Self-Adaptive QoS Control in Cognitive Networks That Is Based on Service Awareness

Release Date:2011-06-21 Author:Chengjie Gu, Shunyi Zhang, Yanfei Sun Click:

This work was funded by the National High Technology Research and Development Planning (“863” Project) under Grant No. 2006AA01Z232, 2009AA01Z212, 2009AA01Z202, and the National Natural Science Foundation Project under Grant No. 61003237.


1 Concept of the Cognitive Network
    Network service types are varied, and network environments are complex and dynamic.  Traditional end-to-end insurance technology lacks intelligent inference and self-learning capabilities. Therefore, it cannot adapt to provide ideal service under dynamically changing network conditions [1].


    The obvious problem is that the network system cannot perceive the service demands of end users and cannot effectively and dynamically change QoS according to variations in the internal and external environment of the network system [2]. Academics have started to integrate cognitive elements from next generation networks (NGNs) into current networks in order to overcome these embedded defects. Consequently, the concept of cognitive networks (CNs) has arisen.


    Research on CN is focused on the cognitive radio (CR). Mitola [3] first put forward CR and the architecture of the cognitive ring. A CR system obtains frequency spectrum through perception. It determines the reconstruction scheme of CR according to the optimization object and can adapt to changes in the frequency spectrum environment.


    CNs based on CR were conceived by the Motorola and Virginia Tech companies [4]. A CN has cognitive processes and perceives the current network condition. It perceives changes in itself and in the environment. It then makes plans and determinations and takes action based on these perceptions. The FOCALE architecture of dual close-loop control is also provided.
At SIGCOMM 2003, Clark [5] and others proposed introducing Knowledge Plane (KP) to the Internet. The key to this concept is that KP can perceive its own behavior. It can analyze problems and adjust its operation to increase reliability and robustness.


    In 2007, Baldo [6] used fuzzy logic to process modularization and inaccuracy effectively in the CN. In 2008, Siebert [7] pointed out that the ability of a CN to implement tasks through autonomous self-management, self-optimization, self- monitoring, self-maintenance, self-protection, and self-healing was an important feature. In 2009, Fortuna [8] suggested that Thomas’s definition of CN was incomplete. Knowledge expression and cognitive ring are the most important elements of the CN.


    The IEEE is currently discussing standardization of the integration architecture of isomerism wireless access networks. In these discussions, the concept of CN is used. CN is seen as a new way of improving overall network and end-to-end system performance as well as simplifying network management. It is the trend of next-generation communication [9].
CN is a new research area and has just taken its first steps in China and in other countries. Therefore, relevant theories and techniques need to be further studied.

 
    The cognitive functions of a CN are implemented by distributed intelligent agents based on AI technology. Agents with learning and reasoning capabilities are deployed on each node in the network to monitor and collect environment information. These agents cooperate and exchange information so that the network can perceive its current status. End-to-end targets can be achieved based on the network status, and network resources can be evaluated, predicted, planned, adjusted, and allocated based on a knowledge library. As a result, the network has self-perception, self-learning, self-optimization, self-healing, and self-configuration capabilities. It can be measured, controlled, managed, and trusted.


    In the QoS control architecture of the CN in this paper, the key concept is the network’s ability to perceive changes in the CN environment and adjust itself in real time. 

 
    Self-adaptive control technology can plan and allocate limited network bandwidth effectively so that network performance is improved. The technology also manages and controls network traffic according to service features in order to improve the revenue of unit bandwidth. Therefore, intensive self-adaptive control is essential to solving network QoS problems in CNs.


2 Key Technologies of Cognitive Network QoS
    According to the definition given by Thomas [4], the structure of a CN is described as “Target-Cognitive Decision-Reconfiguration,” as shown in Fig.1. The target layer reflects the target demands put forward by the application, user, or resource. With Cognitive Specification Language (CSL), the targets are mapped to specific mechanical demands and fed back to one or more relevant CN elements. The cognitive decision-making layer implements status switching of NEs and perceives the current network status according to the requirements of the target layer. It obtains the NE configuration through certain methods.

