Download the Full Version of textbook for Fast Typing at textbookfull.

com

Wireless and Satellite Systems 9th International

Conference WiSATS 2017 Oxford UK September 14 15
2017 Proceedings 1st Edition Prashant Pillai

https://textbookfull.com/product/wireless-and-satellite-
systems-9th-international-conference-wisats-2017-oxford-uk-
september-14-15-2017-proceedings-1st-edition-prashant-
pillai/

OR CLICK BUTTON

DOWNLOAD NOW

Download More textbook Instantly Today - Get Yours Now at textbookfull.com

 Recommended digital products (PDF, EPUB, MOBI) that
you can download immediately if you are interested.

Image and Graphics 9th International Conference ICIG 2017

Shanghai China September 13 15 2017 Revised Selected
Papers Part III 1st Edition Yao Zhao
https://textbookfull.com/product/image-and-graphics-9th-international-
conference-icig-2017-shanghai-china-september-13-15-2017-revised-
selected-papers-part-iii-1st-edition-yao-zhao/
textboxfull.com

Advances in Computer Entertainment Technology 14th

International Conference ACE 2017 London UK December 14 16
2017 Proceedings 1st Edition Adrian David Cheok
https://textbookfull.com/product/advances-in-computer-entertainment-
technology-14th-international-conference-ace-2017-london-uk-
december-14-16-2017-proceedings-1st-edition-adrian-david-cheok/
textboxfull.com

Multi Agent Systems and Agreement Technologies 15th

European Conference EUMAS 2017 and 5th International
Conference AT 2017 Evry France December 14 15 2017 Revised
Selected Papers Francesco Belardinelli
https://textbookfull.com/product/multi-agent-systems-and-agreement-
technologies-15th-european-conference-eumas-2017-and-5th-
international-conference-at-2017-evry-france-
december-14-15-2017-revised-selected-papers-francesco-belardinel/
textboxfull.com

Image Analysis and Processing ICIAP 2017 19th

International Conference Catania Italy September 11 15
2017 Proceedings Part II 1st Edition Sebastiano Battiato
https://textbookfull.com/product/image-analysis-and-processing-
iciap-2017-19th-international-conference-catania-italy-
september-11-15-2017-proceedings-part-ii-1st-edition-sebastiano-
battiato/
textboxfull.com
Image Analysis and Processing ICIAP 2017 19th
International Conference Catania Italy September 11 15
2017 Proceedings Part I 1st Edition Sebastiano Battiato
https://textbookfull.com/product/image-analysis-and-processing-
iciap-2017-19th-international-conference-catania-italy-
september-11-15-2017-proceedings-part-i-1st-edition-sebastiano-
battiato/
textboxfull.com

Virtual Augmented and Mixed Reality 9th International

Conference VAMR 2017 Held as Part of HCI International
2017 Vancouver BC Canada July 9 14 2017 Proceedings 1st
Edition Stephanie Lackey
https://textbookfull.com/product/virtual-augmented-and-mixed-
reality-9th-international-conference-vamr-2017-held-as-part-of-hci-
international-2017-vancouver-bc-canada-july-9-14-2017-proceedings-1st-
edition-stephanie-lackey/
textboxfull.com

Theoretical Computer Science 35th National Conference

NCTCS 2017 Wuhan China October 14 15 2017 Proceedings 1st
Edition Dingzhu Du
https://textbookfull.com/product/theoretical-computer-science-35th-
national-conference-nctcs-2017-wuhan-china-
october-14-15-2017-proceedings-1st-edition-dingzhu-du/
textboxfull.com

Social Media Processing 6th National Conference SMP 2017

Beijing China September 14 17 2017 Proceedings 1st Edition
Xueqi Cheng
https://textbookfull.com/product/social-media-processing-6th-national-
conference-smp-2017-beijing-china-
september-14-17-2017-proceedings-1st-edition-xueqi-cheng/
textboxfull.com

Trusted Computing and Information Security: 11th Chinese

Conference, CTCIS 2017, Changsha, China, September 14-17,
2017, Proceedings 1st Edition Ming Xu
https://textbookfull.com/product/trusted-computing-and-information-
security-11th-chinese-conference-ctcis-2017-changsha-china-
september-14-17-2017-proceedings-1st-edition-ming-xu/
textboxfull.com
Prashant Pillai
Kandeepan Sithamparanathan
Giovanni Giambene
Miguel Ángel Vázquez
Paul Daniel Mitchell (Eds.)

231

Wireless and Satellite

Systems
9th International Conference, WiSATS 2017
Oxford, UK, September 14–15, 2017
Proceedings

123
Lecture Notes of the Institute
for Computer Sciences, Social Informatics
and Telecommunications Engineering 231

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
Prashant Pillai Kandeepan Sithamparanathan
•

Giovanni Giambene Miguel Ángel Vázquez

•

Paul Daniel Mitchell (Eds.)

Wireless and Satellite

Systems
9th International Conference, WiSATS 2017
Oxford, UK, September 14–15, 2017
Proceedings

123
Editors
Prashant Pillai Miguel Ángel Vázquez
University of Wolverhampton Centre Tecnològic de Telecomunicacions
Wolverhampton de Catalunya
UK Barcelona
Spain
Kandeepan Sithamparanathan
RMIT University Paul Daniel Mitchell
Melbourne, VIC University of York
Australia York
UK
Giovanni Giambene
University of Siena
Siena
Italy

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

Lecture Notes of the Institute for Computer Sciences, Social Informatics
and Telecommunications Engineering
ISBN 978-3-319-76570-9 ISBN 978-3-319-76571-6 (eBook)
https://doi.org/10.1007/978-3-319-76571-6

Library of Congress Control Number: 2018934354

© 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 ninth edition of the 2017
European Alliance for Innovation (EAI) International Conference on Wireless and
Satellite Systems (formerly PSATS), which was held in the historic city of Oxford, UK,
during September 14–15, 2017.
The aim of this conference is to bring together researchers, developers, and prac-
titioners from around the world working in the field of wireless and satellite systems.
The theme of WiSATS 2017 was on the means of bringing wireless and satellite
services directly to the user for personal communications, multimedia, and location
identification.
The technical program of WiSATS 2017 comprised 19 full papers, including seven
papers in the main conference track. Other papers were presented as part of a series of
workshops, including: Next-Generation mmWave and Optical Satellite Systems: From
Channel Modelling to System Performance Evaluation (SATPROP: five papers),
Unmanned Aerial Systems (IWUAS: five papers), and Communication Applications in
Smart Grid (CASG: two papers). In addition to the high-quality technical paper pre-
sentations, the technical program also featured two keynote speeches, an invited talk,
and a panel session. The two keynote speeches were given by Dr. Hector Fenech
(Director of Future Satellite Systems at Eutelsat, France) and Vincenzo Pellegrini
(Senior Radiocommunication and SDR Systems Engineer at IDS, Ingegneria dei Sis-
temi S.p.A, Italy). The invited talk was presented by Prof. Fun Hu from the University
of Bradford, UK, and the panel session was on the subject of “SatCom in 5G and
Beyond: Opportunities and Challenges.”
We would like to thank all members of the Organizing and Steering Committees for
their effective cooperation in putting together and delivering the conference. It was a
great pleasure to work with such an excellent team. In particular, the Technical Pro-
gram Committee, led by our TPC co-chairs, Giovanni Giambene and Miguel Ángel
Vázquez, who managed the peer-review process. We are also grateful to our conference
manager, Lenka Bilska, and her colleagues Erika Pokorna and Dominika Kalafutova
for their assistance. We would like to thank all the authors who submitted papers to the
WiSATS 2017 conference and contributed to its success. A special thanks also to all
the sponsors for their support.
We strongly believe that the WiSATs conference provides a good forum for all
researchers, developers, and practitioners to discuss science and technology pertaining
to wireless and satellite systems. We expect that the next WiSATS conference will be
as successful and stimulating as this edition, as indicated by the contributions presented
in this volume.

January 2018 Paul Daniel Mitchell

Prashant Pillai
Kandeepan Sithamparanathan
 Organization

Steering Committee
Imrich Chlamtac (Chair) EAI/Create-Net, Italy
Kandeepan RMIT, Australia
Sithamparanathan
Mario Marchese University of Genoa, Italy
Agnelli Stefano ESOA/Eutelsat, France
Prashant Pillai University of Wolverhampton, UK

Organizing Committee
General Co-chairs
Prashant Pillai University of Wolverhampton, UK
Kandeepan RMIT University, Australia
Sithamparanathan

Technical Program Committee Chairs

Giovanni Giambene University of Siena, Italy
Miguel Ángel Vázquez CTTC, Spain

Local Chair
Shumao Ou Oxford Brookes University, UK

Workshops Chairs
Daniele Tarchi University of Bologna, Italy
Sooyoung Kim Chonbuk National University, South Korea

Publicity and Social Media Chairs

Celimuge Wu The University of Electro-Communications, Japan
Constantinos T. Angelis TEIE, Greece

Publications Chair
Paul Daniel Mitchell University of York, UK

Web Chair
Atm S. Alam University of Surrey, UK
VIII Organization

Panels Chair
Bhavani Shankar University of Luxembourg, Luxembourg

Industry Chair
Alberto Ginesi European Space Agency

Tutorials Chair
Aduwati Sali University of Putra Malaysia, Malaysia

Conference Manager
Lenka Bilska European Alliance for Innovation (EAI)

Technical Program Committee

Claudio Cicconetti MBI, Italy
Atm Alam University of Bradford, UK
Giovanni Giambene Università degli Studi di Siena, Italy
Raed Abd-Alhameed University of Bradford, UK
Yongqiang Cheng University of Hull, UK
Alban Duverdier CNES, France
Gabriele Oligeri University of Rome Tre, Italy
Ana Perez UPC, Spain
Angeles Vazquez-Castro Universidad Autónoma de Barcelona, Spain
George Eleftherakis The University of Sheffield International Faculty,
CITY College, Greece
Konstantinos Liolis SES S.A., Luxembourg
Misha Filip University of Portsmouth, UK
Djuradj Budimir University of Westminister, UK
Robert Ewatson University of Bath, UK
Haitham Cruickshank University of Surrey, UK
Haile-Selassie Rajamani University of Bradford, UK
Ray Sheriff University of Bradford, UK
Alister Burr University of York, UK
Franco Davoli DITEN, University of Genoa, Italy
 Contents

Security Issues in Satellite Networking

Security Aware Virtual Base Station Placement in 5G Cloud Radio

Access Networks. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3
Tshiamo Sigwele, Prashant Pillai, Abimbola Sangodoyin,
and Yim Fun Hu

DoS Attack Impact Assessment on Software Defined Networks . . . . . . . . . . 11

Abimbola Sangodoyin, Tshiamo Sigwele, Prashant Pillai,
Yim Fun Hu, Irfan Awan, and Jules Disso

Analysis of the Suitability of Satellite Communication for Time-Critical IoT

Applications in Smart Grid and Medical Grade Networks. . . . . . . . . . . . . . . 23
Frank Ball and Kashinath Basu

Advanced PHY-MAC Schemes for Future Satellite Systems

Making H-ARQ Suitable for a Mobile TCP Receiver over LEO

Satellite Constellations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 33
Bastien Tauran, Emmanuel Lochin, Jérôme Lacan, Fabrice Arnal,
Mathieu Gineste, Laurence Clarac, and Nicolas Kuhn

Joint Beam Hopping and Precoding in HTS Systems . . . . . . . . . . . . . . . . . . 43

Alberto Ginesi, Emiliano Re, and Pantelis-Daniel Arapoglou

Link Adaptation Algorithms for Dual Polarization Mobile

Satellite Systems . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 52
Anxo Tato, Pol Henarejos, Carlos Mosquera,
and Ana Pérez-Neira

Bandwidth Management Using MPLS Model for Future Mobile

Wireless Networks . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 62
Oba Zubair Mustapha, Ray E. Sheriff, and Felicia L. C. Ong

IWUAS Workshop Session

A Survey on Network Architectures and Applications for Nanosat

and UAV Swarms . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 75
Manlio Bacco, Pietro Cassarà, Marco Colucci, Alberto Gotta,
Mario Marchese, and Fabio Patrone
X Contents

Command and Control of UAV Swarms via Satellite. . . . . . . . . . . . . . . . . . 86

Pietro Cassarà, Marco Colucci, and Alberto Gotta

Toward Decentralised Consensus and Offloading for Area Coverage

in a Fleet of Drones . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 96
Hanna Kavalionak, Emanuele Carlini, Pietro Cassarà,
and Carlo Meghini

How to Support the Machine Learning Take-Off: Challenges

and Hints for Achieving Intelligent UAVs . . . . . . . . . . . . . . . . . . . . . . . . . 106
Patrizio Dazzi and Pietro Cassarà

UAVs and UAV Swarms for Civilian Applications: Communications

and Image Processing in the SCIADRO Project . . . . . . . . . . . . . . . . . . . . . 115
Manlio Bacco, Stefano Chessa, Marco Di Benedetto, Davide Fabbri,
Michele Girolami, Alberto Gotta, Davide Moroni,
Maria Antonietta Pascali, and Vincenzo Pellegrini

CASG Workshop

A Comparative Assessment of Embedded Energy Storage

and Electric Vehicle Integration in a Community Virtual Power Plant . . . . . . 127
Oghenovo Okpako, Haile-Selassie Rajamani, Prashant Pillai,
Ugonna Anuebunwa, and K. Shanti Swarup

Investigating the Impact of Cyber-Attack on Load Profile of Home Energy

Management System . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 142
Ugonna Anuebunwa, Haile-Selassie Rajamani, Prashant Pillai,
and Oghenovo Okpako

SATPROP Workshop Session

Propagation Elements for the Link Budget of Broadband Satellite Systems

in Ka and Q/V Band . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 155
Spiros Ventouras, Paul S. Crawford, and Charilaos I. Kourogiorgas

Monthly and Seasonal CFLOS Statistics for Optical GEO Feeder

Links Design . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 163
Nikolaos K. Lyras, Charilaos I. Kourogiorgas,
Athanasios D. Panagopoulos, and Konstantinos P. Liolis

Large Scale Site Diversity Experimental Campaign Between Greece

and UK Using ALPHASAT: First Results . . . . . . . . . . . . . . . . . . . . . . . . . 174
Apostolos Z. Papafragkakis, Charilaos I. Kourogiorgas,
Athanasios D. Panagopoulos, Spiros Ventouras,
and Pantelis-Daniel Arapoglou
 Contents XI

Architectural Design of the Q/V Band Site Diversity Experiment Between

Austria and Hungary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 184
Michael Schmidt, Laszlo Csurgai-Horvath, Peter Horvath,
Balint Horvath, Antonio Martellucci, and Juan Rivera Castro

Flexible Capacity Allocation in Smart Gateway Diversity Satellite Systems

Using Matching Theory . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 195
Anargyros J. Roumeliotis, Charilaos I. Kourogiorgas,
Argyrios Kyrgiazos, and Athanasios D. Panagopoulos

Author Index . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 205

Security Issues in Satellite Networking
 Security Aware Virtual Base Station
Placement in 5G Cloud Radio Access
Networks

Tshiamo Sigwele1(B) , Prashant Pillai2 , Abimbola Sangodoyin1 ,

and Yim Fun Hu1
1
Faculty of Engineering and Informatics, University of Bradford,
Bradford, West Yorkshire BD7 1DP, UK
{T.Sigwele,a.o.sangodoyin,Y.F.Hu}@bradford.ac.uk
2
Faculty of Technology, Design and Environment,
Oxford Brookes University, Oxford, UK
[email protected]

Abstract. In fifth generation (5G) cloud radio access networks

(C-RAN), baseband processing of base stations (BS’s) will be processed
on virtual machines called virtual BSs (VBS) in the centralized cloud
architecture. The existing researches mostly focus on how to maximize
resource utilization and reduce energy consumption in 5G C-RAN using
VBS placement. However, security issues in the context of VBS place-
ment within 5G C-RAN have been rarely addressed. In this paper, a secu-
rity aware VBS placement (SAV) scheme within 5G C-RAN is proposed
where the placement of VBSs to physical machines (PMs) considers the
security levels of both the VBS and the PM. A rigorous simulation study
is conducted for validating the proposed scheme, which shows a signifi-
cant security improvement of 16% compared to the heuristic simulated
annealing scheme (HSA).

