Collaborative Computing Networking

Applications and Worksharing 11th

International Conference
CollaborateCom 2015 Wuhan November
10 11 2015 China Proceedings 1st
Edition Song Guo
Visit to download the full and correct content document:
https://textbookfull.com/product/collaborative-computing-networking-applications-and-
worksharing-11th-international-conference-collaboratecom-2015-wuhan-november-10
-11-2015-china-proceedings-1st-edition-song-guo/
More products digital (pdf, epub, mobi) instant
download maybe you interests ...

Collaborative Computing Networking Applications and

Worksharing 13th International Conference
CollaborateCom 2017 Edinburgh UK December 11 13 2017
Proceedings Imed Romdhani
https://textbookfull.com/product/collaborative-computing-
networking-applications-and-worksharing-13th-international-
conference-collaboratecom-2017-edinburgh-uk-
december-11-13-2017-proceedings-imed-romdhani/

Distributed Computing and Internet Technology 11th

International Conference ICDCIT 2015 Bhubaneswar India
February 5 8 2015 Proceedings 1st Edition Raja
Natarajan
https://textbookfull.com/product/distributed-computing-and-
internet-technology-11th-international-conference-
icdcit-2015-bhubaneswar-india-february-5-8-2015-proceedings-1st-
edition-raja-natarajan/

Information Security Practice and Experience 11th

International Conference ISPEC 2015 Beijing China May 5
8 2015 Proceedings 1st Edition Javier Lopez

https://textbookfull.com/product/information-security-practice-
and-experience-11th-international-conference-ispec-2015-beijing-
china-may-5-8-2015-proceedings-1st-edition-javier-lopez/

High Performance Computing and Applications Third

International Conference HPCA 2015 Shanghai China July
26 30 2015 Revised Selected Papers 1st Edition Jiang
Xie
https://textbookfull.com/product/high-performance-computing-and-
applications-third-international-conference-hpca-2015-shanghai-
china-july-26-30-2015-revised-selected-papers-1st-edition-jiang-
Rough Sets and Knowledge Technology 10th International
Conference RSKT 2015 Held as Part of the International
Joint Conference on Rough Sets IJCRS 2015 Tianjin China
November 20 23 2015 Proceedings 1st Edition Davide
Ciucci
https://textbookfull.com/product/rough-sets-and-knowledge-
technology-10th-international-conference-rskt-2015-held-as-part-
of-the-international-joint-conference-on-rough-sets-
ijcrs-2015-tianjin-china-november-20-23-2015-proceedings-1s/

Information and Communications Security 17th

International Conference ICICS 2015 Beijing China
December 9 11 2015 Revised Selected Papers 1st Edition
Sihan Qing
https://textbookfull.com/product/information-and-communications-
security-17th-international-conference-icics-2015-beijing-china-
december-9-11-2015-revised-selected-papers-1st-edition-sihan-
qing/

Environmental Software Systems Infrastructures Services

and Applications 11th IFIP WG 5 11 International
Symposium ISESS 2015 Melbourne VIC Australia March 25
27 2015 Proceedings 1st Edition Ralf Denzer
https://textbookfull.com/product/environmental-software-systems-
infrastructures-services-and-applications-11th-ifip-
wg-5-11-international-symposium-isess-2015-melbourne-vic-
australia-march-25-27-2015-proceedings-1st-edition-ralf-denz/

Trust Management IX 9th IFIP WG 11 11 International

Conference IFIPTM 2015 Hamburg Germany May 26 28 2015
Proceedings 1st Edition Christian Damsgaard Jensen

https://textbookfull.com/product/trust-management-ix-9th-ifip-
wg-11-11-international-conference-ifiptm-2015-hamburg-germany-
may-26-28-2015-proceedings-1st-edition-christian-damsgaard-
jensen/

Algorithms and Discrete Applied Mathematics First

International Conference CALDAM 2015 Kanpur India
February 8 10 2015 Proceedings 1st Edition Sumit
Ganguly
https://textbookfull.com/product/algorithms-and-discrete-applied-
mathematics-first-international-conference-caldam-2015-kanpur-
india-february-8-10-2015-proceedings-1st-edition-sumit-ganguly/
Song Guo
Xiaofei Liao
Fangming Liu
Yanmin Zhu (Eds.)

163

Collaborative Computing:
Networking, Applications,
and Worksharing
11th International Conference, CollaborateCom 2015
Wuhan, November 10–11, 2015, China
Proceedings

123
Lecture Notes of the Institute
for Computer Sciences, Social Informatics
and Telecommunications Engineering 163

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
Falko Dressler
University of Erlangen, Erlangen, Germany
Domenico Ferrari
Università Cattolica Piacenza, Piacenza, Italy
Mario Gerla
UCLA, Los Angels, 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 Zomaya
University of Sydney, Sydney, Australia
Geoffrey Coulson
Lancaster University, Lancaster, UK
More information about this series at http://www.springer.com/series/8197
Song Guo Xiaofei Liao
•

Fangming Liu Yanmin Zhu (Eds.)

•

Collaborative Computing:
Networking, Applications,
and Worksharing
11th International Conference, CollaborateCom 2015
Wuhan, November 10–11, 2015, China
Proceedings

123
Editors
Song Guo Fangming Liu
School of Computer Science School of Computer Science
and Engineering and Technology
The University of Aizu Huazhong University of Science
Aizuwakamatsu and Technology
Japan Wuhan
China
Xiaofei Liao
School of Computer Science Yanmin Zhu
and Technology Department of Computer Science
Huazhong University of Science and Engineering
and Technology Shanghai Jiao Tong University
Wuhan Shanghai
China China

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

Lecture Notes of the Institute for Computer Sciences, Social Informatics
and Telecommunications Engineering
ISBN 978-3-319-28909-0 ISBN 978-3-319-28910-6 (eBook)
DOI 10.1007/978-3-319-28910-6

Library of Congress Control Number: 2015959593

© Institute for Computer Sciences, Social Informatics and Telecommunications Engineering 2016
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.

Printed on acid-free paper

This Springer imprint is published by SpringerNature

The registered company is Springer International Publishing AG Switzerland
 Preface

On behalf of the Organizing Committee of the 11th EAI International Conference on

Collaborative Computing: Networking, Applications and Worksharing, we are very
pleased to present the proceedings in which researchers and contributors from the
world share their research results and new ideas. The 11th EAI International Confer-
ence on Collaborative Computing: Networking, Applications and Worksharing serves
as a premier international forum for discussion among academic and industrial
researchers, practitioners, and students interested in collaborative networking, tech-
nology and systems, and applications.
We were also honored to have Prof. Jiannong Cao from Hong Kong Polytechnic
University and Prof. Huadong Ma from Beijing University of Posts and Telecommu-
nications as the keynote speakers. In addition, the technical program also included
several jointly organized international workshops aiming to highlight the latest research
developments in all aspects of collaborative computing.
We would like to thank the authors for submitting their research papers to the
conference and contributing to the quality of the final program. We are also grateful to
all Technical Program Committee members and the reviewers for their time, efforts,
and comments in selecting high-quality papers for inclusion in our great technical
program.

December 2015 Yanmin Zhu

Fangming Liu
 Organization

Steering Committee
Imrich Chlamtac (Co-chair) Create-Net, Italy
James Joshi (Co-chair) University of Pittsburgh, USA
Calton Pu Georgia Institute of Technology, USA
Elisa Bertino Purdue University, USA
Arun Iyengar IBM, USA
Tao Zhang Cisco, USA
Dimitrios Gerogakopolous CSIRO, Australia

Organizing Committee
General Co-chairs
Song Guo University of Aizu, Japan
Xiaofei Liao Huazhong University of Science and Technology,
China

TPC Co-chairs
Fangming Liu Huazhong University of Science and Technology,
China
Yanmin Zhu Shanghai Jiao Tong University, China

TPC Vice Chair

Deze Zeng China University of Geosciences, China

Workshops Chair
Wenbin Jiang Huazhong University of Science and Technology,
China

Local Arrangements Chair

Hong Yao China University of Geosciences, China

Publication Chair
Fei Xu East China Normal University, China

Web Chair
Fei Xu East China Normal University, China
VIII Organization

Technical Program Committee

Shu-Ching Chen Florida International University, USA
Maria Luisa Damiani University of Milan, Italy
Schahram Dustdar TU Wien, Austria
Federica Paci University of Southampton, UK
Julian Jang-Jaccard CSIRO ICT Centre, Australia
Ralf Klamma RWTH Aachen University, Germany
Kun-Lung Wu IBM T.J. Watson Research Center, USA
Ibrahim Korpeoglu Bilkent University, Turkey
Dan Lin Missouri University of Science and Technology, USA
Patrizio Pelliccione Chalmers University of Technology and University of
Gothenburg, Sweden
Agostino Poggi University of Parma, Italy
Ravi Sandhu University of Texas at San Antonio, USA
Shankar Banik The Citadel, USA
Ting Wang IBM Research, USA
Jinsong Han Xi’an Jiaotong University, China
Chao Jing Guilin University of Technology, China
Haiming Chen Institute of Computing Technology, Chinese Academy
of Sciences, China
Chunming Hu Beihang University, China
Haibin Cai East China Normal University
Tianyu Wo Beihang University, China
Dong Li Institute of Computing Technology, Chinese Academy
of Sciences, China
Sabrina Leone Università Politecnica delle Marche, Ancona, Italy
Fei Xu East China Normal University
 Contents

Collaborative Cloud Computing

Adaptive Multi-keyword Ranked Search Over Encrypted Cloud Data . . . . . . 3

Daudi Mashauri, Ruixuan Li, Hongmu Han, Xiwu Gu, Zhiyong Xu,
and Cheng-zhong Xu

A Collaborated IPv6-Packets Matching Mechanism Base on Flow Label in

OpenFlow . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 14
Weifeng Sun, Huangping Wei, Zhenxing Ji, Qingqing Zhang,
and Chi Lin

Traveller: A Novel Tourism Platform for Students Based on Cloud Data . . . . 26

Qi-Ying Hu, Chang-Dong Wang, Jia-Xin Hong, Meng-Zhe Hua,
and Di Huang

Achieving Application-Level Utility Max-Min Fairness of Bandwidth

Allocation in Datacenter Networks . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 36
Wangying Ye, Fei Xu, and Wei Zhang

Cost-Effective Service Provisioning for Hybrid Cloud Applications . . . . . . . . 47

