A utility-based fuzzy TOPSIS method for energy efficient network selection in heterogeneous wireless networks

Abstract

Mobile terminals in 4G heterogeneous wireless networks continuously undergo horizontal and vertical handovers. In order for a mobile terminal to be connected to a network in the best possible way in terms of QoS performance and energy consumption, access network selection as the main decision within the handover process is obviously crucial. This paper presents a novel method that takes into account user preferences, network conditions, QoS and energy consumption requirements in order to select the optimal network which achieves the best balance between performance and energy consumption. The proposed network selection method incorporates the use of parameterized utility functions in order to model diverse QoS elasticities of different applications, and adopts different energy consumption metrics for real-time and non-real-time applications. User preferences are easily configured for different application and situation contexts through the use of linguistic assessments and their representation as triangular fuzzy numbers. The aggregation of multiple criteria for the calculation of the overall rating of the networks is performed through the use of the Fuzzy Set Representation TOPSIS method that resolves the issue of inconsistency related to conflicting decision criteria and is modified through the use of the employed utility functions for the elimination of the ranking abnormality problem. Finally, simulations are conducted in order to demonstrate how the proposed method would work and confirm its suitability and effectiveness.

Introduction

The deployment of a whole range of wireless technologies (2G, 3G, WLAN) in combination with the emergence of innovative wireless systems such as WiMAX (World Interoperability for Microwave Access) and MBWA (Mobile Broadband Wireless Access) have laid the foundations of an IP-based Fourth Generation (4G) wireless communications infrastructure in which all technologies are integrated through a single IP-based core. Apart from promising continuous and pervasive metropolitan network coverage delivering broadband capabilities, this new wireless network solution makes possible the satisfaction of a continuously growing consumer demand for mobile internet and other broadband services, which would not be possible through the use of a single mobile technology such as 3G [2].

Services consumed by users in such wireless environments have diverse QoS requirements, must be delivered in an ubiquitous manner and, most importantly, are characterized by a power consumption that is relatively high. Since 4G mobile terminals roam freely across different wireless systems, they continuously undergo horizontal and vertical handovers. The former occurs when a mobile terminal moves across different access points of the same wireless system whereas the latter occurs when a mobile terminal moves from one wireless system to another. On the one hand, battery lifetime has not been increased in the same degree as the power consumption requirements. On the other hand, the issue of energy efficiency and economy has assumed nowadays a great social importance. Therefore, it is critical for these two reasons to optimize vertical handover decisions in an energy efficient manner balancing at the same time the trade-off between performance and energy consumption [17], [26].

The vertical handover process comprises of three steps: (i) handover initiation, (ii) network selection and (iii) handover execution. Handover initiation is triggered by an external condition such as signal degradation, critical battery level requiring a switch to a more energy efficient network, a command by a rule based system monitoring QoS network characteristics, or a user command. Network selection is the main component of the vertical handover process: the mobile terminal must be connected to a network in the best possible way according to the Always Best Connected principle [10]. This principle specifies that the selection of the best network must take into account both user preferences and a number of different criteria including QoS requirements of diverse applications (available bandwidth, packet loss, latency, jitter, BER, etc.), monetary cost, energy consumption, operator requirements, etc. and has guided the development of various vertical handover management systems [13], [29]. Finally, the last step, handover execution, refers to how the mobile device performs the transition (e.g. break-before-make or make before-break) from one network to another. The handover decision mechanism and process control can be situated in a network entity (Network-Controlled Handover – NCHO) or in the mobile terminal itself (Mobile-Controlled Handover – MCHO).

The above description makes evident that the optimization of energy consumption mainly refers to the network selection step of the handover process. Apart from selecting the most energy efficient network, a significant reduction of energy consumption may also be achieved through the reduction of network scanning operations that occur when the network selection process starts. Such a reduction may be achieved through the use of GPS capabilities, which can be used to store network coordinates, through the use of a Media Independent Information Service (MISS) repository of available networks and their static parameters which is defined by the IEEE 802.21 standard, through adaptive network scanning based on motion detection of the mobile terminal [17], etc. However, investigation of mechanisms reducing network scanning operations is out of the scope of this paper.

Therefore, in this paper, we focus exclusively on the core network selection process and propose a novel network selection method based on a modified fuzzy version of TOPSIS (Technique for Order Preference by Similarity to the Ideal Solution) [3] that takes into account both network conditions and user preferences as well as QoS and energy consumption requirements in order to select the network that achieves the best balance between performance and energy consumption with minimal user involvement. The TOPSIS method has been formerly applied to network selection problems in heterogeneous wireless environments [1], [25]. The major contributions of our approach are the following:

• The use of different energy consumption metrics for real-time and non-real-time applications according to their specific energy consumption characteristics.

• The use of a utility function to model diverse QoS elasticities of network applications. Utility functions for QoS and energy consumption criteria are also used to create the normalized decision matrix of TOPSIS for the elimination of the ranking abnormality problem.

• The use of linguistic assessments and triangular fuzzy numbers for the easy configuration of user preferences in different application and situation contexts.

• The adopted fuzzy modification of TOPSIS (FSR TOPSIS) has the advantage that it resolves the issue of the contradictory nature of the “principle of compromise” of TOPSIS. Network selection criteria may be conflicting (e.g. bandwidth/energy consumption) and therefore the FSR TOPSIS method is utilized in order to aggregate them in a consistent and theoretically sound way.

The remainder of this paper is organized as follows: Section 2 summarizes related work and Section 3 presents the proposed network selection method in detail. In Section 4 the simulation experiment that was conducted in order to evaluate the performance of the proposed method is described and its results are discussed. Finally, Section 5 concludes and indicates some future extensions of this research.