Keywords: 5G · Cloud computing · C-RAN · Cloud security

Virtual machine placement

1 Introduction
The fifth generation (5G) cellular networks will experience a thousand-fold
increase in data traffic with over 100 billion connected devices by 2020 [1]. Such
surge in traffic will be from smart-phones, tablets, machine-machine connec-
tions and the Internet of Things (IoT). In order to support this sky-rocketing
traffic demand, heterogeneous cloud radio access H-CRAN networks has been
proposed where macro remote radio heads (RRHs) are overlaid by smaller cell
RRHs like femto, pico, micro and relay to increase capacity using spatial fre-
quency reuse. The 5G C-RAN uses cloud computing virtualization techniques
to host basestation (BS) functions in virtual machines (VMs) called virtual BSs
(VBSs) in the BS cloud [2]. One of the driving forces in the BS cloud is VBS
c ICST Institute for Computer Sciences, Social Informatics and Telecommunications Engineering 2018
P. Pillai et al. (Eds.): WiSATS 2017, LNICST 231, pp. 3–10, 2018.
https://doi.org/10.1007/978-3-319-76571-6_1
4 T. Sigwele et al.

placement where VBS are migrated among physical machines (PMs) to maximise
BS resource utilization and reduce energy consumption. Nevertheless, security
issues in the BS cloud in C-RAN have been rarely addressed. The introduction
of running BS functions in PMs in the BS cloud brings about security issues.
A prevalent VBS image with known vulnerabilities can be instantiated by an
attacker in BS cloud, therefore it may generate a large number of security holes
for attackers which may include eavesdropping users conversation. The intro-
duction of VBS images with known security vulnerabilities to a PM can lead to
security risks to the co-located VBSs in that PM. This is because of the one to
many mapping between the PM and the VBSs which makes vulnerabilities prop-
agate rapidly across the entire BS cloud. Some of the attacks that an attacker can
introduce include compromising the hypervisor and also side channel attacks to
the VBSs co-located within the same PM [3]. A VBS that has been compromised
can infect other VBS sharing the same hypervisor, memory or CPU. When a
VBS with no security risks is migrated to the PM with a compromised VBS,
that VBS will be compromised too. As such there is need for a security aware
VBS placement in the BS cloud to avoid the security risks.
In this paper, a VBS and PM security evaluation based on their vulnerabili-
ties is first conducted then, based on these security evaluations, a novel security
aware VSB placement is developed which minimises the security risks in the BS
cloud. The VBSs and PMs are grouped into vulnerability levels such that a VBS
with low vulnerability will be allocated to a PM with low vulnerability and a
VBS with high vulnerability will be allocated to a PM with high vulnerabil-
ity. The rest of the paper is organised as follows. Section 2 describes the related
work on VBS placement in C-RAN. The proposed security aware VBS place-
ment framework is described in details in Sect. 3. The results and discussions are
presented in Sect. 4. Then Finally the conclusions are presented in Sect. 5.

2 Related Work
Researchers have proposed several VBS placement strategies to improve resouce
utilization and improve energy consumption within 5G C-RAN. However, to the
best of our knowledge, there are no efforts on VBS placement strategies in 5G
C-RAN to minimize the security risks for the BS cloud platform. The author
in [4] proposed a VBSs virtualization scheme that minimizes the power con-
sumption with a linear computational complexity order. The scheme is based on
a heuristic simulated annealing (HSA) algorithm, which combines a bin pack-
ing algorithm with simulated annealing. Simulation results show that the HSA
effectively decreases system power consumption when compared to standard
approaches. However, the simulated annealing VBS placement scheme does not
consider security in migration of VBSs. Authors in [5] proposed a BBU reduc-
tion scheme for C-RAN that allocates VBSs to RRHs based on the imbalance of
subscribers in office/residential areas. A set of upper limit of VBS utilization is
defined to avoid overloading of the VBS. However, the author did not consider
security in their VBS consolidation scheme. Namba et al. in [6] proposed a base-
band unit (BBU) reduction network architecture called Colony-RAN due to its
 Security Aware 5

ability to flexibly change cell layout by changing the connections of BBUs and
RRHs in respect to traffic demand. However, the proposed method has frequent
ping pong reselections of RRH to BBU. The author in [7] proposed a model for
reducing power consumption in H-CRAN by turning off the BBU in the cloud.
However, the author assumes that all the BBUs in the BS cloud operate at
full load which is unrealistics. The author in [8] proposed a VBS virtualization
scheme that minimizes the power consumption of the BS cloud. The VBS vir-
tualization problem is formulated as a bin packing problem, where each VBS is
treated as a bin with finite computing expressed in million Operations per Time
Slot (MOPTS). The dynamics of the cell traffic load is treated as an item that
needs to be packed into the bins with the size equal to the computing resources
in MOPTS, required to support the traffic load. Nevertheless, security has not
been considered. Authors in [9] proposed a C-RAN system using virtualization
technology on general purpose processors (GPPs) where BBUs are dynamically
provisioned according to traffic load. In [10], Cheng et al. developed an energy
efficient C-RAN system with a reconfigurable backhaul that allows 4 BBUs to
connect flexibly with 4 RRHs using radio-over-fiber technology. The backhaul
architecture allows the mapping between BBUs and RRHs to be flexible and
changed dynamically to reduce energy consumption in the BBU pool. However,
the paper assumes static user traffic whereas in reality BS traffic is dynamic.

3 Proposed Security Aware VBS Placement (SAV)

3.1 System Model

The proposed architecture of SAV is shown in Fig. 1. The system comprise of 5G

heterogeneous network which is made up of macro RRH overlaid by femto, pico,
micro and relay RRHs. The RRHs from the radio side are connected to the BS
cloud via high speed, low latency fiber cables. These connection is called the fron-
thaul [2]. The fronthaul in the BS cloud is first connected to the dispatcher which
route data to their respective VBSs within physical machines (PMs) according
to some VBS placement rules from the SAV module in the controller. The SAV
module in the controller contains the VBS placement framework which will be
explained in details in Sect. 3.3. The SAV takes all VBSs and PMs as inputs and
output the VSB-PM map. The BS cloud also comprise of the PMs which host
VBSs. The PMs contains hypervisors which runs on top of the PM operating
system (OS). The hypervisor is responsible for creating VBS on the PMs. Each
RRH has its own VBS.

3.2 VBS and PM Security Evaluation

The security evaluation procedure consists of evaluating the risks of both the
VBSs and PMs in the BS cloud.
6 T. Sigwele et al.

BS Cloud Fiber fronthaul

Controller
SAV Dispatcher

PM PM PM

VBS VBS VBS VBS VBS VBS

Hypervisor Hypervisor Hypervisor

OS OS OS

Fig. 1. The proposed SAV architecture.

VBS Security Evaluation: First, we quantify each VBS’s vulnerabilities

based on the US National Vulnerability Database (NVD) [11], in which all
vulnerabilities called common vulnerability and exposures (CVE) are scored
according to the Common Vulnerability Scoring System (CVSS) [3]. The NVD
is the repository which provides CVSS scores for all CVE vulnerabilities. At
present, NVD contains information about over 69,000 CVE vulnerabilities (as of
04/27/2015) [3]. NVD was created by the government of United States to help
the Department of Homeland Security to warn public about common computer
vulnerabilities. These vulnerabilities now include latest attacks on cloud com-
puting environments. The CVSS base score is the primary metric and describes
the severity of the vulnerability. The base score uses an interval scale of (0, 10)
to measure the severity of vulnerabilities.
First a check is perfomed with NVD to collect potential vulnerabilities in the
VBSs. Vulnerability scanner tools, such as, Nessus and Qualys are available to con-
duct this job. Since it is possible for a VBS to have more than one vulnerabilities,
it is usually desirable to aggregate the scores of individual vulnerabilities for each
VBS. The VBS vulnerability can be divided into three descrete levels: low compro-
mise, medium compromise and high compromise. Suppose the collected vulnera-
bilities of all the VBSs within P Mi are stored in the set VP Mi = {Vj |j = 1, 2, ...n}.
Then the compromise level of a V BSj denoted CV BSj can be given as:
 Security Aware 7

Vj
CV BSj = n (1)
m=1 Vm

Physical Machine Security Evaluation: The VBS compromise level has

been computed. Next the probability of survivability for each PM is computed
based on the compromise level (CV BSj ) of each hosted VBS. The survivability
probability is the possibility that all owned VBSs of a PM can survive during
the attacks in one or more of the VBSs. If any of the VBSs in the PM is com-
promised, then the PM will also be compromised with high probability. Given
the compromise level of all VBSs in P Mi as {CV BS1 , CV BS2 , ..., CV BSn }, then
the survivability score for P Mi denoted as SP Mi is given as:
n

SP Mi = (1 − CV BSj ) (2)
m=1

The survivability quantifies the PM security level which correspond to three

discrete states: low survivability, medium survivability and high survivability.
Next the VBS placement based on the VBS and PM security levels is described
in the next section.

3.3 The SAV Framework

From the previous discussions, we can learn that the success of attacks highly
depends on the placement strategy of the cloud. Thus, our approach is to find a
systematic solution to place VBSs into PMs which can reduce security risks in
the BS cloud. If a VBS has low compromise, it will be infeasible to place it on
a PM with low survivability. Also, if a PM is of high survivability, it would be
infeasible to place a VBSs with high compromise which may results in a negative
survivability impacts on the PM. In the proposed SAV scheme, the compromise
level of the VBS and the survivability of the PM are taken into consideration
when performing VBS placement. In SAV, it will be reasonable to place a VBS
of low compromise level to a PM with high survivability. Also, in SAV, the VBSs
are placed to PMs such that the CV BSj state match the SP Mi state as follows:
(i) A low compromise VBS is placed on high survivability PM
(ii) A medium compromise VBS is placed on medium survivability PM
(iii) A high compromise VBS is placed on low survivability PM
Figure 2 shows the flowchart of how the proposed SAV scheme operate. The
input to the SAV scheme is all the VBSs and all the PMs in the BS cloud. First
for every VBS, the security evaluation is performed by computing the comprise
level of each VBS using Eq. (1). Then the compromise level (CV BSj ) is used for
calculating survivability of each PM (SP Mi ). A test is then performed on each
VBS to check whether the compromise state of that VBS match the survivability
of the PM. If not, then all the other VBS compromise levels are checked and if
they match the PM survavibility state, the VBS is then allocated to the PM.
The process will continue untill all VBSs are allocated to PMs securely.
8 T. Sigwele et al.

Start

Input: all VBSs, all PMs

VBS Security Evaluation

PM Security Evaluation

For

For

Does state No
match j=j+1
state
Yes
A

Are all VBS No

allocated i=i+1
to PMs?
Yes

End

Fig. 2. The flowchart of SAV framework.

4 Results and Discussion

To analyse the proposed SAV framework, a 2 tier H-CRAN long term evolution
(LTE) layout is considered where macro RRHs are overlaid by small cells RRHs.
Up to 200 RRHs are considered and each RRH has its own VBS in the BS cloud.
As explained before, to the best of our knowledge, there are no security aware
VBS placement schemes in 5G C-RAN at the moment, this paper is the first
to proposed such framework. Therefore, the proposed SAV will be compared
with the HSA scheme in [4] which consider VBS placement for saving energy
consumption in the BS cloud without considering security. Figure 3 shows the
effects of increasing the number of VBSs in the network on the PM survivability
for all the schemes. For both the proposed SAV scheme and the HSA schemes,
as the number of VBSs in the BS cloud increases, the PM survivability decreases
because the chances of comprimised VBSs are higher with the increase of VBSs.
Also it can be observed in the diagram than the SAV scheme performs better than
the HSA scheme by 16% with high chances of survivability of an average of 0.65
compared to the HSA scheme with chances of survivability of 0.5. This is because
 Security Aware 9

0.9

0.8

0.7
PM survavibility

0.6

0.5

0.4

0.3

0.2 SAV
HSA
0.1

0
20 40 60 80 100 120 140 160 180 200
Number of VBSs

Fig. 3. The effects of VBS variation on PM survivability.

