DOCTORA L T H E S I S
Department of Computer Science, Electrical and Space Engineering Division of Computer Science
On Mobility Solutions in Heterogeneous Networking
Environments
Daniel Granlund
ISSN 1402-1544 ISBN 978-91-7583-384-2 (print)
ISBN 978-91-7583-385-9 (pdf) Luleå University of Technology 2015
Daniel Granlund On Mobility Solutions in Heter ogeneous Netw orking En vir onments
On Mobility Solutions in Heterogeneous Networking
Environments
Daniel Granlund
Dept. of Computer Science, Electrical and Space Engineering Lule˚ a University of Technology
Skellefte˚ a, Sweden
August 2015
Supervisors:
Christer ˚ Ahlund, Robert Br¨ annstr¨ om, and Patrik Holmlund
ISSN 1402-1544
ISBN 978-91-7583-384-2 (print) ISBN 978-91-7583-385-9 (pdf) Luleå 2015
www.ltu.se
This thesis is dedicated to my beloved family, for their love, encouragement,
and endless support.
Every step of the way.
iii
Abstract
Mobility support for users and devices, such as sensors, connecting to the Internet is a continuously growing trend. Different types of wireless networking technologies like WiFi, LTE, and ZigBee are available, creating a heterogeneous wireless networking envi- ronment. In general, the technology, ranging from network equipment to communications protocols, available today, tends to be designed with limited flexibility when it comes to supporting heterogeneity. Proprietary technology and highly optimized mechanisms limit the potential use of networking infrastructure for multiple purposes.
Supporting mobility for wireless devices between different networking technologies and administrative domains requires secure and scalable mobility management solutions.
Current research in the field of mobility management and security in heterogeneous net- working environment tends to handle mobility management performance, such as com- munications characteristics, and security related aspects as separate, and more or less mutually exclusive. The work presented in this thesis serves to challenge this trade-off and proposes solutions that seeks to bridge the gap between the two areas.
This thesis proposes using an Authentication, Authorization and Accounting (AAA) system based on the RADIUS AAA protocol that enables a common AAA infrastructure to operate in a heterogeneous networking environment, and that enables a hierarchical interconnection structure between service providers. The proposed AAA architecture offers a highly scalable AAA infrastructure with access technology independent support for user and device authentication as well as mobility on a global scale. Further, to support handoffs between networks, a set of methods for facilitating improved handoff decisions for intra- and inter-operator mobility in heterogeneous networks are presented.
These methods rely on metrics that include indicators on network load conditions to improve service stability and decrease application down time during handoff.
Results show that the proposed AAA infrastructure can be built to scale in order to support a very large number of entities, more than 1000 user or device authentications per second, using industry standard hardware. Further, an AAA architecture compatible sensor authentication protocol is proposed, implemented, and validated that supports sensor mobility and reduces power consumption, on wireless sensor nodes, with 33%
compared to state-of-the-art protocols by reducing CPU and communications overhead.
Using the proposed mechanism for facilitating handoffs in wireless sensor networks, a reduction of 44% in packet loss is achieved during a mobility session compared to a traditional solution.
v
Contents
Part I 3
Chapter 1 – Thesis Introduction 5
1.1 Introduction . . . 5
1.2 Roadmap and Brief Summaries of Included Publications . . . 9
1.3 Chapter Summary . . . 14
Chapter 2 – Background 17 2.1 Heterogeneous Access Networks and Mobility Management . . . 17
2.2 Security, Privacy, and AAA . . . 28
2.3 Chapter Summary . . . 33
Chapter 3 – Related Work 35 3.1 Mobility Management Protocols and Performance Metrics . . . 35
3.2 AAA Handling in Heterogeneous Computer Networks . . . 40
3.3 Inter- and Intra-domain Mobility and AAA in Wireless Sensor Networks 42 3.4 Chapter Summary . . . 44
Chapter 4 – Conclusions and Future Work 45 4.1 Summary of the Thesis . . . 45
4.2 Comparison with Related Work . . . 47
4.3 Future Work . . . 48
Part II 51
Paper A 53 1 Introduction and Background . . . 552 Mobility Management with MultiHomed Mobile IP . . . 57
3 Network Selection Technique and Policy-Based Decision Model . . . 57
4 The M4 Software Architecture . . . 59
5 Evaluation Framework and Scenarios . . . 64
6 Results . . . 65
7 Related Work . . . 67
8 Discussion and Future Work . . . 68
Paper B 69 1 Introduction . . . 71
2 Related Work . . . 72 vii
4 Evaluation Framework . . . 73
5 Results . . . 74
6 Conclusions and Future Work . . . 76
Paper C 79 1 Introduction . . . 81
2 Proposed AAA Management Scheme for Mobility Management Scenarios 82 3 Implementation . . . 84
4 Evaluation and Results . . . 85
5 Conclusions . . . 86
6 Related and Future Work . . . 87
Paper D 89 1 Introduction . . . 91
2 Architecture Description and Background . . . 92
3 AAA Server Performance . . . 94
4 Experimental Results . . . 97
5 Conclusions and Future Work . . . 100
Paper E 103 1 Introduction . . . 105
2 Related Work . . . 108
3 EAP-Swift: A Lightweight Protocol for Sensor Authentication . . . 109
4 Results Analysis . . . 113
5 Conclusions and Future Work . . . 121
Paper F 123 1 Introduction . . . 125
2 Background and Related Work . . . 127
3 Proposed Mechanism . . . 128
4 Implementation and Experimental Setup . . . 131
5 Results Analysis . . . 134
6 Conclusions and Future Work . . . 142
Paper G 143 1 Introduction . . . 145
2 Background and Problem Definition . . . 147
3 Related Work . . . 149
4 Proposed Sensor Authentication and Mobility Protocol . . . 151
5 Results Analysis . . . 156
6 Conclusions . . . 167
List of Abbreviations 169
References 173
viii
Acknowledgments
I would like to thank my supervisors, Professor Christer ˚Ahlund, Dr. Robert Br¨annstr¨om, and Dr. Patrik Holmlund. First off, I am really grateful for this opportunity to do my doctoral studies at Lule˚a University of Technology. Secondly, I would like to thank you for inspiring, encouraging, and sharing your knowledge throughout my study period.
Also, a special thank you to: Associate Professor Karl Andersson, Dr. Karan Mitra, Dr. Muslim Elkotob, Dr. Dan Johansson, and Mr. Stefan Lundberg with whom I have worked closely on several publications and projects. Working with you has been very rewarding and interesting, not to mention has resulted in a lot of laughter. My research activities have been funded by several projects; BasicNet, I2, MOSA, and Sense Smart City. I would of course like to thank the members of those projects for good collaboration and support as well as providing me with the opportunity to conduct research within this interesting field. Most importantly, I would like to express my utmost gratitude to my beloved family for supporting me to the fullest, every step of the way; Karoline, my sons Albin and Max, my parents Tommy and Agneta, my sister Ida with family, my sister-in-law Emelie, and also my grandparents deserve a special thanks for always providing their full support and confidence in me.
