ThC3 – Service Oriented Optimisation

Thursday, 20 June 2019, 14:00-15:30, Room 3

 

Session chairVlatko Lipovac (University of Dubrovnik, Croatia)

 

In-Flight Entertainment and Connectivity in the 5G Era: The 5G ESSENCE Experimental Platform

Elisenda Temprado Garriga (Safran Electronics & Defense – Triagnosys Gmbh, Germany); Leonardo Goratti (Zodiac Aerospace, Germany); Antonino Albanese (ITALTEL S. p. A, Italy); Claudio Meani and Pietro Paglierani (ITALTEL, Italy); Olga Segou (ORION Innovations, Greece); Emmanouil Kafetzakis (Orion Innovations P. C., Greece); Christos Xylouris (Orion Innovations PC, Greece); Maria Spada (Wind Tre SpA, Italy)
Currently, there is a relentless activity to conduct trials of 5G technology in realistic situations, targeting use cases that are driven by vertical stakeholders. Aligned with this strand and capitalizing on the recent standardization efforts of 3GPP that recently froze Release 15, we provide overview of an industry grade 5G experimental platform that targets the next generation of In-Flight Entertainment and Connectivity (IFEC) services, when we consider the airplane is a self-contained edge environment. The platform is developed as part of the 5G ESSENCE project activities. The contributions of this paper are multiple: describe the IFEC market from a general standpoint, show the hardware and software constituents of the 5G platform and present two virtualized services, namely virtualized video transcoding and caching, which are forecast to play an important role for the future of IFEC on commercial flights.

 

A Novel Multi-Factored Replacement Algorithm for In-Network Content Caching

Lijun Dong (Huawei, USA); Richard Li (Huawei USA, USA)
The paper takes multiple new factors into consideration in designing the optimal replacement algorithm for in-network content caching, i.e. requester’s tolerance to common semantics information of requested content, latency requirement on content arrival. The paper proposes a new metric, named importance value, to indicate a content is important enough to be cached in the network. By formulating a 0-1 knapsack optimization problem, an algorithm based on dynamic programming is proposed (named MTT). The performance valuations show the proposed MTT algorithm improves over the existing Least Recently Used (LRU) and popularity based replacement algorithms in achieving shorter response time in various content popularity settings.

 

Application of Deep Learning Techniques to Video QoE Prediction in Smartphones

Carlos Cárdenas Angelat and Janie Baños Polglase (DEKRA, Spain); Cristian Vaca-Rubio and Mari Carmen Aguayo-Torres (Universidad de Malaga, Spain)
This paper evaluates the use of deep learning prediction models for assessing the quality of experience of video applications using input data from a single measuring point at the smartphone. Three architectures for deep neural networks, single task \ac{FC}, multitask \ac{FC} and convolutional, have been implemented. A set of four relevant metrics recommended by ITU-T models has been considered as outputs for the prediction models. The amount of data received by the device, a parameter which can be easily extracted from the majority of smartphones, is used as input to the machine learning algorithms. For the convolutional neural model, the amount of transmitted data is also used as input. The predictions models have been applied to evaluate YouTube applications on smartphones, but could be generalized to other services. The overall achieved performance on the test set is around 90\%, a very promising result as the generation of new labelled datasets is quite feasible at low cost thanks to the fully automated test setup developed.

 

Field Trials for Evaluating Advanced Digital Terrestrial MIMO TV Broadcasting System

Takuya Shitomi (Japan Broadcasting Corporation, Japan & Universitat Politecnica de Valencia, Spain); Akihiko Sato, Noriyuki Shirai, Tsuyoshi Nakatogawa, Masahiro Okano, Shogo Kawashima, Tomoya Ijiguchi, Hiroaki Miyasaka and Shingo Asakura (Japan Broadcasting Corporation, Japan); Tomoaki Takeuchi (Japan Broadcasting Corporation (NHK), Japan); Madoka Nakamura (Japan Broadcasting Corporation, Japan); Kenichi Tsuchida (NHK Science & Technology Research Laboratories, Japan)
The transmission performance of cross-polarized MIMO is being studied to determine how to increase the capacity of terrestrial broadcasting and make it more robust. As an initial work on MIMO transmission, the authors evaluated the performance by field experiments in two scenarios. Results showed that when the transmission capacity is increased by MIMO using transmission parameters with the same robustness as SISO, although the degradation in the required received power from the laboratory test is increased by about 0.4 dB, the transmission capacity can be doubled. We confirmed that the required received power can be improved by 9.3 dB when MIMO is introduced to enhance transmission robustness, i.e., maintaining the same transmission capacity as SISO. The field trials were performed using a transmission system for next-generation terrestrial broadcasting being studied in Japan, but it also shows promise for wireless communication systems that introduce MIMO into the HPHT network using the UHF band in urban areas.

 

Practical Cross-Layer Testing of HARQ-Induced Delay Variation on IP/RTP QoS and VoLTE QoE

Adriana Lipovac, Anamaria Bjelopera, Ivan Grbavac, Ines Obradović and Tomo Sjekavica (University of Dubrovnik, Croatia)
As the LTE PHY/MAC-layer IR HARQ and RLC-layer ARQ protocol, adopted in LTE, may create sudden ramping of the delay when retransmissions occur, the arising question is whether these significantly and dominantly contribute to the overall packet delay or not? With this regard, in addition to the TCP retransmissions, with LTE protocol stack, there are two more retransmitting layers – the PHY/MAC and the RLC, which may impose intolerable additional delay of data blocks and so of the IP and finally the RTP packets, which further on may significantly degrade the end-to-end voice QoE, especially when voice is to be transmitted over LTE. We propose and demonstrate a VoLTE QoS and QoE test procedure based on cross-layer protocol analysis (HARQ/IP/TCP-UDP/RTP) and perceptual speech quality measurements. It has been shown how this concept can be used as an integral QoS criterion during the installation, commissioning and maintenance of VoLTE networks.