35

30
Power Consumption (KW)

25

20

15

10
SAV

5 HAS

0
20 40 60 80 100 120 140 160 180 200
Number of VBSs

Fig. 4. The effects of VBS variation on power consumption.

the SAV scheme considers security of VBSs and PMs before performing the VBS
placement. Figure 4 shows the effects of varying the number of VBSs on the power
consumption in the BS cloud. For both the schemes, as the number of VBSs
increases, the power consumption also increases as more VBSs requires more
PMs which will consume more power. On average, the SAV scheme consumes
15% more energy compared to the HSA scheme. This is because, even though
10 T. Sigwele et al.

the SAV scheme provide some security benefits, this results in the penalty of
more power of 15% being consumed by the SAV scheme as a results of the VBS
placement scheme security overheads which requires more PMs to be deployed.

5 Conclusion
The existing researches mostly focus on how to maximize resource utilization and
reduce energy consumption in fifth generation (5G) cloud radio access networks
(C-RAN) using virtual base station (VBS) placement. However, security issues in
the context of VBS placement within 5G C-RAN have been rarely addressed. In
this paper, a security aware VBS placement (SAV) scheme is proposed where the
placement of VBSs to physical machines (PMs) considers the security levels of
both the VBS and the PM. A rigorous simulation study is conducted for validat-
ing the proposed scheme, which shows a significant security improvement since
the proposed SAV scheme outperforms the heuristic simulated scheme (HSA) by
16% with power consumption penalty of 15%.

References
1. 5G: A Technology Vision, Huawei Tech. White Paper (2013)
2. Chen, K., Duan, K.: C-RAN: the road towards green RAN. In: China Mobile
Research Institute White Paper (2011)
3. Yuchi, X., Sachin, S.: Enabling security-aware virtual machine placement in IaaS
clouds. In: Military Communications Conference, MILCOM 2015, pp. 1554–1559.
IEEE (2015)
4. Qian, M., Hardjawana, W., Shi, J., Vucetic, B.: Baseband processing units virtu-
alization for cloud radio access networks. IEEE Wirel. Commun. Lett. 4, 189–192
(2015)
5. Namba, S., Warabino, T., Kaneko, K.: BBU-RRH switching schemes for central-
ized RAN. In: 2012 7th International ICST Conference on Communications and
Networking in China (CHINACOM), pp. 762–766. IEEE (2012)
6. Namba, S., Matsunaka, T., Warabino, T.: Colony-RAN architecture for future
cellular network. In: Network and Mobile Summit (FutureNetw), pp. 1–8. IEEE
(2012)
7. Domenico, A., Katranaras, E.: Energy efficiency benefits of RAN-as-a-service con-
cept for a cloud-based 5G mobile network infrastructure. IEEE Access 2, 1586–1597
(2014)
8. Qian, M., Hardjawana, W., Vucetic, B.: Baseband processing units virtualization
for cloud radio access networks. IEEE Wirel. Commun. Lett. 4(2), 189–192 (2015)
9. Kong, Z., Gong, C., Xu, Z.: eBase: a baseband unit cluster testbed to improve
energy-efficiency for cloud radio access network. In: IEEE International Conference
on Communications (ICC), pp. 4222–4227. IEEE (2013)
10. Liu, C., Sundaresan, K., Jiang, M., Rangarajan, S., Chang, G.K.: The case for
re-configurable backhaul in cloud-RAN based small cell networks. In: 2013 Pro-
ceedings of IEEE INFOCOM, pp. 1124–1132. IEEE (2013)
11. United State Government: US National Vulnerability Database. https://nvd.nist.
gov. Accessed Mar 2017
 DoS Attack Impact Assessment
on Software Defined Networks

Abimbola Sangodoyin1(B) , Tshiamo Sigwele1 , Prashant Pillai2 ,

Yim Fun Hu1 , Irfan Awan1 , and Jules Disso3
1
Faculty of Engineering and Informatics, University of Bradford,
Bradford, West Yorkshire BD7 1DP, UK
{a.o.sangodoyin,T.Sigwele,Y.F.Hu,i.awan}@bradford.ac.uk
2
Faculty of Technology, Design and Environment, Oxford Brookes University,
Oxford, UK
[email protected]
3
Nettitude Limited, Warwickshire CV31 3RZ, UK
[email protected]

Abstract. Software Defined Networking (SDN) is an evolving network

paradigm which promises greater interoperability, more innovation, flex-
ible and effective solutions. Although SDN on the surface provides a
simple framework for network programmability and monitoring, few has
been said about security measures to make it resilient to hitherto secu-
rity flaws in traditional network and the new threats the architecture is
ushering in. One of the security weaknesses the architecture is ushering
in due to separation of control and data plane is Denial of Service (DoS)
attack. The main goal of this attack is to make network resources unavail-
able to legitimate users or introduce large delays. In this paper, the effect
of DoS attack on SDN is presented using Mininet, OpenDaylight (ODL)
controller and network performance testing tools such as iperf and ping.
Internet Control Message Protocol (ICMP) flood attack is performed on
a Transmission Control Protocol (TCP) server and a User Datagram Pro-
tocol (UDP) server which are both connected to OpenFlow switches. The
simulation results reveal a drop in network throughput from 233 Mbps
to 87.4 Mbps and the introduction of large jitter between 0.003 ms and
0.789 ms during DoS attack.

Keywords: Software Defined Networks · DoS · Network security

1 Introduction

Computer networks have become part of our everyday lives from government to
commercial enterprises to individuals [1]. These networks are built from large
number of devices such as routers, switches and middle boxes with complex pro-
tocols running on them. Network administrators are saddled with the responsi-
bility of configuring these vendor-specific devices and configuration policies are

c ICST Institute for Computer Sciences, Social Informatics and Telecommunications Engineering 2018
P. Pillai et al. (Eds.): WiSATS 2017, LNICST 231, pp. 11–22, 2018.
https://doi.org/10.1007/978-3-319-76571-6_2
12 A. Sangodoyin et al.

implemented on them. As a result, network management and dynamic response

to events and applications are arduous and prone to error.
In addition to the complexity of configuration, operators have little options
or mechanisms to respond to difficulties and enforce the required policies in
dynamic environments [2]. Similarly, in the face of growing traffic and demand
for more data rate from consumers, the service providers need to keep up with
the pace through investments in bigger and faster links and edge routers, even
though revenues are growing quite slowly [3].
In view of the afore mentioned challenges, the need for a cost effective and
programmable network which is robust enough to meet the demand of users is
imperative. Thus, the emergence of Software Defined Networks (SDN). SDN has
created commendable avenues to overcome age-old problems in networking, while
simultaneously enabling the introduction of complex, secure and reliable network
policies for next generation networks [4]. As a revolutionary concept, SDN alters
existing networks by separating the forwarding functionalities of existing devices,
known as data plane from control element, known as control plane [5].
The future of SDN mainly lies in its acceptance and deployment. Technology
and its deployment take years before it can be available to end users due to
standardisation process and Request for Comments (RFCs). Speculations how-
ever remain as to whether same should be expected for SDN or not. According
to [1] a proposal for open and programmable network is presented. The need for
researchers to run experiment on campus network using an OpenFlow switch is
further emphasised in [1]. In line with this, ETHANE, a new network architec-
ture for enterprise was suggested in [6]. For the proposed architecture, in [6],
ETHANE switch does not need to learn addresses, support Virtual Local Area
Networks (VLANs) or check for source-address spoofing and it has been deployed
in a campus environment. The work in [6] was augmented when Google, deployed
B4 using OpenFlow switches in their Wide Area Networks (WAN) data centre [7].
Also, with the advent of a Linux foundation collaborative project, OpenDaylight
(ODL) [8] platform and VMware NSX virtualisation platform, global acceptance
and deployment is envisaged to be no longer far from reach.
In spite of the programmability, flexibility, universal connectivity and decen-
tralised control, which were critical to the success of SDN, these features are
at odds with making it more secure. The SDN platform can bring with it sev-
eral security breaches which include an increased potential for Denial-of-Service
(DoS) attacks due to controller centralisation and flow table limitations in network
devices [9]. Furthermore, abstraction of flows and underlying hardware resources
make it easier for harvesting of intelligence which can be used effortlessly for fur-
ther exploitation and reprogramming entire network by malicious user [4].
In this paper, the impact of DoS attack on SDN is presented. The simulation
has been perfomed using mininet and OpenDaylight controller tools and the sim-
ulation result shows that DoS flooding attack on SDN network can degrade net-
work performance by decreasing network throughput and introduce large jitter.
This paper is structured as follows: Sect. 2 presents related works on the SDN
architecture, vulnerabilities in the SDN architecture and DoS attacks on SDN.
 DoS Attack 13

Fig. 1. SDN architecture illustrating the data, control and application layers.

The experimental method and tools are presented in Sect. 3. Then Sect. 4 shows
the experimental set-up. The results and analysis are presented in Sect. 5. Finally,
the conclusion and future work are presented in Sect. 6.

2 Related Work
2.1 SDN Architecture
SDN architecture encompasses the complete network platform. It is a modular
approach that defines chain of command and interoperability within network.
Unlike traditional network, the intelligence of data plane devices is removed to
a logically centralised control system [10]. Figure 1 presents the SDN architec-
ture showing the data/infrastructure, control and application layer. In an SDN
architecture, there are two main elements: the controllers and the forwarding
devices. A forwarding device is a hardware or software element specialised in
packet forwarding and based on a pipeline of flow tables where each entry of a
flow table has: a matching rule, action to be executed on matching packets and
counters that keep statistics of matching packets [4]. The controller serves as the
brain of the network and it deals with management of network state. Below is a
description of various layers:
Infrastructure layer: This layer is also known as data plane. It consists
of simple forwarding elements without embedded control or software to take
autonomous decisions. It is accessible through the southbound interface and
allows packet switching and forwarding.
14 A. Sangodoyin et al.

Control layer: This layer consists of SDN controllers providing a consolidated

control functionalities through Application Programming Interfaces (APIs). The
crucial value of the controller is to provide abstractions, essential services, and
common APIs to developers. Three communication interfaces allows the con-
troller to interact: northbound, southbound and the east/westbound interfaces.

(i) Southbound Interfaces: Southbound interface allows the controller and for-
warding elements to interact in the infrastructure layer, thus being the
crucial instrument for clearly separating the control and data plane func-
tionality.
(ii) Northbound Interfaces: This interface is the connecting bridge between
application layer and control layer. It enables the programmability of the
controllers by exposing the data models and other functionalities within the
controllers for use by applications at the application layer. The northbound
interface is mostly a software ecosystem, hence, a common northbound
interface is still an open issue.
(iii) East/Westbound Interfaces: This interface is a special communication inter-
face envisioned for distributed controllers to synchronise state for high avail-
ability. Its function include import/export data between controllers and
monitoring/notification capabilities to check if a controller is up or notify
a takeover on a set of forwarding elements.

Application Layer: The application layer consists of end-user business appli-

cations and network services. Example of application that runs here is network
virtualisation. Network policy is also defined here.

2.2 Vulnerabilities in SDN Architecture

A number of security analyses has been carried on the vulnerabilities in SDN.

Adnan et al. in [5] identified the state of art in SDN security solutions with respect
to each layer of SDN architecture. The work focuses on possible security attacks in
SDN which could be executed. However, no solution to identified threats is pre-
sented. A comprehensive survey of security in SDN is presented in [11,12], the
authors identified vulnerabilities introduced by separation of control and data
plane. Sandra et al. in [11] presents an overview of SDN security and itemise
research work coupled with solution to security issues in SDN. In [12] classifica-
tion is done using the STRIDE approach and possible SDN security controls is
proposed. The concept of offering SDN security as a service is presented in [13].
Kreutz et al. in [2] presents a high level security analysis. Seven main potential
threat vectors are presented. Three of the seven identified threat vectors are
specific to SDN and relates to the three planes present in SDN architecture.
The analysis does not present SDN as a less secure network but triggers the
need for innovative ways of responding to the new threats arising from network
programmability. The authors state the consequences of these threats in SDN
and solutions to the seven threat vectors was proposed.
 DoS Attack 15

In [14], a feasibility study on attacking SDN is carried out by fingerprinting to

ascertain usage of SDN/OpenFlow switches by the network. The SDN network is
then subjected to a specifically crafted flow requests from the data plane to the
control plane to exhaust the network resources. Another security vulnerabilities
was analysed in ProtoGENI [15]. The authors explored three potential security
issues as follows:

(i) Resource connection: Once a malicious user obtains access to one exper-
iment node, attacks can easily be launched by utilising the huge ProtoGENI
computing resources as a launchpad to harm existing internet users.
(ii) Wireless Nodes Distribution: Network sniffing or spoofing can be done
here to identify desired node for launching attacks.
(iii) Virtualization Technology: In ProtoGENI virtualisation, ProtoGENI
resources are shared among as many user as possible. Any bug or compro-
mise from a single device will expose other users in sharing resources to
attacks.

The Authors discovered the possibility of using ProtoGENI resources to launch

flooding attack to the wider internet. Also, the possibility of compromising con-
fidentiality and availability of other ProtoGENI users is high.

2.3 DoS Attack on SDN

DoS and Distributed DoS (DDoS) attack remains one of the severe network secu-
rity problems in both traditional network and SDN. Due to separation of control
and data plane, an attacker could saturate the controller with malformed packets
requiring a flow rule decision. On the other hand, the flow table of the infras-
tructure device can be overwhelmed with malicious packets. To address bottle-
necks of potential saturation attack, AVANT-GUARD [16] introduce connection
migration to reduce amount of data-to-control-plane interactions. The method
enables the data plane to shield the control plane from saturation attacks. How-
ever, the data plane itself is subject to attack. Similarly, a backup strategy which
offers resilience against failures in a centralised controlled network is presented
in [17]. This approach is an attempt to solve single point of failure bottleneck
and it provides seamless transition between primary controller to a back-up con-
troller. However, this solution is limited to centralised implementation and it
also raises concern in terms of trust between the east-west interface communi-
cations. In addition, Braga et al. [18] proposed lightweight, a new method for
detecting DDoS. The proposed method boasts of high rate of true positives and
low rate of false alarm using Self Organising Maps (SOM) for flow analysis. The
lightweight method consider median values in training the SOM. The drawback
of this method is that false negatives are reported when the attack parameter is
set to a low value.
The controller has been compared to an operating system capable of manag-
ing applications through programmatic interface [19]. Similarly, ETHANE was
built to provide network-wide fine-grain policy using a centralised declaration
16 A. Sangodoyin et al.

and enforcing it [6]. While the concept of a centralised controller allow the sim-
plification of policy enforcement and management tasks for network managers,
it creates quite a number of bottle necks. In [9], analysis of SDN implementation
key challenges has been carried out. The authors opined deployment of SDN
technology will contribute to the vision of future communications if outstand-
ing challenges were resolved. In [20] the possibility of DoS attacks and poor
rule design that can lead to saturating volumes of controller queries is discussed.
Though OpenFlow vulnerabilities in terms of lack of adoption of Transport Layer
Security (TLS) for controller-switch communication is highlighted in [20], a num-
ber of vulnerabilities proposed was not verified in the work.

