic symptoms. Examination of the
urine and fæces showed that as the amount of urinary excretion of arsenic
diminished, so that in the fæces increased, and in lead poisoning, even in
massive doses swallowed in error, the amount of lead excreted by the urine
rapidly diminishes in quantity, although the patient may be still suffering from
the effects of lead poisoning. The experiments, also, quoted on p. 100
constantly pointed to the elimination of lead by way of the intestine, and in
practically all the animals that had suffered from chronic poisoning well-
marked dark staining of the upper part of the cæcum due to lead was
invariably present. This staining and excretion of lead of the large intestine
undoubtedly takes place in man. In a case described by Little[18], where
diachylon had been administered, the administration of a large enema
containing sulphate of magnesium came away black. A more detailed result
of the experiments and a consideration of the elimination of lead are reserved
for another chapter, but it is impossible to consider the ætiology of the
disease without some reference to the general histological channels of
absorption and excretion.
Cutaneous Absorption of Lead.—A considerable amount of
controversy has centred on the question of the absorption of lead through the
unbroken skin. It has been shown that such drugs as belladonna applied to
the skin alone may produce dilatation of the pupil; an ointment containing
salicylic acid spread upon the skin and thoroughly rubbed in is followed by the
appearance of derivatives of salicylic acid in the urine; mercury may be
applied to the skin, and rubbed in, in sufficient quantities to produce
salivation; and a very large number of other drugs may be cited, all of which
when applied to the unbroken epidermis with friction produce the
physiological action of the drug.
There is no reason to exclude lead from the category of drugs which may
be absorbed through the medium of the skin, and, as several observers have
shown, animals may be poisoned by lead on applying a plaster of lead
acetate to the skin. Amongst these experiments may be quoted those of
Canuet[19] and Drouet[20] on rabbits. Some observers, among whom may be
mentioned Manouvrier[21], have attempted to prove that paralysis of the
hands occurs more often in the right hand in right-handed people, in the left
hand with left-handed people, and from the various experiments showing
absorption of lead through the unbroken skin they seek to connect the lesion
of the nerve with absorption direct through the skin of the hands.
Many objections can be urged against acceptance of this theory. Lead
workers who are constantly manipulating lead in a state of solution with bare
hands do not appear as a class to be more subject to wrist-drop than do
persons who are exposed to inhalation of fumes or dust of lead; in fact,
incidence of paralysis and of nerve lesions generally is more severe among
persons exposed to prolonged inhalation of minute quantities of lead through
the respiratory tract. The greater the exposure to dust, the greater the number
of cases of anæmia and colic, whilst in other industries, as has already been
stated, where lead exists as an oleate on the hands of the workers day in and
day out for many years, paralysis and even colic are of rare occurrence; in
other words, persons especially exposed to the absorption of lead through
their hands show a much smaller incidence of lead poisoning of all types than
do those exposed to lead dust. Further, the pathology of wrist-drop and
similar forms of paresis tends to show that the nerve supplying the affected
muscles is not affected primarily, but that the initial cause is hæmorrhage into
the sheath of the nerve, producing ultimate degenerative change. The
hæmorrhage, however, is the primary lesion.
REFERENCES.
[1] Goadby, K. W.: A Note on Experimental Lead Poisoning. Journal of Hygiene, vol.
ix., No. 1, April, 1909.
[2] Legge, T. M.: Report on the Manufacture of Paints and Colours containing Lead
(Cd. 2466). 1905.
[3] Duckering, G. E.: Journal of Hygiene, vol. viii., No. 4, September.
[4] Meillère, G.: Le Saturnisme, chap. iv. Paris, 1903.
[5] Armit, H. W.: Journal of Hygiene, vol. viii., No. 5, November, 1908.
[6] Tanquerel des Planches: Traité des Maladies de Plomb, ou Saturnines. Paris,
1839.
[7] Stanski: Loc. cit.
[8] Gautier: Intoxication Saturnine, etc. Académie de Médecine, viii., November,
1883.
[9] Thresh, J. C.: The Lancet, p. 1033, October 7, 1905.
[10] Ibid., January 5, 1909.
[11] Thomason: Report of the Departmental Committee on Lead Manufacture:
Earthenware, China, vol. ii., appendices, p. 61. 1910.
[12] Ibid.
[13] Dixon Mann: Forensic Medicine and Toxicology, p. 495. 1908.
[14] Oliver, Sir T.: Lead Poisoning (Goulstonian lectures). 1891.
[15] Schicksal: Die Bekämpfung der Bleigefahr in der Industrie, p. 38. 1908.
[16] Steinberg: International Congress of Industrial Hygiene. Brussels, 1910.
[17] Cloetta: Dixon Mann’s Forensic Medicine and Toxicology, p. 463.
[18] Little: The Lancet, March 3, 1906.
[19] Canuet, T.: Thèse, Paris, 1825, No. 202. Essai sur le Plomb.
[20] Drouet: Thèse, Paris, 1875. Recherches Experimentales sur le Rôle de
l’Absorption Cutanée dans la Paralysie Saturnine.
[21] Manouvrier, A.: Thèse, Paris, 1873, No. 471. Intoxication par Absorption
Cutanée.
CHAPTER III
SUSCEPTIBILITY AND IMMUNITY
A large number of poisonous substances, among which lead may
be included, are not equally poisonous in the same dose for all
persons. It is customary to speak of those persons who show a
diminished resistance, or whose tissues show little power of resisting
the poisonous effects of such substances, as susceptible. On the
other hand, it is possible, but not scientifically correct, to speak of
immunity to such poisonous substances. Persons, particularly, who
resist lead poisoning to a greater degree than their fellows are better
spoken of as tolerant of the poisonous effects than as being partially
immune.
The degree of resistance exhibited by any given population
towards the poisonous influence of lead shows considerable
variation. Thus, in a community using a water-supply contaminated
with lead, only a small proportion of the persons drinking the water
becomes poisoned. There are, of course, other factors than that of
individual idiosyncrasy which may determine the effect of the poison,
as, for example, the drawing of the water first thing in the morning
which has been standing in a particular pipe. But even if all
disturbing factors are eliminated in water-borne lead poisoning,
differing degrees of susceptibility are always to be observed among
the persons using the water.
Lead does not differ, therefore, from any other drugs to which
persons show marked idiosyncrasies. Thus, very small doses of
arsenic may produce symptoms of colic in susceptible persons; a
limited number of individuals are highly susceptible to some drugs,
such as cannabis indica, while others are able to ingest large doses
without exhibiting any sign of poisoning; and it is well known that
even in susceptible persons the quantity of a particular drug which
first produces symptoms of poisoning may be gradually increased, if
the dosage be continued over long periods in quantities insufficient

