CodaQ: a Context-aware Adaptive QoS Middleware

Get Complete Project Material File(s) Now! »

Context-awareness and self-adaptation

Healthcare systems present a growing demand for software applications that are able to adapt dynamically to changing environment and user conditions as well as to the unavailability of a devices or a services.
Context-aware and self-adapting [57, 58] systems are interesting solutions for this demand. The concept of adaptation and context awareness are now open research subjects as well as the application developments.
Context-aware healthcare is making healthcare services available everywhere, anytime by applying context-aware computing technologies for healthcare, health, and wellness management. Using context-aware computing device and network, the doctor would be informed of patients’ context data such as vital sign, and would call his patients to encounter hospital for more accurate examination. To support the doctor for gathering and analyzing the patients’ context-data, some intelligent systems are required.
In context aware computing environment, computing entities is ranging from sensors and actuators to web services and applications. For example, a wearable health monitoring device can constantly examine one’s blood pressure, body temperature, pulse, etc.; the availability of surveillance camera and embedded microphones and home automation systems at home may support remote health and activity monitoring. To fully use the benefits of context-aware system, an infrastructure which enables QoS self-adaptation and protocol interoperability among heterogeneous components is required.

“Maisons Vill’Âge®”: A telehomecare system for elderly [1-5]

The presented requirements and challenges in 1.2 and remaining obstacles make the design of monitoring and assistance tools quite complex, as each pathology or disability generates its own set of requirements and constraints. Such context increases the need for the development of innovative global ICT-based solution allowing to implement personalized and person-centric care process and tools, which can evolve with the users’ medical and social and physical evolution, and which can compile the various collection and analysis of data to ensure a reliable and sound diagnosis, medication and evaluation of the medical and social state of the person.
The first contribution of this thesis is the development of “Maisons Vill’Âge®”, a new concept of building smart home by integrating telemedicine and home automation systems. We have designed and developed a distributed system architecture to support remote medical consultations and elderly management and homecare across global wide area networks and heterogeneous platforms. The system is designed for the elderly who wish to spend their old age at their own home, in order to increase independence and quality of life. This would not only benefit the elderly who want to live at their own home, but also the national health care system by cutting costs significantly. The first houses are being built with basic implements of data acquisition and human-machine interface. A flat is entirely equipped to act as demonstrator and test bed.

QoS support in wireless sensor networks [6-9]

In all of the data acquisition networks, like health monitoring systems, either the data is collected from the network periodically or on the occurrence of an event, in such systems, the data are highly vital to have a stable monitoring and a minimum number of faulty alerts. However, none of them adapts completely themselves neither to the failure of the nodes nor the temporal variations in data delivered by the sensor network.
In addition, limited energy resources in wireless sensor networks and unreliable radio channel, unavailability of the nodes is a constraint to provide a reliable communication. The unavailability of a node may have many reasons like mobility of the node and node or link failure. It is clear that in a real environment, we cannot control or reduce the number of unavailability in a network but it’s possible to manage it. This necessitates the use of a routing algorithm with a QoS support mechanism which readily adapts to the unavailability and mobility of the nodes and changes in the data delivery rate.
The second contribution of this PhD work is the proposition of a fuzzy logic based QoS-aware routing algorithm is proposed. Stable route recovery, high data delivery ratio and good load balancing are the main characteristics that the proposed algorithm adds to the ad-hoc sensor networks. The algorithm uses a multicriteria decision to form the clusters and manage the mobility and unavailability of the nodes and because of this multicriteria decision strategy it uses fuzzy logic to facilitate the decision process.
This protocol adapts also with mobility and unavailability of the nodes. It is especially effective in networks that use sensor nodes for data aggregation and in which the data delivery ratio is important and the nodes are mobile, like health monitoring sensor networks. In such networks health events and information are sensed by several nodes and therefore, this protocol can help the network to deliver sensed events with high successful rate in the network. The simulation results show that the proposed protocol is well suited for such applications.

