In modern day wireless networks, spectrum utilization and allocation are static. Generally, static spectrum allocation is not a feasible solution considering the distributed nature of wireless devices, thus some alternatives must be ensured in order to allocate spectrum dynamically and to mitigate the current spectrum scarcity. An effective solution to this problem is cognitive radio (CR), which seeks the empty spectrum portions and shares them with the neighboring devices. The CR devices can utilize the available spectrum more efficiently if they try to work together. Therefore, in this work, we review a number of dynamic spectrum allocation techniques, especially those using multiagent systems and game-theoretical approaches, and investigate their applicability to CR networks. The distributed nature of these two domains makes them suitable for CR networks. In fact, the idea of dynamic spectrum sharing using these techniques is not entirely new and several interesting approaches already exist in literature. Thus, in our study we try to focus on existing spectrum sharing literature and cooperative multiagent system for CR networks. We are particularly interested in showing how the distributed nature of multiagent system can be combined with cognitive radios in order to alleviate the current static spectrum usage as well as maintaining cooperation amongst the CR nodes. Moreover, our work includes the description of various scenarios in which spectrum sharing is an essential factor and hence must be performed in a dynamic and opportunistic manner. We also explain the working of our proposed spectrum allocation approach using multiagent system cooperation in one of these scenarios and verify its formal behavior using Petri net modeling.
Top1.1 Introduction
Modern day wireless systems are moving from static and centralized control to distributed and autonomous networks, where the devices may work more dynamically and they can opportunistically select the available spectrum/bandwidth by having frequent interactions and information exchanges with their neighboring devices. By autonomous networks (Schmid, Eggert, Brunner & Quit, 2005), we mean that the control and information are fully distributed and the wireless devices have the capabilities of self organization and adaptability to cope with frequent network changes. Most commonly, the devices are meant to be infrastructure independent and are designed to enable inter-device interactions using multi hops heterogeneous elements. The primary and most common objectives of deploying these autonomous infrastructures is to develop systems having autonomous behaviors and to integrate different kinds of devices together in order to permit the technological mobility between the interconnected domains.
The above mentioned opportunistic and autonomous behaviors are now becoming both possible and necessary by the introduction of cognitive radio (CR) technology (Mitola, 2000)(Mitola & Maguire Jr., 1999) in the wireless network domains and through the advances in the field of Distributed Artificial Intelligence (Weiss, 2000) by exploring the concepts of multiagent systems (Wooldridge, 2002). These cognitive radio devices can then be used in a wide variety of network domains (WLAN, WiRAN, MANETs). In addition, an efficiently designed CR with a software agent deployed on it would be capable of interacting with neighboring radios to form a dynamic, loosely-coupled and collaborative network. Then, these radios work together to collectively reach the goals that are difficult to achieve by an individual radio. Further, the deployment of a multiagent system provides an incentive to conceptualize and design new spectrum sharing solutions for wireless networks. This incentive is particularly attractive and equally important to create spectrum sharing solutions that work in dynamic, distributed and open wireless networks domains. To mention one of the many exciting examples, a mobile phone having no license for spectrum usage can still use the unutilized spectrum portion of a licensed (or paid) user via CR capabilities and later can share its spectrum portion with the neighboring users using a multiagent system (MAS) (Jiang, Ivan & Anita, 2007).
In this work, we propose a thorough study of various spectrum sharing techniques available in literature. Our study moves around two parts. In the first part, we try to focus on introducing CR networks along with CR user’s basic functionalities and capabilities. This section also includes the introduction of various concepts related to MAS coordination. Our second part focuses on presenting several spectrum sharing studies using game-theoretical and MAS approaches and capturing the different ways of applying these techniques for cognitive radio networks. Note that to a larger extent, the idea of using MAS over CR networks is not entirely new and some of the already proposed solutions have been suggested in the literature (Jiang, Ivan & Anita, 2007)(Kloeck & Jondra, 2006)(Tonmukayakul & Weiss, 2005). Nevertheless, we not only view the multiagent systems usage over CR networks, but take this view a step further and introduce a framework, where the devices having agents can cooperate with each other in order to have an efficient use of the available spectrum.