 

 
    The reconfiguration layer is also called the adaptive network layer. The decision of the cognitive decision-making layer is sent to the corresponding entity NE through the Application Programming Interface (API). By adjusting the configuration of the NE, demands of the target layer are met. At the same time, the layer sends the network status through a sensor to the cognitive decision-making layer.

 

2.1 Context Perceiving
    The foundation of CN is the rapid perception of network environment. A CN needs to observe current network environment information in appropriate time. The information is used in later planning and decision making to determine whether the current network meets user requirements.  If not, a suitable reconfiguration method is used to meet user requirements.
Environment information perceived by the CN includes network type, network topology, available resources, interface protocols, and network traffic, all of which affect end-to-end transmission performance [10]. Context perception is an important way of improving network intelligence. It   determines changes in context information and adjusts itself accordingly. When the network environment changes dynamically, the network makes relevant self-adjustments. This self- adjustment uses a reflection mechanism and a policy mechanism. From the policy definition, the network can pre-define an adjustment method when the context changes.

 

2.2 Cross-Layer Design
    The essence of cross-layer design is to break the frame of the traditional network system to meet QoS requirements of the communication system. In this design, the status parameters and QoS parameters of the communication system resources are transmitted in the protocol layer. As a result, a joint design combining various protocol layers is achieved, and system resources are fully utilized in order to provide better service for users [11].


    The purpose of CN is to adjust the relevant NE protocol stack or protocol layer parameters on the basis of CN network information. This ensures users receive high quality end-to-end performance.  The cognitive processing layer knows the status of network layers and determines proper actions according to an optimization algorithm. It reconfigures network parameters and protocol stacks to achieve end-to-end communication.

 

2.3 Reconfiguration
    If the network is not meeting the end-to-end requirements of users, the CN adjusts the protocol stack parameters of the relevant NE to meet these requirements. The adjustment process is the reconfiguration of the network [12]. CN emphasizes the end-to-end target, and it should provide end-to-end reconfigurability. Software radio technology is limited to reconfiguring the terminal, but CN involves all layers of the NEs and protocol standards that a stream passes through. It is a scheme with foresight that ensures QoS targets are met. More factors are considered in end-to-end reconfiguration.


    Realization of CN is based on reconfiguration of the NE. The reconfiguration process can also be implemented through software, but the technical level of this reconfiguration is higher. Terminal reconfiguration, network reconfiguration, and service reconfiguration are contained, and this configuration is not limited to a single node. Multiple NEs on the end-to-end path are covered.  This is called end-to-end reconfiguration (E2R). The complexity and importance of E2R is greater than terminal reconfiguration.


3 CN QoS Control Architecture Based on Service Awareness

 

3.1 Service Awareness
    In recent years, many new applications have emerged, including peer to peer (P2P) networks, VoIP, streaming media, interactive online games, and virtual reality. The emergence of these new services is impacting the traffic model and application mode. The rapid development of P2P in particular has caused explosive growth in traffic, and unlimited bandwidth usage has increased the burden on the network.


    As a result, network congestion has become more serious. Simple expansion cannot meet the requirements of increasing services. Therefore, the best way to perceive, analyze, determine, and control transmission service intuitively is by using CN technology.


    CNs are driven by services. The network system intuitively perceives services on the network, including end user service status and NE service status. Intuitive perception and classification based on the service stream is the foundation of service-centered resource configuration, route adjustment, and dynamic self-adaptive traffic control. In service-aware technology, before the CN is introduced, traditional static port method, payload feature method, and stream statistical feature method are used. The methods are effective for perceiving regular services, but they cannot perceive many new services accurately.


    After CN is introduced, the network has intelligence as well as analysis and decision-making capabilities. In this section, an integrated feature-based service awareness model is constructed for perceiving services intuitively and intelligently in real time. As shown in Fig.2, after the model has obtained regular parameters of the CN, it constructs an integrated feature identification model based on the traffic statistics feature, connection mode, topology feature, and content feature. It constructs an identification engine for each feature and triggers and perceives different identification engines intelligently according to policies.

