Building a Big Data Experiment Platform for Lightweight IPv6 Terminals Breaks the Spatial-Temporal Limits of Teaching

The cultivation of talents in data science and big data technology has long been constrained by the limitations of experimental teaching environments. Traditional big data experimental teaching relies on professional cluster equipment in university computer labs, which not only involves cumbersome deployment and configuration of open-source software such as Hadoop, Spark and HBase, but also restricts learners to fixed time and location, making it difficult to meet the diversified learning needs of the mobile internet era. With the popularization of lightweight mobile devices such as smartphones and tablet computers, the demand for building a big data experimental platform that supports remote access and multi-terminal adaptation has become increasingly urgent. Against this background, a research team from Central South University has proposed a new architecture of big data experiment platform supporting lightweight IPv6 terminals, which effectively solves the pain points of traditional experimental teaching and realizes the goal of remote big data experiments on various lightweight terminal devices, marking a new breakthrough in the integration of IPv6 technology and big data experimental teaching in higher education.

Big data experimental teaching is the core link of talent training in related majors, but the construction of its experimental environment has long been a major challenge for colleges and universities. On the one hand, the construction of a big data cluster requires the installation and debugging of a variety of distributed computing and storage software, and the complex configuration parameters make it difficult for beginners to master, which not only increases the teaching burden of teachers, but also hinders the students’ understanding and practice of professional knowledge. On the other hand, most college students use laptops for daily learning, but the performance of ordinary laptops is far from meeting the requirements of building a distributed big data experimental environment. Limited by memory, CPU and other hardware resources, students can only carry out single-machine experiments on virtual machines, and it is difficult to complete multi-machine cluster deployment, which makes it impossible for them to truly experience and master the core technology of distributed big data processing. Although some commercial big data experiment platforms have appeared on the market to solve the problem of cumbersome basic construction of experimental environments, these platforms still have obvious shortcomings in remote access and terminal support. Most of the internal big data experiment platforms of colleges and universities only support intranet access, making it impossible for students to carry out experimental learning outside the campus; at the same time, these platforms are mostly developed for traditional PC terminals, and lack effective adaptation and support for lightweight mobile terminals, which is inconsistent with the development trend of mobile learning in the current education field.

The rise of mobile learning has put forward new requirements for the informatization construction of higher education. In recent years, with the rapid development of mobile internet technology, intelligent terminal devices have been deeply integrated into people’s daily life and learning, and the design of learning resources based on mobile devices has become an important research direction of mobile learning. For big data experimental teaching, if we can make full use of the portability and mobility of lightweight mobile terminals, and build a big data experimental platform that supports such terminals, we can effectively break the spatial-temporal limitations of traditional experimental teaching, allowing students to carry out big data experimental learning anytime and anywhere, and greatly improving the flexibility and efficiency of learning. However, the construction of such a platform faces a key technical problem: how to establish a stable, safe and efficient connection and identification between a large number of lightweight mobile terminals and the big data experiment platform. The traditional IPv4 technology has been unable to meet the demand due to the shortage of address space, and the problems of low data transmission efficiency and poor security also restrict the remote access of lightweight terminals to the big data platform. At this time, IPv6 technology, with its unique technical advantages, has become the optimal solution to this problem, providing a solid technical foundation for the construction of a big data experiment platform supporting lightweight terminals.

IPv6 technology, as the next-generation internet protocol, has three core technical advantages that are highly adapted to the construction of a multi-terminal big data experiment platform, making it possible to connect a large number of lightweight mobile terminals to the big data experiment platform in a safe and efficient manner. First of all, IPv6 has a huge address space, which fundamentally solves the problem of terminal address allocation. Different from the 32-bit address length of IPv4, IPv6 stipulates that the IP address length is 128 bits, and the theoretical number of available addresses reaches 2^128-1. This massive address space can ensure that each lightweight mobile terminal obtains a unique global IP address, which is crucial for the connection and identification between a large number of mobile terminals and the big data experiment platform. With the explosive growth of the number of mobile terminal devices, the address shortage problem of IPv4 has become increasingly prominent, and IPv6 has become the fundamental way to solve the address problem of terminal devices, laying the foundation for the large-scale access of lightweight terminals to the big data experiment platform. Secondly, IPv6 has reliable security performance, which can effectively guarantee the security of data transmission between lightweight terminals and the big data platform. The network layer of IPv6 can encrypt the user’s transmission data and verify the IP message, realize the authentication of the communication end, and greatly improve the security of the network. At the same time, IPv6 generates addresses according to passwords, which can maximize the prevention of address forgery, ensuring that the data transmitted between the lightweight terminal and the big data experiment platform will not be intercepted or lost, and realizing the secure access and data transmission of the lightweight terminal to the big data experiment platform. For big data experimental teaching, a large amount of experimental data and code need to be transmitted between the terminal and the platform, and the security of data transmission is directly related to the smooth progress of the experiment and the protection of experimental resources, and the security advantages of IPv6 just meet this core demand. Thirdly, IPv6 has efficient data transmission rate, which can realize the fast interaction between lightweight terminals and the cloud cluster of the big data platform. Compared with IPv4, which needs to carry a lot of redundant data in the message header, the fixed header of IPv6 is very short, which can effectively reduce the overhead of data transmission and improve the efficiency of network data transmission. In addition, the IPv6 address allocation method based on the clustering principle allows a single record in the routing table to represent a subnet of the router, which greatly reduces the length of the routing table in the router and improves the speed of the router forwarding data packets. For lightweight mobile terminals with relatively limited data processing capabilities, the high transmission efficiency of IPv6 can ensure the fluency of remote access to the big data experiment platform, making the experimental operation on the lightweight terminal as smooth as that on the traditional PC terminal, and solving the problem of slow interaction between the lightweight terminal and the big data platform due to low transmission rate.