Textbook Mobile Networks and Management Jiankun Hu Ebook All Chapter PDF

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Textbook Mobile Networks and Management Jiankun Hu Ebook All Chapter PDF

Uploaded by

Copyright:

Available Formats

Mobile Networks and Management

Data and Energy Integrated Communication Networks Jie

5G Green Mobile Communication Networks Xiaohu Ge

Principles of Chinese Management 1st Edition Haibo Hu

From GSM to LTE-Advanced Pro and 5G: An Introduction to

Mobile Networks and Management 5th International

Mobile Networks and Management 8th International

Machine Learning and Cognitive Computing for Mobile

Technologies and Eco innovation towards Sustainability

Zahir Tari Sheng Wen (Eds.)

ISSN 1867-8211 ISSN 1867-822X (electronic)

Library of Congress Control Number: 2018941835

Printed on acid-free paper

April 2018 Jiankun Hu

Technical Program Committee Chairs and Co-chairs

Publicity and Social Media Chair/Co-chair

Sponsorship and Exhibits Chair

Posters and PhD Track Chair

Technical Program Committee

Paul Pan Unitec, New Zealand

Offloading of Fog Data Networks with Network Coded Cooperative

Persistent vs Service IDs in Android: Session Fingerprinting from Apps . . . . 14

Towards Developing Network Forensic Mechanism for Botnet Activities

Performance Comparison of Distributed Pattern Matching Algorithms

Robust Fingerprint Matching Based on Convolutional Neural Networks. . . . . 56

A Personalized Multi-keyword Ranked Search Method Over

Application of Fuzzy Comprehensive Evaluation Method for Reservoir

Factor Effects for Routing in a Delay-Tolerant Wireless Sensor Network

Estimating Public Opinion in Social Media Content

An Approach for Host-Based Intrusion Detection System Design

A Robust Contactless Fingerprint Enhancement Algorithm . . . . . . . . . . . . . . 127

Designing Anomaly Detection System for Cloud Servers by Frequency

A Variant of BLS Signature Scheme with Tight Security Reduction . . . . . . . 150

Quantum Authentication Scheme Based on Fingerprint-Encoded

Cooperative Information Security/Cybersecurity Curriculum Development . . . 178

An Energy Saving Mechanism Based on Vacation Queuing Theory

Homomorphic Evaluation of Database Queries . . . . . . . . . . . . . . . . . . . . . . 203

A Cache-Aware Congestion Control for Reliable Transport

A New Lightweight Mutual Authentication Protocol to Secure

Fog Computing as a Critical Link Between a Central Cloud

Performance Assessment of Cloud Migrations from Network

A Cloud Service Enhanced Method Supporting

Application of 3D Delaunay Triangulation in Fingerprint

Malicious Bitcoin Transaction Tracing Using Incidence

Cryptanalysis of Salsa and ChaCha: Revisited . . . . . . . . . . . . . . . . . . . . . . 324

CloudShare: Towards a Cost-Efficient and Privacy-Preserving

Probability Risk Identification Based Intrusion Detection System

Anonymizing k-NN Classification on MapReduce . . . . . . . . . . . . . . . . . . . . 364

A Cancellable Ranking Based Hashing Method for Fingerprint

Author Index . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 391

Ben Quinton(B) and Neda Aboutorab

University of New South Wales, Campbell, ACT 2612, Australia

Abstract. Future fog data networks are expected to be assisted by users

Keywords: Instantly decodable network coding · IoT · Full-duplex

Although much work has focused on implementing IDNC in various network

A distributed wireless data network model is considered in this paper and

3.1 Motivating Example

If we consider the system model depicted in Fig. 1, it can be shown by example

in the ﬁrst solution, we assume a conventional IDNC approach without the

– s1 transmits f1 to u1 and f2 ⊕ f3 to u2 and u3 .

This scenario results in consuming three downloads from the servers.

Solution 2. In this solution (depicted in Fig. 3), we show a scenario where a

3.2 Graph-Based Solution