3 Experimental Method and Tools

In this experiments, Mininet is used [21]. Mininet is an open source network

emulator devoted entirely to OpenFlow architecture and SDN implementation.
For the controller, ODL controller is used [22]. ODL integrates open source,
open standards and open APIs to deliver SDN platform to make networks more
programmable and adaptive. DoS attacks usually engage numerous compromised
hosts and a rich topology to launch a successful attack on its victim. While
our scenario is much simpler than what is obtainable in real world attacks, we
deliberately chose such a low-complexity set-up to expose and analyse the impact
of DoS attack on SDN. Common testing tools such as, ping and iperf are used
to generate traffic between host and servers. Figure 2 shows the methodology
flowchart with each step explained.

Step1 Start

Create UDP and TCP

Step2
server

Launch
Hosts make requests
Step3 ICMP attack
from UDP and TCP
servers

Step4 Measure throughput

and jitter

Plot graph of
Step5
throughput and jitter

Step6 End

Fig. 2. Methodology flowchart.

 DoS Attack 17

Step 1: Start Mininet and ODL controller

Sudo mn --custom scenario.py --topo

--controller = remote,ip = x.x.x.x

Where x.x.x.x represents the ip address of the remote controller.

Check connectivity using

$mininet > net

Step 2: Create UDP and TCP server

UDP: iperf -s -u -p 5566 -i1

TCP: iperf -s -p 5566 -i1

The TCP server is made to listen on port 5566 with a default window size of
85.3 KB. Similarly, UDP server is made to listen on port 5566 with a default
UDP buffer size of 208 KB while receiving 1470 bytes datagrams and the result
is monitored every 1 s.
Step 3: Hosts make requests from TCP server and UDP server

TCP: -iperf -c x.x.x.x -p5566 -t100

UDP: -iperf -c x.x.x.x -u -t100 -p5566

Step 4 and Step 5: Results were extracted using AWK file and results plotted
using MATLAB. Then, malicious hosts 5 and 6 launched flooding attack on the
servers (similar to step 3). Legitimate traffic is started at the beginning of an
experiment, and an attack is launched shortly after for a duration of 100 s.
Step 6: End

mininet# ctrl z (end mininet)

Sudo mn -c (clear topology)

4 Experimental Set-Up
In this section, a series of experiments are performed to verify the effects of DoS
attack in the SDN network. The experimental setup is shown in Fig. 3. To create
the scenario in Fig. 3, many software and tools are used as shown in Table 1.
There are two servers and four switches in the network. Each switch has a host
connected to it. The Transmission Control Protocol (TCP) server is connected
to OpenFlow switch1 while User Datagram Protocol (UDP) server is connected
to OpenFlow switch3. ICMP flood attack will be launched against both servers
by malicious hosts 5 and 6.
18 A. Sangodoyin et al.

Fig. 3. Experimental setup.

Table 1. Simulation parameters

Simulation Details
Platform/Enviroment -Oracle Virtual Box as base environment for simulations
-Ubuntu OS as base for Mininet v2.2 emulator
-Ubuntu Server as base for OpenDaylight (Boron) controller
-Host CPU as intel core i7, 12 G RAM
Attack tool Hping tool - Hping3 used for flooding attack

5 Results and Analysis

As discussed in the experimental setup, we simulate for two different scenarios;
TCP and UDP requests under normal operating condition and under attack.
The results for these scenarios are discussed below.

5.1 Effect of DoS Attack on Throughput

Figures 4 and 5 shows a significant drop in throughput due to malicious
behaviour (ICMP flood attack) being executed by two attacking nodes. The aver-
age throughput for requests made from host 4 to the TCP server is 214 Mbits/s
for a total of 2.5 GB of information transferred in 100 s. Similarly, the aver-
age throughput of host2 requests from TCP server is 233 Mbits/s for a total of
2.72 GB of information transferred.
Notice that Host 2 shows a better bandwidth utilisation than Host 4 and the
reason for this is not far-fetched; they are both connected to OpenFlow switch 1.
While the better bandwidth utilisation is seen as an advantage here, it is a major
security risk and attractive honeypot to launch attack against the server. The
impact of this connection is felt when the server is subjected to ICMP flood attack.
During attack, the average throughput dropped to 106 Mbits/s from 214 Mbits/s
 DoS Attack 19

H4 requests from Server1

350
TCP with attack
TCP without attack
300

250

Throughput (Mbps) 200

150

100

50
0 10 20 30 40 50 60 70 80 90 100
Time(sec)

Fig. 4. TCP requests from host 4 to TCP server under ICMP attack.

H2 requests from Server1

350

300

250
Throughput (Mbps)

200

TCP without attack

150
TCP with attack

100

50
0 10 20 30 40 50 60 70 80 90 100
Time(sec)

Fig. 5. TCP requests from host 2 to TCP server under ICMP attack.

recorded for H4 requests from server1 without ICMP flood attack. The impact of
the attack launched by host 5 and 6 became noticeable after 15 s of transmission
and the bandwidth utilisation degraded for the rest of the transmission. The trend
is similar for host2 requests from TCP server as degradation started after 8 s of
transmission and degraded for the rest of the transmission. The average through-
put for h2 requests from TCP server dropped from 233 Mbits/s to 87.4 Mbits/s
when the server is under attack. The degradation is more severe for host 2 when
under attack even though higher throughput is recorded during normal operation.
Hence, the need for better network design, traffic isolation based on priority for
mission-critical network and dynamic proactive ways of addressing DoS attacks
when the system is under serious attack.
Exploring the Variety of Random
Documents with Different Content
 ON A COLLECTION OF BIRDS LATELY
MADE BY MR. F. STEPHENS IN ARIZONA.
BY WILLIAM BREWSTER.

(Continued from p. 94.)

33. Peucedramus olivaceus (Giraud) Coues. Olive-headed

Warbler.—The Olive-headed Warbler, one of Giraud’s famous
“sixteen” Texas species, has found an unquestioned place in our
fauna only on the strength of three Arizona specimens, taken by Mr.
Henshaw at Mount Graham, in September, 1874. Accordingly, the
acquisition of the fine series catalogued below can scarcely fail to be
a matter of much interest. As will appear from the accompanying
data, Mr. Stephens met with the bird in only a single locality in the
Chiricahua Mountains where it was apparently not uncommon in
March: but he writes of a previous specimen (an adult male) taken
among the Santa Catarina Mountains, in February, 1880, a date
which seems to imply that the species winters in the latter range. His
observations throw no light on its still unknown breeding haunts.
The specimens obtained during the past season were found in pine
woods on the mountain sides at an elevation of from ten to twelve
thousand feet. Although individuals often occurred not far from one
another, two were rarely seen in actual companionship. The only
exception to this is noted under date of March 24, when a small flock
was met with on a steep slope near the summit of one of the
mountains. In their actions these Warblers reminded Mr. Stephens
of Dendrœca occidentalis. They spent much of their time at the
extremities of the pine branches where they searched among the
bunches of needles for insects, with which their stomachs were
usually well filled. Occasionally one was seen to pursue a falling
insect to the ground, where it would alight for a moment before
returning to the tree above. The only song heard consisted of “a few
low notes” which were rarely uttered, but a peculiar “cheerp” was
repeated at frequent intervals.
The examples before me illustrate a fact which I do not find
mentioned by previous writers, viz., that during the first year the
males wear a plumage similar to that of the females. I have three in
this condition; two of them, although in unworn dress, are absolutely
undistinguishable from adults of the opposite sex; the third (No. 77),
however, has the throat appreciably tinged with the brownish-saffron
of the adult male. The females show some variation in respect to the
dusky patch on the side of the head. In most of them it is confined to
the auriculars, and even there is much mixed with yellow; but No. 46
has a continuous, dull-black stripe extending from the bill through
the eye, and spreading over the auriculars in a broad, well-marked
patch. Nos. 94 and 101 differ from the others in having the crown so
slightly washed with olive-green that the whole upper surface is
nearly uniform, a condition which I take to be the immature one of
this sex. The adult males show but little individual variation. Both
sexes and all ages have the basal half of the lower mandible light
brown.
44. ♂ ad., Morse’s Mill, Chiricahua Mountains, March 14, Length,
5.10; extent, 9; wing, 3.12; tail, 2.35; culmen, .56; tarsus, .72.
45, ♂ ad., same locality and date. Length, 5.40; extent, 9.20; wing,
3.16; tail, 2.55; culmen, .55; tarsus, .69. Iris dark brown.
72, ♂ ad., Morse’s Mill, March 19. Length, 5.40; extent, 8.90.
91, ♂ ad., Morse’s Mill, March 24. Length, 5.40; extent, 9; wing,
3.08; tail, 2.50; culmen, .55; tarsus, .75.
92, ♂ ad., same locality and date. Length, 5.20; extent, 8.90.
102, ♂ ad., Morse’s Mill, March 25. Length, 5.30; extent, 8.80;
wing, 3.10; tail, 2.44; culmen, .56; tarsus, .75.
77, ♂ im., Morse’s Mill, March 20. Length, 5.20; extent, 8.90;
wing, 3.03; tail, 2.37; culmen, .55; tarsus, .77. In plumage of the ♀.
90, ♂ im., Morse’s Mill, March 24. Length, 5.10; extent, 8.50;
wing, 2.85; tail, 2.30; culmen, .56; tarsus, .71. Same remarks.
103, ♂ im., Morse’s Mill, March 25. Length, 5.10; extent, 8.50;
wing, 2.90; tail, 2.33; culmen, .57; tarsus, .67. Same remarks.
 46, ♀ ad., Morse’s Mill, March 14. Length, 5.20: extent, 8.50; wing,
2.93; tail, 2.35; culmen, .56; tarsus, .73.
47, ♀ ad., same locality and date. Length, 5; extent, 8.30; wing,
2.87; tail, 2.18; culmen, .58; tarsus, .73.
81, ♂ ad., Morse’s Mill, March 21. Length, 5; extent, 8.50; wing,
2.76; tail, 2.35; culmen, defective; tarsus, .72.
93, ♀ ad., Morse’s Mill, March 24. Length, 5.20; extent, 8.80.
94, ♀ ad., same locality and date. Length, 5; extent, 8.20: wing,
2.84; tail, 2.18; culmen, defective; tarsus, .71.
101, ♀ ad., Morse’s Mill, March 25. Length, 5.10; extent, 8.50:
wing, 2.87; tail, 2.22; culmen, .58; tarsus, .75.
34. Dendrœca æstiva (Gmel.) Baird. Yellow Warbler.

210, ♂ ad., Cienega Station, April 16. Length, 5; extent, 7.50; wing, 2.75; tail,
2.20; tarsus, .74. “Iris dark brown; bill dark horn color above, lighter below; legs
pale brown. Common in the migrations.”

35. Dendrœca coronata (Linn.) Gray. Yellow-rumped

Warbler.—Chiricahua Mountains; a single specimen, taken March
26.
From its general dispersion over North America, the Yellow-
rumped Warbler was of course to be expected in Arizona, at least as a
visitor, but I cannot learn that it has been previously detected within
the limits of that Territory. Mr. Stephens, however, sends me an
adult female which must be referred to coronata, although it is in
some respects peculiar, if not intermediate between that species and
auduboni. The wing-bands are as distinctly separated as in coronata
(with females and immature males of both species this character is
not always well-defined), and the throat, generally, is equally white,
but on its left side, adjoining the maxillary line, there is a small patch
of the faintest possible yellow. The light superciliary stripes, which
should be at least indicated in female coronata, are also entirely
wanting.
114, ♀ ad., Chiricahua Mountains, March 26. Length, 5.50; extent,
8.70; wing, 2.98; tail, 2.52. “Iris brown.”
 36. Dendrœca auduboni (Towns.) Baird. Audubon’s Warbler.

343, ♂ ad., Tucson, May 7. Length, 5.80; extent, 9.52; wing, 3.05; tail, 2.75. “Iris
dark brown; bill and legs black.”
37. Dendrœca nigrescens (Towns.) Baird. Black-throated Gray Warbler.
—On April 1, Mr. Stephens secured five males of this species among the Chiricahua
Mountains. The only additional specimens in the collection are two females taken
late in the season (No. 203, ♀ ad., Cienega Station, April 15. No. 357, Santa Rita
Mountains, May 12.).

38. Dendrœca townsendi (Nutt.) Baird. Townsend’s Warbler.

2.98, ♀ ad., Tucson, April 28. Length, 5.10; extent, 7.70; wing, 2.45. “Iris dark
brown; bill and legs black; soles of the feet yellow. Among mesquites.”
373, ♂ ad., Santa Rita Mountains, May 13. Length, 5.30; extent, 8.10; wing, 2.64.
374, ♀ ad., same locality and date. Length, 4.90; extent, 7.40; wing, 2.44. “Iris
dark brown; soles of feet yellowish. Water oaks of foot-hills; very fat.”
Even the most adult males of this species seem to have the throat-patch slightly
sprinkled with yellow. At least I have yet to see one with the black absolutely pure
and unmixed.