 


    As a result, known or unknown, encrypted or plain text traffic can be identified accurately and efficiently. The integrated feature-based intelligent cognitive model distinguishes known or unknown, encrypted or plain text services. This forms the technical basis for a CN self-adaptive QoS control architecture based on service awareness.

 

3.2 Three-Level QoS Control Architecture in a CN
    As shown in Fig.3, the network QoS decision-making and control architecture has a three-level structure, composed of NE (device) cognitive module, autonomous domain cognitive server, and central cognitive server. Each part provides cognitive capability (self-awareness, self-learning, and self-decision making). The NE (device) cognitive module is the basic unit of the CN QoS awareness, analysis, and control system. It provides awareness and decision-making capability and dynamically adjusts NE parameters or configuration. The NE and user-end devices deployed with cognitive module form a cognitive autonomous domain (configured with a domain cognitive server) that is responsible for managing and controlling the NE device, service traffic, and network resources.

 


    At the same time, a central cognitive sever configured in the architecture is responsible for monitoring, awareness, and management of the running status of the entire network. The layered structure reduces the load on the central cognitive server. Even if the server fails temporarily, service QoS guarantee and management throughout the entire network is not affected.


    Distributed networking and communication is enabled between autonomous cognitive servers so that information is exchanged in real time. The reason for using distributed management in the domain cognitive servers is to increase system reliability, flexibility, and expansibility. In the autonomous domain, adjacent nodes communicate so that distributed cooperative monitoring and self-adaptive processing is possible. The architecture integrates the features of centralized architecture and distributed processing technology.

 

3.3 Port and Path Collaboration in Self-Adaptive QoS Control
    CN end-to-end QoS is guaranteed by cognitive NEs. Cognitive NEs are cooperative or independent. The NEs perceive the network condition in real time, bring the trends together, and analyze the network condition. They configure themselves based on existing policies for achieving end-to-end QoS targets.


    The following describes the integration of service source-end QoS control and link QoS control in the CN based on service awareness, resource appointment concept, and control theory. A collaborative port and path policy-based self-adaptive QoS control mechanism is proposed to solve the problem of end-to-end QoS guarantee for service traffic. The mechanism sends real-time network parameters to the autonomous domain server (or central cognitive server) through a feedback control.


    As a result, the self-adaptive QoS control mode is integrated into the terminal NE and routers.  The history of the network condition is compared with the current condition to form a control policy and to update the policy library through self-learning. At this point, the control policy is optimal. The mechanism can ensure the normal operation of a single NE and has the features of CN. The mechanism uses relevant NE devices and reasonably allocates limited resources to improve end-to-end QoE and QoS. In this way, the performance of the entire network is optimized. Fig. 4 shows the awareness-based service source end control layer and distributed awareness-based link control layer.

 


    Awareness-based service control at the source end is implemented through self-adjustment of the source-end transmission rate, intuitive closing of service, and intuitive decrease of QoS target. When the service source end launches a service in the traditional network, the current network condition is not considered. The CN service end has certain cognitive functions; therefore, the cognitive information comes from the domain server or central control server.


    When certain conditions are met—for example, when bandwidth is sufficient—resources in the network can accept the access of other service traffic, and the service traffic can be transmitted to the peer end. When high-priority users need to transmit services, but the current network does not provide sufficient resources as required by the SLA, the central cognitive server (or domain cognitive server) negotiates with users at the service source end. If the user accepts a reduction of QoS, the source end transmits the service traffic according to the negotiated results. If QoS requirements cannot be reduced, the cognitive server recycles the network resources being used according to the resource distribution policy or even forcibly closes certain low-priority services.


    Link control based on distributed awareness is implemented through intuitive control of NE traffic, route management, QoS degradation positioning, and intuitive queue management. By perceiving the network and making decisions based on this information, switches or routers with cognitive functions in the network can intuitively control traffic of different services and ensure the volume of trusted service traffic and key service traffic. They can also limit the volume of unsafe traffic or non-key traffic.


