ABSTRACT
Contents 13.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 394 13.2 Attributes for Network Selection . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 395
13.2.1 Network Attributes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 396 13.2.2 User Attributes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 396 13.2.3 Attribute Preference. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 397
13.3 Network Selection Methods . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 397 13.3.1 MADM-Based Network Selection . . . . . . . . . . . . . . . . . . . . . . . 398
13.3.1.1 Construction of Decision Matrix. . . . . . . . . . . . . . 398 13.3.1.2 Determination of Weighting Vector . . . . . . . . . . 399 13.3.1.3 Calculation of Network Ranking. . . . . . . . . . . . . . 400
13.3.2 Enhancements for MADM-Based Network Selection . . . 402 13.3.3 Simulation Illustration . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 404 13.3.4 Implementation Consideration . . . . . . . . . . . . . . . . . . . . . . . . . . . 407
13.4 Impact of Network Selection on Spectrum Allocation . . . . . . . . . . . . 409 13.4.1 Service Demand from Mobile Users. . . . . . . . . . . . . . . . . . . . . . 410 13.4.2 Competition Model between RANs for Spectrum
Allocation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 412 13.4.3 Numerical Results and Discussion . . . . . . . . . . . . . . . . . . . . . . . 415
13.4.3.1 Capacities of RANs . . . . . . . . . . . . . . . . . . . . . . . . . . . . 416 13.4.3.2 Profits of RANs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 418
13.5 Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 420 References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 420
With deregulation and technology evolution on cognitive radios (CR), mobile users may have the freedom to select and adapt to one preferred network from multiple radio access networks (RANs) that coexist in overlapping areas. To select a preferred serving RAN, the mobile terminal (MT) can first detect and sense the characteristics (attributes) of the network side and the user side. This chapter describes the relevant network attributes, user attributes, and the individual user’s preference. For the convenience of mobile users, the disparity and conflict of these attributes need to be captured by network selection method for automatic decision making. MADM (multiple attribute decision making)-based methods are viewed as a promising approach to achieve this. This approach effectively aggregates the attributes and the individual user’s preference. The procedure of the MADM-based network selection methods and the enhancements for the benefits of users and mobile operators are presented in this chapter. On the other hand, mobile operators can adjust the network attributes (e.g., price) to attract users to access and dynamically request a spectrum from a common spectrum pool to meet the varying service demands. The interaction of network selection of mobile users and spectrum allocation between RANs is analyzed by formulated service demand model and noncooperative competition model between RANs.