Skellefte˚a, August 2015 Daniel Granlund
ix
Publications
This thesis work has resulted in the following publications:
1. D. Granlund, C. ˚Ahlund, P. Holmlund, A Handoff Mechanism for Mobile Wireless Sensor Applications, Submitted for review.
2. D. Granlund, P. Holmlund, C. ˚Ahlund, Opportunistic Mobility Support for Re- source Constrained Sensor Devices in Smart Cities, Sensors, vol. 15, nr. 3, pp.
5112-5135, 10.3390/s150305112.
3. D. Granlund, C. ˚Ahlund, P. Holmlund, EAP-Swift: An Efficient Authentication and Key Management Mechanism for Resource Constrained WSNs, International Journal of Distributed Sensor Networks, 10.1155/2015/460914.
4. K. Mitra, S. Saguna, C. ˚Ahlund,D. Granlund M2C2: A Mobility Management System for Mobile Cloud Computing, The IEEE Wireless Communications and Networking Conference, WCNC 2015, New Orleans, USA, March 2015.
5. D. Granlund, D. Johansson, K. Andersson, R. Br¨annstr¨om, A case study of appli- cation development for mobile and location-based services, The 15th International Conference on Information Integration and Web-based Applications & Services, iiWAS2013, New York, USA, December 2013.
6. D. Granlund, R. Br¨annstr¨om, Smart city: the smart sewerage, The 37th IEEE Conference on Local Computer Networks (LCN): 6th IEEE Workshop On User MObility and VEhicular Networks, ONMOVE 2012, Florida, USA, October 2012.
7. R. Br¨annstr¨om and D. Granlund Sensor monitoring of bridge movement: a sys- tem architecture, The 36th IEEE Conference on Local Computer Networks (LCN):
5th IEEE Workshop On User MObility and VEhicular Networks, ONMOVE 2011, Bonn, Germany, October 2011.
8. D. Granlund and C. ˚Ahlund, A Scalability Study of AAA support in heteroge- neous networking environments with global roaming capabilities, The 10th IEEE Conference on Trust, Security and Privacy in Computing and Communications, TrustCom-11, Changsha, China, November 2011.
1
bility management for heterogeneous wireless networks, The 7th Annual IEEE Con- sumer Communications and Networking Conference, CCNC 2010, Las Vegas, USA, January 2010.
10. M. Elkotob, D. Granlund, K. Andersson, and C. ˚Ahlund, Multimedia QoE Op- timized Management Using Prediction and Statistical Learning, The 35th IEEE Conference on Local Computer Networks, LCN 2010, Denver, USA, October 2010.
11. D. Granlund, K. Andersson, M. Elkotob, and C. ˚Ahlund, A uniform AAA handling scheme for heterogeneous networking environments, The 36th IEEE Conference on Local Computer Networks (LCN): 3rd IEEE Workshop On User MObility and VEhicular Networks, ONMOVE 2009, Z¨urich, Switzerland, October 2009.
12. D. Granlund, K. Andersson, and R. Br¨annstr¨om, Estimating network performance using low impact probing, The 1st Workshop on Wireless Broadband Access for Communities and Rural Developing Regions, WIRELESS4D ’08, Karlstad, Sweden, December 2008.
13. K. Andersson, D. Granlund, and C. ˚Ahlund, M4: MultiMedia Mobility Manager : a seamless mobility management architecture supporting multimedia applications, The 6th International Conference on Mobile and Ubiquitous Multimedia, MUM 2007, Oulu, Finland, December 2007.
14. R. Br¨annstr¨om, C. ˚Ahlund, K. Andersson, and D. Granlund, Multimedia flow mobility in heterogeneous networks using multihomed mobile IP, Journal of Mobile Multimedia, vol. 3, no. 3, pp. 218-234, September 2007.
Papers 2, 3, and 14 are peer-reviewed journal publications. Papers 4-13 are peer- reviewed and published at international conferences and paper 1 is submitted for review to an international journal. Paper 1, 2, 3, 8, 9, 11, 12, and 13 are included in this thesis to form chapters A through G. Formatting is slightly modified in order to improve the presentation. A brief summary of each paper is provided in section 1.2.2.
2
Part I
3
Chapter 1 Thesis Introduction
“There are three types of people: The ones who make things happen. The ones who wait for things to happen, and the ones who wonder what just happened”
Tony Beets - Legendary gold miner This chapter provides an introduction to the thesis, states the research question, and presents the thesis contributions. The order of the included papers is presented in the form of a graphical roadmap and research methodology is covered as well as a summary of the included papers.
1.1 Introduction
The use of smart-phones and other highly portable devices has increased significantly over the last couple of years. In the last five years (2010-2015), the number of active mo- bile broadband subscriptions increased from 807 million to 3 459 million [1]. Applications requiring Internet access have evolved from email and simple web browsing to real-time streaming applications and games, placing high demands on the network connection and its availability. Many mobile devices are equipped with more than one network interface and using these in an optimal way, depending on the context, is a challenge [2]. A major challenge includes determining which network interface that is the most beneficial to use at a time, considering the traffic pattern, taking into account parameters such as network characteristics, cost, and power consumption [3]. Security is also an important aspect that needs consideration when interconnecting networks and service providers across the Internet. Managing security, including authentication and encryption is essential, espe- cially for wireless access networks that are prone to security threats due to the unguided media [4]. Furthermore, switching between networks, or performing a handover [5]1typ-
1Throughout this thesis, the terms handover and handoff are used interchangeably to describe the task of changing network connection.
5
ically involves a number of difficulties such as redirecting traffic to reduce disturbances like packet loss and application down-time [6].
Extending the scope towards sensors and other resource constrained devices places fur- ther challenges on protocols and technologies [7]. Parameters such as power consumption, delay tolerance, and time synchronization might further complicate the overall picture.
Generally, research carried out within the area of mobility in heterogeneous networks tends to focus on optimizing only a subset of the aforementioned parameters. For exam- ple, a large number of papers exist that seek to develop mobility management solutions with focus on minimizing the packet loss while mentioning the other parameters such as security as beyond the scope of the paper. From the other viewpoint, network security protocols, such as authentication protocols, designed to maximize the robustness and se- curity, will typically involve CPU intensive operations and a high number of handshakes between communicating peers, making it unsuitable for performing the rapid connec- tion establishment required to minimize packet loss during handover [8]. The trade-off between security and power consumption in a Wireless Sensor Network (WSN) is an ex- ample where CPU intensive cryptographic operations and excessive communication will cause rapid battery depletion [9].
