Fuzzy logic vertical handover algorithm for heterogeneous wireless networks

We live an age in which we use wireless networks every moment of our lives. Recently, the need for data usage has triggered the development of mobile technology. Wireless networks that are diversified and developed according to their needs have advantages over each other. Bringing these advantages together to providing a more suitable use of technology has created an important working environment for researchers. Wireless mobile communication networks have been improved in a dazzling style by researchers. Improving journey of wireless technology has begun with 1G in 1980s. Because 1G technology supports only analog communication and low bandwith, 1G technology has been replaced with 2G at 1990s. Although 2G technology increases bandwidth and supports digital communication techniques, it does not fulfill increasing demand of high speed data transfer. After 3G technology has come out at 2000s, 2G's speed of data transfer has been improved. Multimedia applications and internet of things technologies have been developed same time with mobile communication networks. Because of increasing data transfer speed hunger, 4G technology improvements have followed 3G networks since 2010s. Internet of things technology provides communication to things via internet. Earth population increases day by day so things that have internet connection increase. 5G technology provides much more data transfer than 4G that prepares suitable network place for internet of things. 5G technology is expected to appear in 2020s. Different wireless technologies can be used together. Network which includes different technologies, is called heterogeneous network. Heterogeneous networks have been developed parallel with wireless communication networks. Heterogeneous networks keep an important development place for researcher. In this work, WLAN and 4G networks are used to create heterogeneous network. Although creating heterogeneous networks is the main purpose, ensuring of seamless working of different technologies is very important. If there is not seamless environment on different network infrastructures, created network will be useless. Considering these approaching, 4G and WLAN are used for network then handover decision algorithm is created for seamless working of different technologies. In the real world, different results emerge for changing situations. The use of fuzzy logic for uncertain situations is a very popular implementation. Classical approach is Aristotelian standard that diversify structures to two situations. These two situations are true or false which equal to {0,1} cluster in math's world. In the other hand, fuzzy logic says this approaching is missing because of uncertain situation. Fuzzy logic defines a lot of situation like maybe true, true, maybe false, false and endless options which equal to [0,1] cluster in math's world. Human age situation can be given as an example of these two approaching. According to classical approach, human's age can be diversified as old or young but fuzzy logic approaching helps to diversifying to child, young, teen, middle aged and old. Light of this example it can be said that nature rules are written with uncertain situation so these situations are needed to take care. In addition to nature example, fuzzy logic has endless usage areas that are digital cameras, washing machines, automobiles and televisions. Fuzzy logic has three steps. In first step, inputs are sent to fuzzifier in which inputs are diversified to different classes. Second step is inference. Rules are written to give logic to system. In second step fuzzy output sets are obtained by using rules. Defuzzifier is the last step. Last outputs come after third step. Uncertainty is also encountered in the heterogeneous wireless network environment. In these uncertain situations, the available standard Aristotelian algorithms do not respond well. Therefore, the fuzzy logic of Lütfi Aliasker Zade needs to be used in the handover algorithms. In wireless network environment, there are uncertain situations like capacity of network, traffic density, RSS, mobile terminal's speed, consumed energy and spending money. Fuzzy logic gives better analyses than Aristotelian approaching when it is used in wireless network. The light of the emerging needs, fuzzy logic algorithm is designed to make more efficient handover by using IEEE 802.21 that enables 4G and WLAN technologies to work together. People are social and mobile creatures. Two features of people bring mobile communication needs that are satisfied by cooperating of same technological infrastructure or different technological infrastructure. In order to mobile users make handover, wireless networks with the same or different technological infrastructure need to work together. If the same technological infrastructure is used, this handover is called horizontal. On the other hand, vertical handover is called when different infrastructure is used. When handovers occur, there is interruption on the mobile user side during data usage. The IEEE 802.21 standard has been designed and started to be used by the IEEE group because of reducing the number of discontinuities. This standard is ideal for using same or different technological infrastructures together. IEEE 802.21 standard is placed between network layer and data link layer. Because of the place, IEEE 802.21 standard is called with two point half layer. There are features of IEEE 802.21 standard like making decision of handover, providing service continuity, finding-selecting network, managing power and securing identity. Clients of IEEE 802.21 are coded in all infrastructure such as mobile terminal, WLAN, 4G or 3G. These clients gives information to central structure to do its job. Central structure process the information. In the backplane of central structure, handover algorithms are run so that the IEEE 802.21 standard can handle handovers. 