Toward a multi-objective and multi criteria approach

As mentioned before, the main objective is proposing a system for forming dynamically the optimized clusters and adaptive routing.
Several dynamic clustering strategies have been proposed in the literature. ZRP [34], LEACH [35] and HEED [36] are some examples of clustering protocols. In [34], the zone routing protocol (ZRP) a hybrid strategy, is proposed by Haas and Pearlman which attempts to balance the trade-off between proactive and reactive routing. The objective of ZRP is to maintain proactive routing within a zone and to use a query–response mechanism to achieve inter-zone routing. In ZRP, each node maintains its own hop-count constrained routing zone; consequently, zones do not reflect a quantitative measure of stability, and the zone topology overlaps arbitrarily.
LEACH [35] is an application-specific data dissemination protocol that uses clustering to prolong the network lifetime. LEACH clustering terminates in a constant number of iterations (like HEED [36]), but it does not guarantee good cluster head distribution and assumes uniform energy consumption for cluster heads. In contrast, HEED makes no assumptions on energy consumption and selects well distributed cluster heads but HEED assumes quasi-stationary nodes.
In brief, the existing clustering protocols use many assumptions which make them not able to address the needs of telehomecare application: Some algorithms assume a network with fixed or low mobility nodes, which is not the case in a telehealthcare network; Some algorithms are based on centralized control that makes them not scalable; Some algorithms use periodic rounds to change cluster head and elect a new one. The new cluster head will be fixed for one round, but in an ad-hoc network with dynamic topology, cluster topology may change during this period, and in this case a new cluster head must be elected. Therefore this type of algorithms will be good for networks with fixed or very low mobility nodes, but not for telehomecare networks.
To balance the load in the network, most of the clustering protocols use different parameters to choose cluster-heads. Cluster ID [37], connectivity degree [38, 39] and periodical cluster heads election [35] are used in order to share the load among all the nodes of the network. By applying cluster ID or highest connectivity methods, the same node may be chosen as cluster-head every time, and that will result in draining its energy very fast.
Regarding the routing, the existing algorithms can be calcified into static (non-adaptive) and dynamic (adaptive) algorithms. Static algorithms do not make routing decisions based on current state and topology of the networks, in contrast, dynamic algorithms change their routing strategy either when the state and the topology of the network changed or periodically. These adaptive routing algorithms are more applicable in the mobile wireless sensor networks because of their dynamic nature.
Numerous parameters have to be considered when choosing a route in sensor networks. The route choice can be characterized by various criteria (See figure 9). Existing algorithms decide the routing path by using mathematical crisp functions [40, 41] which take into account just the crisp and precise parameters (like distance, load, etc.). But real life data are not always crisp, and all descriptions cannot be always expressed or measured precisely. These algorithms also do not take into account the essential criteria of optimal routing like unavailability rate of the node.
One possible solution for this multicriteria problem is using Fuzzy Logic [92]. Fuzzy systems usually describe complex technical systems in a simple way tolerating impreciseness. Linguistic rules are used to describe the model. It is not necessary to know the exact mathematical function which describes the behavior of the system, because fuzzy logic has potential for dealing imprecision in data using human reasoning without needing complex mathematical modeling. Fuzzy logic on the other hand is capable to combine many criteria to decide a route. The values of the real world parameter are mapped to fuzzy values by fuzzification and then processed by the interference machine. These fuzzy results are transformed into crisp values again by defuzzification.

READ  Sr-Microwave Standards Comparisons

Fuzzy Logic and its applications in the networks

As discussed, in our proposal, we use fuzzy logic because it is capable of making real time decisions, even with incomplete information. Moreover fuzzy logic can be used for context by blending different parameters – rules combined together to produce the suitable result. In this section we review some examples of applications of fuzzy logic in the clustering, routing and QoS. For reading comfort, we also give in appendix A a short recall of the basic notions in Fuzzy logic.

FLoR – Fuzzy Logic based Adaptive Clustering and Routing for Wireless Sensor Networks

This section presents FLoR, our adaptive clustering and routing algorithm. As a result of our discussions in the last sections, Mobility and Failure management and Load balancing are main important problems to be addressed in ad-hoc sensor networks. To address these problems, FLoR has 5 main parts: Fuzzy logic decision making, Clustering (Cluster-head election), Mobility management, Load balancing and Unavailability management. Figure 10 shows the main parts of the protocol. As we see in this figure, fuzzy decision making is the basic part of our proposal. That means, the 4 other parts of the protocol, use fuzzy logic to make decision or to process an event.

