Vertical handover between wireless service providers

Abstract

We discuss the dynamics of user handover between two coexisting wireless service providers and analyze the consequent exploitation of the offered diversity by the use of multi-standard terminals. We show that the potential capacity benefits of mobile-initiated vertical handovers (VHO) are substantial, but it is important to choose the correct VHO criteria in order to achieve optimum load balancing and equilibrium states (both globally and socially). A fast-handover scheme, based on replicator dynamics is presented, which exhibits fast convergence to the socially optimal states by allowing only a subset of the necessary VHOs among the air interfaces to take place. I. IGiven the massive deployment of wireless networks, it is common for mobile users to have several choices of collocated WLANs to connect to. This phenomenon is made especially evident by the large scale appearance of 3G systems operated by major networks. In fact, mobile user chips which support multiple standards already exist and, additionally, there has been a significant amount of work in creating flexible radio devices capable of connecting to any existing standard [1]. It is therefore reasonable to expect that in the near future users will have the option to connect to different networks and to switch dynamically between them on a real-time basis, based on the offered throughput and/or price.The dynamics of this process has several interesting aspects. Firstly, with no central authority to moderate and shepherd the users' s selfish behavior, there is the very real danger of total lack of coordination between the users, leading to frustration and suboptimal performance. Moreover, even though users now have more choices to connect to, the (finite) resources of nearby access points (AP's) still remain an object of competition.It is clear from the above that this process can be modelled in terms of a non-cooperative game. Of course, this is not a novel idea in and by itself: an excellent survey of applications of game theory to networking appears in [2]. For example, uncoordinated random access channels have been analyzed by optimizing their transmission probabilities [3], or their power control [4]; another application is in CDMA systems (e.g.[5]-[7]); and, in the direction of connecting to multiple wireless nodes, [8] considered the possibility of connecting to several 802.11 APs using a single WLAN card.In this paper we analyze instead the dynamics of vertical handover between service providers possibly employing different standards. 1 We will be assuming that users can switch air interfaces in the time-scale of hundreds of milliseconds and call this scheme multi-mode operation (MMO), as opposed to single mode operation where users are not capable to handover between air interfaces (AIs). We thus postulate that parallel connections to both AIs are available, allowing each user to switch between AIs at rates faster than the typical session duration. Due to the limited processing capabilities at the terminal, it will only be able to proc...

Distributed Learning-Based Spectrum Allocation with Noisy Observations in Cognitive Radio Networks

“…In the spectrum sharing literature, there are a few works studying log-linear dynamics to learn equilibrium points. For instance, [23] studies log-linear equilibriums (or logit equilibriums) for a stochastic game and [24] presents a log-linear learning algorithm for a strategic game with a unique purestrategy NE. However, in this paper, log-linear learning is introduced and investigated for a strategic potential game with multiple pure-strategy NEs aiming to enable equilibrium selection.…”