39. Siurus nævius (Bodd.) Coues Northern Water Thrush.—A

single specimen taken May 4, at Tucson. It was among willows on the
borders of a stream.
This example differs from New England ones in being darker
above and less yellowish beneath. In these respects, as well as some
minor ones, it resembles a rather peculiar style from West Virginia to
which I once called attention.[68] Mr. Ridgway kindly furnishes the
following opinion regarding its relationship with S. notabilis. “The
Siurus from Tucson is very different in proportions from the type of
notabilis, with which I have compared it, but it may be a small
individual of that form. The wing is about the same length, but the
bill and tail are very much shorter, and the tarsi more slender. The
color above is grayer, the streaks beneath much narrower, and the
spots on the throat much smaller.” Notabilis, based as it is on a
single specimen, and instituted in a species which varies to an
unusual degree in size, color and markings, seems to me, however, to
be, at best, a very doubtful race.
329, ♂ ad., Tucson, May 4. Length, 6.20; extent, 9.50; wing, 3.10;
tail, 2.32; tarsus, .85; culmen, .64. “Iris brown; bill black above,
brown below; legs light brown. Very fat. Stomach contained insects.”
40. Geothlypis macgillivrayi (Aud.) Baird. Macgillivray’s
Warbler.—Two specimens collected at Tucson (♀ April 20, ♂ June
8). “I have not found it common in either Arizona or New Mexico.”
41. Geothlypis trichas (Linn.) Caban. Maryland Yellow-
throat.—Mr. Stephens found this species “abundant along streams,”
an experience at variance with that recorded by Mr. Henshaw, who
met with it but twice while in Arizona.
The only specimen taken agrees closely with some examples from
the Truckee River, Nevada, and differs from my eastern
representatives, in having the upper parts yellowish-olive instead of
olive-green; the crown-band much broader and creamy white in
color; the wings and tail longer; the yellow beneath richer, and
extending more over the abdomen. Mr. Ridgway has already called
attention[69] to some of these differences which, as he now writes me,
would be enough to warrant the varietal separation of the western
bird, were it not that specimens from both sections of the country
occasionally vary in such a manner as to invalidate any characters
that could at present be proposed. With the acquisition of better
series, however, it is probable that the representatives of two regions,
as yet undefined, will be found to present sufficiently constant
characteristics to deserve distinctive names.
219, ♂ ad., Cienega Station, April 17. Length, 5.40; extent, 6.90;
wing, 2.16; tail, 2.40; culmen, .55. “Iris brown; bill black, bluish
beneath; legs pale brown.”
42. Icteria virens longicauda (Lawr.) Coues. Long-tailed
Chat.—This bird was observed only in the vicinity of Tucson. The
first specimen was taken April 30, and it soon afterwards became
abundant.

310, ♂ ad., Tucson, April 30. Length, 7.50; extent, 9.40; wing, 3.12; tail, 3.52.
“Bill and legs black.”
318, ♂ ad., Tucson, May 3. Length, 7.70; extent, 9.60; wing, 3.05; tail, 3.61.
 335, ♂ ad., Tucson, May 5. Length, 7.30; extent, 9.70; wing, 3.12; tail, 3.45.
521, ♂ ad., Tucson, June 11. Length, 7.10; extent, 9.40; wing, 3.15; tail, 3.36.

43. Myiodioctes pusillus pileolatus (Pall.) Ridgw. Pileolated

Warbler.
Although Mr. Henshaw referred all his Arizona Black-capped
Flycatchers to pusillus, mine are absolutely typical of pileolatus; in
fact they are brighter than some specimens from Nicasio (California),
the yellow below being richer, and the upper surface more yellowish,
while the bill is equally

narrow and several shades lighter in color. Compared with eastern examples
they of course present an even greater contrast. Dr. Coues was undoubtedly right
in saying (Birds of the Colorado Valley, p. 327) that pileolatus “is not confined to
the Pacific coast region”; but I cannot agree with him in thinking it an inconstant
form. On the contrary, I find its characters, as proposed by Mr. Ridgway, so well
maintained that any one of my western birds can be separated at a glance when
placed in a series of twenty-one specimens from the Atlantic States.
221, ♂ ad., Cienega Station, April 17. Length, 4.70; extent, 6.80; wing, 2.17; tail,
2.23; width of bill below nostrils, .12. “Iris brown; bill dark above, pale brown
below. Common here in willows and underbrush along streams.”
257, ♂ ad., Tucson, April 21. Length, 4.90; extent, 7; wing 2.27; tail, 2.30; width
of bill below nostrils, .12.

44. Setophaga picta Swains. Painted Redstart.—During the

past season this beautiful species was met with only among the
Chiricahua and Santa Rita Mountains, but in 1876 Mr. Stephens
found it in New Mexico, a Territory from which I believe it has not
previously been reported. In the Chiricahua Mountains it was not
uncommon after March 21, and many specimens were taken near
Morse’s Mill, at an elevation of fully seven thousand feet. They
occurred most numerously among pines, in a cañon where they had
been previously observed in April, 1880. This experience, it will be
observed, differs somewhat from that recorded by Mr. Henshaw,
who says: “It appears not to inhabit the high mountains nor the
extreme lowlands, but to occupy an intermediate position, and to
find the rocky hills covered with a sparse growth of oak most
congenial to its habits.”
In the Santa Rita Mountains, where it was rather common in May,
Mr. Stephens had the good fortune to find its previously unknown
nest and eggs. The nest, which is now before me, is large, flat and
shallow. It is composed of bark, coarse fibres from weed-stalks, and
fine, bleached grasses, the latter, with a few hairs, forming a simple
lining. The cup measures 2.10 inches in width by 1 inch in depth;
while the external diameter of the whole structure is rather more
than 5 inches, and its depth about 1.50. The eggs, which were three
in number, measure respectively .64×.51; .64×.50; and .66×.49. They
are clear, dead white, delicately spotted with light reddish-brown, the
markings being sparsely distributed over the general surface of the
egg, and handsomely wreathed about its larger end. Neither nest nor
egg resembles that of S. ruticilla. But a greater surprise is the
character of the nesting-site, which was “under a projecting stone, in
a bank near a small stream.” This position is so unexpected that,
from an unproved collector, I should hesitate to accept the
accompanying evidence of identification, which is a simple statement
that the parent was sitting, and was distinctly seen. But knowing as
well as I do Mr. Stephens’ unusual accuracy and conscientiousness in
such matters I cannot doubt the correctness of his determination,
especially as the Painted Redstart is a bird of such striking colors and
markings that it could not possibly be mistaken by one who is so
familiar with its appearance in life.[70] After all the case is not more
peculiar than that presented among Helminthophilæ by Lucy’s
warbler which, as has just been shown, departs from the normal
nesting habits of the genus and builds in holes, behind loose bark
and in all sorts of unexpected places. The nest above described was
taken May 18, when the eggs were sufficiently advanced in
incubation to show that the clutch was complete.
Mr. Henshaw comparing the sexes, says: “The adult plumage of
the sexes differs little, though the coloration in the female is quite
perceptibly duller throughout. The black is less lustrous; the wings
are blackish brown instead of pure black; the white on the wing
confined to the coverts, and only just visible on the edges of the
secondaries.” These differences, however, are not always maintained
for one of the two adult females before me is quite as bright as the
average male, while the black is not less lustrous, and the white
edging on the secondaries is even broader. The other is more like
those examined by Mr. Henshaw, but seems to be peculiar in having
the sides, with a broad collar across the nape, fine stone-gray.
45. Vireo gilvus (Vieill.) Bonap. Warbling Vireo.—Found
among all the well-timbered mountains visited, but nowhere as a
common bird.
Of the several characters which are said to distinguish var.
swainsoni from gilvus proper, I can appreciate only the slightly
different shape of the bill. The relative length of the wing-quills is an
absolutely inconstant characteristic with birds from any of the
localities represented in my series, while I do not find that western
specimens—at least California and Arizona ones—are either paler or
grayer than many we get in the Atlantic States. Indeed, nearly the
darkest one in my whole suite comes from Arizona. In view of these
facts I cannot regard swainsoni as worthy of varietal recognition.
46. Vireo solitarius cassini (Xantus) Ridgw. Cassin’s Vireo.—
Common among the foot-hills of the mountains.
Mr. Henshaw has so satisfactorily defined[71] the characters which
respectively distinguish the Cassin’s and Plumbeous Vireos from
solitarius proper, as well as from each other, that there is no room
for any further remarks on what, previous to his examination, was a
very tangled problem. The specimens mentioned below are all
unmistakably referable to cassini, although one or two of them
present slight approaches to plumbeus. It is a singular fact that Mr.
Stephens did not meet with any typical examples of the latter race.
209, ♂ ad., Cienega Station, April 16. Length, 5.40; extent, 8.70.
“Iris brown; bill dark horn-color above, lighter below; legs dark
bluish.”
214, ♀ ad., same locality and date. Length, 5.60; extent, 9.10; wing,
3; tail, 2.44.
236, ♀ ad., Tucson, April 19. Length, 5.60; extent, 8.70; wing,
2.89; tail, 2.41.
316, ♀ ad., Tucson, May 2. Length, 5.30; extent, 8.50; wing. 2.71;
tail, 2.26.
 346, ♀ ad., Tucson, May 7. Length, 5.30; extent, 9; wing, 2.76; tail,
2.23. “Very fat. Would not have laid for a long time.”
354, ♂ ad., Santa Rita Mountains, May 11. Length, 5.10; extent,
8.80; wing, 2.82; tail, 2.27. “Iris brown; bill nearly black, bluish at
base below; legs lead-color.”
47. Vireo huttoni stephensi var. nov. Stephens’ Vireo.

Ch. Sp.— ♂ ♀ Similis V. huttoni sed rostro robustiori, alis longioribus. Supra
griseo-cinereus, infra fusco-albidus. Uropygio et marginibus caudæ sordide
virenti-olivaceis. Alis albo bifasciatis; remigibus albo-marginatis. Loris et orbe
circum-oculari (macula fusco-brunnea in palpebra superiori excepta), cinereo-
albis.
Adult ♂ (No. 5,728, author’s collection—collector’s No. 41—Chiricahua
Mountains, Arizona, March 14, 1881. F. Stephens). Bill stout; wings from .30 to .40
inches longer than tail. Above grayish-ash; the crown, vertex and sides of head and
neck nearly pure; the back faintly tinged with olive; the rump and an edging on the
tail feathers, dull olive-green. Wings with two nearly confluent bands on the
coverts, and the outer edges of the inner secondaries, broadly white; outer quills
edged more narrowly with the same color. Beneath brownish or smoky-white, with
a mere wash of yellowish on the sides and crissum. Upper eyelid dusky brown;
remainder of orbital region, with the lores, ashy-white in decided contrast with the
nearly clear cinereous of the head generally. Lining of wings white.
Dimensions. Length, 5.20; extent, 8.50; wing, 2.90; tail, 2.25; culmen, .50.
Habitat. Arizona and New Mexico.
Four additional specimens offer no variations affecting any of the characters
above detailed.
In its generally dull, grayish coloration, with little trace of olive or yellow shades,
this Vireo is curiously like V. pusillus, but the under parts are obscured with
brownish, while the differences in size and proportions are too evident to require
detailed comparison. From the smaller, much brighter-colored V. huttoni, which is
unmistakably its nearest United States relative, it may be distinguished by the
following diagnoses.
V. huttoni.—Wing, 2.28 to 2.37. Olive-green above and olivaceous-yellowish
beneath. No clear white anywhere.
V. huttoni stephensi.—Wing, 2.55 to 2.90. Grayish-ash above with no decided
olive-green excepting on the rump and tail. Beneath brownish-white, untinged
with yellowish excepting on the sides and crissum. Wing-bands pure white and
nearly confluent.
It will be observed that the above differences are closely parallel to those which
separate Vireo belli and V. pusillus, while they are in no respect less important.
Indeed were I disposed to emphasize certain peculiarities presented in the wing-
formula of my type, it would not be difficult to make out an equally good case of
specific distinctness, but unfortunately, the relative length of the wing-quills
(including the spurious primaries) proves to be quite as variable in V. huttoni and
its Arizona race, stephensi, as I find it to be in V. pusillus and V. belli, and, I might
add, in all closely allied species which I have so far studied. In short, I am
convinced that this feature, if ever of any diagnostic value, is so with only a small
proportion of the birds to which it has been so freely and confidently applied.
In naming this Vireo after its discoverer, Mr. F. Stephens, I have paid but a
deserved compliment to that gentleman’s zeal and energy as a field ornithologist.
He notes the bird as “not uncommon in scrub oaks” among both the Chiricahua
and Santa Rita Maintains. He also writes me that he has taken specimens in New
Mexico, where, near Fort Bayard, a nest with four eggs was obtained in 1876. In
both Territories it seems to be confined to the mountain ranges, where it
undoubtedly breeds in all suitable localities.
41, ♂ ad., Morse’s Mill, Chiricahua Mountains, March 14. Length, 5.20; extent,
8.50; wing, 2.90; tail, 2.25; tarsus, .73; culmen, .50; depth of bill at nostrils, .15.
“Iris brown.”
50, ♂ ad., Morse’s Mill, March 16. Length, 4.90; extent, 8; wing, 2.55; tail, 2.20;
tarsus, .73; depth of bill at nostrils, .15.
118, ♂ ad., Morse’s Mill, March 28. Length, 5; extent, 7.90; wing, 2.68; tail, 2.30;
tarsus, .70; culmen, .50; depth of bill at nostrils, .15.
140, ♂ ad., Chiricahua Mountains, March 31. Length, 5.10; extent, 8.40; wing,
2.65; tail, 2.25; tarsus, .73; culmen, .49; depth of bill at nostrils, .15.
353, ♂ ad., Santa Rita Mountains, May 11. Length, 5; extent, 8.10; wing, 2.74;
tail, 2.25; tarsus, .70; culmen, .48; depth of bill at nostrils, .15.
Seven California specimens of V. huttoni measure as follows:—
1443, ♂, Nicasio. Wing, 2.35; tail, 2.20; tarsus, .75; culmen, .50; depth of bill, .11.
1445, ♂, Nicasio. Wing, 2.31; tail, 2.15; tarsus, .76; culmen, .51; depth of bill, .11.
1444, ♀ , Nicasio. Wing, 2.35; tail, 2.25; tarsus, .76: culmen, .49; depth of bill,
.10.
1446, ♀ , Nicasio. Wing, 2.32; tail, 2.28; tarsus, .74; culmen, .50; depth of bill,
.14.
6800, ♂ , Berkeley Co. Wing, 2.37; tail, 2.30; tarsus, .75; culmen, .46; depth of
bill, .11.
6801, ♀, Berkeley Co. Wing, 2.28; tail, 2.15; tarsus, .75; culmen, .51; depth of bill,
.11.
6339, ♀, Riverside. Wing, 2.34; tail, 2.14; tarsus, .75; culmen, .52; depth of bill,
.14.
 48. Vireo pusillus Coues. Least Vireo.—An abundant summer
species frequenting willows along streams and, near Tucson, thickets
of mesquites. “It is active, restless and very noisy.”
Numerous nests were taken. The only one sent me is a shallower,
but nevertheless rather more elaborate structure, than that of V. belli
to which, however, it bears a strong resemblance. It is mainly
composed of fibrous shreds, apparently obtained from the stalks of
some herbaceous plant. The lining is of delicate, bleached grasses,
which are very neatly arranged. The eggs are white with a cluster of
small black spots about the larger ends. The clutch comprised three,
a number which was not exceeded in any of the other nests. The
notes relating to this set are as follows: “Tucson, June 11. Nest
pensile between the forks of a small mesquite branch, about five feet
from the ground, in a thicket of weeds and brush. Incubation
commenced. Female shot. This species seems to abandon a nest if it
is found before any eggs are laid.”

