A Metric-Based Approach for Quality Evaluation in Distributed Networking Systems

Introduction

A distributed communication network provides transparent services across wired and various wireless networks such as wireless local area networks (WLANs), Wireless Metropolitan Area Networks (WMANs) and Wireless wide area networks (WWANs). From networking perspective, a distributed environment was envisioned as an integration of Wired, IEEE 802.11 WLANs and 3G/2.5G/2G/B3G all these cellular technologies, mobile WiMAX being the major player in the middle. However, the advancement of long-term evaluation (LTE) will play a fundamental role in this integrated architecture and will form the 4th generation (4G) or next-generation of wireless networks. The heterogeneous wireless access, the exclusive all IP-based architecture and the advanced mobility support are the key drivers of this generation (Hossain, 2008).

The distributed heterogeneous network oriented technologizes complement each other (Zahran, Liang, & Saleh, 2006). 3G and 2G-based cellular communication technologies are well-known for wide area coverage, complete mobility, and roaming. However, traditionally these technologies offer low bandwidth, and expensive data traffic solutions (Chalmers, Krishnamurthi, & Almeroth, 2006). LTE technology is developed in response to overcome the limitations of the conventional cellular technologies. Yet, this technology also has its own drawbacks in terms of satisfying service and QoS guarantee (Liu at el., 2015). To overcome such constraints, Third Generation Partnership Project (3GPP) has developed LTE Advanced (LTE-A) (Luis, 2006). This technology has enhanced the performance of current LTE networks by integrating various advanced techniques. A few to name are carrier aggregation, large-scale distributed antenna systems (DASs), various low-power nodes, millimeter-wave, and massive multiple-input multiple-output (MIMO) (Gosh, 2010). On the other hand, WLANs provide high data rate at low cost, but with limited coverage, whereas WiMAX delivers last mile mobile broadband access and backhaul for WLANs (Fangmin, Luyong, & Zheng, 2007).

Hence, from network designing, planning, and troubleshooting perspectives, a QoS evaluation method is required that can bring these domains into a common platform. Generally speaking, the key performance parameters for these technologies are not directly comparable. For example, the delay ranges of UMTS and WLAN are entirely different. Therefore, a high value of delay measured from a WLAN may not be considered elevated in a UMTS environment. On the other hand, the applications running over them have the same QoS characteristics regardless of the communication technology they utilize. Hence, in this context, an application-based QoS evaluation approach is more applicable than a communication technology-based performance evaluation approach. The most current studies aim to evaluate the QoS of each application or radio access network individually. This method is useful in case of homogeneous networks (e.g.: uses single access technology). However, for heterogeneous networks, the presence of multiple types of applications and access technologies make the assessment of the overall performance challenging.

On the other hand, security and QoS mechanisms in distributed networking are closely related to each other (Yogender & Ali, 2006). Security mechanisms in a distributed network can influence the Quality of service and vice versa. Some security mechanisms can reduce the network performance. Similarly, some QoS mechanism can impact the network security. A good QoS management and monitoring mechanism can prevent the network information from being compromised (Luis, 2006).

This paper proposes a unified metric based QoS evaluation method to unify multiple performance evaluation parameters and security metrics into a single measurement metric. The QoS measurement method is further extended. The metrics measure the performance of different entities in a network such as applications and radio access networks (RANs) and represent the measurement with a single value. This approach is evaluated using various simulation scenarios. The specific focus is on multimedia-based applications such as video conferencing (VC), video streaming (VS) and voice. These applications have distinct characteristics, necessitating special attention to their QoS evaluation methods.