This thesis emphasizes the need for a more holistic view when designing protocols and systems to support mobility and aims to bridge the gap between performance ori- ented and security oriented protocols in mobility management applications rather than treating them as separate topics. Further, the approach taken in this thesis is to use a common infrastructure to connect sensors and other equipment, as well as users to the network. This offers several advantages such as reduced cost, increased interoperability, and compatibility [2].
1.1.1 Research Question
This thesis aims at identifying performance issues, related to authentication and mobility management in heterogeneous networking environments. The main research question identified and addressed in this thesis is:
• How to enable a uniform, secure, and scalable roaming support in heterogeneous access networks, considering devices ranging from resource constrained sensors to highly capable personal computers
The term heterogeneous access network in this context refers to a networking en- vironment comprising different access network technologies. Issues related to mobility management in heterogeneous access networks typically include loss of service during handover. Loss of service can be caused by high latencies in the authentication process or network signaling during handover, suboptimal handover decision strategies, and scal- ability problems due to increasing number of users. Further, addressing these issues on a battery powered device must be done in a way that utilizes the available battery capacity in an efficient way. Reducing delays and overhead is an essential issue when looking at a perspective where e.g. tens of thousands of sensor devices in a city uses the public
1.1. Introduction 7 cellular wireless network for sensor data transmission. For a service provider, handling these issues is essential to provide a scalable and reliable service across technologies and in a multi-operator environment.
The aim is to enable an AAA model for mobility between a multitude of access network technologies, wireless as well as wired, without the need for pre-established roaming agreements between included parties. The model should be able to handle authenticating users and devices, ranging from smart phones and laptops to small sensor devices in large numbers using the same basic mechanism and infrastructure. In order to achieve seamless inter-operator mobility, high demands are put on the AAA system to provide means for timely user authentication and admission. Another important parameter is scalability on a global level. Scalability in this context refers to the system or protocols ability to handle a growing number of users, devices and/or usage.
The proposed architecture is studied in terms of handover latencies, battery consump- tion, scalability, security, and implementability which can be used as design guidelines and considerations for building a scalable and mobility enabled network. Further, meth- ods are developed for estimating available network resources with without saturating the network link in order to make handover decisions.
1.1.2 Thesis Contribution
The research carried out to find answers to the stated research question has led to the following key contributions:
• A scalable AAA architecture comprising an authentication protocol and extensions for improved and uniform mobility management support for a wide range of devices
• Methods for evaluating available network resources in order to make timely han- dover decisions based on low-impact measurements
The contributions presented in this thesis are focused on the mobility supporting part of the network. Measurements and other evaluations carried out in this thesis, as well as scalability and implementability discussions are centered around this area. Evalua- tions are carried out in real-world testbeds using hardware and radio communications technology and settings that comply to European radio communication standards and restrictions only. Other limitations that might exist with any access technology such as service outage or monetary cost as well as business and payment models are beyond the scope of this thesis.
A scalable AAA architecture comprising an authentication protocol and extensions for improved and uniform mobility management support for a wide range of devices
Most common AAA handling protocols are typically designed and configured for use with a specific access technology at a time. Since, in a heterogeneous networking environment, a wide range of technologies may be available, there is a need for a uniform AAA protocol that can handle wide range of technologies simultaneously. Also, when interconnecting network service providers and different access technologies the system must be designed
AAA-H
AAA AAA
AAA-L
AAA AAA
AAA
IP tunnel ISP A
ISP C ISP B
ISP D (home ISP)
ISP E ISP F
ISP G (visited ISP)
Figure 1.1: Hierarchical interconnection of service providers.
for handling not only the AAA part but also network related tasks like IP mobility between subnets. When implementing and testing a system it might operate well for a few or a few dozens of users but when deploying a system in a larger scale with tens of thousands of users it might behave completely different. Identifying bottlenecks and weak points is important to, if possible, avoid or dampen the impact of such factors. By applying a holistic systems approach, and gathering information on where the weakest links are, design guidelines may be formed to help build more efficient and scalable systems.
Figure 1.1 shows a conceptual overview of the proposed AAA architecture where a hierarchical model is used. A service provider needs only a pre-established trust rela- tionship with directly connected entities and messages are routed throughout the tree.
This thesis contributes with an authentication protocol extension that provides a mecha- nism for supporting different networking technologies using the same AAA infrastructure.
Using the proposed authentication protocol extension, mobility management, signaling, and configuration parameters are carried within AAA messages. Tunnels are established dynamically between service providers in order to maintain seamless mobility on the net- work layer, in a uniform way, both for Ethernet based (e.g WiFi) and PPP based (e.g.
3G) connections. Also, a scalability study is carried out to ensure that the proposed architecture will handle a growing network.
1.2. Roadmap and Brief Summaries of Included Publications 9 A sensor authentication protocol, EAP-Swift, is designed, implemented, and eval- uated that supports the aforementioned AAA technology proposed to support global roaming of smart phones and other Wi-Fi and 3G/LTE enabled devices. The protocol is reduced in complexity and tailored to run on sensor devices with limited capacity without losing its capability to be compatible with infrastructure that supports global mobility.
Besides authenticating the sensor device, the protocol supports exchanging information needed to generate encryption keys for end-to-end encryption of payload data over unpro- tected communication channels. The protocol is validated and evaluated in a variety of scenarios with both mobile and stationary nodes and parameters such as mobility speed and power consumption are measured.
Methods for evaluating available network resources in order to make timely handover decisions based on low-impact measurements
Regardless of whether the handover decision is carried out manually by the user or by an automated software component, it is useful with an accurate prediction of the expected performance of the intended target network. Measurement methods that have low im- pact the network as well as accurate calculation models are needed in order to estimate available resources prior to taking a handover decision. This thesis contributes with a set of methods and models for estimating available bandwidth for improved handover decisions in different scenarios using only low impact probing packets.
1.1.3 Thesis Organization
The remainder of this chapter provides and an overview of published papers as well as a roadmap tying them together. Chapter 2 gives a more in-depth background to the work presented in this thesis. Chapter 3 discusses related work in the area. Chapter 4 concludes the thesis, discusses how the related work presented in chapter 3 relates to the work presented in this thesis, and discusses openings for future work. Chapters A through G present the included papers.
1.2 Roadmap and Brief Summaries of Included Pub- lications
The work presented in this thesis consists of 14 peer-reviewed publications of which seven are included as chapters in this thesis.
1.2.1 Roadmap
Publications included in this thesis are depicted in figure 1.2. The arrows indicate the logical workflow that led to each publication. The publications within green boxes are the ones included in this thesis, solid line boxes indicate publications with the thesis author as first author.