205, ♂ ad., Cienega Station, April 15. Length, 5; extent, 7.10; wing, 2.21; tail,
2.25. “Iris dark brown; bill dark above, light below; legs dark.”
235, ♀ ad., Tucson, April 19. Length, 5.10; extent, 7.30; wing, 2.23; tail, 2.25.
262, ♂ ad., Tucson, April 22. Length, 6; extent, 7.10; wing, 2.28; tail, 2.34.
275, ♂ ad., Tucson, April 25. Length, 5; extent, 7; wing, 2.21; tail, 2.25.
276, ♀ ad., same locality and date. Length, 4.90; extent, 6.90; wing, 2.18; tail,
2.25.
282, ♂ ad., same locality and date. Length, 5; extent. 7.10; wing, 2.30; tail, 2.30.
461, ♀ ad., Camp Lowell, May 31. Length, 5; extent, 6.90; wing, 2.21; tail, 2.25.
“Laying.”
499, ♀ ad., Tucson, June 7. Length, 5; extent, 6.90. Skin lost.
589, ♀ ad., Camp Lowell, June 24. Length, 4.80; extent, 6.80; wing, 2.21; tail,
2.25.

49. Vireo vicinior Coues. Gray Vireo.—The only individuals met

with were a male and female—apparently a mated pair—which were
taken at Tucson, on April 26. “They were in low brush and were very
shy.”
 286, ♀ ad., Tucson, April 26. Length, 5.60; extent, 8.20; wing, 2.63; tail, 2.67;
tarsus, .80.
287, ♂ ad., same locality and date. Length, 5.60; extent, 8.30; wing, 2.58; tail,
2.70; tarsus, .80. “Iris dark brown; bill plumbeous, darkest above; legs light
plumbeous.”

50. Lanius ludovicianus excubitorides (Sw.) Coues. White-

rumped Shrike.—“Common and generally distributed.”
It is unfortunate that so much prominence has been given to the
white rump of excubitorides as a distinguishing character, for I have
yet to see a good series of Shrikes from any Western locality,
excepting, possibly, Arizona, which did not afford a considerable
percentage of dark-rumped birds; and conversely, it is by no means
difficult to find light-rumped specimens in the East. The same
instability also affects most of the other characters which have been
assigned to excubitorides, as is sufficiently shown by the various
conflicting rulings of the authorities regarding the precise definition
and limits of distribution of this troublesome race. The only
differential points which seem to me to hold good with any number
of specimens, are the lighter, purer ash of the upper parts as
compared with those of ludovicianus, and the smaller and very much
weaker bill. But if these alone are to be depended upon, it becomes
necessary to limit the distribution of ludovicianus proper to the Gulf
States, Georgia and the Carolinas, if not strictly to Florida, and to
refer all representatives from the United States at large, east of
California, to excubitorides: and this course, I believe, will ultimately
have to be adopted. The proper position of the dark California form
which is so curiously like ludovicianus remains to be satisfactorily
determined.
51. Ampelis cedrorum (Vieill.) Baird. Cedar Waxwing.—Met
with but once, at Galeyville, where on January 12, 1881, several were
shot from a small flock. Mr. Henshaw took a single specimen near
Camp Apache, in September, 1873.
52. Progne subis (Linn.) Baird. Purple Martin.—“Common.”

438, ♂ ad., Tucson, May 26. Length, 7.6; extent, 15.7; wing, 5.45. “Iris dark
brown; bill black; legs blackish.”
 53. Petrochelidon lunifrons (Say) Lawr. Cliff Swallow.—At
Yuma. “They were breeding abundantly along a bluff above the
town.”
54. Tachycineta bicolor (Vieill.) Caban. White-bellied
Swallow.—“Common in the migrations.”

195, ♂ ad., Cienega Station, April 15. “Iris dark brown; bill black; legs brown.”

55. Tachycineta thalassina (Swains.) Caban. Violet-green

Swallow. “Common.”

212, ♀ ad., Cienega Station, April 16. “Iris dark brown; bill and legs black.”

56. Stelgidopteryx serripennis (Aud.) Baird. Rough-winged

Swallow.—Common. Breeds.

211, ♀ ad., Cienega Station, April 16. “Iris and legs dark brown.”

57. Pyranga ludoviciana (Wils.) Bp. Louisiana Tanager.—

Santa Rita Mountains. “They frequent oaks, and are not very
common.”

408, ♂ ad., Santa Rita Mountains, May 18. Length, 7.30; extent, 7.60; wing,
3.80; tail, 3.17. “Iris dark brown; bill blackish horn-color above, greenish-yellow
below.”

58. Pyranga hepatica Swains. Liver-colored Tanager.—This

Tanager was not uncommon in the Santa Rita Mountains, where the
first specimen was taken on May 12. “They range from the foot-hills,
through the oaks to the lower pines on the mountains.”

359, ♀ ad., Santa Rita Mountains, May 12. Length, 7.80; extent, 12.10; wing,
3.75. “Bill black above, bluish horn-color below; legs lead-color; iris brown.”
 377, ♂ ad., Santa Rita Mountains, May 14. Length, 8.20; extent, 12.70; wing,
4.20.
380, ♀ ad., Santa Rita Mountains, May 14. Length, 8.10; extent, 12.40; wing,
4.07. “This bird would have laid in about ten days.”
386, ♂ ad., Santa Rita Mountains, May 15. Length, 8.20; extent, 12.80; wing,
4.10.

59. Pyranga æstiva cooperi Ridgw. Cooper’s Tanager.—Mr.

Stephens found this bird rather common at a point about five miles
south of Tucson, where it frequented the cottonwoods along a small
river. He also informs me that in May, 1875, he took several
specimens on the Rio Grande River, between Albuquerque and
Mesilla, and some others on the Gila, in New Mexico, during May
and June.

227, ♂ ad., Tucson, April 19. Length, 8.10; extent, 12.40; wing, 3.83; tail, 3.50.
“Iris brown; bill pale horn-color; legs pale brown. Skin very tender. The first seen
this season.”
268, ♂ ad., Tucson, April 23. Length, 8.20; extent, 12.40.
297, ♂ ad., Tucson, April 27. Length, 7.90; extent, 12.20; wing, 4; tail, 3.60.
515, ♂ ad., Tucson, June 10. Length, 8.10; extent, 12.20; wing, 3.85; tail, 3.60.
522, ♂ im. Tucson, June 11. Length, 8; extent, 12.20; wing, 3.78; tail, 3.46. In
mixed yellow and red plumage.
526, ♂ ad., same locality and date. Length, 8.10; extent, 12.50; wing, 3.89; tail,
3.45.
579, ♂ ad., Camp Lowell, June 23. Length, 8; extent, 11.60; wing, 3.99; tail, 3.58.
339, ♀ ad., Tucson, May 7. Length, 7.90; extent, 12.20; wing, 3.75; tail, 3.39.
 NOTES ON THE SUMMER BIRDS OF THE
UPPER ST. JOHN.
BY CHARLES F. BATCHELDER.

(Concluded from page 111.)

41. Carpodacus purpureus (Gm.) Bd. Purple Finch.—Common.

42. Astraga1inus tristis (Linn.) Cab. Goldfinch.—Common. Though
somewhat beyond the limits of my subject, I quote the following from Mr.
McLeod’s notes: “This winter [1876–77] they have been abundant, although the
season is very severe. I have seen them at this time of year but once before.” The
Goldfinch has been supposed not to winter north of Massachusetts.
43. Chrysomitris pinus (Wils.) Bp. Pine Finch.—Seen in May at Grand Falls.
Mr. H. A. Purdie tells me that he observed it at Houlton in June, 1878.
44. Passerculus sandwichensis savanna (Wils.) Ridgw. Savanna Sparrow.
—Common in the pastures at Grand Falls. At Fort Fairfield it was common. It was
found in grassy fields, especially along the roadsides.
45. Poœcetes gramineus (Gm.) Bd. Grass Finch.—Common at Fort Fairfield.
Some seen in the open fields at Grand Falls.
46. Melospiza fasciata (Gm.) Scott. Song Sparrow.—Abundant at Grand
Falls. It was common at Fort Fairfield.
47. Melospiza palustris (Wils.) Bd. Swamp Sparrow.—“Not common” at
Houlton. Not found at Fort Fairfield or Grand Falls.
48. Junco hyemalis (Linn.) Scl. Black Snowbird; “Bluebird.”—Very
common at Fort Fairfield. At Grand Falls it was very abundant everywhere.
49. Spizella socialis (Wils.) Bp. Chipping Sparrow.—This bird was quite
abundant at Grand Falls. The nests found were not the loose structures they are in
Massachusetts, but were well lined with hair. It was rather common at Fort
Fairfield.
50. Zonotrichia albicollis (Gm.) Bp. White-throated Sparrow.—Very
abundant at Grand Falls wherever there was dead wood on the ground. At Fort
Fairfield also it was very abundant; this bird and Junco hyemalis were the
commonest species. The nests were apt to be in a clearing near the edge of woods,
and frequently were in damp places. They were often under a fallen branch, or at
the foot of a sapling, and were but slightly concealed.
The White-crowned Sparrow is probably only a migrant through this section.
With regard to its abundance, however, I quote the following from Mr. McLeod’s
notes: “These Sparrows make their first appearance from May 10th to 18th. Some
seasons they are very abundant, scores of them at a time feeding in my garden. By
June 1 they have disappeared. In the autumn I have seen but one flock of them.”
51. Zamelodia ludoviciana (Linn.) Coues. Rose-breasted Grosbeak.—
Common in low hard woods at Grand Falls. Rather common at Fort Fairfield,
apparently more so than in eastern Massachusetts. Rather common at Houlton.
52. Dolichonyx oryzivorus (Linn.) Swains. Bobolink.—Apparently not rare
at Fort Fairfield. Found in grassy fields and meadows near the river. Not observed
at Grand Falls. At Houlton “arrives by the 25th of May, common by June 15.” July
2, on our return from Fort Fairfield, Mr. Dwight and I saw them at several places
along the St. John River above Fredericton.
53. Agelæus phœniceus (Linn.) Vieill. Red-winged Blackbird.—“Quite
common at Eel River, ten miles from Houlton” (R. R. McL.). It does not occur at
Fort Fairfield or Grand Falls.
54. Quiscalus purpureus æneus Ridgw. Crow Blackbird.—Common at
Fort Fairfield, in the town, along the river, and about a small pond back in the
woods. At Grand Falls it was not uncommon about the town. “Very common” at
Houlton.
55. Corvus corax Linn. Raven.—Rare at Grand Falls. Not met with at Fort
Fairfield. “Very rare” at Houlton.
56. Corvus americanus Aud. Crow.—Common.
57. Cyanocitta cristata (Linn.) Strickl. Blue Jay.—Common at Grand Falls. At
Fort Fairfield it was rather common, but shy and seldom seen.
58. Perisoreus canadensis (Linn.) Bp. Canada Jay.—At Houlton: “very
common. These birds do not often appear in the thickly settled part of the town,
but are very abundant around the lumber camps in this vicinity.” This no doubt
explains the fact that the species was not seen by any of us at Grand Falls and Fort
Fairfield.[72]
59. Tyrannus carolinensis (Linn.) Bd. Kingbird.—Rather common at Fort
Fairfield. At Grand Falls several were seen, but it was not common.
60. Myiarchus crinitus (Linn.) Caban. Great Crested Flycatcher.—In
June, 1878, Messrs. H. A. Purdie and Ruthven Deane observed a pair nest-building
at a point in New Brunswick about six miles east of Houlton.
61. Sayornis fuscus (Gm.) Bd. Pewee.—One was observed at Fort Fairfield,
June 28. “Very rare” at Houlton.
62. Contopus borealis (Swains.) Bd. Olive-sided Flycatcher.—Common in
the woods at Grand Falls. This species was rather common at Fort Fairfield. We
usually saw them perched on the tops of tall dead trees in clearings. They were
rather shy.
63. Contopus virens (Linn.) Caban. Wood Pewee.—At Fort Fairfield it
appeared to be not uncommon. It was not met with, however, at Grand Falls.
64. Empidonax flaviventris Bd. Yellow-bellied Flycatcher.—At Fort
Fairfield this species was rather common in wet evergreen woods, especially in
those that had small streams flowing through them. It was not observed at Grand
Falls. Messrs. Purdie and Deane found it rather common at Houlton in June, 1878.
[73]