Based on the technical advantages of IPv6, the research team from Central South University has designed a complete big data experiment platform architecture supporting lightweight IPv6 terminals, which integrates cloud platform technology and IPv6 technology, and divides the whole platform into three core modules: lightweight terminal, internet transmission layer and server cluster end, forming a closed loop of data interaction and experimental operation from the terminal side to the platform side. The lightweight terminal module mainly includes various mobile terminal devices such as smartphones, tablet computers and laptops, as well as IPv6 routers used to send terminal transmission information. These terminal devices are the operation entrance of users, and through the dedicated application client, users can complete all experimental operations such as code editing, data submission and experimental result viewing on the lightweight terminal, without the need to install complex big data experimental environment locally, which greatly reduces the hardware requirements for the user terminal. The internet transmission layer is the core bridge connecting the lightweight terminal and the server cluster end, and the most important feature of this layer is the adoption of a dual-stack network transmission mode supporting both IPv4 and IPv6. This design not only ensures the compatibility of the platform with the traditional IPv4 network environment, but also gives full play to the technical advantages of IPv6, realizing the stable, safe and efficient transmission of information between the lightweight terminal and the server cluster end, and ensuring the integrity and security stability of the transmitted experimental data and code. The server cluster end is the core processing and storage center of the big data experiment platform, which is composed of Master nodes, Slave nodes, distributed resource managers and IPv6 routers for receiving terminal transmission information. The cluster nodes are configured with mainstream big data processing software and frameworks such as Spark, Hadoop, Kafka and MySQL, which can provide a complete distributed big data experimental environment for users, supporting both single-machine experiments and multi-machine cluster deployment experiments, and meeting the experimental teaching needs of different levels and different stages.

In order to further optimize the user experience and functional applicability of the platform, the research team has refined the platform architecture into three more detailed parts: App application client, data transmission layer and server end, each part with clear functional positioning and perfect technical support, forming a systematic and hierarchical big data experiment platform system. The App application client is the direct operation interface for users on the lightweight terminal, and its functional modules are divided into code submission module, data submission module and data visualization module according to the actual needs of big data experiments. The code submission module supports the selection of programming languages and code input, allowing users to edit and submit big data experimental code on the lightweight terminal; the data submission module is responsible for the import of experimental data, supporting the upload of various types of experimental data sources; the data visualization module can display the experimental results in a visual way, and users can choose different visualization methods according to their needs, making the experimental results more intuitive and understandable. Users only need to download the dedicated big data experiment application client on various lightweight terminals supporting IPv6, complete user registration and login, and then establish a connection with the big data experiment platform through the data transmission layer based on IPv6 technology, realizing real-time data interaction with the big data cluster, and completing the whole process of big data experiments on the lightweight terminal.

The data transmission layer, as the communication link between the client and the server end, is built based on the core technology of IPv6, including IPv6-based data transmission, address conversion and tunnel technology, which fully guarantees the security and efficiency of data transmission between the two ends. The working process of the data transmission layer is highly standardized: the client sends an access request and its unique IPv6 address to the network, the IPv6 router receives and parses the client’s request data, and sends it to the internet transmission layer; the transmission layer finds the network interface where the big data platform equipment of the server end is located according to the access address IP, and then the IPv6 router of the server end receives and parses the access request, and finally establishes a stable and exclusive connection between the client and the server end. This set of transmission process fully utilizes the address uniqueness, security and high transmission efficiency of IPv6, ensuring that the experimental code and data submitted by the user on the lightweight terminal can be transmitted to the server cluster end quickly and safely, and the experimental results processed by the server end can be fed back to the client in real time, realizing the seamless connection of the whole experimental process.

