Social big data management through collaborative mobile, regional, and cloud computing.

Area of Science:

• Computer Science
• Data Science
• Network Engineering

Background:

• Smart devices generate vast amounts of social big data (SBD), overwhelming mainstream networks and degrading performance.
• Current network infrastructures struggle with the high volume and localized nature of SBD, leading to increased costs and delays.
• A significant portion of SBD originates from local areas, not requiring immediate remote server migration.

Purpose of the Study:

• To propose a collaborative computing framework integrating mobile edge computing (MEC), regional computing (RC), and cloud computing (CC).
• To address performance degradation and cost issues caused by SBD on social media platforms (SMP).
• To optimize data access and transfer strategies for SBD.

Main Methods:

• Developed a hybrid computing model combining MEC, RC, and CC for SBD management.
• Implemented a tiered data storage and access strategy, prioritizing nearest available resources.
• Scheduled large data transfers to off-peak hours to alleviate network congestion during peak times.

Main Results:

• Demonstrated significant reduction in network delay and response times for social network platforms.
• Achieved substantial cost savings by optimizing data handling and reducing reliance on remote cloud servers.
• Effectively mitigated the performance impact of SBD on mainstream and social network infrastructure.

Conclusions:

• The proposed collaborative MEC, RC, and CC structure effectively manages SBD challenges.
• This hybrid approach enhances social media platform performance and reduces operational costs.
• Optimizing data locality and transfer schedules is crucial for handling the demands of smart device ecosystems.