Bin Luo, Yipei Niu, and Fangming Liu

Architecture and Evaluation

On Rule Placement for Multi-path Routing in Software-Defined Networks . . . 59

Jie Zhang, Deze Zeng, Lin Gu, Hong Yao, and Yuanyuan Fan

Crowdstore: A Crowdsourcing Graph Database. . . . . . . . . . . . . . . . . . . . . . 72

Vitaliy Liptchinsky, Benjamin Satzger, Stefan Schulte,
and Schahram Dustdar

An ARM-Based Hadoop Performance Evaluation Platform:

Design and Implementation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 82
Xiaohu Fan, Si Chen, Shipeng Qi, Xincheng Luo, Jing Zeng,
Hao Huang, and Changsheng Xie

Research on Service Organization Based on Decorator Pattern . . . . . . . . . . . 95

Jianxiao Liu, Zaiwen Feng, Zonglin Tian, Feng Liu, and Xiaoxia Li

Personalized QoS Prediction of Cloud Services via Learning

Neighborhood-Based Model . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 106
Hao Wu, Jun He, Bo Li, and Yijian Pei
X Contents

Collaborative Application

Multi-core Accelerated Operational Transformation for Collaborative

Editing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 121
Weiwei Cai, Fazhi He, and Xiao Lv

NFV: Near Field Vibration Based Group Device Pairing . . . . . . . . . . . . . . . 129

Zhiping Jiang, Jinsong Han, Wei Xi, and Jizhong Zhao

A Novel Method for Chinese Named Entity Recognition Based on

Character Vector . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 141
Jing Lu, Mao Ye, Zhi Tang, Xiao-Jun Huang, and Jia-Le Ma

Android Apps Security Evaluation System in the Cloud. . . . . . . . . . . . . . . . 151

Hao Wang, Tao Li, Tong Zhang, and Jie Wang

Sensor and Internet of Things

Protecting Privacy for Big Data in Body Sensor Networks: A Differential

Privacy Approach . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 163
Chi Lin, Zihao Song, Qing Liu, Weifeng Sun, and Guowei Wu

Characterizing Interference in a Campus WiFi Network via Mobile Crowd

Sensing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 173
Chengwei Zhang, Dongsheng Qiu, Shiling Mao, Xiaojun Hei,
and Wenqing Cheng

On Participant Selection for Minimum Cost Participatory Urban Sensing

with Guaranteed Quality of Information . . . . . . . . . . . . . . . . . . . . . . . . . . . 183
Hong Yao, Changkai Zhang, Chao Liu, Qingzhong Liang, Xuesong Yan,
and Chengyu Hu

k-CoAP: An Internet of Things and Cloud Computing Integration Based on

the Lambda Architecture and CoAP . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 195
Manuel Díaz, Cristian Martín, and Bartolomé Rubio

A Framework for Multiscale-, QoC- and Privacy-aware Context

Dissemination in the Internet of Things . . . . . . . . . . . . . . . . . . . . . . . . . . . 207
Sophie Chabridon, Denis Conan, Thierry Desprats, Mohamed Mbarki,
Chantal Taconet, Léon Lim, Pierrick Marie, and Sam Rottenberg

Security

SSG: Sensor Security Guard for Android Smartphones . . . . . . . . . . . . . . . . 221

Bodong Li, Yuanyuan Zhang, Chen Lyu, Juanru Li, and Dawu Gu
 Contents XI

Fast Secure Scalar Product Protocol with (almost) Optimal Efficiency . . . . . . 234
Youwen Zhu, Zhikuan Wang, Bilal Hassan, Yue Zhang, Jian Wang,
and Cheng Qian

Efficient Secure Authenticated Key Exchange Without NAXOS’ Approach

Based on Decision Linear Problem . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 243
Mojahed Ismail Mohamed, Xiaofen Wang, and Xiaosong Zhang

Towards Secure Distributed Hash Table. . . . . . . . . . . . . . . . . . . . . . . . . . . 257

Zhe Wang and Naftaly H. Minsky

An Anomaly Detection Model for Network Intrusions Using One-Class

SVM and Scaling Strategy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 267
Ming Zhang, Boyi Xu, and Dongxia Wang

Short Paper

Layered Consistency Management for Advanced Collaborative Compound

Document Authoring . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 281
Johannes Klein, Jean Botev, and Steffen Rothkugel

On Ambiguity Issues of Converting LaTeX Mathematical Formula to

Content MathML . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 289
Kai Wang, Xinfu Li, and Xuedong Tian

LTMF: Local-Based Tag Integration Model for Recommendation . . . . . . . . . 296

Deyuan Zheng, Huan Huo, Shang-ye Chen, Biao Xu, and Liang Liu

A Privacy-Friendly Model for an Efficient and Effective Activity

Scheduling Inside Dynamic Virtual Organizations . . . . . . . . . . . . . . . . . . . . 303
Salvatore F. Pileggi

A Discovery Method of Service Bottleneck for Distributed Service . . . . . . . . 309

Jie Wang, Tao Li, Hao Wang, and Tong Zhang

A Collaborative Rear-End Collision Warning Algorithm in Vehicular

Ad Hoc Networks . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 317
Binbin Zhou, Hexin Lv, Huafeng Chen, and Ping Xu

Analysis of Signaling Overhead and Performance Evaluation in Cellular

Networks of WeChat Software . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 323
Yuan Gao, Hong Ao, Jian Chu, Zhou Bo, Weigui Zhou, and Yi Li

Exploration of Applying Crowdsourcing in Geosciences: A Case Study

of Qinghai-Tibetan Lake Extraction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 329
Jianghua Zhao, Xuezhi Wang, Qinghui Lin, and Jianhui Li

Author Index . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 335

Collaborative Cloud Computing
 Adaptive Multi-keyword Ranked Search Over
Encrypted Cloud Data

Daudi Mashauri1, Ruixuan Li1(&), Hongmu Han1, Xiwu Gu1,

Zhiyong Xu2, and Cheng-zhong Xu3,4
1
School of Computer Science and Technology, Huazhong University
of Science and Technology, Wuhan 430074, Hubei, China
[email protected],
{rxli,hanhongmu,guxiwu}@hust.edu.cn
2
Department of Mathematics and Computer Science,
Suffolk University, Boston, MA 02114, USA
[email protected]
3
Department of Electrical and Computer Engineering,
Wayne State University, Detroit, MI 48202, USA
[email protected]
4
Shenzhen Institute of Advanced Technology, Chinese Academy of Science,
Shenzhen 518055, Guangdong, China

Abstract. To preserve data privacy and integrity, sensitive data has to be

encrypted before outsourcing to the cloud server. However, this makes keyword
search based on plaintext queries obsolete. Therefore, supporting efficient
keyword based ranked searches over encrypted data became an open challenge.
In recent years, several multi-keyword ranked search schemes have been pro-
posed in trying to solve the posed challenge. However, most recently proposed
schemes don’t address the issues regarding dynamics in the keyword dictionary.
In this paper, we propose a novel scheme called A-MRSE that addresses and
solves these issues. We introduce new algorithms to be used by data owners
each time they make modifications that affects the size of the keyword dic-
tionary. We conduct multiple experiments to demonstrate the effectiveness of
our newly proposed scheme, and the results illustrates that the performance of
A-MRSE scheme is much better that previously proposed schemes.

Keywords: Cloud computing
Searchable encryption
Multi-keyword query

Ranked search
Encrypted data

1 Introduction

Cloud computing is a model for enabling ubiquitous, convenient, on-demand network

access to a shared pool of configurable computing resources, such as networks, servers,
storage, applications and services, which can be rapidly provisioned and released with
minimal management effort or service provider interaction [1]. Cloud computing
provides affordable and convenient ways to store and manage huge amounts of data
generated by data owners. However, even with all of its advantages, cloud computing
still faces great challenges following a serious threat posed to data owners about

© Institute for Computer Sciences, Social Informatics and Telecommunications Engineering 2016
S. Guo et al. (Eds.): CollaborateCom 2015, LNICST 163, pp. 3–13, 2016.
DOI: 10.1007/978-3-319-28910-6_1
4 D. Mashauri et al.

security and privacy, especially when it comes to sensitive data. Encryption of data
before shipping to the cloud server offers a viable solution to data owners regarding
data integrity and confidentiality. However, keyword searches based on plaintext
queries became obsolete on encrypted data, and data owners have to download the
entire database they own and start decrypting files one by one, looking for those of
interest. It goes without saying that this is too much to bear, especially in today’s
pay-as-you use fashion.
Searchable encryption is the earliest scheme that utilizes keyword search over
encrypted data [2, 3]. Solutions proposed by these schemes resolved the issues con-
cerning security and privacy of data. However, they also introduced more obstacles as
they encountered huge computation and communication overheads. These overheads
resulted from the fact that the proposed solutions do not offer any ranking mechanism
after performing a keyword search as they based on disjunctive searches. To resolve
this issue, another era of ranked keyword search schemes over encrypted came to the
rescue. Single keyword ranked search [4] was among the first published works pro-
viding a practical implementation. It fulfilled its designed goals as far as ranked
searches are concerned. However, supporting only a single keyword search from
thousands of encrypted files was not an efficient solution that data owners anticipated
for a while.
Recently, a number of research works, such as [5, 6], have been done in order to
facilitate multi-keyword queries over encrypted cloud data and they also support result
ranking. MRSE [1] is one of the earlier works crowned in supporting multi-keyword
ranked queries over encrypted data. It also provides a viable solution that works under
practical implementation. Most of the recent works on multi-keyword ranked searches
don’t address the issues regarding any future modifications that will affect the size and
content of the keyword dictionary. There is a large computation and communication
overhead posed to data owners each time they modify their keyword dictionaries. As a
privacy requirement, two query vectors resulting from similar set of keywords cannot
be the same. Hence, the cloud server won’t be able to determine if they come from the
same set of keywords. However, the cloud server can still determine due to the fact that
finally they will result into similar sets of ranked files, although they look different
upon submission.
In this paper, we propose a new scheme called A-MRSE in order to resolve the
resulting issues of modifications on keyword dictionaries. In our newly proposed
scheme, we consider both real life scenarios where the data owners can either insert or
remove certain keywords from the dictionary. We propose new algorithms that can be
used each time the data owner makes these changes. They present minimum com-
munication and computation overhead.
The contributions of this paper can be summarized as follows:
• We design a novel scheme that is adaptive and supports any modification made on
the keyword dictionary, either inserting or removing keywords with minimum
overhead.
• We improve security of the ranked results by sealing the cloud server from any form
of statistical attacks.
 Adaptive Multi-keyword Ranked Search Over Encrypted Cloud Data 5

The rest of this paper is organized as follows. Section 2 introduces our A-MRSE
scheme, which is followed by results and a discussion in Sect. 3. Section 4 describes
related works, and we conclude with future works in Sect. 5.