    As service requirements and network resources are changing in real time, bottlenecks or QoS-degrading parts of the end-to-end network can be detected by cognitive route management and QoS degradation positioning. In addition, analysis and decision making can be performed, and service traffic can be re-routed. An intelligent and intuitive queue management algorithm can also be used to determine congestion in the CN.


    Awareness-based intuitive queue management is oriented to the server’s collaborative drive policy. This policy is integrated into the intuitive queue management method to improve the resource appointment algorithm and router buffer management mode. Resources of the router or end system can then be reserved.


4 Conclusions
    Owing to the complexity, isomerism, and ubiquity of the access mode and network applications, current networks cannot meet the QoS requirements of users. CN is considered a new way of improving entire network performance and end-to-end system performance as well as simplifying network management. It is the trend of next-generation communication. CNs are important for ensuring performance in complex and isomerism networks.
This paper proposes a self-adaptive QoS control architecture for CNs. With service awareness, self-adaptive control can be implemented in a CN. This architecture is a new approach for solving the problem of NGN end-to-end QoS. Some techniques have been applied in the experimental platform of the national project 863 "Key Techniques in Network Behavior Model-Based Cognitive Network QoS." Monitoring devices are also created and applied for optimizing the network of a carrier. The technique demonstrates sophisticated applicability and stability.

References
[1] C.Lin, Zhichang Shan and Fengyuan Ren, QoS of Computing Networks, Beijing: Tsinghua Uni. Press, 2004.
[2] C.Lin, Zhichang Shan, Fengyuan Ren, "QoS of next generation networks," Chinese Journal of Computers, vol.31, no.9, 2008, pp.1525-1535.
[3] J. Mitola and G. Maguire, "Cognitive radio: Making software radios more personal," IEEE Personal Commun., vol.69, no.8, 1999, pp.13-18.
[4] R. Thomas, Cognitive Networks. Blacksburg, VA: Virginia Polytechnic and State University, 2007.
[5] D. Clark, C. Partrige, J. Ramming et al., "A knowledge plane for the Internet," in Proc.Conf. on Applications, Tech., Architectures, and Protocols for Comput. Commun.(SIGCOMM‘03), New York, pp.3-10.
[6] N. Baldo and M. Zorzi, "Fuzzy logic for cross-layer optimization in cognitive radio networks," IEEE Commun. Mag., vol.46, no.4, 2008, pp.64-71.
[7] M. Siebert, "Self-X control in (future) mobile radio networks," in Proc.European-Chinese Cognitive Radio Syst. Workshop, Beijing,2008.
[8] C. Fortuna and M. Mohorcic, "Trends in the development of communication networks: cognitive networks," Computer Networks, vol.53, no.9, 2009, pp.1354-1376.
[9] F. Shao and Lifeng Wang "Cognitive network structure and approach based on cognitive level," J.Beijing Univ. Tech., vol.35, no.4, 2009, pp.1181-1187.
[10] P. Balamuralidhar and R. Prasad, "A context driven architecture for cognitive nodes," Wireless Personal Communications, vol.45, no.1, 2008, pp. 423-434.
[11] V. Srivastava and M. Motani, "Cross-layer design: A survey and the road ahead," IEEE Commun. Mag., vol.43, no.12, 2005, pp. 88-95. 
[12] M. Pitchaimani, B. Ewy, J. Evans, "Evaluating Techniques for Network Layer Independence in Cognitive Networks," Proc.IEEE Int. Conf. on Commun. (ICC’07), Glasgow, 2007, pp.6527-6531.

[Abstract] This paper analyzes a self-adaptive Quality of Service (QoS) control architecture for cognitive networks (CNs) that is based on intelligent service awareness. In this architecture, packets can be identified and classified using an intelligent service-aware classification model. Drawing on Control Theory, network traffic can be controlled with a self-adaptive QoS control mechanism that has side-road collaboration. In this architecture, perception, analysis, correlation, feedback, decision making, allocation, and implementation QoS mechanisms are created automatically. These mechanisms can adjust resource allocation, adapt to a changeable network environment, optimize end-to-end performance of the network, and ensure QoS.

[Keywords] cognitive network; service-awareness; self-adaptive control; QoS