Article Preview
Top1. Introduction
Nowadays, 3G networks (e.g. UMTS and IEEE 802.11 a/b/g/n) and 4G networks (e.g. LTE and IEEE 802.16) communications have been already deployed for ensuring ubiquitous mobile Internet. The 5G networks are expected to reach market around 2020 (I-Scoop, 2008). As a result, different real time applications such as smart automation, e-commerce transactions, video streaming, online gaming and mobile TV can be delivered to the users. Moreover, by using more intelligent devices which are equipped with multi-interfaces, the users have the possibility and the opportunity to use different wireless communications at anytime and anywhere.
The coexistence of various access points driving from different wireless communications building a heterogeneous wireless network environment (HWNE). The major advantage of this HWNE lies on the complementary characteristics between different access networks. This advantage permits to satisfy the needs of the users in terms of real time applications. Nevertheless, one of the major challenges concerns this HWNE is achieving ubiquitous connectivity for the users. To cope with this issue, the vertical handover protocol (Lahby et al., 2012) represents a promising solution in this HWNE. Indeed, this protocol has been deployed in order to achieve always best-connected service (Gustafsson & Jonsson, 2003) when users roam from one access point to another. The vertical handover protocol has a three different phases: handover initiation, handover decision, and handover execution.
In this paper, we focus on the handover decision phase, because, it plays a crucial role in vertical handover protocol. In this phase, which is called also network selection, the mobile terminal should be make a decision in order to select from the reachable wireless networks, the best serving network in terms of QoS. However, the major trend of this phase, is how evaluating the available networks, for making a decision according to different criteria such as battery, velocity, QoS level, security level, users preferences, perceived QoS, etc. According to literature review (Yan et al., 2010) different algorithms can be used to deal with the network problem. All of these existing algorithms have considered only one access network when the mobile terminal roam from one network to another. To the best of our knowledge, up to the present day, there exists no network selection algorithm which allows the selection of the best path in terms of QoS from a source node to a target node during the network selection decision. In this paper, our major contributions are as follows:
- •
We have reviewed current existing related work proposed in the literature to model the network selection issue;
- •
We have introduced the graph theory to model the network selection problem for heterogeneous wireless networks;
- •
We have developed a new cost function based on dynamic and static criteria to quantify the weight edge between two nodes;
- •
We have developed a new network selection policy based on Dijkstra's algorithm and AHP method to choose the best path;
- •
We have tested and evaluated the performance of our policy in real testbed by using Mininet platform.
The remaining of this paper is organized as follows. The related work concerning vertical problem is presented in Section 2. Our proposed policy to model and to solve the vertical handover problem is described in Section 3. Section 4 includes our Testbed and the experimental results. Finally, conclusion and future work are presented in Section 5.
TopThe network selection decision is extremely important in such HWNE, because it allows ensuring ubiquitous mobile Internet for the users. For this reason, many algorithms have been proposed in the recent years to model and to optimize this process. The authors (Wang, & Kuo, 2013) have surveyed different mathematical theories proposed in the literature for modeling the network selection issue. These theories can be classified into seven approaches: utility function approach, cost function approach, fuzzy logic approach, Markov chain approach, combinatorial optimization approach, and multiple attribute decision making (MADM):