2 Adaptive Multi-keyword Ranked Search Over Encrypted

Cloud Data

In this section, we will present our adaptive multi-keyword ranked search over
encrypted cloud data (shorted as A-MRSE) scheme. In order to quantitatively evaluate
the coordinate matching like in MRSE, we adopt “inner product similarity” in our work
as well. Also like in MRSE [1], we define an index vector for each file based on the
keywords it contains from the dictionary; two invertible matrices and a bit vector are
also used for index vector encryption and trapdoor generation. However, our work
solves the issue with MRSE in a sense that it allows more keywords to be added in the
dictionary as well as some of them to be removed from it. The detailed design of
A-MRSE scheme includes the following six aspects.
(1) InitialSetup: The data owner selects a set of n keywords from the sensitive
plaintext dataset F, and u dummy keywords to be inserted in the indexing vector in
order to strengthen security and maintain privacy. The index vector mandates any
future modifications that can be made on the keyword dictionary. In MRSE, the index
vector has three parts which are first n locations used to indicate presence or absence of
real keywords, followed by u locations for dummy keywords, and terminated by the
constant 1 at the last position as shown in Fig. 1.

Keywords (n) Dummy (u) 1

Fig. 1. MRSE index vector structure.

MRSE vector structure makes any modification on the keyword dictionary

unworkable since positions of the keywords are fixed. However, in A-MRSE, we
mirror the existing structure and derive a new vector structure having security locations
(last constant dimension and dummy keywords locations) at the beginning, followed by
n locations of real keywords as shown in Fig. 2.

1 Dummy (u) Keywords (n)

Fig. 2. A-MRSE vector structure.

6 D. Mashauri et al.

With this vector structure, A-MRSE supports any future modifications of the
keyword dictionary size. For any file in the dataset, if it has real keyword Wj, then in the
corresponding index vector p[1 + u+j] = 1, otherwise 0.
(2) KeyReduce: since A-MRSE supports keyword dynamics in the dictionary as
compared to MRSE, the data owner calls this algorithm with number of keywords to be
reduced as an input parameter to generate a new secret key SK k2 from the previously
generated SK k1. Previously generated matrices M1 and M2 will then be resized into
new matrices M10 and M20 each having (d-r) × (d-r) dimension. That is accomplished by
removing the last r-rows and r-columns, which finally yields a d-r square matrix.
Modification of the splitting vector S demands special attention due the role played
of each bit position in it. Basically, the dictionary size will change from n to (n-r) after
removing r keywords. This algorithm inspects the new dictionary. If a keyword still
exists in both dictionaries (the old and new sized), then that particular bit is copied into
a new vector S0 and omitted otherwise. The process continues for all n locations and
finally gives a ðd  r ÞS0 bit vector. Algorithm 1 shows how KeyReduce works.

Algorithm 1. KeyReduce(k1, r)

Input: number of reduced keywords and original secret key.

Output: new secret key SK k2.
Method: the key reduce algorithm works as follows.
1: Receive the integer input parameter r;
2: Retrieve the original secret key SK k1 components;
3: Resize matrices M1, M2 to M'1, M'2 by applying dimension reduction;
4: Read the old dictionary as file f1 and new dictionary as file f2;
5: for each line in file f1 and file f2
6: if (keyword in f1 exists in f2)
7: copy the value of bit position for this keyword from S into S';
8: else
9: skip bit position;
10: end if
11: end for
12: Give SK k2 with 3-tuple as {S', M'1, M'2};

(3) KeyExtend: after adding more files on the cloud server, the data owner will
definitely need to include new keywords in the dictionary. In this case, MSRE cannot
work any longer as it suffers huge computation overhead as well as bandwidth inef-
ficiency. This is where A-MRSE comes into account as it allows easy expansion of the
secret keys relative to the increase of keywords in the dictionary.
 Adaptive Multi-keyword Ranked Search Over Encrypted Cloud Data 7

If z keywords are added, this algorithm generates two new z × z invertible matrices,
Mz1, Mz2, and a new z bit vector Sz. These newly created matrices will be added to
original matrices M1 and M2 and finally gives two modified matrices M10 and M20 having
(d + z) × (d + z) dimension according to block diagonal matrix theorem [7].
On the other hand, splitting vectors Sz and S will be joined and make a new vector
S0 by copying all elements of vector S into S0 then followed by appending elements of
Sz. Algorithm 2 shows how KeyExtend works.

Algorithm 2. KeyExtend(k1, z)

Input: original secret key and number of newly added keywords.

Output: new secret key SK k3.
Method: the key extend algorithm works as follows.
1: Receive the integer input z;
2: Retrieve original key SK k1 components;
3: Generates two invertible matrices Mz1, Mz2 and a bit vector Sz;
4: Add Mz1 to M1, Mz2 to M2 by using diagonal block matrix operation and produces
M'1 and M'2 having (d+z) (d+z) dimensions;
5: for each bit position in vector S
6: copy it into a new vector S';
7: end for
8: for each bit position in vector Sz
9: copy and append it into new vector S';
10: end for
11: Give SK k3 with 3-tuple as {S', M'1, M'2};

(4) BuildIndex: this algorithm builds an encrypted searchable index for plaintext
files in the original set F. Initially, the data owner applies similar procedures as in
MRSE [1] before addition or reduction of keywords from the dictionary.
For each file, a bit vector pi is set. Then starting with security positions, p[1] is set
to 1, and values in dummy keyword positions between p[1] and p[2 + u] are set to a
random number e. The remaining positions will be filled, indicating whether the file
contains keywords from the dictionary. Therefore, p[2 + u] to p[1 + u+n] will be set to
1 if the file contains a dictionary keyword and 0 otherwise. After setting all bit positions
in vector pi, splitting procedures will then follow as in secure kNN computation [8]
except that the index structure is reversed. This implies in A-MRSE, we start with
security locations then followed by real keyword locations. The BuildIndex algorithm
is shown in Algorithm 3.
8 D. Mashauri et al.

Algorithm 3. BuildIndex(F, SK)

Input: the secret key SK, and the file set F.

Output: the encrypted searchable index.
Method: the build index algorithm works as follows.
1: Receive the file set F;
2: for each Fi F
3: Generate a bit vector pi;
4. Set p[1] = 1, and p[2] – p[1+u] = i;
5. for j = (u + 2) to (1 + u + n)
6. if Fidj • W
7. Set p[j] = 1;
8. else set p[j] = 0;
9. end if
10. for j = 1 to (1 + u + n)
11. if S[j] = 1
12. p1[j] + p2[j] := p[j];
13: else p1[j] := p2[j] := p[j];
14: Run p1 = M1Tp1, p2 = M2Tp2, and set Ii = {p1, p2};
15: Upload encrypted files {Fi} • C and I = {Ii} to the cloud server;

(5) TrapdoorGen: Having a set of interested keywords, an authorized data con-

sumer calls this algorithm to generate a secure trapdoor in order to search and retrieve a
number of encrypted files from the cloud server. For a multi-keyword query q, a query
vector is generated using the same strategy as in MRSE with v number of dummy
locations set to 1 and all remaining locations set to 0.
A score is used to determine the location of the file in the matching result set.
In MRSE, this score was calculated by using Eq. 1 for index file pi.
 X 
pi
q ¼ r x i þ eðvÞ þ ti ð1Þ

Xu et al. [5] discovered the impact of values of the dummy keywords inserted in the
final score, which causes “out-of-order” problem. This happens when a file with
popular keywords obtaining lower score and finally will not be included in the returned
list to the data consumer.
To ameliorate the in-order ranking result while maintaining privacy-preserving
property, all locations containing real keywords are multiplied by random number r,
and all locations containing dummy keywords are multiplied by another random
number r2 which is obtained by using Eq. 2. Finally, the score is calculated by using
Eq. 3.
 Adaptive Multi-keyword Ranked Search Over Encrypted Cloud Data 9

Randomð0; r Þ
r2 ¼ ð2Þ
ðv  MAX ððu  cÞ; ðu þ cÞÞÞ
X
pi
q ¼ Ii
T ¼ r  xi þ r2  e ð vÞ þ t i ð3Þ
 
Finally, the trapdoor T will be generated as M11 q0 ; M21 q00 .
(6) Query: After receiving T from the data consumer, the cloud server calls this
algorithm to calculate the score for each file in the encrypted index I. The data con-
sumer also includes parameter K so that the cloud server will return a list of only top-K
files after a ranked search over the encrypted index. The trapdoor generated from a
similar set of keywords will be different each time. This prevents the cloud server from
performing statistical attacks; however the cloud server can still determine the trap-
doors came from the same keyword set since finally they all retrieve identical top-K
files though with different scores.
To resolve this issue, in A-MRSE we designed a new way to obfuscate the cloud
server from performing statistical attacks. We modify the total number of retrieved files
by adding K 0 files randomly such that K 0 \K. The value of K 0 is obtained by using
Eq. 4.

K0 ¼ e  K ð4Þ

The value of e is used as a security parameter, and it grows from 0 % depending on

the number of files to be retrieved from the cloud server. If it set to 0 %, the implication
is that the data owner prefers efficiency over security. For instance, when K is less than
10, the value of e can be set to 25 %, when K lies between 10–20 it can be set to 20 %,
and when K reaches 50 % it can be 15 %. By doing so, the cloud server cannot
determine whether two queries originated from the same keyword set.

3 Performance Evaluation

In this section we present the results obtained after performing multiple experiments
with different settings. We selected a real life dataset, Enron Email Dataset [9], various
numbers of emails were randomly selected from the dataset for each test. The work-
bench was a Dell Latitude E-5520 machine with an Intel CoreTM i7 CPU @
2.20 GHz × 8 with 8 GB of RAM. The operating system is Linux Mint 15 (Olivia x64),
simulation codes were implemented by using Java programming language and ele-
ments in invertible matrices were double, generated randomly by using a Jama-1.0.3.jar
package [10]. For each test taken, we observed results for both A-MRSE and MRSE
scheme and then we compared their performances.
(1) Key Generation and Editing: the total time in key generation includes the time
to generate bit vector S, as well as the two invertible matrices which then followed by
adding the time to transport and compute the inverse of these matrices. Figure 3 shows
how A-MRSE outperforms MRSE during key generation starting from 1000 keywords
dictionary size and keeps growing up to 8000. In this phase, the previously generated
10 D. Mashauri et al.