The server end is the core functional body of the big data experiment platform, providing comprehensive resource services and technical support for big data experiments, and its server cluster is divided into Master main nodes and Slave slave nodes with clear division of labor and close cooperation. In terms of resource management and data processing, the server end supports a complete set of big data technology stack, including the resource management system based on YARN, the stream data processing based on Kafka, the data processing based on Spark and the machine learning based on SparkML, which can meet the experimental needs of big data processing, analysis and mining at different levels, from basic data processing to advanced machine learning experiments. In terms of data storage, the platform supports a variety of storage methods and databases such as HDFS file system, HBase, Redis and MySQL, which can store and cache various types of experimental data, ensuring the reliability and scalability of data storage. At the same time, the server end is equipped with a data retrieval mode based on ElasticSearch, which can quickly retrieve the experimental resources and learning content in the platform, greatly improving the efficiency of users in obtaining resources, and ensuring the richness of big data experimental resources and learning content. The perfect technical support and rich functional configuration of the server end make the platform able to meet all the needs of teachers’ experimental teaching and students’ experimental learning, and provide a solid technical guarantee for the smooth development of big data experimental teaching.

The construction of the big data experiment platform supporting lightweight IPv6 terminals has important practical significance and far-reaching educational value for the big data talent training in colleges and universities, which not only solves the practical problems in traditional big data experimental teaching, but also conforms to the development trend of informatization and mobile learning in higher education. First of all, the platform effectively solves the problem of equipment performance constraints in the big data experimental environment. Traditional big data experimental teaching has high requirements on the hardware performance of the user terminal, and most students cannot build a complete experimental environment due to the limitation of terminal performance. The new platform transfers the core experimental environment to the cloud server cluster, and the lightweight terminal only needs to complete the simple operation of code editing and data submission, which completely gets rid of the dependence on the performance of the user terminal, making every student with a lightweight mobile terminal able to carry out big data experimental learning, greatly reducing the threshold of big data experimental teaching. Secondly, the platform completely breaks the spatial-temporal limitations of big data experimental teaching. The traditional big data experiment can only be carried out in the university computer lab with fixed time and location, which restricts the students’ learning autonomy and flexibility. The new platform supports the remote access of lightweight terminals based on IPv6 technology, allowing students to carry out big data experimental learning anytime and anywhere as long as there is a network, whether in the dormitory, library or outside the campus, which makes the experimental learning more flexible and personalized, and effectively improves the learning efficiency and enthusiasm of students. Thirdly, the platform greatly reduces the teaching burden of teachers and optimizes the experimental teaching management. The platform has a perfect functional design for teachers and students: the student end can view learning materials released by teachers, communicate with peers about problems encountered in the experiment, and complete experimental tasks online; the teacher end can release electronic learning materials and arrange experimental tasks through the platform, and the platform can automatically view the students’ experiment completion, making the experimental teaching management more efficient and standardized. Teachers no longer need to spend a lot of energy on the construction and debugging of the experimental environment, and can focus more on the design and guidance of experimental teaching content, improving the quality of experimental teaching. Finally, the platform provides an important guarantee for the cultivation of applied big data talents, and meets the diversified and multi-terminal needs of teachers and students in the new era. With the rapid development of the digital economy, the society has an increasing demand for applied big data talents with practical operation ability. The platform provides students with a convenient and efficient experimental practice platform, allowing them to carry out a lot of big data experimental operations in their daily learning, which helps to improve their practical operation ability and innovative thinking ability, and cultivate more high-quality applied big data talents for the society. At the same time, the multi-terminal adaptation feature of the platform meets the learning habits of modern students who are used to using mobile terminals, and the personalized and flexible learning mode is more in line with the development characteristics of contemporary higher education.

The integration of IPv6 technology and big data experimental teaching is an important exploration of the informatization reform of higher education, and the construction of the big data experiment platform supporting lightweight IPv6 terminals is a concrete practice of this exploration. In the future, with the continuous popularization of IPv6 technology and the deep development of mobile learning in higher education, such a platform will be further optimized and improved in terms of functional expansion, user experience and technical integration. On the one hand, the platform can further integrate artificial intelligence and other emerging technologies, add intelligent experimental guidance and error correction functions, and provide more personalized experimental teaching services for students; on the other hand, the platform can expand the scope of terminal support, adapt to more types of lightweight intelligent terminals, and further improve the portability and flexibility of experimental learning. At the same time, the platform can also carry out inter-school resource sharing, realize the sharing of big data experimental resources between different colleges and universities, and improve the utilization efficiency of educational resources, promoting the balanced development of big data talent training in colleges and universities across the country.

The research and construction of the big data experiment platform supporting lightweight IPv6 terminals not only solves the practical pain points in the current big data experimental teaching of colleges and universities, but also provides a new idea and method for the informatization construction of experimental teaching in other engineering majors. In the era of digital economy, the integration of internet technology and higher education is an inevitable trend of educational development. Taking technology as the driving force, breaking the traditional teaching limitations, and building a more flexible, efficient and personalized teaching environment is the core direction of the reform and development of higher education experimental teaching. The successful exploration of this platform provides a valuable reference for the integration of next-generation internet technology and experimental teaching in higher education, and will surely promote the further development and innovation of informatization construction of higher education experimental teaching in China.

Author Information: Gao Jianliang, Gao Jun, Duan Guihua (School of Computer Science, Central South University, Changsha 410083, Hunan, China) Journal Name: Industry and Information Technology Education DOI: 10.3969/j.issn.2095-5065.2021.10.011

(Word count: 3896)