Multimedia flow mobility in heterogeneous networks using
multihomed mobile IP
M4: MultiMedia Mobility Manager : a seamless mobility management architecture supporting multimedia
applications
Estimating network performance using low impact
probing
A uniform AAA handling scheme for heterogeneous networking
environments
A Scalability Study of AAA support in heterogeneous networking environments with global roaming
support Bandwidth efficient mobility
management for heterogeneous wireless networks Multimedia QoE
Optimized Management Using
Prediction and Statistical Learning
EAP-Swift: An Efficient Authentication and Key Generation
Mechanism for Resource Constrained WSNs
A Handoff Mechanism for Mobile Wireless Sensor
Applications
Opportunistic Mobility Support for Resource Constrained Sensor
Devices in Smart Cities M2C2: A Mobility
Management System for Mobile Cloud
Computing
Sensor monitoring of bridge movement: a system architecture
Figure 1.2: Publication Roadmap. The publications within green boxes are the ones included in this thesis, solid line boxes indicate publications with the thesis author as first author.
1.2. Roadmap and Brief Summaries of Included Publications 11
1.2.2 Summaries of Included Publications
Paper A: M4: MultiMedia Mobility Manager: a seamless mobility manage- ment architecture supporting multimedia applications This paper presents an implementation and proof-of-concept evaluation of a versatile mobility management pro- totype based on multi-homed mobile IP with soft handovers. A network layer metric for access network selection is presented, based on round-trip time and jitter. In a graphical user interface, the user may enter preferences regarding network performance, battery consumption and monetary cost which will be the base for the access network selection policy. Also, an asymmetric decision model for vertical handover is implemented when switching between access networks with significantly different performance in order to reduce packet loss and erratic behavior. An experimental evaluation using CDMA2000 and IEEE 802.11 networks is presented running a Voice over IP (VoIP) application on top. Results are really convincing and in line with previously simulated results. Dur- ing VoIP sessions, 0% packet loss was achieved when moving between WiFi and UMTS networks. My contribution was prototype implementation, carrying out evaluations and writing.
Paper B: Estimating network performance using low impact probing This paper presents a model for estimating available bandwidth on a network link using very low impact probing packets. The purpose of the proposed bandwidth estimation tech- nique is to improve access network selection when using bandwidth demanding applica- tions. The model uses statistical information about the continuously measured network delay and jitter to calculate an estimated available bandwidth. During evaluation, in WLAN and CDMA networks, the available bandwidth is estimated with a 92% signifi- cance. Furthermore, since some network links may vary significantly in up- and down-link delay causing measurement inaccuracies, a method is proposed for determining the differ- ence in these delays separately by sending return-traffic alternate ways. My contribution was developing the idea behind the paper, carrying out the evaluation and writing most of the paper.
Paper C: A uniform AAA handling scheme for heterogeneous networking environments This paper discusses problems related to interconnecting multiple differ- ent access technologies using one common AAA system. Using a RADIUS-based AAA architecture, IEEE 802.1x is used in combination with the DHCP protocol to provide a AAA server originated configuration for Ethernet-based connections in the same way as for PPP-based connections. The solution is implemented as a plug-in which is in- stalled in local AAA servers that communicate with a DHCP server on the same subnet.
Evaluation results of a real-world implementation show that the authentication and con- figuration is carried out efficiently, both for PPP and Ethernet based connections, using the same AAA architecture. A connection setup time for WiFi was measured to be 0,47s including IP address assignment. My contribution was developing the main concept for the paper, implementing the software prototype, and writing most of the content.
Paper D: A Scalability Study of AAA support in heterogeneous network- ing environments with global roaming support This paper presents an insight into key performance issues in a large AAA architecture. The AAA server performance
and network traffic during AAA handling are studied in an experimental setup. Results show that AAA server performance suffers from different parameters depending on the context. Performance degradations may caused by cryptographic calculations and user database lookups if the authentication rate is high and the server is located close to the supplicant. However, if the server is located farther away, the network performance and server RAM memory comes to play a larger role. Experimental results combined with analytical calculations provides a model and method for determining AAA system scal- ability and design guidelines to handle different scenarios. My contribution was carrying out experimental studies, results analysis, and writing the paper.
Paper E: EAP-Swift: An Efficient Authentication and Key Management Mechanism for Resource Constrained WSNs This paper describes the design, implementation, and evaluation of EAP-Swift: a resource efficient authentication proto- col for authenticating sensors and other devices with limited capabilities. The protocol uses one-way hashes to perform mutual authentication of both the sensor node and AAA server. During the authentication process, information is also exchanged that enables the parties to, upon successful authentication, generate a secret key that can be used to protect the sensor data transmission. Since the protocol is designed to operate in an infrastructure of interconnected AAA servers, authentication and key management is only carried out between the sensor and the AAA server to which the sensor belongs.
No intermediate entity in the AAA infrastructure needs to be pre-configured, or allowed to take part in the key generation, enabling encryption of data to be carried out in an end-to-end fashion. In the studied scenario, network delay, rather than CPU utilization, is identified as the primary reason for excessive authentication delays. While many tradi- tional protocols require at least three round-trips to the AAA server in order to perform a full authentication, the presented protocol uses only two round-trips to complete both authentication and key generation. An example is presented showing a 29% improvement in terms of the maximum number of nodes that can be installed within a single wireless sensor network using EAP-Swift, compared to the baseline protocols. My contribution was designing, implementing and evaluating the protocol as well as writing the paper.
Paper F: A Handoff Mechanism for Mobile Wireless Sensor Applications This paper presents the implementation and evaluation of a mobility management scheme for sensor devices with limited resources. The mobility manager is implemented as an addition to the sensor authentication protocol discussed in the previous paragraph. The mobility manager mechanism running in the sensor records received signal strength along with measured round-trip time to available sensor gateways. Measurements are combined to form a composite metric called a policy value similar to the one described in previous papers on mobility management. The policy value is then used to compare gateways and choose which gateway to connect to. Evaluation is carried out in a simulator as well as in a real-world experiment where mobile sensor devices are configured to transmit data at a fixed rate while moving between two candidate gateway access points. Application down-time and packet loss is recorded during handover which is done both as normal transitions as well as forced. Furthermore, using extensive simulations, a set of optimal parameters are determined to calibrate the proposed handoff mechanism to operate under
1.2. Roadmap and Brief Summaries of Included Publications 13 different WSN load conditions. Using the proposed mechanism, a 44% decrease in packet loss during a mobility scenario was observed. Typical handover latencies are measured to be approx. 600 ms. My contribution was designing the mobility management component, implementing and carrying out experiments as well as writing the paper.
