Xavier De Foy

Mobile Edge Cloud System: Architectures, Challenges, and Approaches

Mobile edge cloud (MEC) is a model for enabling on-demand elastic access to, or an interaction with a shared pool of reconfigurable computing resources such as servers, storage, peer devices, applications, and services, at the edge of the wireless network in close proximity to mobile users. It overcomes some obstacles of traditional central clouds by offering wireless network information and local context awareness as well as low latency and bandwidth conservation. This paper presents a comprehensive survey of MEC systems, including the concept, architectures, and technical enablers. First, the MEC applications are explored and classified based on different criteria, the service models and deployment scenarios are reviewed and categorized, and the factors influencing the MEC system design are discussed. Then, the architectures and designs of MEC systems are surveyed, and the technical issues, existing solutions, and approaches are presented. The open challenges and future research directions of MEC are further discussed.

Experimental Study on Deployment of Mobile Edge Computing to Improve Wireless Video Streaming Quality

Due to the rapid increase in wireless network traffic, especially video traffic, innovative network architectures and algorithms need be developed to reduce congestion and improve the quality of service (QoS). Multi-access edge computing or mobile edge computing (MEC) is a new paradigm that integrates computing and storage capabilities at the edge of the wireless network. In this paper, we design and implement a wireless access network-aware video streaming system based on the MEC concept, called Edge-Controlled Adaptive Streaming (ECAS). ECAS employs in-network video bitrate adaptation to improve data delivery efficiency. In our design, a MEC function intercepts HTTP requests from the client, and a rate adaptation mechanism is employed to decide the best video representation for the client. Updated HTTP requests, with modified video rates based on the adaptation decision, are forwarded to the video server. We also design a resource allocation and streaming bitrate adaptation algorithm to achieve the overall optimization of multiple video streams, with fairness, subject to wireless transmission capacity constraint. This network-assisted adaptive streaming approach allows more accurate estimation of wireless network states and can enhance the QoS of multiple video streams. A prototype is implemented to prove the concept, and the experimental results demonstrate that the proposed scheme significantly improves video streaming performance compared to the de facto standard video streaming technique, Dynamic Adaptive Streaming over HTTP (DASH).

Use of Bloom Filters to convey identity of surveillance targets

Local data offload performed at a small cell is not commercially deployed as a result of lawful interception issues. Previous work has described a solution to address these issues where a local node performs the same lawful interception functions as is done within the mobile core network for traffic that is offloaded at that local node. However, one issue with this solution is how to transfer the identities of those under surveillance to the local node. Passing the actual identities of targets of surveillance could comprise the required secretive nature of the surveillance. Another issue with the original solution is that by examining the traffic through a small cell, an unauthorized person could determine that a users traffic is not being locally offloaded; thereby perhaps indicating that person is a target of surveillance. In this paper, we propose the use of Bloom Filters to convey the identities of those subscribers who are the target of surveillance. The solution is presented and an analysis is included to demonstrate the benefits of the solution. This paper demonstrates that the use of the Bloom Filter hides the identities of the subscribers who are under surveillance. As well, the paper demonstrates that the false positives that occur with a Bloom Filter are actually a benefit from a perspective of obfuscating who is the actual target of surveillance.