What is Regional Computing? How Can it Minimize the Social Big Data Effects?
Regional Computing is similar to edge or fog computing, placed near the users’ end to process and store the data at the regional level, rather than migrating it to the cloud.
Social Media Platforms (SMP) commenced with the birth of Web 2.0 and are widely used nowadays (Hurlburt, 2012). Currently, over 3.78 billion people use social media. Statistics show that by 2025, this number will increase to over 4.41 billion users (Social Networks Statistics, 2021). It means that one in every two persons will be connected to SMP (Fig. 1 shows the social media users) (Social Networks Statistics, 2021). China, the current leading country in social media usage, is expected to reach 1.1 billion by 2025. Today, the technological world is rapidly shifting toward Web 3.0, which introduces Artificial Intelligence (AI) and Virtual Reality (VR) to the internet, especially SMP. Since SMP is widely used in marketing, this integration will greatly increase the number of users, in terms of AI robots (Lassila & Hendler, 2007).
The CISCO (Computer Information System Company) annual report states that by 2023, 66% of the population will be connected to the internet, and the landline average speed will increase to up to 110 Mbps. They also predict that the mobile average speed will increase to up to 44 Mbps (CISCO annual internet report, 2021). Major internet companies such as SpaceX and Google are working to minimize the turnaround time for big data. SpaceX has completely changed this game. Rather than depending on cables and mobile players, they provide internet connection directly from space (with low-orbit satellites).
Social big data is described by the 5 V’s: volume, velocity, variety, veracity, and value. The volume specifies the massive data; the velocity refers to the speedy generation of the data; variety shows the various types of data (i.e., text, audio, video and photos etc.); veracity is the accuracy of data; and the value refers to the importance of the data (Abkenar et al., 2021, Devi et al., 2020). SMP allows the upload and download of high-quality photos and videos. These platforms also allow live high-quality streaming. AI is emerging as a new trend that is expected to publish regularly in the near future (Appel et al., 2020, Capatina et al., 2020, Tan et al., 2020).
The aforementioned services of social media create two major challenges for the network and social media applications.
- Firstly, the massive use of social media puts an extensive workload on the network and increases the overall network delay, transfer time and cost. This overloads the network for other essential communications such as those from the government and the public streams (Li, Yu, Si, Wu, & Zhang, 2020).
- Secondly, dealing with this large number of users and the workload it puts on the social cloud servers is also a challenge. Due to this enormous workload, the cloud servers are overloaded, resulting in a loss of performance and high cost (Xiao, Liu, Li, & Li, 2020).
Pursuant to the forenamed problems, we conducted a survey to ask social media users about their connections. The survey questions were: (i) How many of your social network friends are from the local district? (ii) How many of your social network connections are from the local province/state? (iii) How many of your social network connections are from the country where you live? There were four options for each question: (a) 1%–25% (b) 26%–50% © 51%–75% (d) 76%–100%.
About 500 people participated in this survey. 30% responded that 100% of their connections were from the local district. 50% responded that 100% of their connections were from the local state, and 75% responded that 100% of their connections were from the country where they live. These statistics reveal that a large portion of the data does not need to migrate from the country/region. However, as a result of the unavailability of local servers, this burden is placed on the mainstream network. This massive workload uses the network and the SMP cloud servers without having any positive impact, resulting in low performance and high cost. This fact indicates that social big data must be filtered at the edge before being transferred to the cloud servers. To address these challenges, further investigations need to be carried out to explore these new trends.
In connection to the above survey and to build our hypothesis, the distance effects on delay and cost need to be calculated. For this purpose, we used Microsoft Azure, a leading online tool to check the network between regions around the world for a number of parameters such as delay and upload time of a specific workload. We created a workload of 100 KB and calculated the download time and the delay around the region. Result illustrates that the delay and upload time rises as the distance between the servers increase (Micro Soft Azure Speed, 2021).
As a result of the rapid expansion of smart devices, big data, and massive use of social media, there is a burning challenge to minimize the social network burden on the mainstream network to enhance the service quality of these platforms (Xu et al., 2020). Social big data minimizes these platforms’ performance. To cope with this, we need to engage their data on regional servers.
Therefore, the objectives of this study are to:
- Minimize the effects of the social media workload on the mainstream network.
- Minimize the effects of the workload on social media cloud computing servers to optimize the application performance and costs.
To meet the aforementioned objectives, this article proposed an RC paradigm, similar to edge or fog computing, placed near the users’ end to process and store the data at the regional level, rather than migrating it to the cloud (Aslanpour, Gill, & Toosi, 2020). The difference between regional and edge computing is that RC typically has a one-hop distance from the client. However, in the case of regional computing, it is placed at a country/state level and has a larger distance than edge computing. Regional computing is thus the intermediate level of edge and cloud computing. Furthermore, in comparison to edge computing, regional computing servers have a higher capacity for storage and processing, although this capacity is much lower with respect to cloud servers. This is because regional computing only stores and processes limited data for a limited time. It is transferred to the cloud on off-timing or when the post has become popular. Fig. 3 shows the proposed structure.
Regional computing drastically minimizes turnaround time, transfer time and costs. Social media cloud servers are located at a long distance, resulting in delays and large workload, leading to low performance and high costs. Regional computing can be used to solve this problem, which already has about 104 billion global investments in the form of edge and fog computing (Badshah et al., 2020, Badshah et al., 2019, Social Networks Statistics, 2021).
Results show that this significantly minimizes the workload on the mainstream network and the additional load on the social media cloud, resulting in a smooth performance of the network and social media platforms. The preliminary simulations revealed the great potential of this idea, which requires additional work. Our future plans include analyzing this idea on a large scale by making a prototype or using other powerful simulation tools to confirm the supremacy of this work. Furthermore, mobile cloud computing with the regional and cloud is looking very promising. We plan to shift social media toward mobile cloud computing.