Table of contents :

Chapter 1 Introduction
1.1 Context
1.2 Challenges
1.2.1 QoS support
1.2.2 Context-awareness and self-adaptation
1.3 Thesis Contributions
1.3.1 “Maisons Vill’Âge®”: A telehomecare system for elderly [1-5]
1.3.2 QoS support in wireless sensor networks [6-9]
1.3.3 Context-aware adaptive QoS Middleware [10]
1.4 Thesis Outline
Chapter 2 “Maisons Vill’Âge®”: Smart Use of Sensor Networks for Healthy Aging
2.1 Introduction
2.2 System architecture
2.3 Maisons Vill’Âge®
2.4 Discussion
2.5 Conclusion
Chapter 3 FLoR: a Fuzzy logic based Adaptive Clustering and Routing
3.1 Introduction
3.2 Challenges and objectives
3.3 Toward a multi-objective and multi criteria approach
3.4 Fuzzy Logic and its applications in the networks
3.4.1 Clustering
3.4.2 Routing
3.4.3 QoS
3.5 FLoR – Fuzzy Logic based Adaptive Clustering and Routing for Wireless Sensor Networks
3.5.1 Definitions
3.5.2 Mobility management and new definition for mobility
3.5.3 Load Balancing and our definition of load
3.5.4 Unavailability management
3.5.5 Cluster head election
3.6 How does FLoR work?
3.6.1 Different processes in the nodes
3.6.2 Some examples
3.7 Evaluation
3.7.1 Overview of AODV and LEACH
3.7.2 Evaluation metrics – QoS requirements in wireless sensor networks
3.7.3 Simulation environment
3.7.4 Simulation results
3.8 Conclusion
Chapter 4 CodaQ: a Context-aware Adaptive QoS Middleware
4.1 Introduction
4.2 Definitions
4.2.1 Context
4.2.2 Context awareness and self-adapting
4.2.3 QoC – Quality of Context
4.2.4 Middleware
4.2.5 Context ontology
4.3 Challenges in designing context-aware middleware
4.4 Reference model
4.5 Related works
4.6 Discussion
4.7 System architecture
4.8 Data modeling
4.8.1 Raw Event (RE)
4.8.2 VirtualSensor
4.8.3 Deduced State
4.8.4 Zone
4.8.5 Query and Query Reply
4.9 CodaQ – Context-aware Adaptive QoS Middleware
4.9.1 Context Collector
4.9.2 Data management
4.9.3 Context process
4.9.4 System Observer
4.9.5 Context abstraction
4.10 Context-based Adaptive QoS
4.10.1 Embedded QoS
4.10.2 Run-Time State-based QoS
4.10.3 Spatial and Temporal Consistency
4.11 Discussion and concluding remarks
Chapter 5 Implementation
5.1 Introduction
5.2 General view
5.3 XML-based data presentation
5.3.1 Sensor installation and configuration
5.3.2 Raw event
5.3.3 Context information providing – deduced state
5.4 Sub-systems of the prototype
5.4.1 CodaQ Middleware
5.4.2 Application layer
5.4.3 Sensor nodes side
5.5 Evaluation
Chapter 6 Conclusion
6.1 Summary of Contributions
6.1.1 FLoR – Fuzzy Logic based Adaptive Clustering and Routing for Wireless Sensor Networks
6.1.2 CodaQ – a Context-aware Adaptive QoS middleware
6.1.3 Others
6.2 Future work and concluding remarks
Appendix A Fuzzy Logic
Introduction
What is Fuzzy Logic?
Crisp sets
Fuzzy sets and membership functions
Fuzzy sets operators
Linguistic variables
Hedges
If-Then rules
How does it work?
Fuzzy Inference Systems
Appendix B Hardware and Software environment of development
Hardware overview
Imote2 Processor Board
ITS400 Sensor Board
Software overview
Imote2.Builder SDK
Microsoft .NET Micro Framework
References

GET THE COMPLETE PROJECT

Related Posts