1000 key size was used to create new expanding keys. For instance, for a 5000 key,
A-MRSE uses 32.94 % of total time, and for a 6000 A-MRSE uses 37.85 % of total
time as compared with MRSE.

Fig. 3. Key reduction and generation Fig. 4. Index building with different key sizes.

A remarkable gain is observed when the dictionary size is reduced. It took much
less time to edit the key with A-MRSE than to regenerate with MRSE. As shown in
Fig. 3, A-MRSE remains almost flat during key regeneration while MRSE raises as the
keywords grow in the dictionary. For example, it took 0.33 % of the total time for
A-MRSE to generate a 5000 key from an existing 6000 in case 1000 keywords are
dropped from the dictionary where by MRSE took 298 times more than A-MRSE.
(2) Index Building: the time taken to build a searchable index I for all documents is
the sum of the individual time taken to build indexes I for each document. This
includes mapping of keywords extracted from file Fi to a data vector pi, followed by
encrypting all the data vectors and finally builds a searchable index that will be
uploaded to the cloud server. The cost of mapping or encrypting primarily depends on
the dimension of the data vector which is tied up to the dictionary size. Also, the cost of
building the whole searchable index I depends on the number of sub indexes which
implies the total number of documents in the dataset.
Figure 4 shows a comparison of the time taken to build the searchable index for
both MRSE and A-MRSE with different numbers of keywords in the dictionary where
the documents in the dataset were fixed to 3000. A-MRSE outperforms MRSE, for
instance building index with a 6000 key size, it took 44 % of the total time as compared
with MRSE. Figure 5 shows comparison of the total time taken to build the searchable
index between A-MRSE and MRSE for different numbers of files in the dataset ranging
from 1 K to 8 K inclusive with the key size fixed to 2 K. Again, A-MRS saw off MRSE
in terms of efficiency, for example it took 37.14 % of total time to build index for a
4000-files dataset.
As we can see, in all settings whether the key is fixed with increasing documents in
the dataset or the other way round, A-MRSE still outperforms MRSE. This is because
the keys in A-MRSE contain many zero-valued elements as a result of key expansion.
This makes multiplications during index building to be much faster in A-MRSE than in
MRSE.
 Adaptive Multi-keyword Ranked Search Over Encrypted Cloud Data 11

Fig. 5. Index building with different number Fig. 6. Trapdoor generation.

of files in the dataset

(3) Trapdoor Generation: trapdoor generation involves two multiplications of a

matrix and a splitting query vector. Figure 6 shows the trapdoor generation cost for
both A-MRSE and MRSE. Again, as seen from the graph, A-MRSE saw off MRSE as
it takes less time to complete the whole process. For example, with 5000 keywords
dictionary size, A-MRSE took 45.31 % of the total time as compared with MRSE
scheme. As more keywords are added, more performance gain can be achieved with
our A-MRSE. We are certain that this gain is highly contributed with many zero-values
elements in A-MRSE matrices compared with the ones in MRSE.
(4) Query: the cloud server is responsible for performing query execution by using
querying algorithm with which it computes and ranks the similarity scores for all
documents in the dataset. We conducted several experiments for both A-MRSE and
MRSE with fixed number of documents in the dataset while varying the dictionary size.
The performances were almost comparable between A-MRSE and MRSE. The key
point of our scheme is that, it enables data owners to make use of existing keys and
generate new searching and indexing keys without the need of running key generation
again from scratch. This is for both scenarios of increasing and decreasing number of
keywords in the dictionary.

4 Related Works

Cloud computing is delivered as a product from both software and hardware evolutions
as well as the Internet. It offers actual realization of the long waited utility computing
service. However, with all its benefits, it still comes with a number of security chal-
lenges posed to both individual data owners as well as organizations which use cloud
technology [11]. Enabling keyword search over encrypted data in cloud computing
environment became an open question.
Song et al. [2] was among the earliest researchers to present a practical solution of
keyword search over encrypted data. They offered a solution in which each word in the
plaintext is encrypted with a two-layer encryption scheme that uses stream ciphers.
12 D. Mashauri et al.

Boneh et al. [12] introduced a public keyword encryption scheme with keyword search.
Similar works were also presented in [13] and [14] that put forward searchable
encryption schemes. However, solutions based on public key encryption are usually
very computation expensive. Furthermore, keyword privacy is not protected in public
key setting since the server could encrypt any keyword with the known public key and
then use the received trapdoor to evaluate its ciphertext.
Wang et al. [4] presented the earlier work that explores user’s capability of ranked
search over encrypted cloud data. Ranked search improves systems efficiency and
usability by returning all matching files in a ranked order depending on predefined user
rules, and for this case is its file length and document frequency. Other solutions are
presented in [3] and [15], but all these works presented solutions based on single
keyword search only. Singe keyword ranked search is computational inefficiency when
the number of documents is quite large as the final result can include almost all
documents in the set C as long as they contain a single searched keyword.
Multi-keyword ranked searches came to puzzle out this issue, as initially presented by
Cao et al. [1]. The scheme is semantically secure. However, it lacks actual
implementation.
None of the above schemes addressed the challenges of varying keyword dictionary
size following addition or reduction of files on cloud server. A-MRSE presents a novel
scheme with new algorithms that address this issue and can support any dynamic in
keyword dictionary size with minimum communication and computation overhead.
A-MRSE achieves its desired goals and leaving no drawback as in [1].

5 Conclusion and Future Works

In this paper, we present a novel scheme that is adaptive and it supports multi-keyword
ranked search over encrypted cloud data. We present a novel scheme, called A-MRSE,
which uses new algorithms to solve the existing challenges on multi-keyword keyword
searches over encrypted data. We also strengthen security and privacy by preventing
the cloud server from performing statistical attacks based on the results to be returned
to data consumers after performing a ranked search. A-MRSE can be easily deployed in
many cloud scenarios, such as UniDrive, a synergizing multiple consumer cloud
storage service [16]. We conducted multiple experiments under different settings and
the results illustrates that our new A-MRSE scheme is much better than MRSE.
In the future, we are looking forward to building schemes that can work under
stronger security threats especially when the cloud server is capable of colluding. We
are working on building adaptive schemes that can work for fuzzy keyword searches.

Acknowledgments. This work is supported by National Natural Science Foundation of China

under grants 61173170, 61300222, 61433006 and U1401258, Innovation Fund of Huazhong
University of Science and Technology under grants 2015TS069 and 2015TS071, and Science and
Technology Support Program of Hubei Province under grant 2014BCH270 and 2015AAA013, and
Science and Technology Program of Guangdong Province under grant 2014B010111007.
 Adaptive Multi-keyword Ranked Search Over Encrypted Cloud Data 13

References
1. Cao, N., Wang, C., Li, M., Ren, K., Lou, W.: Privacy-preserving multi-keyword ranked
search over encrypted cloud data. In: Proceedings of INFOCOM, pp. 829–837. IEEE (2011)
2. Song D.X., Wanger, D., Perrig, A.: Practical techniques for searches on encrypted data. In:
2000 IEEE Symposium on Security and Privacy, pp. 44–55. IEEE (2000)
3. Kiltz, E., et al.: Searchable encryption revisited: consistency properties, relation to
anonymous IBE, and extensions. In: Shoup, V. (ed.) CRYPTO 2005. LNCS, vol. 3621,
pp. 205–222. Springer, Heidelberg (2005)
4. Wang, C., Cao, N., Li, J., Ren, K., Lou, W.: Secure ranked keyword search over encrypted
cloud data. In: Proceedings of IEEE 30th International Conference in Distributed Computing
Systems (ICDCS), pp. 253–262. IEEE (2010)
5. Xu, Z., Kang, W., Li, R., Yow, K., Xu, C.: Efficient multi-keyword ranked query on
encrypted data in the cloud. In: The 18th International Conference on Parallel and
Distributed Systems (ICPADS), pp. 244–251. IEEE (2012)
6. Jiad, Y., Lu, P., Zhu, Y., Xue, G., Li, M.: Towards secure multi-keyword top-k retrieval over
encrypted data. IEEE Trans. Dependable Secure Comput. 10(4), 239–250 (2013)
7. Feingold, D.G., Richard, S.V.: Block diagonally dominant matrices and generalizations of
the Gerchgorin Circle theorem. Pacific J. Math. 12(4), 1241–1250 (1962)
8. Wong, W.K., Cheung, D.W., Kao, B., Mamoulis, N.: Secure kNN computation on encrypted
databases. In: Proceedings of the 2009 ACM SIGMOD International Conference on
Management of Data, pp. 139–152. ACM (2009)
9. Cohen W.W.: Enron Email Dataset. http://www.cs.cmu.edu/*enron
10. Hicklin, J., Moler, C., Webb, P., Boisvert, R.F., Miller B., Pozo, R., Remington, K.: JAMA:
Java Matrix Package. http://www.math.nist.gov/javanumerics/jama/
11. Mather, T., Kumaraswamy, S., Latif, S.: Cloud Security and Privacy: an Enterprise
Perspective on Risks and Compliance. O’Reilly Media Inc., Sebastopol (2009)
12. Boneh, D., Di Crescenzo, G., Ostrovsky, R., Persiano, G.: Public key encryption with
keyword search. In: Cachin, C., Camenisch, J.L. (eds.) EUROCRYPT 2004. LNCS, vol.
3027, pp. 506–522. Springer, Heidelberg (2004)
13. Mitzenmacher, M., Chang, Y.-C.: Privacy preserving keyword searches on remote encrypted
data. In: Ioannidis, J., Keromytis, A.D., Yung, M. (eds.) ACNS 2005. LNCS, vol. 3531,
pp. 442–455. Springer, Heidelberg (2005)
14. O’Neill, A., Bellare, M., Boldyreva, A.: Deterministic and efficiently searchable encryption.
In: Menezes, A. (ed.) CRYPTO 2007. LNCS, vol. 4622, pp. 535–552. Springer, Heidelberg
(2007)
15. Li, J., Wang, Q., Wang, C., Kao, N., Ren, K., Lou, W.: Fuzzy keyword search over
encrypted data in cloud computing. In: Proceedings of INFOCOM, pp. 1–5. IEEE (2010)
16. Tang, H., Liu, F.M., Shen, G., Jin, Y., Guo, C.: UniDrive: synergize multiple consumer
cloud storage services. In: ACM/USENIX/IFIP Middleware, Vancouver, Canada (2015)
 A Collaborated IPv6-Packets Matching
Mechanism Base on Flow Label in OpenFlow

Weifeng Sun, Huangping Wei, Zhenxing Ji,

Qingqing Zhang, and Chi Lin(&)

School of Software, Dalian University of Technology, Dalian, Liaoning, China

{wfsun,c.lin}@dlut.edu.cn,
{1299856803,jasonjee1990,1165331393}@qq.com

Abstract. Software Defined Networks (SDN), the separation of a network

device’s control and data planes, do not need to rely on the underlying network
equipment (routers, switches, firewall).It is a new network which collaborated IP
and a lot of relevant technical content. The control of SDN is completely open,
the user can customize any rule strategy to achieve network routing and
transmission, which is more flexible and intelligent. Internet Protocol version 6
(IPv6), with the 128-bit address, is the next generation Internet. In this paper, we
present a Flow Table structure by using Flow Label and a matching approach
which use Flow Label within IPv6 protocol to decrease the size of Flow
Table with OpenFlow and the time of forwarding IPv6 packets in SDN based on
OpenFlow. The simulations and analyses show that this flow table mechanism
performs better.

