A high‐accurate content popularity prediction computational modeling for mobile edge computing using matrix completion technology

Abstract

Mobile edge computing (MEC) is envisioned as a promising platform for supporting emerging computation-intensive applications on capacity and resource constrained mobile devices (MDs). In this platform, the task with high computing resource demand can be offloaded to edge nodes for computing. Moreover, the computing result can be cached to edge nodes. When other MDs request the task that has been cached, the edge nodes can directly return the result to MD. However, the storage capacity of edge nodes is limited, the effective task prediction and caching scheme is one of the key issues for MEC. In this article, a matrix completion technology based content popularity prediction joint cache placement (MCTCPP-CP) scheme is proposed to tackle this issue for MEC. On the one hand, the MCTCPP-CP scheme is the first scheme using matrix completion (MC) technology to content popularity prediction. It proved by experiments that the accuracy of using MC technology to estimate caching content is improved compared with the previous methods. On the other hand, a cache placement decision approach based on the benefit of unit storage is proposed. Extensive numerical studies demonstrate the superior performance of our MCTCPP-CP scheme. The key performance indicators such as task duration, hit rate, estimated error are better than previous schemes by about: 0.13% to 14.01%, 17.28% to 37.65%, and 8.17%. INTRODUCTIONIn the past decades, the number of mobile devices (MDs) has grown at an unprecedented, 1-3 billions of devices (eg, MDs, wearable devices, sensors, and other IoT computing nodes) are connected to the internet for a wide variety of mobile applications. [4][5][6][7] One of the important applications is computation-intensive tasks, such as in-vehicle videos and virtual reality (VR). 8 To process such computation-intensive tasks, many computing capacity and energy supply are required. [9][10][11] Unfortunately, due to the limited computation resources such as limited battery life and insufficient computing capacity, the users' quality of experience will be reduced when executing the computation-hungry applications in MDs. [12][13][14] To mitigate the excessive computing, task offloading has been widely recognized as an efficient Abbreviations: ANA, antinuclear antibodies; APC, antigen-presenting cells; IRF, interferon regulatory factor.

Computation offloading through mobile vehicles in IoT-edge-cloud network

“…In order to provide low processing delay for devices, the MEC computing is proposed to improve the performance of the system [16][17][18][19]. The MEC servers can process many tasks.…”

An active and verifiable trust evaluation approach for edge computing

“…For example, the computing and storage capacity of mobile phones now exceeds that of personal computers more than 10 years ago [7][8][9]. These huge changes have led to extremely huge computing and storage capacity of the edge network, so the current network computing center is transferred from the network center to the edge of the network, forming an emerging computing model such as Edge Computing (EC) or Fog Computing [1,[10][11][12]. Because most applications are based on the data sensed and acquired by IoT devices [13,14], many emerging applications sensitive to latency and bandwidth, such as virtual reality, augmented reality and infrastructure for smart cities, benefit from edge computing.…”

Energy-aware MAC protocol for data differentiated services in sensor-cloud computing

“…However, the development of big data networks is attributable to the development of numerous applications with several sensing devices [7][8][9]. A sensor-cloud system (SCS) combines cloud computing [10][11][12] with sensor nodes [13][14][15] to sense, collect, process, analyze, store and share data for various applications; this combination has largely promoted the development of big data networks [16][17][18].…”