4G and WLAN IEEE 802.11 have technologically different characteristics. These technologies can be combined with the IEEE 802.21 standard to create a heterogeneous wireless network infrastructure that will be widely used in the future. As a result of creation of heterogeneous wireless network infrastructure; features of IEEE 802.11 technology, such as high bandwidth, cheapness can be combined with 4G technology's wide coverage area feature. In order to efficiently utilize the heterogeneous structure, it is necessary to use an algorithm that determines the handovers between different technologies. In academic world researchers focus on what a technology needs. Heterogeneous wireless technology needs handover logic algorithms generally. Literature investigation is made on handover algorithms according to this work's subject. There are a lot of handover algorithms which are designed for working on heterogeneous networks. Algorithms can be varied according to handover criterias, performance metrics or strategies. Common handover criterias are RSS, available capacity, money cost, power consumption and mobile terminal's specifications. When the results are compared, some parameters are important such as handover latency, handover counts, probability of failure handover, throughput, number of lost packets, the total amount of data received and battery usage. Algorithms can be diversified according to their decision strategic features that are classical, function based, user centric based, fuzzy logic, neural network, context aware and combined different systematics. All algorithms need to be making performance tests after designing procedures. There are performance metrics which show performance of algorithms. Comparing performance metrics results point out which algorithms are better. Classic algorithm research must include algorithm design, experiment and performance analysis. This work is made in the light of researches with articles' positive and negative features. Thesis includes designing heterogeneous network, creating fuzzy logic based vertical handover algorithm, simulating of classical and fuzzy logic based algorithms and analyzing performance of two different algorithms. Fuzzy logic handover algorithm works on heterogeneous network that is designed for this thesis. MATLAB program is used for fuzzy logic design step. Machine learning is the most important specification of fuzzy logic. MATLAB provides results that show how designed algorithm's machine learning feature works. Algorithm's fuzzy logic rule series have 27 rules. By using MATLAB, fuzzy logic algorithm is evaluated with 225 different entrance values. High count entrance is fulfilled by low count rule series. In the evaluation, results show that designed algorithm works well for machine learning. After designing step, fuzzy logic based handover algorithm needs a simulation on a simulation program. Therefore, performance analysis can be done with simulation results. Network Simulator program's second version is used for simulation step. WiMAX technology is chosen as 4G technology because WiMAX technology is the only 4G technology that is supported by Network Simulator program version 2.29. When simulator program support LTE or LTE advanced, all simulations can be made once again. It is supposed that same results will be given by NS2. NIST's patch is added to NS2 because of adding IEEE 802.21 standard. After that network topology is designed for simulations. Designed algorithm is coded and evaluated on designed network topology. Testing an algorithm is not sufficient for performance analysis. Therefore, received signal strength feature based classic handover algorithm NA is evaluated with the same simulation experiments of designed algorithm BMA. Although fuzzy logic has 27 rule series, 54 difference entrances are used for testing machine learning skills. Handover numbers, received data amount and total packet loss results are gathered from simulation experiments. Same performance metrics' results are gathered for classical algorithm NA. In the end of the simulations, the simulation results are compared with each other. Comparing two different algorithms' results show that fuzzy logic based handover algorithm BMA gives better results than classic based algorithm NA. Briefly, the use of the fuzzy logic handover algorithm is necessary for some ambiguous situations that exist in parameter usage. If fuzzy logic based algorithm is used for heterogeneous network, handover numbers will be optimized that will decrease packet loss and increase the amount of received data.