65. Empidonax trailli (Aud.) Bd. Traill’s Flycatcher.—Not common at

Grand Falls. They were to be found mostly where there were scattered dead trees.
We did not find it at Fort Fairfield. Mr. H. A. Purdie informs me that it was not
uncommon at Houlton in June, 1878.
66. Empidonax minimus Bd. Least Flycatcher.—Very abundant in hard
woods at Grand Falls. At Fort Fairfield it was rather common.
67. Caprimulgus vociferus Wils. Whip-poor-will.—Mr. McLeod notes that
there are a few at Houlton during the summer. The species was neither seen nor
heard at Fort Fairfield and Grand Falls.
68. Chordeiles popetue (Vieill.) Bd. Nighthawk.—Very abundant at Grand
Falls. At Fort Fairfield it was common; they frequented burnt lands.
69. Chætura pelasgica (Linn.) Bd. Chimney Swift.—At Fort Fairfield they
were common, breeding both in chimneys and in hollow trees. Common in the
burnt country at Grand Falls. Not many were breeding in chimneys, the people
disliking to have them there.
70. Trochilus colubris Linn. Ruby-throated Hummingbird.—Common at
Grand Falls. At Fort Fairfield it was apparently rather common—we saw several.
71. Ceryle alcyon (Linn.) Boie. Belted Kingfisher.—Rather common at Fort
Fairfield. At Grand Falls it was to be seen wherever there was good fishing ground.
72. Picus villosus Linn. Hairy Woodpecker.—Common.
73. Picus pubescens Linn. Downy Woodpecker.—At Fort Fairfield this
species was much less common than P. villosus. It was not uncommon at Grand
Falls.
74. Picoides arcticus (Swains.) Gray. Black-backed Three-toed
Woodpecker.—Common at Grand Falls in burnt cedar swamps. At Fort Fairfield
we shot two, all we saw.
75. Sphyropicus varius (Linn.) Bd. Yellow-bellied Woodpecker.—Common
—the commonest Woodpecker—at Fort Fairfield. They were generally found about
recent clearings, or in the more open mixed woods. At Grand Falls they were
common in hard woods.
76. Hylotomus pileatus (Linn.) Bd. Pileated Woodpecker.—At Grand Falls
half a dozen pairs were seen. Probably there is too little of the heavy forest left in
the immediate neighborhood of Fort Fairfield to suit their tastes, as we did not
meet with them. “Common” at Houlton.
77. Colaptes auratus (Linn.) Sw. Golden-winged Woodpecker.—Rather
common at Fort Fairfield. Not common at Grand Falls.
78. Coccyzus erythrophthalmus (Wils.) Bd. Black-billed Cuckoo.—Mr.
McLeod records this bird in his notes, but without comments. It was not seen at
Fort Fairfield or Grand Falls.
79. Strix nebulosa Forst. Barred Owl.—“Very common” at Houlton. We were
shown a mounted specimen by Mr. Frank P. Orcutt at Fort Fairfield. He
considered it the commonest Owl.
80. Nyctale acadica (Gm.) Bd. Saw-whet Owl.—This bird is not uncommon
at Houlton. Mr. Frank P. Orcutt told us that it was tolerably common at Fort
Fairfield.
81. Bubo virginianus (Gm.) Bd. Great Horned Owl.—“Very common” at
Houlton. Mr. Orcutt said it was rather common at Fort Fairfield.
82. Circus hudsonius (Linn.) Vieill. Marsh Hawk.—Rare at Houlton. One
seen at Fort Fairfield.
83. Accipiter cooperi Bp. Cooper’s Hawk.—Not common at Grand Falls. Not
observed at Fort Fairfield or Houlton.
84. Accipiter fuscus (Gm.) Bp. Sharp-shinned Hawk.—“Not common” at
Houlton.
85. Falco sparverius Linn. Sparrow Hawk.—Commonest Hawk at Grand
Falls. Not met with at Houlton or Fort Fairfield, though Mr. Orcutt considers it
common at the latter place.
86. Buteo borealis (Gm.) Vieill. Red-tailed Hawk.—Not common at Grand
Falls. Not observed at Fort Fairfield. “Common” at Houlton.
87. Buteo pennsylvanicus (Wils.) Bp. Broad-winged Hawk.—Not common
at Grand Falls. It was found breeding at Houlton, but not met with at Fort
Fairfield.
88. Haliæetus leucocephalus (Linn.) Savig. Bald Eagle.—“Not common” at
Houlton.
89. Ectopistes migratorius (Linn.) Sw. Wild Pigeon.—Breeding at Grand
Falls, but not common.
90. Canace canadensis (Linn.) Bp. Spruce Partridge.—At Houlton “mostly
found in the deep fir thickets, or in the swamps of firs and cedars.” Not met with at
Fort Fairfield and Grand Falls, though of course it occurs there.
91. Bonasa umbellus (Linn.) Steph. Ruffed Grouse.—Rather common at Fort
Fairfield. At Grand Falls only a few were seen—in the hard woods.
92. Ardea herodias Linn. Great Blue Heron.—“Common” at Houlton.
93. Nyctiardea grisea nævia (Bodd.) Allen. Night Heron.—“Not common” at
Houlton.
 94. Botaurus lentiginosus (Montag.) Steph.—Bittern.—“Common” at
Houlton. One seen at Grand Falls.
95. Philohela minor (Gm.) Gray. Woodcock.—One seen on Little River Flats
near Grand Falls. At Fort Fairfield we saw a specimen in the collection of Mr.
Frank P. Orcutt, who considered it rare in that neighborhood. “A few breed in the
vicinity” of Houlton.
96. Rhyacophilus solitarius (Wils.) Cass. Solitary Sandpiper.—At Grand
Falls some were seen along the river June 9 (J. A. J.).
97. Tringoides macularius (Linn.) Gray. Spotted Sandpiper.—At Fort
Fairfield it was very numerous along the Aroostook River, and was also noticed in
one or two other places. It was abundant along the rivers at Grand Falls. At
Houlton too it was very common.
98. Porzana carolina (Linn.) Bd. Carolina Rail.—One seen at Fort Fairfield,
June 20, in a wet meadow partly grown up with alder bushes (J. D.).
99. Anas obscura Gm. Black Duck.—“Very common, breeding” at Houlton.
100. Aix sponsa (Linn.) Boie. Wood Duck.—“Quite common” at Houlton.
101. Clangula glaucium americana (Bp.) Ridgw. Golden-eye.—A few seen
at Grand Falls.
102. Mergus merganser americanus (Cass.) Ridgw. Sheldrake.—Not
uncommon at Grand Falls.
103. Mergus serrator Linn. Red-breasted Merganser.—“Very common,
breeding,” at Houlton.
104. Larus argentatus smithsonianus Coues. Herring Gull.—At Houlton
it is common on the neighboring lakes, where it breeds.
105. Podilymbus podiceps (Linn.) Lawr. Pied-billed Grebe.—Rare, breeds,
Houlton.
A SKETCH OF THE HOME OF HYLOCICHLA
ALICIÆ BICKNELLI, RIDGWAY, WITH
SOME CRITICAL REMARKS ON THE ALLIES
OF THIS NEW RACE.
BY EUGENE P. BICKNELL.

That there remained unrecognized at this late day a bird regularly

inhabiting one of the must populous portions of our country; or,
indeed, that a species of eminently boreal habitat during its breeding
season, and not known to occur at all at such time within the limits
of the United States, should have a representative race regularly
breeding in our midst, are facts for which we were little prepared.
Mr. Ridgway’s recent paper[74] announcing these facts being
necessarily of a technical nature, and confined to a formal
description of the new Thrush, it has been thought well on the
present occasion to allude more particularly to the character of the
locality inhabited by the bird, and to some of its associates there, in
connection with other sequential considerations. As the general
physical character of the Catskill Mountains and the faunal features
of the region will be treated by the writer elsewhere, it will be
unnecessary to extend the range of the present relation from the
summit of Slide Mountain in Ulster Co.,[75] where the new race was
discovered.
On June 15, 1881, nearing the summit of this mountain, the forests
of a more northern latitude were forcibly suggested. A shower had
fallen during the ascent, and the sun was still obscured, while a sharp
wind from the northwest piercing the wet woods and sighing among
the balsams, blasted and weather beaten, heightened an impression
of remoteness and desolation. The evergreens, constituting the
principal arboreal growth, extended off on all sides, clothing the
rocky and moss-grown slopes, and presenting the striking contrast of
a young and fragrant second growth clustering about the branchless
and spiny trunks of their sires tottering in decay; or, with tangled
and matted branches outlined here and there, as we approached the
summit, against a gray and cheerless sky. Owing to the
comparatively short life of these trees, that high portion of the
mountain where their tribe had pitched was brought into grim
contrast with its surroundings. Old age and death, continually
present invading their ranks, had everywhere left their traces;
flourishing clusters had been stricken in their fellowship, groups and
gatherings had been divided and scattered, and like a contagion the
destroyer had spread among their hosts. But the younger generations
are continually forming their associations, and with green and
fragrant grouping filling in deserted chambers and screening the
devastation that has gone before, although only to furnish material
for its continuance in the future. All this, with an occasional
undergrowth of greater or less luxuriance, gave a diversified and
somewhat open character to the surroundings, entirely dissimilar to
that of the environing forest; conditions, which, in conjunction with
humidity and elevation, have brought this mountain top into some
relation with the swampland of a more northern region.
Reaching a more elevated portion of the ridge where the ground
was more level and the surface less rocky, that north-woods tree, the
Paper Birch (Betula papyracea) occasionally appeared, and more
abundantly the Mountain Ash. Almost the only remnant of the dense
mountain forests below was the Yellow Birch (Betula lutea) which,
joining the undergrowth, persisted with small and stunted stature to
the summit. On all sides were to be seen the white blossoms of
Viburnum lantanoides which, though also found in the valley
woodlands, had there long since flowered and was now bearing green
fruit. Another characteristic shrub was Amelanchier canadensis
oligocarpa; lower down had been found the var. botryapium, but
here, the northern form was well marked, seeming almost
specifically distinct. In the deep, damp moss, covering and filling in
the rocks beneath the balsam growth, and relieving the ruggedness of
the slopes, northern plants were growing in greater or less profusion.
The Dwarf Cornel (Cornus canadensis) grew in such close luxuriance
in congenial spots, that its snowy bracts imparted an almost uniform
whiteness to whole beds. With, or near it, blossomed the Wood
Sorrel (Oxalis acetosella) with delicately violet-veined petals, and the
appropriately-named Gold-thread (Coptis trifolia) of evanescent
bloom but shining evergreen leaves, and the little Star Flower
(Trientalis americana) were often also associates. Excepting the pale
yellow bells of Clintonia borealis, and the purplish tinge, or veining,
of the blossoms of several other species, all the plants noticed in
bloom at this time upon the mountain bore flowers of some shade of
white. The more open ground about our course along the ridge
supported a luxuriant and graceful growth of that lovely fern
Aspidium spinulosum, and with it, in openings about the summit,
grew abundantly the Mountain Golden-rod (Solidago thyrsoidea)
which, although yet many weeks from bloom, heralded a royal
emblem to light the mountain’s brow ere the white locks of winter
should again possess it.
At the elevation where these plants first appeared the trees
nowhere attained more than a medium stature, those which seemed
best to have surmounted the difficulties of their situation, the Balsam
and the Paper Birch, never rising to a height of more than, perhaps,
twenty-five feet. This growth completely encompassed the range of
vision, but an occasional scantiness in the foliage permitted glimpses
of surrounding mountains rolling off like huge green billows into the
blue distance.
From these evergreens came the leisurely call of the Canada
Nuthatch (Sitta canadensis), and on closer approach the low,
plaintive notes of the little Yellow-bellied Flycatcher (Empidonax
flaviventris). The brief warble of the Black-and-Yellow Warbler
(Dendrœca maculosa) told of the presence of its unseen author in
the surrounding trees, while among the undergrowth the less
frequent, but louder and more sustained song of the Mourning
Ground-warbler (Geothlypis philadelphia) showed that this species,
which had been left at the foot of the mountain, had here reappeared.
At intervals, faintly mingling with these songs, from some hidden
fastness below, came the fantasia of the Winter Wren, a melody that
seemed to pass from the spirit of unclaimed nature, voicing some
mystery of the mountains. The clamor of a party of Blue Jays
occasionally arose and died away in the forest, but here, in this
mountain solitude, their screams seemed more subdued than in less
primitive regions, and lacked that suggestion of consciousness which
individuals constantly within human hearing, seem to acquire. Busily
roaming Chickadees (Parus atricapillus) at times came about our
path, and the Snowbird (Junco hyemalis) was present with its simple
song. Olive-backed Thrushes (Hylocichla ustulata swainsoni) too,
were constantly to be heard, and finally, guided by its near song, one
was followed up and secured. A moment later another Thrush darted
across the path, and disappearing through a young balsam growth,
immediately began to sing a few rods off. The song was different
from that of the bird which had just been shot, so much so, in fact, as
to be remarked even by my guide. It seemed to be more uniform in
character, with less variation and definition of the notes; as I wrote
in my note-book at the time—more suggestive of the song of H.
fuscescens. A conspicuous point of difference was that it was more
subdued in tone, in fact of a somewhat ventriloquous nature. On
examining the bird, in hand, although I had thought myself familiar
with all our eastern Hylocichlæ, I must confess to having been
puzzled. It was obviously neither the Olive-backed nor the Hermit
Thrush, the only species of our own smaller Thrushes which from the
distribution of their group (as then understood) could possibly be
expected to occur. I at once noted its general resemblance to the
Gray-cheeked Thrush, but it seemed impossible that this Hudsonian
bird could be found so far south at this season; and though a second
specimen pointed more strongly toward it, it was not until I had
reached home and made actual comparisons, that I could feel
satisfied that its true relationship was with that species. I had long
noticed certain somewhat constant differences between examples of
aliciæ occurring at New York on their migrations, and incited by
these specimens went carefully over my series of seventeen examples
and found them separable into two forms, characterized by slight
differences in coloration and a notable difference in size. The
examples from the Catskills were more closely allied to the smaller of
the two forms, and these, with, subsequently, my entire series, were
submitted to Mr. Ridgway, the result being the recognition of a new
bird, belonging to our eastern fauna.
But to return to the mountain. It would hardly be justifiable to
make a positive statement about a difficult song that had been but
once identified, but I feel positive that the Thrushes which were last
heard that evening about our camp on the extreme summit of the
mountain were of the new form. Night was rapidly falling, and the
valleys were in darkness, when one sang several times near the camp,
and for some time afterwards a single call-note was occasionally
heard, and the varying distance of the sound showed that the birds
were still active. Excepting these sounds, the last bird notes heard
were those of the Yellow-bellied Flycatcher.
The sharp northwest wind continued late, and the night became
clear and cold. Shortly after midnight the bright moon showed the
temperature, by a thermometer which I had hung beside the camp,
to be 35°, and at sunrise it stood at 32°. Before daylight I was
standing on a boulder of conglomerate on the dim mountain’s brow
listening for the awakening of the birds. The first songs heard were
those of the Hermit Thrush, Snowbird, and Yellow-bellied
Flycatcher, which began almost simultaneously, followed a little later
by those of the Olive-backed Thrush and the Mourning Warbler, but
H. bicknelli was not heard, or at least not near enough to be
distinguished among the other species.
The increasing light upon the mountain seemed to attract the birds
from below, whither, perhaps, they had retired for the night, and
soon many different notes were to be heard about the camp; not,
however, in that boisterous chorus with which the day is often
announced about our homes, in which the notes of many individuals
of many species are blended in such confused medley that separate
voices are almost indistinguishable, but simply the association of a
few vocalists, the very isolation of whose position endowed their
voices with an additional interest and charm.
After those already mentioned the Black-poll Warbler (Dendrœca
striata) began its unpretending notes, which always to me suggest a
short dotted line, and this song, with that of the Black-and-Yellow
Warbler, occasionally alternated about us in agreeable contrast. Now
and then a Canada Nuthatch, on its morning tour, tarried to inspect
some dead trunk or thinly clothed tree, upon the projecting apex of
which, or that of some companion, a solitary Purple Finch
occasionally alighted, and with a few wild fugitive notes was gone, to
other mountain tops or the forests of the descending slopes.
But to revert to the Thrushes. The two specimens of the new form
which were obtained were both males, and were unquestionably
breeding,[76] though no nest known to belong to their species was
found.
 It remains to briefly consider some facts furnished by the birds’
occurrence as narrated. These facts bear directly on the long
contested question of the relationship which H. aliciæ and H.
swainsoni bear to one another, and it can scarcely be denied that the
present evidence on this point is conclusive. Not only have we
representatives of both birds preserving their respective identities at
the same locality, under identical conditions of environment, but
examples of each taken under these circumstances, display, except in
size, even a greater dissimilitude than those which occur together on
their migrations. There is but one tenable interpretation of these
facts: the birds—Hylocichla aliciæ and H. ustulata swainsoni—are
wholly and entirely distinct. Any theory of dichromatism which
might be advanced, aside from its extreme unlikelihood, would be
shown inadequate by the relative differences in proportions of parts
which the two birds exhibit. These differences, as well as those of
color are illustrated by the Catskill birds. A specimen of H. swainsoni
taken at the top of Slide Mountain was in every way typical of its
species, and conspicuously unlike the examples of bicknelli taken at
the same time. Aside from differences in the proportions of parts, the
two birds were strikingly different in color, the decided grayish olive
tinge of the superior surface of swainsoni contrasting strongly with
the much darker brownish cast of its congener. One example of the
latter instead of showing indications of a buffy tinge about the sides
of the head and on the breast, which under the circumstances we
should expect to be the case, were it in any way specifically related to
swainsoni, has absolutely no indications whatever of this shade
about the sides of the head, and actually less on the breast than any
specimens of true aliciæ that I have seen, and this little most evident
low down where the corresponding shade in swainsoni begins to
pale. It seems probable that this newly recognized race of aliciæ is
responsible for much of the ambiguity which the discussion of both
species by different writers has occasioned. Indeed, it seems to
occupy the same position relative to aliciæ proper which, by some,
swainsoni was supposed to hold, viz., the more southern-born
individuals of the species, but that it represents a link specifically
connecting the two, the facts already presented refute. As it occurs
with true aliciæ on the autumn migration most specimens of the new
form are paler and more brownish in color above, and their general
size is nearly that of swainsoni,[77] and these differences may be
regarded by some as approaches towards the latter species. In both
species there is a wide individual variation, but the closest approach
of each towards the other never exceeds that limit within which each
may vary without its specific distinctness being compromised. I have
yet to see a specimen of either which would admit of the slightest
question as to its identity. I speak thus of adult birds. In such closely
related species the young must almost necessarily approximate, and
to these we must appeal for light on the things that have been—on
the question of origin—whether one has been derived from the other,
or both species from a common ancestor. Such obscure insight into
this point as I have been permitted seems to indicate that the latter
alternative will be found to be the more correct, but, for the present,
from lack of the necessary data this important subject is proscribed.
It is unnecessary here to repeat the diagnosis of the new form of
Hylocichla aliciæ given by Mr. Ridgway in the paper before cited. As
this writer states, the race breeds “probably in other mountainous
districts of the northeastern United States” than the single locality
where it was discovered, and it seems very singular that up to the
present time we have no knowledge of its occurrence in the summer
season elsewhere, even in regions where the two congeneric species
with which it was here associating—H. nanus[78] and H. swainsoni—
are well known to be common summer residents. The occurrence of a
representative of H. aliciæ in the United States at all during its
breeding season is a matter of surprise, especially when we recollect
the boreal distribution of the typical form during that period, and
read[79] that so far towards the north as the Yukon and the Great
Slave Lake it occurs “only as a bird of passage to and from more
northern breeding grounds.” Additional information respecting the
distribution of the new race will be awaited with great interest.
 SHORT NOTES ON THE BIRDS OF BAYOU
SARA, LOUISIANA.
BY CHARLES WICKLIFFE BECKHAM.

