Cooperative user–network interactions in next generation communication networks

Abstract

Next Generation Communication Networks employ the idea of convergence, where heterogeneous access technologies may coexist, and a user may be served by anyone of the participating access networks, motivating the emergence of a Network Selection mechanism. The triggering and execution of the Network Selection mechanism becomes a challenging task due to the heterogeneity of the entities involved, i.e., the users and the access networks. This heterogeneity results in different and often conflicting interests for these entities, motivating the question of how they should behave in order to remain satisfied from their interactions. This paper studies cooperative user–network interactions and seeks appropriate modes of behaviour for these entities such that they achieve own satisfaction overcoming their conflicting interests.

Introduction

Converged Networks [1] allow different access networks, terminals and services to coexist bringing forth a new communication paradigm, which is user-centric [2], i.e., the user is no longer bound to only one access network but may indirectly select the best available access network to support a service session [3]. Upon a new service request or even any dynamic change affecting the session (e.g., mobility) one of the participating access networks needs to be selected in order to support the session. Next Generation Communication Networks – controlled by an IP Core Network – need to be equipped with a Network Selection mechanism to assign the best access network to handle service activation, or any dynamic session change. Such decision may result in a single access network or even multiple networks cooperating to handle a service (the multiple networks case is treated in [4]). This paper studies the resulting interaction between a user and a network, and seeks the best behaviour for each entity such that their conflicting interests are overcome and satisfaction is achieved.

Interactions between entities with conflicting interests, follow action plans designed by each entity in such a way as to achieve a particular selfish goal and are known as strategic interactions. Game Theory is a theoretical framework that studies strategic interactions, by developing models that prescribe actions in order for the interacting entities to achieve satisfactory gains from the situation. In this paper, we utilise Game Theory in order to model, analyze and finally propose solutions for the user–network interactions arising due to the Network Selection mechanism. Initially, we utilise Game Theory in order to model the relationship between the user and the network as a strategic game. Particularly, we utilise the notion of Present Value (PV) for a sequence of payoffs in a repeated game [5] in order to reach a decision that will be both user-satisfying and network-satisfying.

In an attempt to improve the user–network relationship according to the knowledge acquired during the interaction history, we define a notion of adaptivity for the user as a player in the access network selection decision game, and based on that, we propose a new user strategy for the repeated user–network interaction game. The proposed adaptive strategy, together with the rest of the strategies studied in this paper, are evaluated through simulations that explore their behaviour. The numerical results verify that the players gain more when cooperating, as expected from the theoretical analysis. Moreover, simulations show that the proposed adaptive user strategy results in profitable payoffs, also reinforcing the conclusions of the theoretical analysis.

The paper is organized as follows. Section 2 describes related work and motivates the subsequent work. Section 3 investigates the user–network interaction both as a one-shot and as a repeated interaction. The paper continues by a study of access network selection (Section 4) leading to a proposal of a new adaptive user strategy. A numerical evaluation of all strategies analyzed in the paper is found in Section 5 and finally, conclusions are drawn in Section 6.