Improving Fairness in Load Balancing Among Radio Access Technologies in Heterogeneous Wireless Network
Chapter One
AIM AND OBJECTIVES
The aim of the work is to develop an efficient decision-making algorithm to perform fair load balancing of radio resource among heterogeneous wireless technologies through the following objectives:
- To adopt a heterogeneous wireless network model of seven different co-existing seven wireless technologies; EDGE, HSPA, WiMax, HSPA+, WiFi G, WiFi and LTE from the work of Donoso et al (2014)
- To develop a two-step load balancing scheme comprising Randomized algorithm(RA) which is Monte Carlo based and Load leveling algorithm (LLA)
- To validate the performance of the developed load balancing scheme against Min-
- Max strategy as presented in Donoso et al’s work in terms of fairness index.
CHAPTER TWO
LITERATURE REVIEW
INTRODUCTION
To carry out the research, literature review was carried out, which served as a guide towards achieving the set goals. The review of literature is categorized in to two parts namely: review of fundamental concepts, and review of similar works, which are further discussed as follows.
REVIEW OF FUNDAMENTAL CONCEPTS
This is an overview of concepts relevant to the study. This includes the overview of Heterogeneous Wireless technologies, a brief on the candidate technologies considered in the research work, Load balancing (LB) as it applies to HWN, LB algorithms pertinent to this work.
Heterogeneous Wireless Technologies
This subsection presents a brief description of the different wireless communication technologies that are considered in this dissertation.
Wireless technologies can be grouped into four categories according to their ranges of coverage, namely: Wireless Personal Area Network (WPAN), Wireless Local Area Network
(WLAN), Wireless Metropolitan Area Network (WMAN) and Wireless Wide Area Network
(WWAN)(Ali, 2012)
In the June, 1999 issue of IEEE Personal Communications in titled “EDGE: Enhance Data
Rates for GSM and TDMA/136 Evolution” GSM is recognized as one of the second generation wireless technologies with worldwide success(FALL Report, 1999)
Today, three quarters of GSM networks support EDGE, representing more than 350 networks in approximately 150 countries. Because of the very low incremental cost of including EDGE capability in GSM network deployments, virtually all new GSM infrastructure deployments are also EDGE-capable and nearly all new mid- to high-level GSM devices include EDGE radio technology (Rysavy Research, 2008).
The UMTS is a third generation (3G) mobile communications system that provides a range of broadband services to the world of wireless and mobile communications. The UMTS delivers low cost, mobile communications at data rates of up to 2 Mbps. It preserves the global roaming capability of second generation GSM/ General Packet Radio Service (GPRS) networks and provides new enhanced capabilities. The UMTS is designed to deliver pictures, graphics, video communications, and other multimedia information, as well as voice and data to mobile wireless subscribers. UMTS also addresses the growing demand of mobile and Internet applications for new capacity in the overcrowded mobile communications sky. The new network increases transmission speed to 2 Mbps per mobile user and establishes a global roaming standard. UMTS allows many more applications to be introduced to a worldwide base of users and provides a vital
link between today’s multiple GSM systems and the ultimate single worldwide standard for all mobile telecommunications such as International Mobile Telecommunications–2000 (IMT– 2000).
Figure 2.3 gives the general network architecture of the UMTS, where CN, RNS and UE stand for the core network, radio network subsystem and user equipment respectively.
CHAPTER THREE
METHODOLOGY
INTRODUCTION
This chapter describes the detailed procedure carried out in developing the heterogeneous wireless network model and the algorithms used in solving the resource allocation problem in HWN.
METHODOLOGY
The following methodology was adopted in carrying out the research work: I. Adoption of:
- A heterogeneous wireless network model comprising seven different Radio Access technologies co-existing to serve many mobile users. TheseRadio Access Technologies are: EDGE, HSPA, WiMax, HSPA+, WiFi G, WiFi and LTE.
- Development of A two-step load balancing scheme constituted by Randomised algorithm which is Monte Carlo based and Load levelling algorithm.
- Implementation of (i) using system level simulator on Matlab 2013b.
- Evaluation and comparison of the system level performance of the developed algorithm in terms of Load balancing against RRA, LCA and
Min-Max strategy using the Jain’s fairness index.
SYTEM MODEL
As a result of limited nature of resources, an improper allocation would impair seriously on the performance of the network and consequently give a wrong perception to the users. So, there is a need to define a mathematical model that encodes user requirements and factor in some constraints to make it perform well when deployed in real life scenario.