Keywords: IPv6
OpenFlow
Flow label
Compression of flow table

1 Introduction

The Software Define Networks (SDN) [1] is an approach or architecture to not only
simplify computer networks but also to make it more reactive to the requirements of
workloads and services placed in the network. Originally, the control planes and data
planes are consolidated into network equipments such as the IP routers and the Ethernet
switches. SDN collaborates a lot of network transmission protocols and functions. It
allows for a centrally managed and distributed control, management, and data plane,
which policy that dictates the forwarding rules is centralized, but the actual forwarding
rule processing are distributed among multiple devices.
Internet Protocol version 6 (IPv6) [2] is the latest version of the Internet Protocol. It
uses a 128 bit address, allowing 2128 times as many as IPv4, which uses 32 bit
addresses. IPv6 is bringing many unique benefits when it is combined with emerging
technologies such as Clouds Virtualization, Internet of Things, etc. These benefits
include host automation, scalability in addressing, route aggregation, forwarding and
traffic steering functions [3]. IPv6 enables the transformation that occurs at the net-
working infrastructure level which can make SDN and network easily easy to scale.
There is no doubt that most of the SDN solutions are based on the OpenFlow protocol
[4], which was defined by Open Networking Foundation (ONF) [5]. OpenFlow was

© Institute for Computer Sciences, Social Informatics and Telecommunications Engineering 2016
S. Guo et al. (Eds.): CollaborateCom 2015, LNICST 163, pp. 14–25, 2016.
DOI: 10.1007/978-3-319-28910-6_2
 A Collaborated IPv6-Packets Matching Mechanism 15

originally imagined and implemented as part of network research at Stanford University.

And Nick McKeown firstly descripts detail the concept OpenFlow [6]. However, as the
IPv6 support in OpenFlow, It is a problem that how to reduce the size of the flow table
faced with OpenFlow. According to the OpenFlow standard, the flow table can be
achieved with TCAM, Ternary Content Addressable Memory that is based on the con-
tents of the address to find. In the traditional network equipment, TCAM mainly for FIB,
MAC, MPLS Label and ACL table because of the length of the match fields for each table
vary, so it can be separately designed, and has the maximum capacity limits in order to
achieve a minimum overhead. Although OpenFlow designed multi-flow table, in order to
reduce the overhead stream table, forming a handle in the form of pipeline, reducing the
total number of records in the flow table. With the increasing of the matching fields, the
space of flow table will be limited. This problem will become more and more prominent
with the increasing of hosts. The size of flow table and the matching processing of
OpenFlow that be more complex restrict the IPv6 development in SDN.
In this paper, we propose the flow table structure by using Flow Label within IPv6
protocol, in order to decrease the bite of flow table and the time of forwarding IPv6
packets in SDN based on OpenFlow. OpenFlow is designed forwarding of L1-L4 layer
forwarding entries. We use Flow Label of L3 layer to match, instead of the address of
IPv6 and the port of TCP/UDP, the OpenFlow match fields of L3-L4 Layer.
The main contributions of this paper are:
(1) We propose the flow table structure by using Flow Label in IPv6 feature, instead
of L3-L4 Layers elements of matching, in order to decrease the time of matching
and the IPv6 flow table structure of OpenFlow.
(2) We analysis and compare the latency, the jitter and the size of flow table between
flow table with flow label and flow table without flow label, and proved that the
former can perform OpenFlow better in SDN.

2 Related Work

The first edition of OpenFlow focused on IPv4 and did not support IPv6 flow [7]. ONF
started to consider how IPv6 flows could be accommodated. Then the ONF published
their OpenFlow1.2 which is the first version that supports IPv6 packet matching [8].
The OpenFlow1.2 provides basic support for IPv6, OpenFlow1.2 compliant switch can
match on IP protocol number (Ethernet type 0 × 86dd = IPv6), IPv6 source/destination
addresses, traffic class, flow label, and ICMPv6 types. This is a start at allowing IPv6
unicast and multicast traffics to match and OpenFlow flow table in a switch. When the
ONF published the OpenFlow1.3 [9]. There is a few more IPv6 functions added in
OpenFlow1.3 which expands the IPv6 Extension Header handling support and
essential features including Hop-by-hop IPv6 extension header, Router IPv6 extension
header, Router IPv6 extension header, access control, quality of service and tunneling
support and had the same ability to match on IPv6 header fields such as
source/destination addresses, protocol number (next header, extension header),
hop-limit, traffic class, flow label, and ICMPv6 type as in Fig. 1. When the ONF release
OpenFlow1.4 nothing changed regarding IPv6 support [10].
16 W. Sun et al.

Ethernet (L2) IP (L3) Tranport (L4)

Ingress
Port VLAN
Src MAC Dst MAC Type Src IP Dst IP Protocol Src Port Dst Port
ID

Fig. 1. Ten-tuple for matching in OpenFlow

With the development of OpenFlow and IPv6, Chia-Wei Tseng combines the
existing IPv6 protocol and the SDN (SDNv6) for future network, to provide the smarter
and reliable network communication architecture. The SDNv6 [11] motivates a net-
work architecture composed of reliable virtual entities, plug-and-play access with
auto-configures and flexible service clouds over a physical network. Wenfeng Xia
proposed a Software Defined Approach to Unified IPv6 Transition which unifies the
variety of IPv6 transition mechanisms [12]. Xiaohan Liu proposes an IPv6 Virtual
Network Architecture (VNET6) to support flexible services in IPv6 network. IPv6 is a
critical protocol in VNET6 [13]. The VNET6 is adaptive to video service with high
bandwidth and low tendency and improves quality of experiences to users. Batalle [14]
proposes the IPv4 and IPv6 routing separation, and provides a different OpenFlow
controllers conducted by the inter-domain routing OpenFlow network management
method. Rodrigo, Fernandes and Rothenberg in Brazil and Hungary started to leverage
these new features in OpenFlow1.3 [15]. Ivan Pepelnjak also uses the new features in
IPv6 [16] to describe how OpenFlow can be used to help secure the IPv6 Neighbor
Discovery Protocol (NDP) because it suffers from many of the same vulnerabilities as
IPv4 ARP. William Stallings present about the various elements of an OpenFlow table
[17] that includes the IPv6 header fields that can be matched. Araji and Gurkan present
ESPM, Embedding Switch ID, Port number and MAC Address within IPv6 protocol
and SDN technology, to decrease CAM table entries on the switch by forwarding the
packets [18]. David R. Newman goes through the installation steps required to set up an
OpenFlow protocol network using Mininet with OpenFlow1.3 support for IPv6, to
evaluate IPv6 unicast and IPv6 multicast [19].

3 IPv6-Packets Flow Label Matching Mechanism

In this section, the matching approach by using Flow Label within IPv6 and the
structure of flow table by using Flow Label within IPv6 is proposed and defined. Flow
Label is a new field in the IPv6 protocol and it can be used for flow classification. And
then, we shows the matching approach by using Flow Label in OpenFlow instead of
L3-L4 Layer elements, the IPv6 source address, the IPv6 destination address, the
protocol version, the TCP/UDP source port, the TCP/UDP destination port and Flow
Label within IPv6 is introduced in Flow Table as the IPv6 required match fields to
reduce the size of flow table. This paper considers the matching approach which flow
label is used for match process and flow table can reduce the rate of IPv6 flow matching
and the size of IPv6 Flow table, to provide the faster and more reliable network
communication in OpenFlow.
 A Collaborated IPv6-Packets Matching Mechanism 17

3.1 Flow Label

Flow Label is a new field in the IPv6 protocol proposed and defined as the length of 20
bits. RFC defines the flow label which the source node can use the flow label of IPv6
header marked packets, and the source node request for specially treatment, such as, QoS
services and other real-time transactions. Flow label can be resolved to meet the needs and
rules by checking the flow label to determine which stream it belongs to. According to the
forwarding rules, the routers and hosts which do not support the flow label need to flow
label field to all zeros, and the receiving packets do not modified the value of flow label.
The traditional traffic classification process is as follows:
(1) Find a destination IP address;
(2) To compare the protocol number;
(3) Compare the destination port number (transport layer performed);
(4) Comparing the source IP address and filter address (here mainly is to determine
whether the packet matches the filtering rules);
(5) Comparing the source port number.
However, a uniqueness flow must have the same attributes, including the source
address, destination address, flow label. Therefore, we propose a matching method of
using Flow label within the IPv6 protocol for OpenFlow which supports IPv6.

3.2 The Matching Approach by Using Flow Label Within IPv6

Based on OpenFlow
Flow is an important concept not only in the data flow communication network but also
in OpenFlow. The Open Flow Switcher consists of one or more flow tables, which
perform packet lookups and forwarding. The switcher communicates with the con-
troller and the controller manages the switcher by the OpenFlow switch protocol. Using
the OpenFlow protocol, the controller can add, update, and delete flow entries in flow
tables. Each flow table in the switcher contains a set of flow entries; each flow entry
consists of match fields, counters, and a set of instructions to apply to matching
packets. A flow table entry consists of a set of L2/L3/L4 match conditions.
In the matching process of L3-L4 Layer which don’t use the flow label, the
OpenFlow agreement will check the packets which are the IPv4 packets or IPv6
packets, and compare the destination IP address. After the comparison of TCP/UDP
source and destination port numbers, using multi-stream pipeline processing table can
effectively enhance the flow table processing efficiency, especially in matching IP
layer, IPv4 needed source IP address, destination IP address, and IPv6 IP address will
be added to 128 bits, greatly increasing the flow table match delay. Before flow label
field is introduced, the common matching process of L3/L4 layer is as follows:
(1) Compare the protocol number;
(2) Compare the source IP address and destination IP address which has 128 bits in
IPv6;
(3) Compare the source port number and the destination port number (transport layer
performed);
18 W. Sun et al.