Paper G: Opportunistic Mobility Support for Resource Constrained Sen- sor Devices in Smart Cities This paper covers the implementation of the protocols described in the two previous papers applied in a real-world scenario, namely a Smart City environment. Ten battery powered sensors that measure ambient temperature and uses a passive IR detector to count pedestrians are placed along roads in a city environ- ment. The concept is to have fixed sensors and mobile gateways rather than the opposite as presented in the previous paper. In a smart city environment, this is an effective solu- tion since it might eliminate the need for cellular network communication capabilities for small battery powered sensors. Five mobile gateways are placed in cars that randomly pass sensors and gathers data. Sensors and mobile sensor gateways are divided into two administrative domains with five sensors in each domain.
Each administrative domain has its own AAA server and database which, in the backend, are interconnected in a tree like AAA infrastructure. Authentication delay and packet loss during handover is recorded and presented. Sensor gateways use the cellular LTE/3G network available in the city as a data uplink which accounts for a large portion of the overall authentication/handover delay. The results show that in a real- world scenario with more uncontrolled variables, such as competing traffic, and therefore, higher delay and jitter in the cellular network, authentication delays are higher, but still within acceptable limits. The performance is compared to the same experiment carried out in a controlled radio lab with an isolated LTE cell and no competing traffic. This part of the experiment is carried out in order to provide ideal circumstances and represents a scenario where low volume sensor traffic might be prioritized in the cellular network.
My contribution was developing the concept, designing the experiments and writing the paper.
1.2.3 Research Methodology
The research methodology used throughout this thesis is based on a iterative workflow where previous work has played a large role in the development of new ideas. Comprehen- sive literature studies were carried out to position the work with regard to the state-of-the art within the area. Close collaboration with colleagues and fellow researchers as well as both local and global companies has also been very important to capture real world problem statements, positioning the research, and obtaining feedback on proposed ideas.
Inter-departmental activities such as recurring PhD student seminars that discussed the latest research as well as joint courses helped keeping up with the research within the field and inspired collaboration among colleagues.
The work presented in this thesis was carried out mainly in two phases where the first phase, which led up to the first two contributions listed in section 1.1.2, was carried out in the area of heterogeneous access networks mainly intended for personal computer
communication. In the second phase, the results achieved, mechanisms developed, and experiences gained in the first phase, were applied to the area of WSNs, and in partic- ular WSNs in smart city environments which proved to share many properties with the environments studied in the first phase.
The workflow and research methodology for each publication was the same for all publications included in this thesis. In the first step, a problem definition was determined along with a hypothetical reasoning on how e.g. a system would benefit from a solution to the defined problem. In the second step, a solution to the previously defined problem was proposed and validated analytically as well as compared to the state-of-the art within the area. In the third step, the proposed solution was implemented and evaluated. In most cases (papers A, C, D, E, and G) a real-world implementation was carried out on physical hardware. In papers B and F, the real-world, experimental, prototype evaluation was complemented with a simulator evaluation and in papers D and G with analytical evaluation. In the fifth and final step, results were collected, analyzed, and evaluated.
Based on the results analysis and, in some cases, feedback from reviewers, the process was re-iterated from step two (or sometimes the first step).
The approach taken in this thesis has been to perform real-world experiments and, when deemed necessary, in combination with simulations and analytical work. In a het- erogeneous wireless networking environment where a magnitude of devices are interacting, there are many variables that are hard to model in a simulator which might lead to an over-simplified simulator model and inaccurate results. Using a real-world experimental validation, however, introduces variables that are hard to control and that might have a significant impact on the result. For example, if an experiment is carried out in a public commercial cellular network, the network load might vary significantly over short time periods and cause unpredictable results.
A common argument for preferring simulations over real-world experiments is that it might be costly, more time consuming, and impractical to carry out large-scale studies in a real-world environment, since it will typically require large amounts of hardware.
However, since real-world experimentation provides a proof of concept in a real-world setting, it provides a strong validation of tested ideas and mechanisms. Further, in a real-world experiment, it might be hard to isolate a certain set of parameters to study due to the large number of variables and difficulty to exactly re-produce experiments (especially if a certain mobility pattern is followed). In the experiments presented in paper D and G, large scale system scalability was evaluated by breaking down integral components into smaller pieces that could be stress-tested as single units rather than as parts of a larger system. This way, a lot of the added noise from external factors could be eliminated.
1.3 Chapter Summary
This chapter introduced the thesis, stated the research question, and listed the contri- butions and outcomes. Further, it provided an insight to, and a brief summary of, all the publications included in the thesis. A roadmap was provided that depicted how
1.3. Chapter Summary 15 the included publications relate to each other. The research methodology was also pre- sented. The next chapter will provide background and motivation to the thesis work including discussions around performance in mobility management and AAA handling in heterogeneous networking environments.
Chapter 2 Background
This chapter presents the background to the thesis work, with focus on mobility management, performance issues in heterogeneous networks, AAA solutions, and ar- chitectures. The chapter covers a broad range of devices ranging from smartphones, laptops, tablets, and other mobile devices typically designed for providing high speed Internet access for personal use to low-power sensor devices with limited communication requirements. This range is throughout this chapter divided into two basic categories called mobile computing devices and sensor devices respectively.
2.1 Heterogeneous Access Networks and Mobility Man- agement
When talking about mobility in the context of computer networking and Internet access, we often refer to the ability to move around freely and maintain connectivity. Following the development and increasing popularity of mobile devices such as smartphones and laptops, it is easy to see and realize a growing request for access networks with good coverage and support for mobility. Today, we can see a number of wireless technologies developed to support this need. Technologies exist with widely varying characteristics depending on their intended use. The most noticeable differences to the end user are network performance, cost, and coverage. Satellite based communication, for example, offers world-wide coverage but limited performance and at a very high cost. On the other hand there are solutions like IEEE 802.11 or WiFi that offers high capacity at low cost, but with limited communication range (<200 m). In between, there are technologies like WiMAX and cellular network technologies like LTE.
The term heterogeneous networks, which is used frequently throughout this thesis refers to networks where more than one access networking technology is available at the same time and location. Because of the aforementioned variations in access network characteristics, it is likely that combining a set of technologies and using the one that is most suitable at the time would be the most beneficial method [2]. In fact, many mobile
17
devices manufactured today are equipped with multiple wireless interfaces e.g. LTE and WiFi. However, there are a number of special issues related to combining communication technologies, IP subnets, and service providers in order to provide seamless mobility, ranging from low-level technological challenges to high-level economical models for billing.
If we shift the context away from personal devices with high requirements on bandwidth and connectivity towards small sensor devices with limited capacity, the same basic issues arise but within a different scope. Wireless communications technologies specifically designed for sensor devices, such as ZigBee [10], Bluetooth [11], Z-Wave [12]. provide low-power, limited range communication among sensors or between sensor and gateway.