Network Model
Let N, M and S be the sets of nRATs, m mobiles and s services that compose a Cellular System, respectively as shown in Equation 3.1. Additionally, let yj K, ∈[0,1] be a binary parameter that indicates if the service k of the mobile j is activated or not. We calculate the load of the network i (ai ) as the sum of demanded bandwidth ( Dk ) of each connected service (k),for each mobile (j) over the total capacity of the network channel (Ci ) (Donoso et al., 2014)
CHAPTER FOUR
RESULTS AND DISSCUSSION
INTRODUCTION
In this chapter, the achieved results in resource load balancing in terms of fairness is presented after been measured with the Jain’s fairness index. The achieved results are compared to those of Round Robin, Least connected algorithms and the results of Donoso et al (2014). In addition, the errors of the algorithms relative to the ideal fairness index value of unity were presented.
The behavior of the proposed algorithm is observed from 10 mobiles to 1000 mobiles. The sample size of the number of mobiles considered is from 10 to 1000 only.
Distribution of mobiles among the wireless technologies
This subsection presents a number of mobiles with the 7 wireless technologies (EDGE, HSPA, WiMax, HSPA+, WiFi G, WiFi, LTE) offering three services, namely; Voice, Data and Video. The mobiles are randomly distributed within the network.
CHAPTER FIVE
CONCLUSION AND RECOMMENDATION
INTRODUCTION
This chapter presents the conclusion and shortcomings encountered during the period of this research work as well as the direction for future research.
CONCLUSION
This research work studied the problem of radio resource load balancing in the field of Heterogeneous Wireless Network. It was established that even though many research works in the field are ongoing, the issue of fairness in balancing the loads across the coexisting wireless technologies is still demanding more attention as many previous research works are deficient in addressing that. `
The developed algorithm which is a two-step heuristic strategy made up of Monte Carlo based algorithm and Load Leveling Algorithm has proved to be a viable solution when compared to Round Robin, Least Connected Algorithms and Min-Max strategy.
The average Jain’s Fairness Index for RRA, LCA, Donoso et al., (2014) and DA respectively are: 0.7980, 0.8115,0.8874 and 0.9119. While their Average relative errors are 22.37%, 19.73%, 11.23% and 8.51% respectively. This proves novelty of the DA which is very close to the optimal value of 1.
SIGNIFICANT CONTRIBUTIONS
The significant contributions of the work are:
- The developed algorithm outperformed the Round Robin, Least connected algorithms and Donoso’s work with an achieved a fairness index close to unity (0.9119).
- The work measured the deviation of the achieved fairness from unity (1) which is the ideal value and got a relative error of 8.51%.
LIMITATIONS AND FUTURE WORK
Even though the research is promising in the area of load balancing in heterogeneous wireless network, it must be acknowledged that it is limited to a maximum of 1000 mobiles. The research work also assumes that all mobile equipment in the model support all the wireless access technologies, but in reality, some mobiles only support older technologies.
For future work, it is recommended that mobile equipment that do not support latest wireless technologies are considered in the design, therefore given the work more real life perspective as these equipment are unlikely to be phased out soon.
Furthermore, the scope of the research could be expanded to accommodate more user equipment beyond the number considered in this research.
REFERENCES
- Abduljalil, M. A. (2014). Resource Scheduling Algorithms in Long Term Evolution ( LTE ). IOSR Journal of Electronics and Communication Engineering (IOSR-JECE), 9(6), 50– 53.
- Alam, F., Thayananthan, V., & Katib, I. (2016). Analysis of Round-robin Load-balancing Algorithm with Adaptive and Predictive Approaches. UKACC 11th International Conference on Control (CONTROL) Belfast, UK.
- Ali, M. (2012). Load Balancing in Heterogeneous Wireless, Optimized load aware vertical handovers in satellite-terrestrial hybrid networks.
- Andersson, K. (2010). On access network selection models and mobility support in heterogeneous wireless networks. Retrieved from http://swepub.kb.se/bib/swepub:oai:pure.atira.dk:publications/87a0ca80-a551-11dfa707-000ea68e967b?tab2=abs&language=en
- Aruna, L., Tamilnadu, S., & Aramudhan, M. (2014). A Novel Survey on SLA based Load leveling in Cloud Computing Abstract : International Journal of Research in Computer and Communication Technology, 3(6), 658–662.
- Crow, B. P., Widjaja, I., Kim, L. G., & Sakai, P. T. (1997). IEEE 802.11 Wireless Local Area
- Networks. IEEE Communications Magazine, 35(September), 116–126. https://doi.org/10.1109/35.620533