In the pipeline of multi-flow table, the matching process of flow table match is very
complex as it is in Fig. 2, so that the delay of flow table match of OpenFlow Switcher
will increases obviously, especially in the match of IPv6 address, which has 128 bits.

Matching in OpenFlow
Switch
Packet
Packet In Src MAC,Dst Proto,Src TCP/UDP
Ingress Port Out
MAC,Type,Vla IPv6,Dst IPv6 Src,TCP/UDP
(L1)
n (L2) (L3) Dst (L4)

Matching by using Flow Label

in OpenFlow Switch
Packet
Packet In Src MAC,Dst
Ingress Port IPv6 Flow Out
MAC,Type,Vlan (L2)
(L1) Label

Fig. 2. IPv6 packets are matched against flow table in OpenFlow switcher

We propose a matching approach by using Flow Label within IPv6 feature in IPv6
based on OpenFlow instead of L3-L4 matching fields, such as IPv6 source address,
IPv6 destination address, TCP/UDP source port, TCP/UDP destination port. In the
IPv6 with flow label mechanism, we can use this field, flow label (20 bits), quintuple
information will be combined to generate a random flow label serial number for each
quintuple information, and generates the corresponding entries in the flow state, it can
find the entry for traffic classification according to the value of each stream flow label.
The different source node may randomly send the same flow label value, but the
probability is very small (about 106 level), which can be considered that the source
node will correspond the only flow label value in simple OpenFlow network. The
possible Hash algorithm can be used in the matching of flow label. Thus, the process
based on traffic classification IPv6 flow label can be simplified to a one-step process:
Find the value of the stream flow state table based on the Flow Label. Obviously, in
such a mechanism, matching process of match fields is simplified a lot when the IPv6
packets are matched against flow table.

3.3 Flow Table with Flow Label Within IPv6 Based on OpenFlow
Flow table is an important concept in OpenFlow, Flow Table are a pieces of the
forwarding tables, as MAC table, IP table, ACL in tradition network. In OpenFlow
agreement, each flow table composed by many Flow entries. Flow table entry is the
smallest unit of flow table matched to each flow in network transmission. According to
OpenFlow standard, A flow table consists of many flow entries, and each flow entry
include: Match Fields which match packets, Priority matching precedence of flow
entry, Counters which update when packets are matched, Instructions which can
modify the action set or pipeline processing, timeouts that can maximum the amount of
time or idle time before flow is expired by the switch, cookie of which the data value
chosen by the controller. And a flow table can support L1-L4 matching, so the match
 A Collaborated IPv6-Packets Matching Mechanism 19

fields can consists of L1-L4 matching element, such as L1 (Ingress port, physical port),
L2 (VLAN ID, VLAN PCP, Ethernet source address. Ethernet destination address,
Ethernet type), L3 (IP protocol number, IPv4 or IPv6 source address, IPv4 or IPv6
destination address, IP DSCP, IP ECN, Flow Label), L4 (TCP or UDP source port,
TCP or UDP destination port), just as it is in Fig. 3.

Match
Priority Counters Instructions Timeouts Cookie
Fields

Ethernet (L2) L3-L4(IPv6)

Ingress
Port VLAN
Src MAC Dst MAC Type Flow Label (20 bit)
ID

Fig. 3. Main components of a flow entry in a flow table with flow label

In this paper, we propose to use Flow Label fully within IPv6 protocol in place of
the L3-4 match fields in IPv6 network match fields based on OpenFlow. For example,
in IPv6 flow table based on OpenFlow, when a IPv6 host1 connect to a IPv6 host2, the
Flow Labe will be assigned according to the Packet-in packet, not the match fields of
L3-4 Layer, such as IPV6 source address, IPv6 destination address and TCP/UDP
source port, TCP/UDP destination port. Especially in the match fields of IPv6 address,
which has 128 bits, will increase the size of flow table obviously. Thus the controller
automatically and naturally determines the flow label number of the flow based on its
location in the IPv6 network.

3.4 The Analyses of Flow Label Matching Mechanism Based

on OpenFlow
According to OpenFlow standard, OpenFlow defines the flow table entries of matching
L1-L4 Layers match fields to lookup the match fields and forward packets. A flow table
consists of flow table entries is designed to match against packets, which consists of the
L1-L4 Layer metadata of matching flow table entry. When the packets are matched
against multiple tables in the pipeline, the process of flow table match is very complex,
the complexity of match is expressed:

TðnÞ ¼ Oðn!  að2l ÞÞ ð1Þ

And n is the number of flow tables, a is the number of actions and l is the length of
match fields. As the number of flow tables, the number of actions and the length of
match fields is increasing, it will lead to the increasing of complexity in the pipeline
process. So we analyze the flow label matching mechanism from theory and imple-
mentation in IPv6. The latency of packet forwarding is an important indicator of
network performance. For the network performance of forwarding, we can use the
arithmetic average to evaluate, and is shown in Eq. (2).
Another random document with
no related content on Scribd:
 encystment, 147
Lafoea, 280;
L. dumosa, 280
Lafoëina (Campanulariidae, 280), 277
Laganum, 548, 549
Lagena, 59, 63, 66 f.;
fossil, 70
Lagenaceae, 59
Lagoon, 390 f.
Lamblia, 111;
L. intestinalis, conjugation, 116 n.
Lampetia, 418
Lancet-plate, 599
Lang, on syngamy, etc., 34 n.;
on Protozoa, 46;
on distinctions of pseudopodia, 47 n.
Lankester, on Protozoa, 45 f.;
on classification of Protozoa, 49 n.;
on Proteomyxa, 89;
on Sporozoa, 94;
on Haemosporidae, 102;
on pigment of Stentor coeruleus, 154 n.;
on Torquatella, 155 n.;
on chlorophyll of Ephydatia, 175;
on Limnocodium, 292
Lankesterella, 97, 102
Lankesteria, 97;
L. ascidiae, life-cycle of, 95
Lantern-coelom of Echinus esculentus, 524;
represented by buccal sinus of Holothuria nigra, 566
Lanuginella pupa, 198
Lar, 273;
L. sabellarum, 266, 267, 268
Larcoidea, 77
Larva, of sponges, 180, 226, 227;
"asexual," 228;
of Tubularia, 271;
 of Stylasterina, 284;
of Trachomedusae, 290;
of Narcomedusae, 295;
of Velella, 302;
of Scyphozoa, 317;
of Alcyonaria, 341;
of Renilla, 360;
of Zoantharia, 373;
of Zoanthidae, 405;
of Cerianthidea, 411;
of Ctenophora, 416, 419;
of Cribrella, 462;
of Luidia, 605;
of Asterina gibbosa, 463, 610, 611, 612;
of Ophiuroidea, 606;
of Echinus, 607;
of Synapta, 608;
of Antedon rosacea, 618, 619, 620
Larval brain, of Echinopluteus, 607;
of Antedon rosacea, 619
Larval type of development, 601
Lateral mouth-shields of Ophiothrix fragilis, 485
Lauterborn, on sapropelic organisms, 48;
on budding in Rhizopods, 56 n.
Laveran, on Sporozoa, 94;
on Acystosporidae, 102
Laverania, 97;
parasite of bilious or pernicious fever, 104 f.
Lebrunia, 382;
L. coralligens, 373
Lecqueureusia, 52;
L. spiralis, test of, 55
Lee, A. Bolles, on action of clearing reagents, 11 n.
Leech, host of Haemogregarina, for sexual process, 102
Leeuwenhoek, on organisms of putrefaction, 42 f.
Léger, on Protozoa, 45;
on Sporozoa, 94;
 on sperms of Pterocephalus, 99 n.;
and Duboscq, on Sarcocystis tenella, 108 n.
Leidy, on Protozoa, 46
Leiopathes, 368, 409
Leiopathidae, 409
Leiosella, 225
Leipoldt, on the madreporic vesicle of Echinoidea, 528
Lelapia australis, 192
Lembadion, 137;
caudal cilia of, 141 n.
Lembus, 137;
caudal cilia of, 141 n.
Lemnalia, 349
Lendenfeld, von, 218, 220, 220 n.
Lepidogorgia, 355
Leptobrachiidae, 325
Leptodiscus, 110, 132, 134
Leptogonaster, 456, 467
Leptogorgia, 340, 357
Leptopenus, 404
Leptophyllia, 404
Lesser, Hertwig and, on Heliozoa, 71
Lesueuria, 419
Lesueuriidae, 419
Leucandra, 191, 192, 209, 221
Leucilla, 191, 191
Leucin, 15
Leuckart, 245
Leucocyte, 4 f.;
movements of, 7 f.
Leucophrys, 137
Leucosin, 115
Leucosolenia, 221;
collar-cell, 186;
larva, 227 f.;
spicule, 232;
L. botryoides, buds, 228, 229
Leucosoleniidae, 185 f.
Levander, on Caenomorpha, Metopus, etc., 154 n.
Leydenia, 90 f.
Liberation of sporozoites of Acystosporidian parasite in relation to fits
of fever, 103
Lice, supposed spontaneous generation of, 42
Lichen compared to Radiolarian with symbiotic holophytic
organisms, 86
Lichnophora, 138;
adoral wreath, 138 n.
Lieb, Calkins and, on rhythm in life-cycle of Ciliata, 148 n.
Lieberkühn, 167, 178, 237 n.
Lieberkühnia, 59, 61
Life-cycle, life-history, of Trichosphaerium sieboldi, 54, 56;
of Polythalamic Foraminifera, 67 f.;
of Lankesteria ascidiae, 95 f.;
of Coccidium schubergi, 99 f., 101;
of malarial parasites, 103, 104 f.;
of Flagellata, 116 f.;
of Ciliata, 147 f.
Light, stimulus of, 19, 21 f.;
function of, in carbohydrate formation, 36;
effect on Euglena, 125
Lillie, on regeneration in Protozoa, 35 n.
Limicolous Protozoa, 48
Limit of growth, Herbert Spencer's, 23 f., 31
Limnocnida, 293;
L. tanganyicae, 293
Limnocodium, 293;
L. sowerbyi, 292
Linantha, 322
Linckia, 459, 471
Linckiidae, 461, 471
Lindström, 346
Linerges (allied to Atollidae, 322), 316
Linin, 6, 24 f.;
of ovum of Sea-urchin, 7
Linuche, 322
Lionotus, 137, 152 n.
Lipochrome, 39
Liriantha appendiculata, 291, 295
Liriope, 288, 290, 295;
L. rosacea, 289
Lissodendoryx, 224
Lissomyxilla, 225
Lister, A., on Myxomycetes, 93 n.
Lister, J. J., on Foraminifera—reproduction, 67 f.;
dimorphism, 67;
 palaeontology, 70;
classification, 58 f.;
on Astrosclera, 194 n.
Lithistida, 194, 212, 215
Lithobius forficatus, host of Coccidium schubergi, 99
Lithocercus, 78;
L. annularis, 82
Lithoninae, 193 f.
Lithostrotion, 394
Littoral Protozoa, 48
Lituaria, 364
Lituola, 59
Lituolidaceae, 59
Living beings, characters of, 16 f.;
criterion of, 11
Lizzia, 265, 270
Lobata, 414, 416, 418 f.
Lobophytum, 333, 347, 349
Lobopodia, 47 n.
Locomotion, in Heliozoa, 73
Loeb, Jacques, on "chemical" fertilisation, 32 n.;
on polarity in regeneration, 229 f.
Loftusia, 59, 70, 283
Lohmann, on Silicoflagellates, 114 n.
Loisel, 237, 238
Longitudinal band of cilia of Dipleurula, 604;
of Tornaria, 616
Longitudinal fission of Eutreptia viridis, 124;
of Flagellates, 109, 115;
of Bodo saltans, 117 f.;
of Craspedomonadidae, 122
Longitudinal flagellum and groove in Dinoflagellata, 130, 131
Longitudinal section, of a young Asteroid, 445;
of a young Ophiuroid, 486;
of a Holothuroid, 563;
of Antedon, 584;
of free-swimming larva of Antedon, 618
Lophocalyx philippensis, 229
Lophoctenia, 418
Lophohelia, 399;
L. prolifera, 399
Lophomonas, 111, 123
Lophophore, 579
Lophophyllum, 406 f.
Louse, host for sexual process, etc., of Haemosporidian, 102 n.
Loxodes, 137, 144, 152 n.
Loxophyllum, 137, 152
Lucernaria, 320, 321;
L. campanulata, 321
Lucernariidae, 320
Ludwig, on the blood-system of Asteroidea, 449;
on the axial sinus of Ophiuroidea, 487;
on the classification of Holothuroidea, 570
Lühe, figures of Lankesteria, 95
Luidia, 467 f., 471, 477;
fossil, 475;
larva of L. ciliaris, 605
Luminosity or phosphorescence of sea, due to Cystoflagellata, 132,
134 f.;
to Dinoflagellata, 132
Lunule, 548
Lychnorhiza, 325
Lychnorhizidae, 325
Lytocarpus (Plumulariidae, 279), 277