Technologies like these typically offer wireless communication in the range of hundreds of meters.
In most cases, it is desired to communicate sensor data to a backend system, for ex- ample requiring connectivity via a wireless gateway connected to the Internet [13]. While sensor devices typically have only one radio communications interface, the capability to move between and establishing a connection to a new gateway is crucial when envision- ing scenarios with either mobile sensors and/or gateways. In order to support this, the sensor needs to be aware of the situation in terms of available connection points and to be able to detect when it is suitable to perform a handover to a different access point.
Considering these arguments, it is evident that many parallels can be seen between the basic requirements on a mobility management scheme regardless of whether it deals with a personal computer connected to the Internet or a coin-size, battery powered sensor device that communicates sparsely.
Proprietary protocols for sensor communication exist that offers solutions to some mobility related issues, using different strategies, depending on the actual use of the sensor. Such protocols are typically designed for only one radio technology and hardware, and thereby constraining the network design to a single vendor solution. If the sensor application is limited to what can be provided by the single vendor solution, it can be a feasible, albeit not a very scalable design. A more general mobility management and connectivity approach that can be applied on a wide range of hardware from different vendors allowing them to inter-operate is likely a more attractive approach [14]. To further complicate the mobility management need, there might be cases where a sensor device or device owner would like its device to be capable of roaming, not only within its own network or administrative domain but between administrative domains enabling the device to be mobile on a more global scale [15]. This scenario places even higher requirements on interoperability and compatibility between networking technologies.
Authentication, security, and trust for mobile devices also become a larger issue, when expanding the scope from mobility within a single administrative domain, towards a more global, inter-domain mobility scenario. When a device visits a foreign network or administrative domain, it needs to verify that the network can be trusted and that communication can be carried out without compromising security. Equally important, the visited network needs to authenticate the device and ensure that it belongs to a known and trustworthy entity. Furthermore, there might be cases where there are economic aspects involved, e.g. that the visited network owner bills the device owner for providing
2.1. Heterogeneous Access Networks and Mobility Management 19 network access to the visiting device. This introduces the need for detailed tracking of type and duration of services provided. A subset of these problems from a more technological perspective will be discussed throughout the remainder of this chapter.
2.1.1 Mobility Management for Mobile Computing Devices
In this thesis, the term mobility is an overarching concept, describing users and devices ability to move between different network access points and access technologies. This sec- tion will begin with providing an overview of current solutions to mobility related issues, for mobile computing devices, on different layers of the OSI model. In the next section, a similar overview will be provided, covering both proprietary and non-proprietary mo- bility management solutions, designed for use with wireless sensor devices. First off, a set of terms commonly used in the context of mobility management will be described.
The most important entity is the Mobile Node (MN) which typically is a computer or handheld device that is subject to the mobility. A Correspondent Node (CN) is a node or a peer that the MN communicates with during a mobility session. The term multi-homing describes the case where a MN has the ability to have multiple network connections active simultaneously. Mobility management can be split into two distinctly different categories:
terminal based and network based. In terminal based mobility management, the critical parts of the mobility management, such as handover decision making, is performed in the MN. For network based mobility management, it is performed in the network and thereby is transparent to the mobile device. The main advantages of placing the management functionality in the network include: the network can optimize performance by taking a set of nodes rather than a single node into account when making decisions and that the MN does not necessarily need to be aware of the mobility scenario, since it can be abstracted by the network [16]. The drawbacks of network based mobility management include major changes in the network infrastructure, to provide support for mobility at all locations and prevents the user/MN from affecting mobility management related tasks.
The task of changing the network attachment point for a mobile device is generally referred to as performing a handover (or a handoff) [5]. There are two basic types of handovers, horizontal and vertical, each of which can be further sub-categorized into hard and soft handover. Horizontal handover refers to changing the point of attachment within the same access technology. An example of this would be a WiFi network where access points are interconnected to form an Extended Service Set (ESS). The network interface will re-associate with a new access point while maintaining connectivity on the same interface. Vertical handover, on the other hand, means switching between different technologies. This type of handover is performed when switching between e.g. WiFi and UMTS networks and often requires changing network interface and IP configuration.
The difference between hard and soft handover is that during a hard handover, the network connectivity is dropped while establishing a new connection. While during a soft handover, the old connection is maintained while the new session is established.
Horizontal handovers are typically handled on the data-link layer where the network interface changes its point of attachment and frames are sent via different paths in the
Application Presentation
Session Transport
Network Data-link Physical
Application Presentation
Session Transport
Network Data-link Physical
MN CN
Figure 2.1: Mobility management in the OSI model.
subnet. This type of handover may still involve procedures causing handover delay like radio scanning and DHCP re-configuration, however, the same IP address is still valid which, in most cases, will keep on-going connections alive.
During vertical handovers this is typically not the case. IP addresses are often said to have a dual nature in that they can be seen as both a location identifier as well as an endpoint identifier. The location identifier (network portion) of the IP address is used during transit throughout the network in order to find the endpoint location. When a mobile node changes subnet, the IP address can therefore not be migrated without changing the location identifier and thereby breaking the ongoing connections.
Many solutions exist to address this issue in a variety of ways, ranging from application layer to network layer solutions. Figure 2.1 depicts the OSI model and the layers of the model typically responsible for mobility management, in existing solutions. At the network layer, the most widespread approach is Mobile IP (MIP) [17]. In Mobile IP (see figure 2.2), the mobile node, MN is assumed to belong to a Home Network where it is configured with a fixed IP address that is always valid. This address is referred to as a Home Address (HoA). When the mobile node leaves the home network, a server entity called a Home Agent (HA), is responsible for intercepting incoming traffic (at the home network) destined for the MN and tunnel it to the MNs current location. Returning traffic is then tunneled back to the HA and sent from the home network to the destination (CN).
Tunneling is done, either directly to the MN, or to a server entity called a Foreign Agent (FA) in the visited network. If a FA is available, it will then act as a tunnel endpoint, and de/encapsulate packets to/from the MN. Otherwise, the FA functionality may be incorporated in the MN (co-located FA). Connections are established and maintained
2.1. Heterogeneous Access Networks and Mobility Management 21
(FA2)
HA (FA1)
Home network MN
CN
Figure 2.2: Mobile IP scenario.
by sending Binding Update (BU) and Binding Acknowledgment (BA) messages between the included entities to signal changes in the topology. Using this mechanism, the MN can keep the same IP address and maintain IP connectivity while roaming. The Mobile IP protocol is designed to be transparent to overlying layers abstracting the mobility handling from the application perspective.
