Big Data Challenges in 5G Networks Salman Rashid Faculty of Computing and Engineering Universiti Teknologi Malaysia Johor Bahru, Malaysia rashid.salman@graduate.utm.my Shukor Abd Razak Faculty of Computing and Engineering Universiti Teknologi Malaysia Johor Bahru, Malaysia shukorar@utm.my Abstract— Now a days, world is witnessing a huge flood of data due to ever growing heterogeneous traffic, mobile network subscribers and online services. This trend is evolving continuously at a rapid pace and diversely in the form of big data. Wide range of use-cases scenarios with diverse requirements brings huge challenges for 5G. One of the most important requirements for use cases is high scalability, ubiquitous connectivity with low latency and high data rate with optimal energy, are equally important in 5G. Big data analytics is required to process this huge amount of raw data and extract small sized and useful information. This information can be used by arbiters to make consistent decisions. The paper presents challenges of 5G technical scenarios, big data perspective and emerging technologies of 5G. This paper also provides a detailed overview of big data challenges, imminent in achieving 5G goals. Keywords—connection density; big data analytics; traffic volume I. INTRODUCTION Characteristics of 5G include extremely low latency, high speed data transfer and world-wide connectivity. These characteristics makes 5G agile to perform broad set of application scenarios: from high user mobility to pervasive video, from ultra-reliable to tactile Internet communication, from broadcast like services to massive internet of things (mIOT), from lifeline communications to broadband access everywhere. For enabling such diverse applications, ruthless improvements in 4G network is required: more connected devices (10-100 times), higher data rates (10-100 times), less latency (1ms), availability (99.999 %) [1], longer battery life (10 times), higher mobile data volume (1000), network management operational expenses (5 times less), energy consumption (10 times less) [2]. Field trial design and classic manual approaches are impractical due to new eco system where heterogeneous networks are expected to increase tremendously [3]. The 5G radio access and core networks need orchestration of its resources and to control the network efficiently, flexibly and with scalability, 5G networks will be based on SDN/NFV infrastructures. Currently, during normal operations by management and control functions a huge amount of data is already generated in 4G networks and expected more to come in 5G networks due to heterogeneity in layers, diversification process and technologies, the additional management and control complexity increases in SDN and NFV architectures, due to the initiation of internet of things and machine to machine (M2M) paradigm [4]. With the advent of such technologies, increase in variety of services and applications with different traffic patterns and quality of service/ quality of experience requirements. The time sensitive multimedia and highly bandwidth intensive services over the 5G networks bring new challenges, constraints and problems not encountered in 4G networks. A. Potential of 5G The functionalities and architecture of future 5G networks are expected to be agile in order to accommodate the heterogeneous requirements (latency, reliability and bandwidth) of 5G applications [5]. The expansion of the service scope of cellular networks includes a wide range of services such as mission critical machine type communication, Internet of things and mobile broadband has evolved towards 5G and beyond systems [6]. In 5G internet protocol (IP) is expected to reach about 20 Gbps per sector with ultra large content traffic. This ultra large content traffic will travel fast on the wired and wireless network. The low band spectrum is essential for various scenarios such as high mobility and seamless coverage as required in massive Machine Type Communication (mMTC) and ultra-Reliable Low Latency Communication (uRLLC). When enhanced Mobile Broadband (eMBB) is mature it will offer data rates of up to 20 Gbps and high number of users will experience data rates of 100 Mbps [7]. To handle vertical services, with storage, connectivity and communication solutions, 5G must be transformed radically, designed for a precise digital business case of different industries (e.g. energy, automotive, health care, and multimedia). To provide services to users keeping the same cost that user is paying today is set to be a serious challenge for mobile operators and vendors. As the expected data traffic is 1000x in the coming era. Wireless industry is facing a strange challenge. It has to improve infrastructure to 1000 times, without increasing the CAPEX and OPEX [8]. Data rate requirement for various applications shown in Fig 1, are different in 5G networks. 5G support verity of use case where as demand of data rates is also variable in each scenario. Data rate requirement for smart grids for switching windmills ON/OFF and control automation is between 10 Kbps to 1.5 Mbps. For factory automation where E2E data rate requirements are 1 Mbps for smooth and efficient working. As far as intelligent transport system is concerned the data rate expected from 5G is between 10 Mbps to 700 Mbps. The reason for such a high data rate is due to its safety and high connectivity with extremely low latency. Healthcare services include tele-diagnosis, tele-surgery, tele-rehabilitation its 152 978-1-7281-1340-1/19/$31.00 ©2019 IEEE ICUFN 2019 Authorized licensed use limited to: UNIVERSITY TEKNOLOGI MALAYSIA. Downloaded on January 22,2022 at 06:55:47 UTC from IEEE Xplore. Restrictions apply.