Maas, 168, 189 n., 228 n., 230, 231 n., 232 n., 233, 324
MacBride, E. W., on Echinodermata, 425 f.
MacBride, Massee, on Myxomycetes, 93 n.
MacCallum, on malarial parasites, 103
M‘Dougall, on motile reaction of Protozoa, 19 n.
M‘Intosh, 370 n.
MacMunn, 169
Macrocnemic, 405
Macrocneminae, 405
Macrogonidia of Volvox, 126, 127
Macro-, prefix misused to mean "large," usually replaced here by
"mega," q.v.
Madrepora, 368, 373, 387, 389, 395;
M. forma cervicornis, 395;
M. forma palmata, 395;
M. forma prolifera, 395
Madreporaria, 369, 371, 384 f.
Madrepores, 326 = Madreporaria, q.v.
Madreporic vesicle (or right hydrocoel) of Asterias rubens, 448;
of Ophiothrix fragilis, 490;
of Echinus esculentus, 528;
development in Dipleurula, 609
Madreporidae, 395
Madreporite, 428;
of Asterias rubens, 434;
of Ophiothrix fragilis, 487;
of Ophiuroidea, 493;
of Cladophiurae, 493;
of Echinus esculentus, 512, 517;
of Echinocardium cordatum, 562;
of Holothuria tubulosa, 564;
of Elasipoda, 571;
of Pelagothuriida, 572;
in older fossil Pelmatozoa, 583;
in Thecocystis sacculus, 596
Magosphaera, 89
Maidenhair tree, spermatozoa of, 38
Malacogorgia, 334, 357
Malarial fever produced by Acystosporidae, 103 f.
Mal de Caderas (= falling sickness of cattle), 119
Male gamete, 33;
motile in Lower Plants, Higher Cryptogams, Cycads, and Ginkgo,
38;
of Pandorina, 128 f.;
of Eudorina, 129;
of Peritrichaceae, 151
—see also Sperm, Spermatozoon
Malignant tumour, associated with Leydenia, 91
Mammals, syngamy in, 34;
contain Sarcosporidiaceae in muscles, 108
Man, host of Amoeba, 57;
of Coccidiaceae, 102 f.;
of Sarcocystis tenella, 108 n.;
of Trichomonas vaginalis, 119;
of Trypanosomes, 119 f.;
of the Ciliata Nyctotherus and Balantidium, 152
Manicina (Fam. Astraeidae, 399), 373;
M. areolata, 370, 389
Mann, on function of nucleus, 24 n.
Manson, on relation of Filarial disease to gnats or mosquitos, 103
Manson, the subject of inoculation experiments with malarial
parasites, 106
Manubrium, 251
Margelis ramosa, 269
Margelopsis, 270, 274 f.
Marginal, anchors, 320;
cirrhi, 139 f.
Marginaster, 464
Marine, Foraminifera, 60 f.;
Heliozoa, 75
Marrow, red, of bones, habitat of resting states of malarial parasites,
106 n.
Marshall, on amphidiscs, 179
Marshall, on Pennatulacea, 359 n.;
on the physiology of the nervous system of Antedon rosacea, 585
Marsigli, 167
Marsupifer valdiviae, 379
Marsupites, 588
Maryna, 137;
M. socialis, tube, 152
Massee, on Myxomycetes, 93 n.
Mastigamoeba, 109, 112
Mastigophora (Bütschli's name for Flagellata), 109
Mastigophrys, 71, 75 n.
Maturation of schizont of Acystosporidian parasite in relation to
fever-fit, 103
Maupas, on Protozoa, 45;
on life-cycle of Ciliata, 147 f.
Maupasia, 111, 124
Mayer, 312
Meandrina, 388, 401;
M. labyrinthica, 370
Mechanical stimuli, 19 f.
Median dorsal process of ciliated band of Bipinnaria, 606
Medium gametes in Pandorina, 128 f.
Medusa, 250;
of Millepora, 259 f.;
of Gymnoblastea, 262 f.;
of Calyptoblastea, 277 f.;
in Trachomedusae, 288 f.;
in Narcomedusae, 295 f.;
fresh-water, 292 f.;
of Siphonophora, 302, 309;
in Scyphozoa, 310 f.
Medusome, 250, 251, 252
Megagamete, 33;
see also Female gamete, Oosphere
Megalactis griffithsi, 384
Megalosphere, megalospheric, 67 f.
Megamastictora, 183, 184 f.
Meganucleus, 136, 139 f., 144, 149 f.;
degeneration of, in conjugation, 148 f.;
new formation of, in conjugation, 148, 151;
of Stylonychia mytilus, 139 f.;
of Carchesium, 146;
of Paramecium caudatum, 148, 151;
of Trachelius ovum, 153;
of Stentor, 154;
of S. polymorphus, 156;
of Suctoria, 160 f., 162;
 conjugation of, in Dendrocometes, 161
Megazooid of Vorticella, 157
Megazoospores, 85
Meissner, on classification of Spatangoidea, 554 n.
Melanin, 103
Melitodes, 333, 351, 353;
M. chamaeleon, 338;
M. dichotoma, 338, 351
Melitodidae, 337, 351
Mellita, 548, 549
Melobesia, 422
Melonitidae, 557
Membrana reticularis, 199, 200
Membrane, undulating, of Flagellata, 110, 115, 123;
of Trypanosoma, 115;
of Trichonymphidae, 123;
of Ciliata, 137, 139 f., 145, 156 f.;
of Stylonychia mytilus, 139;
of Pleuronema, 145;
of P. chrysalis, 153, 154;
of Caenomorpha uniserialis, 155;
of Vorticella, 156, 157
Membranella, 137, 139 f., 145;
of Stylonychia mytilus, 139 f.;
of Metopus sigmoides, 154;
of Caenomorpha uniserialis, 155;
of Vorticella, 156
Meresjkowsky, 238, 274
Merozoite, 97;
of Coccidium schubergi, 99 f., 101;
of Haemosporidae, 102;
of Acystosporidae, 103, 104 f.
Mertensia, 417;
M. ovum, 417;
stage of Lobata and Cestoidea, 414
Mertensiidae, 417
Mesenchyme, 604
Mesenteric filaments, Alcyonaria, 331, 333;
Zoantharia, 369
Mesenteries, of Alcyonaria, 329, 334;
of Zoantharia, 329, 366 f., 368;
of Asterias rubens, 439;
of Holothuria nigra, 562;
of Antedon rosacea, 585, 586
Mesnil, on Sporozoa, 94;
Caullery and, on Actinomyxidiaceae, 98 n.
Mesodinium, 137, 152
Mesogloea, 246;
of Alcyonaria, 330
Metabolic, metabolism, 13
Metacnemes, 367
Metacrinus, 588, 591
Metallogorgia, 355
Metamorphosis, of Insects, 44;
of Dipleurula, 610 f.
Metamp, 168, 169
Metaphytes, 41
—see also Plants, Higher
Metazoa, 40 f.;
rheotaxy of, 21;
origin of, from Protozoa, 40 f.;
flagellate sperms of, 109;
hosts of Polymastigidae, 111
—see also Animals, Higher
Method of study of the life-cycle, of organisms of putrefaction, etc.,
44;
of Flagellata, 116;
of Ciliata, 147
Metopus, 137;
M. pyriformis, 154;
M. sigmoides, 154
Metridium, 381 (= Actinoloba, q.v.)
Metschnikoff, 167, 178, 237 n., 296
Microbes, 44
Microciona, 225
Microgamete, 33;
of certain Coccidiaceae, 101
—see also Sperm, Spermatozoon
Microgromia socialis, 59 f., 60
Microhydra, 256
Micromastictora, 183, 195 f.
Micronuclei, micronucleus, 136, 139, 144 f., 148 f., 151 f., 155, 157,
159, 160 f.;
of certain Flagellata, a blepharoplast, 109 n.;
relations of Trypanosomic blepharoplast to, 121;
of Stylonychia mytilus, 139;
of Paramecium caudatum, 148, 151;
in conjugation, 148 f.;
numerous, of Stentor, 154;
of Vorticella, 157;
of Suctoria, 160 f.;
of Podophrya, 160;
of Acineta jolyi, 160
Micropyle, 230
Microscleres, 176
Microsolena, 404
Microsphere, microspheric, 67 f.
Microzooid of Vorticella, 157
Microzoospores, 85
Miescher's tubes, 108
Migratory pairing nucleus, 149 f.;
of Peritrichaceae, 151 f.
Miliola (Quinqueloculina), 65
Miliolidaceae, 59
Miliolina, 59, 66, 70
Milleporina, 257 f., 258, 260, 282;
nematocysts of, 247, 259
Milleporina, 257
Mimicry among Gymnostomaceous Ciliata, 152 n.
Minchin, on Sporozoa, 94 f.;
on Sponges, 168, 172 n., 185, 186 n., 227 n., 232, 316 n.
Minnows prey on Anopheles, 106
Minous inermis, 268
Minyadidae, 328, 366, 377, 383
Miserly cells, 32 f.
Mithrodia, 464
Mithrodiidae, 464
Mitosis, 25, 26, 27 f.;
functions of, 28 f.;
of micronuclei in Ciliata, 144
Mitrophanow, on trichocysts, 142 n.
Mitrophyes, 306
Miyajima, 273
Mnemia, 420
Mnemiidae, 420
Mnemiopsis, 420
Mnestra (position undetermined), 269
Mohl, von, on protoplasm, 3
Mole-cricket, host of Lophomonas, 123
Molluscs, hosts of Gregarines, 98
Molluscum contagiosum, 102
Molpadiida, 568, 569, 575, 576, 577, 578
Monacanthid, 457
Monadidae, 111
Monadineae, applied to Proteomyxa by Cienkowsky and Zopf, 89
Monads, a name for the lowest, simplest Flagellata, 109, 116 n.
Monas, 111, 119;
M. dallingeri, gametes of, 116 n.
Monaxon, 183, 184, 232
Monaxonic (= symmetrical about one single axis), 76
Monaxonida, 211, 216 f.
Moniliform meganucleus of Stentor, 156
Monobrachiidae, 274
Monobrachium, 274
Monocaulus, 263, 273;
M. imperator, 269, 273
Monocyclica, 594
Monocystis, 97 f.
Monograptus, 282
Monophyes, 303, 306
Monophyidae, 306
Monoprionidae, 282
Monopylaea (= Nassellaria), 76
Monorhaphis, 197
Monosiga, 111, 122
Monosiphonic, 275
Monotrypasta (= Osculosa), 76
Monoxenia darwinii, 342
Monstrous Foraminiferal shells, possible formation of, 69
Montipora, 390, 396
Moore, 293
Mopsea, 353
Morgan, on regeneration, 35 n.
Morphological contrast of Animals and Plants, 38 f.
Mortensen, on classificatory value of pedicellariae, 532;
on classification of Cidaridae, 534;
of Echinothuriidae, 536
Moseley, 258, 333, 338, 345, 411
Mosquito (= gnat), 103 f.;
dappled-wing-, intermediate hosts of Acystosporidae, 103
Mosquito-netting, a prophylactic against malarial fever, 103
Moss-dwelling Protozoa, 48
Mosser, F., 418 n.
Motile organs, 17 f.
Motile reactions of Protozoa, 19 f.
Motility, 9
Motion, ciliary, 18;
gliding, of protoplasm, 47 n.
Moulting of cuticle or cell-wall in Dinoflagellata, 130;
of Dendrocometes, 161
Mouth, of Flagellata, 113;
absent from Opalinidae, 123;
of Maupasia, 124;
excreta expelled by, in Noctiluca, 133;
of Ciliata, 137, 145 f.;
 of Gymnostomaceae, 137, 143, 145, 152;
of Stylonychia mytilus, 139 f.;
of Dysteria, 145;
of Pleuronema, 145;
of P. chrysalis, 153;
of Paramecium caudatum, 148, 151;
trichocysts of, in Gymnostomaceae, 143;
of Trachelius ovum, 153
Mouth-angle of Ophiothrix fragilis, 482
Mouth-frame, of Asteroidea, 436, 483;
of Ophiothrix fragilis, 482;
of Ophiuroidea, 483, 492 f.;
of Ophiarachna incrassata, 484;
of Ophiacantha, 492;
of Ophioscolex, 492;
of Ophiothrix, 492
Mouth-papilla, of Ophiuroidea, 483, 492;
of Ophiocoma, 493
Movements, amoeboid, 5 f., 125 n.;
of Protista, 16 f.;
of Higher Plants, how produced, 38;
springing, of Bodo saltans, 114;
of Euglena, 124 f.;
euglenoid, 125 f.;
metabolic, 125 n.;
of Sporozoa, 125 n.;
of Stylonychia, 138;
springing, of tailed Ciliata, 141 n.;
of Halteria, 155;
of Suctorian tentacles, 159 f.
Muggiaea, 306;
M. atlantica, 304;
M. kochii, 303
Müller, J., on recognition of Echinoid larva, 518;
on the name Pluteus, 607
Müller, O. F., on Protozoa, 45
Multicilia, 109
Multinucleate Amoeba (Pelomyxa), 16;
Protozoa, regeneration of, 35
Multiple budding, in Suctoria, 160 f.
Multiple fission, 30 f.
—see also Brood-division
Murbach, 248 n., 288
Muricea, 356
Muriceidae, 330, 355
Muscle of Vorticella, 157
Muscle-cell, 19
Muscular contraction, physical explanation of, 19
—see also Myonemes
Mussa, 401
Mycetozoa (= Myxomycetes, q.v.), 50, 90 f.;
in relation to Fungi, 40;
studied by botanists, 45;
relations of, 49
Myocyte, 96, 98
Myonemes, 19, 96;
of Trypanosoma, 120 f.
of Stylonychia, 140;
of Ciliata, 142;
of Vorticella, 157
Myophrisks, 80
Myriophrys, 71
Myriothela, 269, 274
Myriothelidae, 274
Myxaster, 466
Myxasteridae, 464
Myxastrum, 70, 73
Myxidium, 98;
M. lieberkühnii, 107
Myxilla, 225
Myxobolus, 98;
spores of, 107
Myxobrachia, 83
Myxogasteres, Myxogastres, 90 f.
Myxoidea, 89
Myxomycetes, 90 f.;
rheotaxy of plasmodium in, 21
—see also Mycetozoa
Myxospongiae, 196
Myxosporidiaceae, 98, 106 f.;
spores, 107