When running on an IPv6 platform, new possibilities are introduced to MIP. As seen in [18] route optimization can be used to direct traffic originating from the MN directly to the CN (without going through the HA). This mechanism increases the efficiency by reducing overhead and suboptimal routing, especially if the MN and CN are topologically close. At the transport layer, the Stream Control Transmission Protocol (SCTP) [19], is an example of a mobility management solution that supports multi-homing. The SCTP protocol replaces the otherwise commonly used UDP and TCP protocols. In the multi- homed case, SCTP informs the CN about all of its valid IP addresses and establishes a transmission path through each network connection. These paths are then monitored and maintained by HEARTBEAT chunks which are acknowledged by the peer with a HEARTBEAT-ACK chunk. This monitoring traffic can also be used to measure network performance parameters such as delay and jitter to use as a decision criteria for path selection.
The SCTP protocol also has some significant advantages over the ordinary TCP and UDP protocols when transferring data, including higher security level and better performance due to the multi-stream capability [20]. Multi-streaming means that several streams can be transferred simultaneously in a single SCTP association. Using this mechanism, for example in a HTTP web request, the entire web page is transferred in parallel, rather than making new TCP connections for each component on the page, which reduces signaling overhead and server load. A special version of SCTP that has additional
support for mobility is ”mobile SCTP” or mSCTP [21]. mSCTP is an extension of SCTP in that has an option called ADDIP. This option enables each endpoint to add and delete IP addresses to or from an established association, and thus enabling a new network to be added and chosen as primary path, during on-going data transmission.
In [22] a handover latency performance study is carried out between the mSCTP and MIPv6 protocols. mSCTP outperforms MIPv6 with 67 milliseconds as compared to 1841 milliseconds in average vertical handover latency. This is mainly explained by the large binding update delays for MIPv6. SIGMA [23] is yet an example of an implementation for supporting SCTP based mobility. Like in mSCTP, SIGMA outperforms MIPv6 when it comes to handover latency, packet loss rate and throughput. It is also proven to be more network friendly when it comes to handling TCP slow start than MIP, which can be an important factor in modern networking. Common for all transport layer mobility supporting protocols is that they have to be supported at both end-points. Therefore, the aforementioned transport layer mobility solutions are rarely used in applications available today and there is generally a limited support in modern operating systems (TCP and UDP are the prevalent transport layer protocols).
The Session Initiation Protocol (SIP) [24] is a text-based application layer protocol that is designed for initiation, modification, and termination of interactive multimedia sessions such as video, games, and conferencing. A user is identified by a SIP identity in the form of a Uniform Resource Identifier (URI) which looks similar to an email address.
A central SIP server or registrar keeps track of users and their last known IP address.
Sessions are initiated using an INVITE message. SIP also implements a specialized re-INVITE message that can be used to inform a CN that the MN has changed some characteristics, e.g. the IP address, mid-session. When a re-INVITE message is received, a new connection is immediately established with the new address and the session can continue. This protocol works only for SIP enabled applications and running multiple applications would imply running multiple SIP sessions in parallel.
SIP is also used by the IP Multimedia Subsystem (IMS) [25] standardized in 2002 by 3GPP. The IMS architecture creates two distinctly separated planes; service control plane and transport plane. This separation provides the ability to introduce new services into the service control plane regardless of the underlying transport plane. The purpose of IMS is to provide an overarching architecture common for all operators and technologies that will enable a user to roam freely between access technologies and different operators.
Since the service layer is separated from the transport layer, all services are available at all locations. IMS uses only standardized IETF protocols that run over IP, like SIP.
Charging a consumer for a used service is much more complex when the service has been carried over a set of different operators and access technologies with different pricing etc. rather than by a single operator. Because IMS decouples the access network from the service being used, IMS is sometimes a less appealing architecture from a business perspective. While IMS was originally designed for 3G networks it now supports other access technologies like WiMax, WLAN and also fixed networks.
The Host Identity Protocol (HIP) [26] is a protocol that, via an extension, supports multihoming and mobility by decoupling the host identity from the hosts location in-
2.1. Heterogeneous Access Networks and Mobility Management 23 formation (IP address). By replacing the IP addresses in a HIP-enabled network with cryptographic end-point identifiers, and letting the network (securely) handle mapping between the end-point identifiers and the current location, HIP can support mobility at the transport layer that is secure and transparent to overlying applications. HIP uses a Rendezvous server that acts similar to a home agent in MIP, and is responsible for keeping track of at which IP address a certain host is reachable at, and forwards traffic to it. Unlike a MIP home agent, the HIP Rendezvous server only routes the HIP sig- naling traffic, while the regular (payload) traffic is sent directly between communicating peers, and thereby reducing the risk for suboptimal routing. The major drawbacks of HIP include the introduction of a whole new layer in the TCP/IP model and the need for extensive support in the operating system of both communicating peers [27].
2.1.2 Mobility Management for Wireless Sensor Devices
As mentioned earlier, a large number of wireless communications technologies and pro- tocols exist for low power sensor devices. Examples of the most commonly used once include ZigBee [10], Z-Wave [12], Bluetooth Low Energy [28], and IEEE 802.15.4. When discussing mobility in the context of wireless sensors, mobility usually refers to a nodes capability do be mobile within a single wireless sensor network using multi-hop routing protocols to find new routes. Since this thesis mainly focuses on inter-domain mobility, no in-depth study will be carried out on intra-domain mobility. However, it is important to consider that if a sensor node is mobile within a single sensor network and changes its point of attachment, depending on the protocol and topology, there might be a need for re-authentication and key exchange to maintain security even within the same network.
Mobility management in wireless sensor networks can, as previously mentioned, be carried out in several ways and on different levels depending on the wireless technology used. In this section, a number of protocols will be described and compared in terms of their mobility support. ZigBee specifies a set high-level open standards communication protocols used to form small Personal Area Networks (PAN’s) with a limited transmission distance of approximately 10-100 meters (depending on radio and antenna). The ZigBee protocol is based on the underlying IEEE 802.15.4 standard for MAC and PHY layers and supports mesh networking as well as star and tree topologies. Most ZigBee devices use the 2.4 GHz ISM frequency for radio communication.
Security within the ZigBee network is enabled by 128-bit symmetric encryption using a pre-shared key. A network can have one of three distinctive roles, namely: coordinator, router and end-device. A network requires at least one coordinator node that accounts for automatically configuring the network, regardless of topology, and acts as a central hub for routing. A router is a device that can be used to extend the network by allowing other nodes to use it as a relay point. The end-device is a device that typically acts only as a sensor and communicates sensor information. Mobility within the network is handled using the AODV [10] routing protocol. In a tree topology with hierarchical addressing, the node will need to change its address if changing connection point to another branch in the tree which may cause service disruption [29].