As the avian fauna of the lower Mississippi Valley is now receiving

some attention,[80] it seems well that I should contribute my mite of
information to the general fund.
Bayou Sara and the adjoining town of St. Francisville, in the parish
of West Feliciana, are situated on the east bank of the Mississippi
River, 170 miles above New Orleans by that stream and about 80
miles in an air line northwest of it. It is 30 or 40 miles north of Baton
Rouge, near which place Dr. Langdon made his observations in April,
1881. The following notes were made principally on and near
“Wyoming,” two miles from the river, the plantation of Ex-Gov. R. C.
Wickliffe, a place possessing peculiarly agreeable ornithological
associations on account of its former owner, Gen. Dawson, having
entertained Audubon as his guest for several months. It will be
remembered that the type specimen of Buteo harlani was captured
here.
The topography is much more interesting, and is quite different
from that farther south and that immediately opposite on the west
side of the river. A level plateau, 100 feet above the levee, begins
about a quarter of a mile from the river and extends back into the
State of Mississippi. This plateau is deeply cut by numerous creeks
and ravines, the banks of which are generally densely wooded, with
water oak, sweet-gum, cedar, prickly ash, magnolias, etc. All of the
level ground on top is in a state of cultivation; cotton being the
principal crop. A few miles farther up the high ground does not
extend so near the river, the intervening space being occupied by
several small lakes and swamps—a great resort for water birds of all
kinds. On account of the high water I did not have an opportunity of
visiting this interesting field.
My observations extended only over a period of five days from
April 15th to 19th, 1882, inclusive, but a great deal of ground was
canvassed in that time; nearly the whole of each day being spent in
the field. A good many birds were shot, but few were preserved, as
taxidermy was necessarily subordinated to field-work. Dr. Langdon
in his interesting paper particularly remarks the absence of the
Catbird, Black-and-White Creeper, White-browed Yellow-throat,
Kentucky Warbler, Large-billed Water Thrush, and the Redstart, but
I found all of these at “Wyoming,” together with many others not
noted by either him or Mr. Hay, the Catbird and Kentucky Warbler
being particularly abundant.
The writer was greatly assisted in his work by Mr. Robert
Wederstraudt of “Wyoming,” a young man whose unusually close
and accurate observations of birds and bird life rendered his help
peculiarly valuable. Many of the following notes are credited to him
entirely. I have followed the nomenclature of the Smithsonian list of
1881.

1. Hylocichla mustelina (Gm.) Bd. Wood Thrush.—Common in woodland,

and several seen in the yard near the house.
2. Merula migratoria (L.) Sw. and Rich. American Robin.—Not observed.
They appear here in large numbers early in February to feed on the fruit of the
“wild peach,” and hundreds are shot for the table. They leave early in March.
3. Mimus polyglottus (L.) Boie. Mockingbird.—Very abundant, both in the
town about gardens and yards, and in the country. Frequenting open ground
exclusively. Four sets of eggs were taken; two perfectly fresh, and two about half
incubated. Mr. Wederstraudt called my attention to a curious foraging habit of this
bird. We noticed one hopping along the ground in an open grassy place, pausing at
every three or four hops to extend and close its wings. It repeated this several times
until a grasshopper was flushed, when the bird immediately “reached” for it, and
having captured it, made off to a neighboring bush to eat it. Mr. Wederstraudt says
that he has observed this interesting performance many times.
4. Galeoscoptes carolinensis (L.) Cab. Catbird.—Abundant in the shrubbery
in the creek bottoms. None were seen near the dwellings.
5. Harporhynchus rufus (L.) Cab. Brown Thrasher.—Abundant in same
places as the last. Three clutches of three eggs each were taken, in one of which
incubation was very far advanced, and on the 19th a nest was found containing two
young nearly able to fly.
 6. Sialia sialis (L.) Haldem. Bluebird.—Observed several pairs in town and in
the country. Not as common as in Kentucky.
7. Polioptila cærulea (L.) Scl. Blue-gray Gnatcatcher.—A common, and, on
account of its active and noisy habits, conspicuous bird.
8. Lophophanes bicolor (L.) Bp. Tufted Titmouse.—Not very common.
Frequenting principally the tops of trees.
9. Parus carolinensis Aud. Carolina Chickadee.—But few observed. A pair
bred in a hole in a cedar post within twenty yards of the house last year.
10. Thryothorus ludovicianus (Gm.) Bp. Carolina Wren.—Very abundant
everywhere. A clutch of three eggs was taken on the 19th from a nest in a small
recess formed by the junction of several timbers, under the piazza, which was
frequented at all times of the day. The nest was empty on the 16th, one egg was
deposited on the 17th, one on the 18th, and one on the 19th. I saw neither of the old
birds about the place at all, and it was only by capturing the female on the nest at
night, that the eggs were positively identified. A pair have bred about this piazza
for many years, I am informed.
11. Mniotilta varia (L.) V. Black-and-white Creeper.—A male, the only one
seen at all, was captured in a dense wood on the 17th.
12. Parula americana (L.) Bp. Blue-yellow-backed Warbler.—Very
abundant. A persistent but weak vocalist.
13. Dendrœca æstiva (Gm.) Bd. Summer Yellowbird.—Common in open
places.
14. Dendrœca blackburniæ (Gm.) Bd. Blackburnian Warbler.—Common
in large trees about open ground.
15. Dendrœca dominica albilora Bd. White-browed-yellow-throated
Warbler.—A male, the only one seen, was shot out of a magnolia tree on the 10th.
In all of my Kentucky specimens of this bird the anterior portion of the superciliary
line has a trace of yellow. In this one no yellow is perceptible.
16. Dendrœca pinus (Wils.) Bd. Pine-creeping Warbler.—Apparently not
uncommon. Preferring open ground. In song.
17. Siurus auricapillus (L.) Sw. Golden-crowned Thrush.—One specimen
captured in a thicket on the 15th.
18. Siurus motacilla (V.) Coues. Large-billed Water Thrush.—Heard one
singing in a densely wooded ravine on the 17th. Mr. Wederstraudt has often seen
them in pairs along the smaller water-courses.
19. Oporornis formosa (Wils.) Bd. Kentucky Warbler.—One of the most
abundant inhabitants of the dense growth along the ravines. Two or three were
often heard singing at the same time.
20. Geothlypis trichas (L.) Cab. Maryland Yellow-throat.—Abundant in
the usual places.
21. Icteria virens (L.) Bd. Yellow-breasted Chat.—Very abundant. In full
song.
 22. Myiodioctes mitratus (Gm.) Aud. Hooded Warbler.—Found in same
places, and almost as abundant as the Kentucky Warbler. An inhabitant of the
undergrowth principally. In song; its note being uttered at intervals of 15 or 20
seconds as it hops from branch to branch in pursuit of insects.
23. Setophaga ruticilla (L.) Sw. Redstart.—A single specimen, a male,
captured in a swamp. It was in company with a female.
24. Vireosylvia olivacea (L.) Bp. Red-eyed Vireo.—Very abundant
everywhere.
25. Vireosylvia gilva (V.) Cass. Warbling Vireo.—Heard one singing in a
shade tree in Bayou Sara on the 15th.
26. Vireo noveboracensis (Gm.) Bp. White-Eyed Vireo.—Very abundant
and voluble everywhere.
27. Lanius ludovicianus L. Loggerhead Shrike.—Not observed. Mr.
Wederstraudt says that they are not uncommon here in the fall. He once saw one
kill and devour a small bird in a thorn tree.
28. Ampelis cedrorum (V.) Bd. Cedar Waxwing.—Observed several small
flocks. Said to be very abundant here in winter when numbers are shot for the
table. Known here as the “ortolan”—the fourth bird, I believe, embraced under that
comprehensive name.
29. Progne subis (L.) Bd. Purple Martin.—Common about Bayou Sara and
St. Francisville.
30. Stelgidopteryx serripennis (Aud.) Bd. Rough-winged Swallow.—Very
abundant. Beginning to breed. Several holes examined but no eggs found. One was
shot out of a dead tree.
31. Pyranga æstiva (L.) V. Summer Redbird.—Abundant about dwellings and
open ground. In song.
32. Passerculus sandwichensis savanna (Wils.) Ridgw. Savanna Sparrow.
—Common in old wet fields. One individual captured, a female, had a very large
tumor on the bill and several smaller ones on the toes.
33. Zonotrichia albicollis (Gm.) Bp. White-throated Sparrow.—Abundant
in parties of six or eight in the undergrowth about open places in the low lands.
34. Peucæa æstivalis illinoensis Ridgw. Oak-woods Sparrow.—Two
specimens of this interesting form were taken; both males. One was shot from the
top of a small bush near the edge of an old corn field; the other from the top of an
isolated pine on the edge of a cotton field. Both were singing when shot. No others
were observed. This, I believe, is the most southeasterly “record” of the form.
35. Melospiza palustris (Wils.) Bd. Swamp Sparrow.—Not uncommon in the
usual places.
36. Pipilo erythrophthalmus (L.) V. Chewink; Towhee.— Abundant. Locally
known as the “Joree.”
37. Cardinalis virginianus (Briss.) Bp. Cardinal Grosbeak.—Very abundant.
Took a set of three fresh eggs on the 17th. Nest as usual.
Welcome to our website – the ideal destination for book lovers and
knowledge seekers. With a mission to inspire endlessly, we offer a
vast collection of books, ranging from classic literary works to
specialized publications, self-development books, and children's
literature. Each book is a new journey of discovery, expanding
knowledge and enriching the soul of the reade

Our website is not just a platform for buying books, but a bridge
connecting readers to the timeless values of culture and wisdom. With
an elegant, user-friendly interface and an intelligent search system,
we are committed to providing a quick and convenient shopping
experience. Additionally, our special promotions and home delivery
services ensure that you save time and fully enjoy the joy of reading.

Let us accompany you on the journey of exploring knowledge and

personal growth!

textbookfull.com