Nagana disease of hoofed quadrupeds, 119

Naked Protozoa, 51 n.
Narcomedusae, 288, 295 f.
Nardoa, 455, 456, 471
Nassellaria, 76, 78;
skeleton of, 83;
geological occurrence of, 88
Nassoidea, 78
Nausithoe, 322;
Scyphistoma of (= Spongicola fistularis), 317;
N. punctate, 322;
N. rubra, 322
Nectocalyces, 297, 298, 305
Needham on spontaneous generation, 43
Nematocyst, 246 f., 247;
of Actinomyxidiaceae, 98;
of Myxosporidiaceae, 98, 107;
of Myxobolus mülleri, 107;
of Polykrikos, 131, 249;
of Epistylis, 249;
of Aeolis, 248;
of Hydra, 247;
of Millepora, 247, 259;
of Siphonophora, 300;
of Scyphozoa, 312;
of Alcyonium, 247;
of Sarcophytum, 248;
of Cerianthus, 247
Nematodes parasitic in blood, 103
Nematophores, 277
Nemocera (= gnats or mosquitos), 103 n.
Neohelia, 399
Neolampas, 554
Neosporidia, 97, 106
Nephthya, 341, 349
Nephthyidae, 349
Neresheimer, on neurophane fibrils in Ciliata, 143 n.
Nerve-ring, of Asterias rubens, 444, 447;
of Ophiothrix fragilis, 488;
of Echinus esculentus, 518, 521, 527;
of Antedon rosacea, 583;
outer, of A. rosacea, 585
Nervous fibrils in Ciliata, 143
Nervous system, in Animals, not in Plants, 39 f.;
of Asterias rubens, 444 f.;
of Ophiothrix fragilis, 488;
of Echinus esculentus, 518 f.;
of Holothuria nigra, 566;
of Antedon rosacea, 583 f.
Neuron, 444
Neurophane (= supposed nervous fibrils in Ciliata), 143 n.
Newts, Trichodina parasitic in, 158
Nitriles in relation to nutrition, 36
Noctiluca, 110, 132 f.;
endosarc, 144;
N. miliaris, 133
Nodosaria, 59, 63, 66 f.
Nosema, 98;
N. bombycis, 107;
organism of pébrine, 107
Nubecularia, 59
Nuclear apparatus, of Infusoria, 48, 136;
of Ciliata, 139 f., 144 f.;
of diffused granules, in marine Ciliata, 144 n.;
of Suctoria, 159
—bipartition in Trichosphaerium, 54
—reduction of Actinosphaerium, 75 n.;
of Monocystis, 96;
of Coccidiaceae, 100, 104 f.;
of Acystosporidae, 104 f.;
of Myxosporidiaceae, 107;
of Flagellates, 116 n.
—divisions, in spores of Lankesteria, 95
—see also Mitosis, Karyokinesis
Nuclearia, 70
Nuclein mass (= karyosome), 24
Nucleinic acid, 7 n.
Nucleole, nucleolus, 7, 24, 25 f., 27;
of Sea-urchin ovum, 7;
of Sphaerella, 126
Nucleolidae, 554
Nudeolites, 554
Nucleoplasm, 6
Nucleoproteids, 12
Nucleus, 6;
of cell, 6 f.;
of Amoeba, 5 f.;
of A. polypodia, 10;
resting, function of, 24 n
—in mitosis, 25 f., 27 f.;
—of Euglypha, 29;
of Paramecium caudatum, 148
—pairing state of, 34;
of Ciliata, 150 f.;
of Paramecium caudatum, 148
—of Rhizopods, 52;
of Pelomyxa, 52;
of Microgromia socialis, 60;
of Foraminifera, 62;
of mega- and microspheric forms of Foraminifera, 68 f.;
of Heliozoa, 71, 72, 74;
of Clathrulina, 74;
of Radiolaria, 76;