Another commonly used sensor technology is Z-Wave [12], which supports lower bit rates than ZigBee, and communicates on a narrow ISM band around 900 MHz (the exact frequency depends on country). Much like a ZigBee network, a Z-Wave network can support multi-hop communication in order to increase range. Z-Wave networks are limited to 232 devices which may be a problem in large scale implementations. Since Z- Wave networks use source-routing, the network does not perform well for mobile devices that may change their position. While ZigBee is likely the most prevalent protocol among the open standards, a number of proprietary solutions exist as well. DigiMesh [30] by Digi is one example of a protocol that builds on the same basic topology as ZigBee with coordinator, routers and end-devices, but with the big difference that it allows routers to be sleeping periodically, and therefore enables them to be battery powered. However, this functionality requires synchronized sleep among devices and when a mobile sensor connects to a new network it needs to synchronize with the network coordinator, which is a power consuming task. The DigiMesh protocol is only implemented on the X-Bee and X-Bee PRO series of hardware provided by Digi and operates on the 900 MHz or 2.4 GHz frequency.
Bluetooth Low Energy (BLE) or Bluetooth Smart [28] is a re-engineered version of the well-known Bluetooth technology commonly used in e.g. mobile phones to communi- cate with peripheral equipment such as headsets and speakers. Compared to the classic Bluetooth protocol, BLE reduces the energy consumption while still maintaining roughly the same communication range of approx. 100 m. This makes it attractive for sensor device connectivity. Since BLE devices typically only communicate using a single hop, there is no explicit mobility handling within the protocol. However, a connection latency of less than 10 ms as compared to 100 ms for classic Bluetooth, enables a slave (sensor) to detect and connect to a new master device (gateway) quickly, if being mobile. Figure 2.3 depicts a general (OSI inspired) protocol stack and the span in which the different technologies and protocols operate.
Another important technology in this area is 6LoWPAN [31], which defines a way to encapsulate and compress a standard IPv6 header into a much smaller frame, such as the ones used in IEEE 802.15.4 compliant MAC protocols. This enables IP connectivity all the way to the sensor device and allows the sensor to be identified with a globally valid IPv6 address. This, in turn, enables sensors to use IP-based protocols, such as Mobile IPv6 [32] to handle IP mobility on the network layer on a global scale.
Ordinary IPv6 requires a Maximum Transmission Unit (MTU) of at least 1280 bytes to operate. This requirement is generally met by for example IEEE 802.3 based tech- nologies such as Ethernet which has a MTU size of 1500 bytes. IEEE 802.15.4 frames, on the other hand, are limited to an MTU of 127 bytes. This MTU with a maximum IEEE 802.15.4 frame overhead of 25 bytes results in a remaining 102 bytes for upper layers.
Since an ordinary IPv6 header requires at least 40 bytes, using no compression, would result in a comprehensive header overhead in relation to the frame size. To reduce the IP header overhead, a compression mechanism used in 6LoWPAN is described in RFC 6282 [33]. Using this mechanism, the IP header can be reduced to a few bytes depending on, for example, address types used.
2.1. Heterogeneous Access Networks and Mobility Management 25
Application
Application Interface
Security
Network Layer
MAC/Data-link Layer
Physical Layer
Custom appIEEE 802.14.2 ZigBee Bluetooth Low Energy
Z-Wave
6LoWPAN
Figure 2.3: Sensor Networking Protocols.
2.1.3 Handover Decisions and Metrics
Usually, performing a handover is related to some kind of service disruption, albeit very briefly in some cases. Radio frequency scanning, access point association, automated configuration and mobility management protocol signaling are all examples of procedures that will impact the handover latency. Wireless network interface cards typically only have one radio, which means that the interface will have to drop the ongoing session and free the radio, in order to scan for other access points. When the intended target access point has been identified, an initial handshake and protocol parameter negotiation is typically needed to establish a contact with the access point. Negotiation parameters may include data-link layer protocol to use, header compressions, user authentication, and authorization.
The next step involves IP configuration and establishing a path to the network. If a mobility management protocol is used, signaling in the form of re-/registration messages will cause additional delays. The ITU-T states that in order for a VoIP call not to be affected, the handover latency should stay below 50 ms [34]. During horizontal handovers in e.g. WiFi and cellular networks these constraints are typically met without greater effort. However, completing the steps mentioned in the previous paragraph, vertical handovers will often cause service outages in the range of hundreds of milliseconds up to several seconds in worst cases. For real-time applications such as VoIP and video, this will lead to a significant and noticeable service disruption.
Another factor that may affect the overall network performance is extensive signaling by the mobility management protocol. Further, depending on the protocol, a certain amount of signaling may be needed to establish new associations between entities when the network topology changes (for example with BU-messages in MIP). Also, to detect changing network conditions and react timely, the MN may be required to probe available resources at a high frequency. For example, probing the available network resources periodically may place high momentary load on the network which will affect other users and disrupt ongoing packet flows.
In cases where the MN is located at the edge between two adjacent wireless networks, an undesired situation may arise where the mobility management protocol is not able to make a good decision and oscillates back and forth between the two networks [35].
If the target network is predicted to be only slightly better than the currently active, switching over and placing load on the target network may cause the prediction to be inverted since the target network will decrease its available resources while the available resources in the original network will increase, causing an oscillation between the two.
This so called ping-pong effect will cause extensive signaling and service disruption for the MN. Network characteristics and performance are likely to fluctuate over time, especially in wireless environments. Therefore the network has to be continuously monitored in order to make decisions regarding access network selection. Measurement methods as well as metrics need to be carefully chosen to provide a fair and accurate decision making.
The term network performance can have multiple meanings depending on the applica- tion. A VoIP call, for instance, has high demands on delay and jitter at low bandwidths while a FTP download only needs the highest possible throughput [34]. With this in mind, the decision making entity should take into consideration, not only the measur- able network parameters, but also the types of applications used and their requirements.
Other parameters could also be included in the decision process such as monetary cost and power consumption. For some wireless technologies such as LTE, the output power is adjusted according to the signal distance to the base station. In such cases the power consumption of a mobile device can be severely affected by the distance to the base sta- tion and the MN can benefit from turning the unused radio interface off while a another interface is used.
When it comes to measuring and determining the network performance, it can be carried out in a multitude of ways. Two basic categories are passive and active probing.
Passive probing includes measuring the Signal-to-Noise Ratio (SNR) and other radio pa- rameters that can be monitored without putting any strain on the network. On the other hand, active probing involves sending traffic that is used to carry out the measurement.
It is usually preferred to have minimum impact on the network while making good and frequent measurements. A good tradeoff between the mentioned methods is needed to provide the optimum information to the handover decision process and this tradeoff is hard to define [36].
While giving a valuable indication on radio performance, low level parameters such as signal strength and radio resource availability, are very technology specific, and therefore hard to use when comparing technologies with each other. Network layer parameters
